>
Vol. XXIV, No. 6 PSYCHOLOGICAL REVIEW PUBLICATIONS November, 1917
Psychological Review
EDITED BY
HOWARD C. WARREN, PRINCETON UNIVERSITY
JOHN B. WATSON, JOHNS HOPKINS UNIVERSITY (J.ofExp. Psychol.)
JAMLS R. ANGELL, UNIVERSITY OF CHICAGO (Monographs}
SHEPHERD I. FRANZ, GOVT. HOSP. FOR INSANE (Bulletin) AND
MADISON BENTLEY, UNIVERSITY OF ILLINOIS (Index)
ADVISORY EDITORS
R. P. ANGIER, YALE UNIVERSITY; MARY W. CALKINS, WELLESLEY COLLEGE ; H. N.
GARDINER, SMITH COLLEGE; JOSEPH JASTROW, UNIVERSITY OF WISCONSIN; C. H.
JUDD, UNIVERSITY OF CHICAGO ; ADOLF MEYER, JOHNS HOPKINS UNIVERSITY ; W. B.
PILLSBURY, UNIVERSITY OF MICHIGAN; C. E. SEASHORE, UNIVERSITY OF IOWA ; G. M.
STRATTON, UNIVERSITY OF CALIFORNIA; MARGARET F. WASHBURN, VASSAR COLLEGE.
VOLUME XXIV, 1917
PUBLISHED BI-MONTHLY BY
PSYCHOLOGICAL REVIEW COMPANY
41 NORTH QUEEN ST., LANCASTER, PA.
AND PRINCETON, N. J,
Entered as second-class matter July 13, ^897, at the post-office at Lancaster, Pa., under
Act of Congress of March 3, 1879
P7
1 PRESS or
THE NEW ERA PRINTING COMPANY
LANCASTER. PA.
CONTENTS OF VOLUME XXIV
January.
The Relation of Intelligence to Social Status. JAMES W. BRIDGES and LILLIAN E.
COLER, I.
Mental Tests with Delinquents and Australian Aboriginal Children. S. D. POR-
TEUS, 32.
The Psychological Concept of Clearness. E. B. TITCHENER, 43.
Compound Substitution in Behavior. S. B. RUSSELL, 62.
The Delayed Reaction in a Child. W. S. HUNTER, 74.
March.
The Laws of Relative Fatigue. RAYMOND DODGE, 89.
More Concerning the Temporal Relations of Meaning and Imagery, EDWARD C.
TOLMAN, 114.
Experiments on the Relative Efficiency of Men and Women in Memory and Reason-
ing. ARTHUR I. GATES, 139.
Individual Differences in Judgments of the Beauty of Simple Forms. EDWARD L.
THORNDIKE, 147.
Preliminary Report on the Relative Intensity of Successive, Simultaneous, Ascend-
ing and Descending Tones. A. P. WEISS, 154.
Discussion :
A New Method of Heterochromatic Photometry— A Reply to Dr. Johnson.
C. E. FERREE and GERTRUDE RAND, 159.
The Stanford (1915) and the Vineland (1911) Revisions of the Binet Scale.
SAMUEL C. KOHS, 174.
. May.
/ The Nature of Mental Process. HARVEY CARR, 181.
A Reformulation of the Law of Association. WALTER S. HUNTER, 188.
The Scientific Productivity of American Professional Psychologists. SHEPHERD
IVORY FRANZ, 197.
> The Psychology of Thinking in the Case of Reading. EDWARD L. THORNDIKE, 220.
The Similarity of Brothers and Sisters in Mental Traits. DANIEL STARCH, 235.
A Method of Recording Errors in Form Board Tests. E. K. STRONG, JR., and ED-
WARD P. GlLCHRIST, 239.
Discussion :
./Introspection versus the Subconscious. LILLIEN J. MARTIN, 242.
The Mnemonic Feat of the ' Shass Pollak.' GEORGE M. STRATTON, 244.
July.
TWENTY-FIFTH ANNIVERSARY OF THE AMERICAN PSYCHOLOGICAL ASSOCIATION:
Varieties of Psychological Experience. JOSEPH JASTROW, 249.
' The Need for Social Psychology. JOHN DEWEY, 266.
The Case of Self against Soul. MARY WHITON CALKINS, 278.
Relation between Structural and Behavior Psychology. A. P. WEISS, 301.
iii
iv CONTENTS.
Discussion :
Meaning and Imagery. THOMAS V. MOORE, 318.
Some Experiments in Motor Reproduction of Visually Perceived Forms.
GEORGE R. WELLS, 322.
September.
An Attempted Formulation of the Scope of Behavior Psychology. JOHN B. WAT-
SON, 329.
Relation between Functional and Behavior Psychology. A. P. WEISS, 353.
The Relation between Emotion and Its Expression. HARVEY CARR, 369.
The Theory of the Social Force*. H. G. KENAGY, 376.
The Mental Work Curve. DANIEL STARCH and I. E. ASH, 391.
Individual Differences in a Normal School Class. ROBERT A. CUMMINS, 403.
November.
Advance Adaptation in Behavior. S. BENT RUSSELL, 413.
Relevant and Irrelevant Speech Instincts and Habits. P. F. SWINDLE, 426.
A Preliminary Report on * Work with Knowledge versus Work without Knowledge
of Results.' GEORGE F. ARPS, 449.
The Behavior of the Human Infant During the First Thirty Days of Life. MAR-
GARET GRAY BLANTON, 456.
Discussion :
A Critique of the Yerkes-Bridges -Hard wick Comparison of the Binet-Simon
and Point Sea1™. FRANK N. FREEMAN, 484.
VOL. XXIV. No. i January, 1917
THE PSYCHOLOGICAL REVIEW
THE RELATION OF INTELLIGENCE TO SOCIAL
STATUS
BY JAMES W. BRIDGES AND LILLIAN E. COLER
Ohio State University
HISTORICAL
Among the many reports of investigations with the Binet-
Simon scale incidental references to the influence of social
status on intelligence are occasionally found; but heretofore
the study of this extremely important and interesting topic
has been relatively neglected, although Binet with his usual
acumen did not overlook the problem.
In 1910 Decroly and Degand1 tested forty-five children
of both sexes in a private school at Brussels. They found
that none of the children tested were below age, nine were
at age, and the rest from one to three years above the level
of their age. These results were very significant, since the
Binet-Simon scale is theoretically supposed to rate equal
numbers retarded and advanced with the mode and the
average at age.
Decroly's and Degand's results were carefully studied by
Binet.2 He thought the best explanation of the difference
between his results and theirs was found in the fact that the
Belgian children came from a private school in Brussels and
represented children of the well-to-do and largely professional
class while the Paris children were from a rather poor section
1 Decroly, O., et Degand, Mile. J., 'La Mesure de 1'intelligence chez des enfants
normaux d'apres les tests de MM. Binet et Simon,' Arch, de psychol, 1910, 9, 81-108.
2 Binet, A., et Simon, Th., 'Nouvelles recherches sur la Mesure du Niveau In-
tellectual chez les enfants d'ecole,' UAnnee Psychol., 1911, 17, 145-201.
I
JAMES W. BRIDGES AND LILLIAN E. COLER
of the city. The instruction in the Belgian school was also
more individual for the classes were very small.
In analyzing the Brussels results by the individual tests,
Binet found that the children from the better social class
scored higher in tests involving thought in the higher sense —
apprehension, criticism, comparison, etc. They also scored
higher than the less favored children in the tests which put a
premium on linguistic readiness — such as the description of
pictures, abstract definitions, comparison of objects, absurdi-
ties, and giving words for three minutes. Binet concluded
that social status must be closely correlated with mentality
and reckoned from the results in the two schools an average
difference of about one and one half years between children
of the better and poorer social classes. He does not take
into account, however, that this difference may vary with
different chronological ages.
Binet, moreover, criticized1 the work of Katharine John-
ston, who had examined two hundred pupils of the Sheffield
schools in England, because she had drawn her subjects from
at least three distinct social groups and had not kept these
groups separate in her averages.
Another study which seemed to confirm Decroly's and
Degand's results was the testing done by M. Morle2 in a
school in a poor part of Paris and compared with the results
from a school situated in a wealthy section. The study was
on a rather small scale as only thirty children were taken,
at random, from each school. The results were as follows:
Retarded
At Age
Advanced
2Yr.
i Yr.
i Yr.
2Yr.
Unfavored school
I
I
II
3
13
10
4
IO
I
6
Favored school
Thus sixteen children out of the thirty tested were ad-
vanced in the favored school while only five were advanced
in the unfavored school. The children from the poorer
1 UAnnee psycho!., 1911, 17, 195 196.
2 Morle, M., 'L'influence de 1'etat social sur le degre de 1'intelligence des enfants,'
Bull. Soc. libre Educ. Psychol. de I'enfant, 1911, 12, 8-15.
INTELLIGENCE AND SOCIAL STATUS 3
section were on the average about one fourth year behind
the level of their age, while those of the favored school
averaged from one fourth to one half year advanced, or a
difference of about three fourths of a year between the two
social classes.
In 1910 the teachers of the Breslau schools in Germany
made a comparative study of children of different social
classes there. The demand for a common school for all
classes to replace the Vorschule and Volkschule had arisen
in Germany. The Volkschule is the elementary public school
attended by the children of the laboring and lower business
classes, while the Vorschule is attended by the children of
the higher social classes. In Prussia the children could enter
the Gymnasium, which has a nine-year curriculum preparing
for the university after three years of preparation in the
Vorschule but only after four years in the Volkschule. The
purpose of this investigation was to find whether the mental
maturity of the child, as well as the curriculum, justified this.
One hundred and fifty-six boys were tested from the two
schools. The Binet-Simon scale modified by Bobertag was
used. The boys tested were seven and nine years of age.
from the Vorschule and seven, nine, and ten years from the
Volkschule. It was found that the nine-year Volkschule
pupils scored 10 percent lower than the pupils of the same
age in the better school, while the ten-year-old Volkschule
boys attained only the average of the nine-year-old Vor-
schule pupils. The difference in average was due largely to
the fact that the Vorschule pupils did nearly twice as well
as the Volkschule pupils of the same age in tests above their
age level. The tests at the age level were passed about
equally well by both schools. This raises the question of
whether children of higher social classes mature earlier than
those of the lower levels.1
Children from three very different environments were
tested in 191 1 by J. and R. Weintrob.2 There were about sev-
1 Hoffman, A., "Vergleichende Intelligenzpriifungen an Vorschulern und Volk-
schiilern," Zsch. f. Angew. PsychoL, 1914, 8, 102-120.
2 Weintrob, J. and Weintrob, R., "The Influence of Environment on Mental
Ability as Shown by the Binet-Simon Tests," /. of £ due. PsychoL, 1912, 3, 577-583.
4 JAMES W. BRIDGES AND LILLIAN E. COLER
enty children of both sexes in each group tested. Group A
consisted of children from a school attended by children of
the wealthy class, with every opportunity for travel, etc.
Group B was composed of children whose fathers were wage-
earners or small business men. Group C was composed of
children from a Hebrew Orphan Asylum with no real home
environment. The schools were compared as to the number
of children, testing above, at, or below the norm for their
age, using the Binet-Simon scale. The A group was found
to rank highest, the C group next and the B group last. The
investigators state: "Judging from the results environment
does not seem to affect greatly mental capacity, if at all."
Instead of the schools ranking A and B and C as might have
been expected, the C school or Jewish Orphanage ranked a
close second to the wealthy school. However, the question
of race enters very largely in this study as the children of the
Asylum were all Jewish, while those of Group A were pre-
dominantly American with a few Germans, Jews and Italians,
and Group B was largely composed of Germans, Italians and
some American children. It is very evident, as the investi-
gators say, that in order to judge fairly differences in environ-
mental influences among groups, the conditions within each
group must be uniform, and the same races must be judged.
A study involving social status, incidentally, was made in
Columbia, S. C. by Miss Strong.1 Her primary purpose was
to investigate the difference between the white and negro
children but in order to make a fair comparison she tested
white children in both the city schools and in the mill district.
The results show that less than six percent of the city school
children were retarded while eighteen percent of the mill
district children were mentally over a year below the level of
their age. None of the mill district children were above
their age level, although ten percent of the city children
scored above their years. Approximately the same percent
of the children in each district were at the level of their age;
eighty-four percent in the city schools and eighty-one per-
1 Strong, A. C., '350 White and Colored Children Measured by the Binet-Simon
Measuring Scale of Intelligence; A Comparative Study,' Fed. Sem., 1913, 20, 485-513.
INTELLIGENCE AND SOCIAL STATUS
cent in the mill district schools. Practically the same course
of study was used in the schools of both districts.
One of the most recent investigations on the subject of
social status was made by Yerkes and Anderson1 in Cam-
bridge, Mass. In this investigation the Yerkes-Bridges Point
Scale2 was used. Fifty-four children in the kindergarten and
first grade of school A were compared with children of the
same sex and approximately the same age in school B.
School A is located in a good neighborhood and the socio-
logical status of almost all the pupils is very good. School B
on the contrary is located in a medium to poor section of the
city and the majority of its pupils live in a rather poor en-
vironment. The children compared were all of English-
speaking parents.
The average number of points scored in the two schools
is indicated below:
Age
4 Yrs.
5 Yrs.
6 Yrs.
7 Yrs.
8 Yrs.
School
A..
JC
27
4.2
AQ
s6
School
B
17
22
29
35
4i
The favored school averages much higher except in the
four-year group. The very young children of the unfavored
group seem to have the advantage here, probably because
they are less timid. The results show that there is a difference
of from twenty percent to thirty percent in mental ability
which maybe associated with differences in sociological status.
The authors point out that in view of a difference so
marked between children of different sociological levels, it is
very unfair to judge them by the same norm and that further
investigating should be done with the view of establishing
norms for different social levels.3
1 Yerkes, Robt. and Anderson, Helen, 'The Importance of Social Status as Indi-
cated by the Results of the Point Scale Method of Measuring Mental Capacity,'
/. of Educ. Psychol, 6, No. 3, Mar., 1915.
2 Yerkes, Bridges and Hardwick, 'A Point Scale for Measuring Mental Ability,'
Warwick and York, 1915. Hereinafter referred to as the "Point Scale."
3 In a book published since this paper was written, Prof. L. M. Terman discusses
the influence of social status. He reports a difference of one to two years between
the superior and inferior classes — a result in close conformity with those mentioned
above, ' The Measurement of Intelligence,' pp. 72 and 115, Warwick and York, 1916.
6 JAMES W. BRIDGES AND LILLIAN E. COLER
EXPERIMENTAL
The Yerkes-Bridges Point Scale was used in this investi-
gation and three hundred and one children were tested in
two schools situated in very different localities of Columbus,
Ohio. The children from school A situated in the better
district will be designated as the favored group, while those
of school B will be designated as the unfavored group.
School A is in a very good residence section near the uni-
versity. The majority of the people own their homes, which
are surrounded by well-kept lawns. A portion of the uni-
versity campus, as well as many wooded lots afford ample
playground for the children. The school building is modern
in every respect, having been completed about six years ago.
The children of the first and second grades spend alternate
half hours in a well-equipped Portable where their play is
supervised by a teacher who has specialized in this work.
This school is considered one of the most desirable in which
to teach in the city and only well qualified teachers obtain
the positions.
The chief occupations of the fathers of the children in this
district were managers, proprietors and officers of manu-
facturies and stores, traveling salesmen, real estate and
insurance agents, and a professional group composed of pro-
fessors, doctors, lawyers, architects, and ministers. A more
complete analysis and grouping of the data by the occupa-
tions of the fathers will be given later in this paper. All the
children of the first and second grades were tested, and as
time did not permit completing the third grade, the children
were taken alphabetically. All the children in the grades
tested were American born and of English-speaking parents.
The testing was done in a hall where occasionally some one
passed but otherwise there was no disturbance; and no third
party was present when the examination was made. The
child's name, date of birth and father's occupation were
recorded in every case and checked by the teacher's record.
School B is situated near the railroad in a poor factory
district of Columbus. The houses average about four or five
rooms and are usually built very near the street. They are
INTELLIGENCE AND SOCIAL STATUS 7
often in very bad repair, having been in that part of Columbus
which was flooded three years ago. The usual rent is from
38 to ?io a month. Where there are yards they are so ill-
kept and muddy that the street is the common playground
f r the children. The school house is old and has no inside
toilet facilities or up-to-date equipment. There is one saloon
on the corner opposite the school building and two others
within a block and a half of it. The fathers, if still in the
family, are receiving low and irregular wages and almost one
half of the fathers of the children in this district belong to the
unskilled and casual labor group. The remainder were in
the more skilled mechanical trades or were teamsters or
delivery men. The mothers are often away all day working
in the factories or doing laundry work to supplement the
husband's income, or in many cases to support the family
entirely. The parents have little idea of the value of educa-
tion and the children often stop school and go to work as
soon as they can secure their working papers. The Asso-
ciated Charities say that probably 50 percent of the families
in this district are registered with some kind of philanthropic
organization. The children are often very poorly clad, far
from clean and frequently undernourished. Many of the
teachers in this school are young and have not taught a great
while.
In the case of several children scoring lowest, other mem-
bers of the family were tested. In five cases the mothers
scored only from 47 to 54 points or a mental age of about
eight years, and sisters and brothers were far below their age
level.
The negroes and children of non-English speaking parents
were excluded in this school, but every other child in the
first, second and third grade was tested, making a total of
136 children. The testing was done in a small room free
from all disturbing elements.
In both schools the tests were given during school hours
by one examiner1 and in all cases doubtful credits were dis-
cussed and decided upon by the authors jointly. The child's
1 Miss Coler.
8
JAMES W. BRIDGES AND LILLIAN JE. COLER
age to the nearest month was determined by subtracting the
date of the birth from the date the tests were given. A given
age group includes all children from the middle of the year
below to the middle of the one above. Thus, in the group
of six-year-olds are included all boys (or girls) from five
years seven months to six years six months, inclusive.
RESULTS
The results for the total 301 Columbus school children
will first be considered and their scores compared with the
scores for Cambridge, Mass., school children of the same
ages. These results are presented in Table I. The first
TABLE I
Age
Columbus
Cambridge
English
Non-English
No.
A.
M.
No.
A.
No.
A.
6
8
9
10
ii
12
37
97
81
59
H
7
5
34-i
42.8
54-7
57-3
55-9
49.6
54-8
33
45
55
59
56
50
55
55
48
47
43
53
55
40
29
35
4i
56
62
65
77
16
25
14
3i
23
24
20
27
31
37
48
I6
62
67
A. — Average.
M. — Median.
column gives the ages, the second the number of Columbus
children tested at each age, the third and fourth the average
and the median scores for Columbus children at each age,
while columns five and six give for comparison the number
and average scores for Cambridge children of English-speaking
parents and columns seven and eight the same for children
of non-English-speaking parents.1
The results are shown graphically in Fig. I. Graph A is
for Columbus children, both schools combined; Graph B for
Cambridge, English-speaking group, and Br for Cambridge,
non-English-speaking group. After nine years the number
of Columbus children at each age is very small and after ten
years composed of children from the unfavored school alone.
1 Point Scale, pp. 66-67.
INTELLIGENCE AND SOCIAL STATUS
This selection is clearly shown by the drop in the curve after
nine years, and is due to the fact that in our present school
system a child is behind if he is over nine years of age and in
the third grade.
Points
80
75
70
65
60
55
50
45
40
35
30
Ages 6
9 10 ii
FIG. i.
12
The most striking feature in these results is the evident
superiority of the Columbus children up to ten years; but it
will be shown in the sequel that their superiority is entirely
due to the influence of the favored group and that when the
10
JAMES W. BRIDGES AND LILLIAN E. COLER
groups are separated the correspondence between the children
of the two cities is even closer than might have been expected.
TABLE II
Age
6 Yr.
7Yr.
8 Yr.
9Yr.
10 Yr.
Cambridge Heterogeneous
Total Columbus
Number. . . .
Score
Number. . . .
71
29
37
73
34
07
6l
39
Si
74
52
CQ
76
59
Combined Average
Score
Number
34-1
108
42.8
1 7O
54-7
142
57-3
i>ii.
55-9
QO
Score .
30.7
39-0
48
54-4
?8.S
The Columbus results from the sixth to the tenth year
inclusive may be combined with the Cambridge results for
the same ages to help in the standardization of the scale.
Points
60
Httey
Ages 6
9
FIG. 2.
ii
12
At first the Columbus results were combined with the Cam-
bridge heterogeneous group1 in the following way. The
number of children at each age in the group was multiplied
by the average score of that age for Columbus and Cambridge
1 Point Scale, pp. 64-65.
INTELLIGENCE AND SOCIAL STATUS
ii
separately. Then these results were added and divided by
the total number at that age. Table II. shows the average
score and number of pupils at each age in each group and the
combined averages or new norms obtained. The latter are
also shown graphically in Fig. 2. The solid line represents
the combined norm for the total 643 children of Columbus and
Points
60
55
45
40
35
Ages 6
8
FIG. 3.
Cambridge; and the dotted line is the original norm from six
to ten years for the 355 Cambridge children alone.
As all the children tested in the Columbus schools were of
English-speaking parents it seemed fairer to use as a standard
for judging the Columbus pupils a norm made from the Co-
lumbus results combined in the same way with the Cambridge
group of children of English-speaking parents,1 instead of the
heterogeneous group.
These norms are shown in Table III. and graphically in
Fig. 3. In this case the solid line represents the combined
1 Point Scale, pp. 66-67.
12
JAMES W. BRIDGES AND LILLIAN E. COLER
norm for 245 Cambridge children of English-speaking parents
and 288 Columbus children, or a total of 534 children. The
broken line shows the original norm for the Cambridge
English-speaking group alone. Selection entering after nine
years in the Columbus results will explain the slight drop in
the solid curve from nine to ten years. This curve was used
as a norm for determining the mental age and coefficient of
TABLE III
Ages
6Yr.
?Yr.
8 Yr.
9Yr.
10 Yr.
Cambridge English-speaking
Number
re
4-8
47
4-1
C-2
Score
20
•2C
4.1
56
11
Total Columbus
Number
37
Q7
81
CQ
14
Score
24 I
4.2.8
C4..7
cy.-j
ce.o
Combined Average
Number. . . .
Q2
I4C
128
IO2
67
Score
3I-I
40.2
49-7
56.8
60.7
mental ability of the Columbus children. For scores above
or below the six, seven, eight, nine and ten year averages the
norm for the English-speaking group of Cambridge alone
had to be used.
COMPARISON OF DIFFERENT SOCIAL GROUPS
The Columbus results from Schools A and B will now be
compared. Tables IV. and V. show the individual scores
by sex and age for the favored and unfavored schools re-
spectively. The average score for each age and each sex
and also for both sexes combined are likewise given in each
table.
These results are shown graphically by Fig. 4, curves A
and B, The favored school is from 21 percent to 32 percent
superior to the unfavored school varying with the chrono-
logical age. The curve drops at nine years in School A and
at ten years in School B> showing that the selection previously
mentioned enters earlier in the former. After ten years the
numbers in the unfavored group are very small and the
scores are not representative of these age groups.
As these results are very similar to those found in the
Cambridge investigation it is interesting to compare them
INTELLIGENCE AND SOCIAL STATUS
TABLE IV
FAVORED A
6
r
3
<
>
i
0
ff.
V.
fc?
(g)
(M6)
(J7)
(M6>
(F4?
it
w
29
31
22
32
49
—
44
54
66
39
30
36
51
4i
55
68
36
39
34
35
38
51
52
45
5°
52
62
49
64
39
35
40
52
49
56
63
75
40
43
52
49
60
63
42
42
40
45
53
Si
61
65
47
43
41
45
53
Si
62
47
42
49
55
55
63
67
48
52
43
49
55
55
64
68
54
45
49
55
55
64
68
47
49
56
56
66
71
A *7
C*7
60
4/
bf
°v
47
50
r T
58
58
rr\
69
72
48
51
51
60
59
59
73
73
48
51
62
77
49
52
63
60
79
52
64
60
So
52
65
62
50
53
66
63
52
53
66
63
54
55
67
65
54
59
69
66
54
72
67
54
V
67
55
75
69
55
76
56
73
56
58
58
59
60
'64'
64
66
72
Ave 39 8
Q
., Q 7
co 8
61 i
64.8
67
56.8
41 »9
43 • -I
40. /
by-0
5 / *9
Uj.l
u/
40.9
48
•4
58
.8
6
4
6c
>.2
= 25 percent above or below general norm.
= 25 percent above or below group norm.
JAMES W. BRIDGES AND LILLIAN E. COLER
TABLE V
UNFAVORED B
6 Yr.
7 Yr.
8 Yr.
9Yr.
10 Yr.
W
¥!
s?
(K>
&>
(F4)
5S?
(F8)
fc!
V.
ffi
V
Si!
V2.'
is>
16
21
22
23
23
17
23
14
21
25
2*
26
26
27
27
28
29
29
29
19
22
22
24
'28*
22
25
27
'36"
39
44
45
45
46
48
50
52
35
35
35
'38'
24
42
36
48
46
46
53
57
55
53
43
45
38
42
44
45
48
49
49
5i
52
57
57
57
54
57
60
50
52
61
60
50
II
72
28
29
29
42
44
2
47
50
50
5i
53
54
54
58
f?
61
V
75
43
47
49
52
53
58
58
59
25
25
11
29
33
34
39
55
34
37
37
43
44
45
45
54
58
32
32
35
37
40
4i
4i
42
43
44
62
62
68
50
54
Ave 28.3
27.2
33-2
36.6
43.1
494
47-9
52.2
53-5
52
53-2
46
52
57-5
_53_
53
28.3
34-4
46.2
50.1
52.8
52.1
54-2
= 25 percent above or below general norm.
=25 percent above or below group norm.
with the two similar groups there.1 Table VI. shows these
results and they are also shown graphically in Fig. 4. Only
the kindergarten and the First Grade were tested in the
Cambridge favored school so the curve is short.
TABLE VI
Favored School
Unfavored School
Age
6Yr.
7Yr.
8Yr.
9Yr.
10 Yr.
6 Yr.
7Yr.
8 Yr.
9Yr.
10 Yr.
Cambridge . .
Columbus . . .
Score . .
Score . .
42
40.9
49
48.4
56
58.8
29
28.2
35
34-4
&
56
SO.I
62
52
64
60.2
1 Point Scale, p. 74, p. 66.
INTELLIGENCE TO SOCIAL STATUS
Points
80
75
70
65
60
55
5°
45
40
35
30
Ages 6
9 10 ii
FIG. 4.
TABLE VII
12 13
School A
School B
M.
F.
M.
F.
Cambridge
Columbus
No
Score
II
42
8
39-8
13
40
9
41.9
29
29
14
28.3
26
36
27.2
No
Score
These differences between the favored and unfavored
groups and the striking similarity between the Columbus
i6
JAMES W. BRIDGES AND LILLIAN E. COLER
and Cambridge results are even more clearly shown by the
data for the six-year-old groups given in Table VII.
COMPARISON OF SEXES
Tables IV. and V. should be consulted for the individual
scores. The averages for boys and for girls in both schools
combined are shown in Table VIII.
TABLE VIII
6 Yr.
7Yr.
8 Yr.
9Yr.
10 Yr.
Males
•72 C
42 2
CA
c6 Q
t;8
Females
36.3
44.1
55
57-7
544
Fig. 5 shows these results graphically. The superiority
of the girls over the boys seems marked up to ten years. By
examining Table IX. and Fig. 6, which give the averages for
Points
60
55
45
40
35
Ages 6
10 II
FIG. 5.
12
each sex in each school separately it will be seen that this
superiority of the girls is brought about almost entirely by
the girls of the unfavored school.
INTELLIGENCE AND SOCIAL STATUS
TABLE IX
6 Yr.
7Yr.
8Yr.
9Yr.
10 Yr.
M.
F.
M.
F.
M.
F.
M.
F.
M.
F.
Unfavored
Favored
28.3
39-8
27.2
41.9
33-2
48.1
36.6
48.7
43-1
59-8
49.4
57-9
47-9
63.1
52.2
64.8
53-5
67
56.8
The difference between the sexes is particularly marked
in the eight-year group of the unfavored school, where there
is an equal number of boys and girls. At the same age in
the favored school the boys have the advantage. These
Points
Ages 6
results seem to bear out the theory held by Thorndike1 and
others that females deviate less from the norm than males;
for there is less difference between the performances of the
girls in the two schools than there is between the performances
of the boys.
1 'Educational Psychology', second edition, 1910, pp. 33-43.
1 8 JAMES W. BRIDGES AND LILLIAN E. COLER
TABLE X
First Grade
Second Grade
Third Grade
No.
M.
A.
Age
No.
M.
A.
Age
No.
M.
A,
Age
Favored ....
Unfavored . .
Difference . . .
57
52
47
27-5
19-5
45-3
29.1
15-7
6-75 mo.
6-9! mo.
2 mo.
55
44
55
45-5
9-5
56.2
46.2
IO
7-8 mo.
8-5 mo.
4^ mo.
53
40
64
54-5
9-5
63.5
55-i
8.4
8-8£ mo.
9-9 mo.
i yr.
M. — Median.
A. — Average.
Age — Chronological age in years and months.
TABLE Xa
First Grade
Second Grade
Third Grade
Favored
Unfavored
Favored
Unfavored
Favored
Unfavored
M
F.
M.
F.
M.
F.
M.
F.
M.
F
M.
F
5
7
5
2
5
5
4
4
5
3
4
o
22
31
17
44
36
29
24
50
49
43
44
29
32
16
19
49
39
35
35
52
49
45
45
30
36
21
22
50
41
36
37
52
48
46
34
36
21
22
Si
45
37
42
53
54
50
46
35
38
22
23
51
49
39
42
53
55
53
50
35
39
22
24
52
49
40
43
56
56
53
50
36
39
23
28
52
49
42
44
58
57
54
50
40
40
23
28
54
51
43
44
60
58
55
51
40
25
29
54
51
43
44
60
61
57
42
25
29
55
51
44
45
61
63
57
52
42
43
25
34
55
52
44
47
62
63
58
53
42
43
43
45
25
25
37
38
55
55
53
55
I
49
49
62
63
64
65
58
58
54
54
45
47
45
49
25
26
42
45
56
55
55
47
52
54
64
64
65
65
fo
55
57
47
49
26
45
56
55
48
57
64
67
62
58
47
50
26
48
58
59
49
57
66
67
72
59
47
51
27
58
59
50
58
66
68
60
47
52
27
58
60
50
67
67
68
61
48
52
27
59
60
52
68
71
61
48
52
28
63
62
52
69
61
48
53
29
64
62
53
69
72
68
48
54
29
65
63
54
69
73
75
49
50
3
29
32
66
66
66
67
11
72
73
73
75
52
32
66
67
76
54
33
72
69
77
54
34
75
79
55
35
60
35
64
36
38
39
41
Ave 45-3
45-4
28.1
31-2
57-7
54-8
45-6
46.8
63.8
63.2
554
54-8
45-3
29.1
56.3
46.2
63-5
55
INTELLIGENCE AND SOCIAL STATUS 19
COMPARISON OF GRADES
Table X. gives in convenient form the number of pupils,
the median and average scores and the average chronological
ages for each grade in each school. The individual scores
for each school and each grade are given on Table X<z.
It will be seen that the superiority of the favored school is
much greater for the first than for the second or third grades.
By examining the Figs. 7, 8, and 9, showing the modes
for the three grades, the same thing will be found to be true.
The favored school, first grade, has a mode of from 45 to 50
points and the unfavored school from 25 to 30 points or a
difference of 20 points. The difference in the second grade is
15 points and in the third the difference lies between 10 and
15 points as the mode is not distinct here.
These results seem to show that the difference between the
children of the two schools is greatest when they enter the
school and that it becomes less from grade to grade. This
would mean that the school work tended to equalize some of
the original difference.
It is necessary, however, to examine further to see if there
may be another explanation before accepting this conclusion.
The average chronological ages for the three grades are also
given in Table X. The ages are nearest the same in the
first grade where there is only two months difference, but this
difference increases to over a year in the third grade. This
age difference is brought about by the fact that in the un-
favored school so many children are in a low grade for their
years. In our present school system a child enters at the
age of six and would therefore be in the third grade at the
age of eight to nine years if regularly promoted. Out of the
forty children in the third grade of the unfavored school there
were eight in the ten-year group, five in the eleven-year group,
four in the twelve-year group and one in the fourteen-year
group. That is, about 45 percent of the children in that
grade are older than they should normally be in the third
grade. On the other hand there are only five older than the
nine-year group in the third grade of the favored school and
none of these is above ten years six months.
20
Number
20
10
JAMES W. BRIDGES AND LILLIAN E. COLER
10
FIRST GHAJ>E MOPES.
A = FAV°REP
8__ . av
-!B \ — ;A
&RADE MODES
B --- * U/VCAVOR£J>
j i i
p0;nts .-.
5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85 90 95
FIGS. 7, 8, AND 9.
INTELLIGENCE AND SOCIAL STATUS
21
If we now compare the average scores for each age group
in the two schools we find that the differences from year to
year are fairly constant. These figures are shown in Table
XL
TABLE XI
6 Yr.
7Yr.
8 Yr.
9Yr.
10 Yr.
Favored
Unfavored . .
40.9
28.3
48.4
34'4
58.8
46.2
64
CO. I
60.2
52.8
Difference
"^•3
12.6
14
12.6
13-9
74
The difference is practically the same for the six- and
seven-year groups and the decrease at ten years is probably
due to selection entering in both schools at this age. More-
over there are too few ten-year-old children for a fair com-
parison. There is thus very little diminution of the initial
difference with age; and the fact that the difference in mental
ability between the two schools is less in the second and
third grades than in the first, must therefore be largely due to
the older children in the second and third grades of the un-
favored school who tend to raise the average of points scored
in that grade and so make it approach nearer to the favored
school's results.
A coefficient of mental ability for each child was obtained
by dividing the number of points actually scored by the
number of points which should be scored at his chronological
age. The standard used was the one obtained by combining
the total Columbus results with the results of the Cambridge
children of English-speaking parents.1 The mental age of
the child can also be obtained by referring his score to the
age at which this score should be obtained.
The average coefficient of mental ability, average mental
age and average chronological age for each grade in each
school are shown in Table XII.
There is a uniform difference of about a year in the
average chronological age of the children from grade to grade
in the favored school. On the other hand there is a differ-
ence of almost two years between the average chronological
1 See Table III.
22
JAMES W. BRIDGES AND LILLIAN E. COLER
age of the children in the first grade and those of the second
grade in the unfavored school. This indicates that the
children of the first grade in the unfavored school remain
on the average two years in that grade before being promoted.
Moreover the mental age for the children of the unfavored
school is about a year lower than their actual chronological
age. It would therefore seem that the children of the un-
favored group are not mature enough mentally when they
enter school at the age of six to be able to do the prescribed
TABLE XII
Favored
Unfavored
Differenc
e
Grade
I
III
I
II
ill
I
11
III
C A .
6— j\ mo
7—8 mo
8-8J mo
G— ot mo
3— c mo
Q— a mo
2 mo.
4.4 mo.
I yr. 5 mo
M.A
CM. A...
7-9
1.24
9-3
1.26
10-4
1.18
:S^9
•77
7-9
.98
8-1 1
•93
2 yr.
•47
i yr. 6 mo.
.28
i yr. 5 mo
•25
C. A., chronological age.
M. A., mental age.
C. M. A., coefficient of mental ability.
•work. The children of the favored group, on the contrary,
are mentally a year ahead of their chronological age and so
these children might equally well enter school at an earlier
age and be able to do the required work. It is also inter-
esting to note in this table that although the difference in the
mental age between the two schools decreases from grade to
grade, the difference in their chronological age increases, thus
keeping a fairly constant difference of two years between the
schools in the grades tested. Binet estimated that there
was a difference of about one and one half years between
children of different social classes, and M. Morle found about
a year's difference in the groups he studied in Paris.
COMPARISON OF SEPARATE TESTS
Table XIII. shows the average scores for each of the
twenty tests making up the Point Scale. Scores for the
favored and unfavored boys and girls are given separately
and the combined average and the difference between the
schools is also given. The difference, although sometimes
very small, is always on the side of the favored school.
INTELLIGENCE AND SOCIAL STATUS
TABLE XIII
Test
Fav. Boys
Unfav. Boys
Fav. Girls
Unfav. Girls
Total
Diff.
Fav.
Unfav.
I
2.90
2.62
2.98
2.91
2.94
2.77
•17
2
3-73
3-18
3-88
3-22
3-8
3-2
.60
3
2.97
2.77
2.98
2.71
2.98
2.74
•34
4
3.18
2.48
3-19
2.67
3-19
2-55
•64
5
3-27
2.02
3-24
2.67
3.26
2-35
.91
6
2.22
1.87
2.20
1.94
2.21
1.91
•30
7
6.76
5-90
6.76
6.
6.76
5-95
.81
8
i-73
1.46
1-54
i-S7
1.64
1.52
.12
9
4.58
3.06
4.28
3-59
4-43
3-33
I.IO
10
5-42
3-98
5.18
4-64
5-30
4-31
•99
ii
2.09
I.48
2.25
1.32
2.17
1.40
•77
12
2-43
2.06
2.27
2.25
2-35
2.16
.19
13
1-51
.78
1.64
1.40
1.58
1.09
•49
14
1.66
•52
1.16
.98
1.41
•75
.66
IS
3-40
2.27
3-So
2.40
345
2-34
i. ii
16
1.65
1.09
1.38
1-45
1.52
1.27
•25
17
2.05
•79
2.OI
.86
2.03
•83
i. 20
18
1-38
•39
i-37
.88
1.38
.64
•74
19
.89
.42
1.19
1.05
1.04
•74
•3°
20
1.30
•75
i-3S
.90
i-33
•83
.50
The five tests in which the greatest superiority of the
favored school is shown are given here in the order of the
amount of difference.
Test number 17, absurd statements, is described by the
authors1 of the Point Scale as primarily a test for 'logical
judgment based on imagination, analysis and reasoning.'
Number 15, comprehension of questions, tests ' practical
judgment involving memory and imagination.'
Number 19, comparison of familiar objects involves
f analysis and comparison of remembered objects and atten-
tion.'
Number 10, concrete definitions, tests ' ideation (associa-
tion) and analysis.'
Number 5, which shows the fifth greatest difference
between the two schools, consists of counting backward
from 20 to i. Here the mental traits involved are 'memory,
imagination and attention.' Probably the reason for the
higher average score in the favored school is the fact that
games involving counting backwards were played by the
children of this school and when they were given this test
1 Point Scale, p. 8.
24 JAMES W. BRIDGES AND LILLIAN E. COLER
they knew what was expected, with little explanation; while
counting backward seemed a new process for most of the
younger children of the unfavored school.
The results of test number 19, abstract definitions, would
probably have shown a greater difference between the two
schools but for the fact that the average for the unfavored
school was raised because a greater number of unfavored
children were able to define * charity.' The familiarity with
this term is easily understood, as charity in some form is
extremely common in the unfavored district.
With the exception* of number 5 it seems, then, that the
greatest difference in the performances of the children of the
two schools lies in the tests involving primarily analysis and
abstraction. This agrees with Binet's analysis of the differ-
ence between the scores of the children in the private school
at Brussels and those from the poorer section of Paris. Binet
found the Belgian children superior in tests involving criti-
cism, comparison, abstract definitions, absurdities, or in those
involving thought in the higher sense. He pointed out
that some of these tests probably put a premium on the
ready use of language and that the children from the higher
social class had the advantage in this respect.
The five tests which show the least difference between
the two schools will now be considered. Number 8, arranging
weights, is described as having to do with ' kinsesthetic judg-
ment, ideation and attention.'
In number I, aesthetic judgment, the difference between
the two schools is also very slight. This is described as
having to do with 'aesthetic judgment involving perception,
association and analysis.' As this is probably the easiest of
all the tests, it was seldom missed by either school, which
probably accounts for the small difference in the average
scores.
'Motor coordination and visual perception' are the traits
involved in Number 12, copying a square and a diamond.
Number 3, comparison of lines and weights, tests Mis-
crimination' of the visual type in the first part and of the
kinaesthetic type in the second.
INTELLIGENCE AND SOCIAL STATUS
Number 16, drawing of designs from memory, involves
'visual memory, perception, attention and motor coordina-
tion.'
With the exception of number I the above tests all involve
sensory-motor functions to a great extent and have to do
primarily with kinsesthetic judgment and motor coordination.
To summarize: The results from the single tests show the
greatest difference in tests involving analysis and abstraction
and the least difference in those involving primarily motor
coordination and kinaesthetic judgment. This agrees with
Thorndike's view that individuals differ least in sensory motor
functions and most in analysis and abstraction.1
INDIVIDUAL SCORES
The individual scores will now be considered as to the
number of children who are twenty-five percent above or
below the norm of their age, when judged by the general
norm and when judged by the average of their own school.
Tables IV. and V. show the individual scores by sex and age
groups for each school. In each group the number of indi-
viduals whose scores depart by twenty-five percent or more
TABLE XIV
No.
Percent
of Total
Using general norm
Favored group, 25% below the norm. . .
Favored group, 25% above the norm. . .
3
73
1.8
44.2
Unfavored group, 25% below the norm.
44
32.4
Unfavored group, 25% above the norm.
ii
8.1
Using separate norms for
each school
Favored group, 25% below the norm. . .
10
6.2
Favored group, 25% above the norm. . .
10
6.2
Unfavored group, 25% below the norm.
19
16.5
Unfavored group, 25% above the norm.
16
13-9
from the norm for that age is indicated. The solid lines
indicate the individuals who deviate twenty-five percent or
more from the general norm for that age group. The dotted
lines indicate the number who deviate twenty-five percent or
more from their own group norm. The exact numbers are
shown in convenient form in Table XIV.
irrhorndike, E. L., 'Educational Psychology,' second edition, 1910, pp. 218-223.
26 JAMES W. BRIDGES AND LILLIAN E. COLER
As will be seen, nearly one third of the children of the
unfavored school have a coefficient of mental ability of .75
when judged by the general norm. This has been suggested
by Dr. T. H. Haines1 as a criterion of feeble-mindedness.
He has shown that it is a more lenient criterion than four
years' retardation above the thirteenth year. For the ages
here considered it is roughly equal to about two years' re-
tardation. The number of the favored group twenty-five
percent or more below the general norm for their age is only 3.
On the other hand the favored school shows over 44 percent,
twenty-five percent or more above the norm, while the un-
favored school has only eight percent. If these results are
compared with those of Cambridge2 where practically the
same numbers are twenty-five percent above and below the
average, it appears that the unfavored school is greatly
weighted by subnormals and the favored school by super-
normals. The question now arises whether it is fair to judge
both schools by the same standard. Let us see what the
results would show if the unfavored and the favored schools
were judged by their own norm or average. In this case the
standard will of course be lowered for the unfavored group
and raised for the favored group. The dotted lines indicate
this in Tables IV. and V. The averages of the pupils were
only used up to and including nine years, however, for beyond
nine years the scores were so low and there were so few cases
that a fair average could not be obtained. Up to the ten-
year group, then, the number twenty-five percent above and
below the norm for each school is about equal, as was the
case in the Cambridge schools.
Considering the great number which must be classed as
very inferior intellectually, if not feeble-minded, in the un-
favored school if the same standard is used for judging both
schools, it seems unfair that groups of children from such
different social classes should be judged by the same norm.
If the sociological factor is not considered in clinical diagnosis,
it seems probable that too high a standard will be expected
1 Haines, T. H., 'Relative Values of Point Scale and Year Scale Measurements
of 1,000 Minor Delinquents,' /. Exp. PsychoL, i, 51-82.
2 Point Scale, p. 55.
INTELLIGENCE AND SOCIAL STATUS 27
of the unfavored individuals and so the degree of mental
deficiency which might exist would be overestimated.
STUDY OF OCCUPATIONS
The children will now be grouped in the two schools by
the occupation of the father. The 165 children in the
favored school were classified as follows:
(1) Professional group 32
Professors 17
Doctors 6
Lawyers 3
Editors 3
Architects 2
Ministers I
(2) Proprietors, officers and managers of manufacturies and stores 32
Proprietors 16
Managers and officers 1 1
Building contractors 5
(3) Traveling salesmen, insurance agents and real estate dealers 39
Traveling salesmen 33
Salesmen 5
Insurance agents 6
Real estate dealers 6
(4) Clerical workers 21
Clerks 13
Bookkeepers and accountants 5
Cashiers 3
(5) The remaining forty-one children were classified in a miscellaneous group.
The main groups in which the 136 children of the un-
favored school were classified are as follows:
(1) Laborers, unskilled 60
This group includes odd-job workers and all unskilled and casual laborers.
(2) Skilled mechanical trades 45
Railroad engineers and mechanics 13
Metal workers 12
Building trades 10
Electricians 3
Shoe cutters 3
Miscellaneous skilled workers 4
(3) Teamsters and delivery men 19
The remaining twelve children in the unfavored school
were put in a miscellaneous class.
In comparing the groups in the favored school, the children
from one group were matched with children as nearly as
28
JAMES W. BRIDGES AND LILLIAN E. COLER
possible the same age in the other group. The results are
here shown for the professional group compared with the
traveling salesmen. Thirty children from each group were
matched in this comparison.
Av. Chron. Age
Av. Mental Age
C. M. A.
Professional group
Traveling salesmen
7 yr. 7 mo.
7 yr. 7 mo.
9 yr. 8 mo.
9 yr. 3 mo.
1.42
1.26
The average chronological age is thus the same, but the
professional group averages 5 months superior to the traveling
salesmen group mentally.
The clerical workers were rather a mixed group. Many
of the clerks were chief clerks and the group as a whole is
small. Comparing seventeen from this group with seventeen
from the manager group of corresponding ages, the following
results were obtained.
Av. Chron. Age
Av. Mental Age
C. M. A.
Clerical .
7 yr. 10 mo.
9 yr. i mo.
1.22
Managing class
7 yr. 8 mo.
9 yr. 5 mo.
1.24
The managing class averages two months younger
chronologically, but shows about four months' superiority
mentally.
In the unfavored school thirty-six children were matched
from the skilled and unskilled laboring classes.
Av. Chron. Age
Av. Mental Age
C. M.A.
Skilled .
7 yr. 1 1 mo.
7vr. 6 mo
Q7
Unskilled
8 yr. I mo.
6 yr 1 1 mo
8O
The unskilled group here has the advantage of being two
months older; nevertheless its average mental age is seven
months less and its coefficient of mental ability .13 less than
the average for the skilled group.
When a group was selected from laborers to match the
ages of the eighteen children in the teamster group, the
following results were obtained.
INTELLIGENCE AND SOCIAL STATUS
29
Av. Chron. Age
Av. Mental Age
C. M. A.
Teamsters
7 yr. 10 mo.
7 yr<
.83
Unskilled laborers
7 yr. 10 mo.
7 yr. 2 mo.
.88
The teamsters appear to have as a class an even lower
mentality than the unskilled laborers, but this is a very small
group and results might be different if larger numbers were
compared.
The following is a summary of the results for the various
occupation groups, irrespective of schools and ages:
No.
Av. Chron. Age
Av. Mental Age
C. M. A.
Professional
•J2
7yr. 3 mo.
9 yr. 8 mo.
1.4.2
Traveling salesmen
Proprietors, etc
39
34
7 yr. 6 mo.
7 yr. 10 mo.
9 yr. 2 mo.
9 yr. I mo.
J.26
1 .21
Skilled
6<?
8 vr
7 yr 10 mo
I 12
Unskilled
60
v yt,
8 yr.
7 yr. i mo.
.83
It is noteworthy in this table that although the chrono-
logical age increases from group to group, the mental age
decreases.
CONCLUSIONS
Our results corroborate the conclusions of Binet in France,
Hoffman in Germany and Yerkes et al. in United States that
there is a very considerable dependence of intelligence upon
sociological condition. We have further shown that when
children are classified according to the occupations of their
fathers, a striking correlation is shown between intelligence
quotient and occupation group. Hence, if mental age rather
than chronological age were used to determine the time for
beginning school, the children of the professional group, for
example, would begin school two years earlier than the child-
ren of the unskilled labor group; for the former mature
intellectually much earlier than the latter.
Incidentally the results have shown that the correlation
of intelligence and social status is probably higher for boys
than for girls. The girls of the poorer school are considerably
superior to the boys; but the boys of the better school are
only at one age noticeably superior to the girls.
30 JAMES W. BRIDGES AND LILLIAN E. COLER
The superiority of the better classes is most evident in
tests that involve higher mental processes like analysis and
abstraction; but it is also shown to a lesser extent in sensory
motor functions.
We have not discussed the causes of this relation of
intelligence to social status for the very good reason that
our data do not contribute anything towards a solution of
the problem. They aim merely to establish the fact and
amount of the difference, and could be used by adherents of
the "Environment Theory" as well as by advocates of
"Inheritance." Thus, the former could emphasize the
quite evident differences in home and school environments,
teaching staff, etc.; while the latter would point to the just
as evident differences in the character and intelligence of the
parents. It is worth noting that in the few cases where the
mothers were tested, they showed a mental age about equiva-
lent to that of their children. If intelligence quotient could
be obtained for a number of successive generations with
different environments, such data might contribute to a
solution of the problem.
We have also omitted discussion of the percentage of
feeble-minded in the different social groups, and have con-
cerned ourselves only with the variations in intelligence; for
we consider the diagnosis of feeble-mindedness and the
measurement of intelligence two quite distinct though re-
lated problems that had better not be confused. Diagnosis
depends upon a number of other considerations as well as
the psychological. The physical aspect can not be wholly
neglected; and the importance of the patient's life history is
generally acknowledged, especially if amentia is to be dis-
tinguished from dementia.
There are also sociological considerations which the advo-
cates of a purely psychological concept of feeble-mindedness
must acknowledge as soon as they consider the problem of
the dividing line between normal and feeble-minded intelli-
gence. The various criteria: two to four years' retardation,
an intelligence quotient below .75, the * lowest three percent'
INTELLIGENCE AND SOCIAL STATUS 31
of the population,1 etc., are all ultimately based upon socio-
logical, or socio-legal considerations. They are merely
statements of the limits below which an individual fails to
attain certain social standards of living.
Now, since these standards of living vary greatly from
group to group, it seems only reasonable that the above
mentioned limits (and intelligence norms) used in diagnosis
should vary too. Otherwise, we might be obliged to classify
whole races as feeble-minded. All Hottentots would prob-
ably be feeble-minded, if judged by Anglo-Saxon intelligence
norms; and similarly the majority of the children of the
unskilled labor group might be classed feeble-minded if
judged by norms for the professional group. The facts
discussed in this paper should therefore find a place among
the various considerations upon which careful diagnosis
depends; but we have preferred to confine ourselves to the
strictly psychological problem: the measurement of the
intelligence of different social groups.
1 Pintner, R. and Paterson, D. G., 'A Psychological Basis for the Diagnosis of
Feeble-mindedness,' /. of Crim. Law and Crim., 7, May, 1916.
MENTAL TESTS WITH DELINQUENTS AND
AUSTRALIAN ABORIGINAL CHILDREN
BY S. D. PORTEUS
School for Mental Defectives, Fitzroy, Australia
During 1915 two groups of delinquent boys were ex-
amined by a new series of mental tests.1 The first group
consisted of boys ranging in age from nine to fourteen years
who had been committed to the care of the Boys' Home at
Burwood, Victoria. Some few were merely neglected chil-
dren but the majority had been before the children's courts
for various minor offences, chiefly truancy and petty thieving.
Under ordinary environment these may be considered crimi-
nals in the making.
The tests given are based on the maze plan and the
subject is required to find a way through the maze in a given
number of trials under certain conditions.2 Success requires
the exercise of prudence in action, forethought, and general
mental alertness. The tests are graded for the mental ages
from three to thirteen years. It is not claimed that they
enable us to arrive at the general mental age of the subject,
though in the majority of cases there is a close correlation
between results by these tests and by the Binet-Simon.
In another investigation, out of one thousand normal children
examined 70.6 percent passed by these tests within one year
of their Binet ages.
Mental age per the Porteus tests means that, in the capaci-
ties of foresight, prudence, resistance to suggestion, and sustain-
ing the attention, the child has reached the average develop-
ment of the age assigned to the tests passed under the given
conditions. Children below ten years of age are most often
placed about one year higher in mental age by the Porteus
tests than by the Binet. This difference is accounted for
1 Tests published by C. H. Stoelting Co., Chicago.
2 See /. of Exp. Fed., June, 1915, or Amer. J. of Psycho-Asthenics, June, 1915.
32
MENTAL TESTS WITH DELINQUENTS 33
mainly by the fact that the former are motor tests and
therefore make a more universal appeal to child interest and
secondly because they were arranged so as to permit of a
comparison between the mentally deficient and the dull
normal child. This explanation is necessary to meet the
possible objection that the tests were too difficult.
Since impulsive and ill-considered action had been char-
acteristic of the delinquent boys' social behavior it was
thought likely that their performances in the tests would
reflect to some degree the same faults of disposition. How
this expectation was fulfilled may be seen by reference to
Table I. below. In some cases where foresight was shown in
the subject's preliminary study of the problem before be-
ginning its working, failure resulted through a too imprudent
trust in the memory. Not a few looked at the maze before
beginning the test and remarked "I see the way out," and
then went impulsively to work only to find that they had
lost the plan and had taken a wrong turning, which, of
course, meant instant failure, since corrections are not
allowed. It was very rare indeed to find a child with the
most intelligent method of attack, viz., a preliminary sizing
up of the problem and then a careful and deliberative work-
ing— prudence and forethought in combination.
The following is a summary of the scores of the boys in
the tests as compared with their chronological ages.
TABLE I
Passed Test
Above age o
At age 5
1 year below chron. age 3
2 years " " " 5
3 years " " 5
4 years " 3
5 years " " " j_
Total 22
Average chronological age 12 years 8 months
Average test passed 10 " 4 "
Average deficiency 2 " 4 "
It is significant that no boy passed a test above his
chronological age while less than 25 per cent, passed 'at age.'
34
S. D. PORTEUS
In Table II. some interesting individual records are given
together with a brief report by the superintendent of the
Home as to each boy's character.
Those in whose personal reports there is a favorable
entry are placed in Section A. Those whose social disposi-
tions are not satisfactory are shown in Section B.
TABLE II
SECTION A
Case
Chron. Age
Test Passed
Deficiency
Superintendent's Report
CD.
II yrs.
10 yrs.
I yr.
Quick wilted, fairly reliable, and
moderately intelligent.
G.T.
12* yrs.
12^ yrs.
—
Truant, good open disposition
fairly intelligent, conduct good.
A. A.
II yrs. 9 mos.
12 yrs.
—
Fairly intelligent, good disposi-
tion.
E. S.
ii yrs.
iify-re.
—
Rather dull, but improving won-
derfully.
W. G.
13 yrs.
12^ yrs.
lyr-
Splendid memory, intelligent for
his age.
C. C.
12 yrs.
115 yrs.
£yr.
Truant. Conduct good, fairly
intelligent.
N. S.
12 yrs. 10 mos.
ii yrs.
i* yrs.
Truant, easily led. Quiet dispo-
sition, fairly intelligent. Con-
duct now good.
SECTION B
J.D.
14 yrs.
II yrs.
3 yrs.
Associated with bad companions.
Habit of petty thieving.
J.R.
I2i yrs.
8 yrs.
4i yrs.
Illegitimate. Rather untrust-
worthy and dull.
P.M.
13! yrs.
9 yrs.
4! yrs.
Truant. Fairly intelligent but
of a sly disposition.
R. F.
u| yrs.
9 yrs.
2* yrs.
Petty thief. Rather sullen dis-
position.
W. P.
niyrs.
9 yrs.
2\ yrs.
Neglected at home. Fairly in-
telligent but sly disposition.
C.J.
12 yrs.
10 yrs.
2 yrs.
A truant and of plausible dispo-
sition. Fairly advanced in
school.
It will be seen that the tests have brought out to a rather
remarkable degree the differences in character in the boys
of the two groups. The total deficiency by the tests is for
Section A — 3^ years; for Section B — 19 years. This accounts
for 13 boys out of 22. In the remaining nine cases there is
nothing noteworthy in the superintendent's reports other
than the references to their intelligence.
MENTAL TESTS WITH DELINQUENTS
35
It should be stated that these personal reports must be
considered thoroughly independent. None of the results
of the testing were known to the superintendent at the time
the reports were furnished.
REFORMATORY BOYS
The next investigation was undertaken at the Royal
Park Reformatory, Melbourne. In the light of their records
the subjects must be considered as youthful criminals.
Table III. summarizes their results and speaks for itself.
TABLE III
Passed Test
At aee o
2 years be'
3 years
4 years
5 years
6 years
7 years
8 years
Total
ow chr
on. aj
>e*. . .2
•i
4
3
5
2
i
. .20
Average chronological age 15 years^6 months.
10
Average test passed 9
Average deficiency 5 {< g "
Note. — In this table 15 years was taken as the upper limit of chronological age
so as to provide a basis of comparison with the mental age. As a matter of fact the
average age will be seen to have been above 15 years.
By comparison with Table I. it will be seen that, though
the average age of the Reformatory boys was higher than
that of the boys at the Home, they passed on the average a
lower test. Some of the individual records are so interesting
that I may be pardoned for giving some details.
Some of the offences charged against these lads were
house-breaking (6) ; serious assault (causing grievous bodily
harm) (3); larceny (7); obscene language (i); street gambling
(i); obtaining money under false pretences (i); murder (i).
The numbers in brackets here indicate the number of such
offences.
The housebreakers showed on the whole better success
in the tests than the others, the mental ages of the six being
36 S. D. PORTEUS
12, ii, ii, ii, 10 and 8 years respectively, whilst their total
deficiency amounted to 23! years. Those who had been
convicted of serious assaults had the mental ages of n, 9, and
8 years respectively, whilst the murderer's mental age was
9 years. Those four, however, who had committed crimes of
violence had a total deficiency of 23 years, an average of
5! years as against the average of 4 years for the house-
breakers. It would seem as if the more talented in this
group of youthful criminals were already qualifying for a
place in the aristocracy of crime, the skilled burglars. The
clumsy insensate crimes of violence were left to boys of a
much lower mental level. Three of the four mentioned
above were sentenced for kicking a companion nearly to
death. The murderer had decapitated his victim with an
axe when they were returning from a joint hunting expedition.
It is most significant that only four of this group of 20
lads were suspected by the penal authorities to be at all
deficient mentally. Among these was the murderer. But
the point that should be emphasized is that there were other
lads of a lower mental level who were considered fully
responsible. When they have committed equally atrocious
.crimes, their mental deficiency too may be realized.
Provided that the offenders show an amount of low
cunning in the commission of their crimes the police appear
very loath to admit the fact of mental deficiency. An inter-
esting instance of this was the case of S. B., aged 15^ years,
examined by me at the Melbourne Gaol. He had been
committed for stealing letters during their transit from the
post office to the city railway station. His plan was to
represent himself to the driver of the mail van as an em-
ployee of the post office, sit at the back of the van, cut open
the mail bags and abstract the contents. The police appeared
very loath to agree to a theory of mental deficiency in his case
and he was examined.
The highest test that he could pass in the Porteus series
was that for 8 years and he required the two trials allowed
for tests for 6 and 7 years. When examined per the Binet,
his mental age was also eight years. He failed to tell the
MENTAL TESTS WITH DELINQUENTS 37
differences between a butterfly and fly, etc., he could not
reckon the value of the stamps nor repeat five numerals
correctly. In the nine year tests he could not give change
nor arrange the weights and he failed in ten year tests through-
out.
The dynamometer records were 17 kilos right hand and
10 kilos left hand — very low for a youth of 16 years of age.
In the Fernald-Healy construction test he required two
minutes thirty seconds.
The diagnosis of feeble-mindedness was, by the foregoing
results, indisputable.1
ABORIGINAL CHILDREN
For comparison with the delinquent groups I am taking
the examination results of some aboriginal children examined
by myself at the Mission Station, Point MacLeay, South
Australia.
Few of them were full-blooded, though in many cases
they were the offspring of marriages between full-blooded
aborigines and half castes. Results are summarized in
Table IV.
TABLE IV
Passed Test
3 years above chron. age I
2 years " " " I
i year " " " 4
At age 12
1 year below chron. age 7
2 years " " " I
3 years " " " _2
Total 28
Average age 10 years 2 months
Average test passed 9 " 9 "
Average deficiency 5 "
1 Since writing this article, I have learned that this boy was examined by a medical
officer at the gaol. He was found to be syphilitic but the equally important fact of
his mental deficiency was unnoticed by the doctor, or was, at any rate, unmentioned!
This implies either carelessness in examination or ignorance of modern methods of
psychological examination. The question of the lad's mental and moral responsibility
was not raised at the trial and he was sentenced to two years' imprisonment. The
diagnosis of deficiency was not disputed but ignored.
38 5. D. PORTE US
On examining the individual records it was found that
the younger children succeeded, relatively, far better than
the older ones. Of 12 children who were over n years of
age, 9 are retarded by the tests, the total deficiency being
16 years. Of the 16 children under n years of age, 6 passed
tests above their chronological age, 9 passed 'at age' and
I was retarded one year.
The increased difficulty of the tests for the upper years
is not sufficient to explain this failure of the older children.
This result bears out the view that the period of mental
development is, in aboriginal children, comparatively short.
This also accords with the experience of the teacher of the
school, who says that he finds little difficulty in bringing his
pupils through the lower grades but finds it very difficult
indeed to educate them beyond about the fourth grade, or
midway through the school course.
Dr. Gertrude Halley, chief schools' medical officer, who
examined these children physically found them well de-
veloped for their ages, weight and height being above the
average. Puberty appeared to be rather early established.
The prepubescent period seems to be the most favorable for
mental development, but is succeeded during early adoles-
cence by a period in which the common racial characteristics
of indolence, shiftlessness, and lack of foresight become
apparent. This view was strengthened by the results of
the tests. It would be interesting to discover by continuing
the tests whether the white race's superiority over the
aboriginal is mainly due to the shorter period of mental
development in the latter, force of heredity in the white
enabling him better to withstand the physical strain associ-
ated with the onset of adolescence, and allowing mental
development to proceed equally with that of the body.
By comparing Tables I., III. and IV. it will be seen that
relatively speaking, the aboriginal children passed far better
than the Reformatory boys and considerably better than the
boys of Section B of Table II. By the use of these tests and
of similar ones which no doubt will soon be developed, it is
hoped that we may reach a stage when we can definitely state
MENTAL TESTS WITH DELINQUENTS 39
by means of an examination that a child is abnormal not
only in intelligence but in disposition and that he thus lacks
the potentialities of good character forming. If, in addition,
psychological tests will enable us even approximately to
assess the importance of these deficiencies in their bearing
on conduct, then we will have achieved a decided step forward
towards the ideal of making training and education ensure
the conservation of the child.
APPLICATION OF TESTS TO DEAF CHILDREN
Interesting results were obtained by the application of the
tests to the children at the school for deaf and dumb, Mel-
bourne.
For children above nine years of age the plan adopted
was to illustrate the working of the six and seven year tests,
the subjects beginning with the test for eight years. It was
rarely found necessary to make any further explanations or
to give further instructions.
On the whole, the deaf children found the tests somewhat
difficult. Temperamental peculiarities were often apparent.
In some cases the subjects gave up the task, signing that it
was beyond their powers. Very many showed extremely
quick perceptions, and worked the tests at a great pace with
a resultant tendency to error through impulsiveness. Gen-
erally they were very quick to notice their errors. A fairly
large proportion were definitely feeble-minded, and a number,
judged by ordinary standards, must be considered dull. In
every case in which the tests showed the child to be feeble-
minded the teacher's judgment concurred. As regards the
boys, there was a close agreement between the teachers'
estimates of intelligence and the verdict of the tests. That
is to say that the tests, considered as intelligence tests, were
satisfactory. In the case of the girls, however, the agree-
ment was not so close, many failing to reach the standard
that the teachers expected.
This partial failure of the tests may be explained in two
ways. The first explanation is that girls on the whole do
not test as high as the boys. Five hundred and eight girls
40 S. D. PORTEUS
(normals) were tested in a recent investigation and their
Binet ages compared with their ages per my tests. By the
latter, 36 percent tested above their Binet age whilst 43
percent tested below, the remaining 20 percent testing the
same. For 492 boys the corresponding figures were: Above
the Binet = 44 percent; Below, = 40 percent; Same, =15
percent.
It will be seen that girls test, on the average, lower than
boys. This is probably due to greater impulsiveness and to
a lesser development of foresight in girls than in boys.
The second factor influencing results is that the training
of the boys at the school tends to accentuate the above-
mentioned sex differences. Organized games and sport
generally are made a special feature of the boys' training at
the school. This is not the case with the girls. Undoubtedly
the effect of this training on the boys is to make them more
self-reliant, resourceful, and more mentally alert than the
girls. It was noticed that the boys who played best, scored,
as a rule, highest in the tests.
Notwithstanding the foregoing objections, I am convinced
that the tests do enable us to make a useful estimate, within
fairly correct limits, of the deaf child's native intelligence.
They certainly allow him to display his ability in a much
fairer light than any other tests do. Even if the Binet test
is adapted to the deaf child's examination, it remains, princi-
pally, a language test, a test of comprehension. It must be
admitted, however, that the successful application of the
Porteus tests depends on the subject's thorough under-
standing of what is required of him. Once this understanding
is gained the application is easy.
The dynamometer records were also taken. The right-
hand and left-hand grips were added together and the rank-
ing of the children in the tests and their ranking in individual
grip records were correlated by the Spearman Footrule
method.
The correlation was:
R = 0.65 (Boys),
R = 0.66 (Girls).
MENTAL TESTS WITH DELINQUENTS
41
Considering the dissimilarity existing between motor intelli-
gence tests of a merely physical measurement, such as power
of grip, I think this correlation may be considered fairly
high. A noticeable feature was the large number of cases in
which the grip of one hand approximated the grip of the
other.
INDIVIDUAL RECORDS: DEAF AND DUMB CHILDREN
Section A
Case
No.
Actual
Age
Age per
Test
Dynamometer
Case
No.
Actual
Age
Age per
Test
Dynamometer
R.
L.
Total
R.
L.
Total
I
I72/12
13
47
48
95
19
9Vl2
9
21
21
42
2
I7Vl2
13
45
48
93
20
H7/12
12
20
17
37
3
4
12
12
46
43
43
44
89
87
21
22
I010/12
98/12
10
10
18
20
18
16
36
36
Is1/!!
13
40
37
77
23
83/12
IO
17
16
33
6
l610/12
13
38
36
74
24
II4/12
II
17
16
33
7
i7g
II
32
42
74
910/12
8
16
16
32
8
II
37
36
73
26
H10/12
8
17
IS
32
9
5710/i2
12
33
38
71
27
I08/12
ii
14
14
28
10
II
32
37
69
28
I22/12
9
14
13
27
ii
I58A2
II6/12
35
33
68
29
I09/12
10
13
12
25
12
l610/l2
II6/12
34
31
65
30
88/12
ii
13
12
25
13
I55/12
13
26
28
54
31
9
9
13
IO
23
14
I25/12
10
26
24
So
34
I010/12
10
10
II
21
15
l61/i2
13
26
23
49
35
73/i2
8
II
10
21
10
13/12
12
23
21
44
36
6U/12
7
8
7
IS
17
I36/12
II
20
23
43
37
62/12
7
7
8
IS
18
I29/12
12
23
20
43
38
65/12
66/12
5
7
12
Section B
Mentally Deficient
Case
No.
I
2
3
4
Actual
Age
Age per
Test
Dynamometer
Case
No.
Actual
Age
Age per
Test
Dynamometer
R.
L.
Total
R.
L.
Total
ltf/12
I410/12
19
I5U/12
H4/i2
I33/12
10
9
II
IO
8
9
35
31
23
25
19
19
28
26
27
24
20
20
63
57
So
49
39
39
8
9
IO
II
12
H2/12
H
95/12
II'/ll
H3/12
88/12
8
\
7
7
4
11
12
II
9
5
13
10
14
10
9
5
26
26
26
21
18
10
A close approach to ambidexterity appeared in about
sixty percent of cases, a difference of less than four kilos.
in strength in each hand appearing in that proportion. In
this particular the deaf appear to resemble closely the men-
tally deficient, a recent investigation by the author amongst
feeble-minded revealing a similar condition. It may be said
42
S. D. PORTEUS
of the latter that they are generally equally weak in each
hand, while of the normal deaf it may be said that they are,
generally speaking, equally strong in each hand.
In the following summary the normal boys' results have
been separated from those of the boys who were considered
feeble-minded. In the table showing individual records the
figures for the latter are given in Section B.
TABLE V.
NORMAL BOYS
Average age i27/i2 years
Average test passed io7/i2 years
Average deficiency 2 years
FEEBLEMINDED BOYS
Average age i32/i2 years
Average test passed 8 years
Average deficiency S2/i2 years
GIRLS
Average age I !6/i2 years
Average test passed 9B/i2 years
Average deficiency 2l/n years
INDIVIDUAL RECORDS: GIRLS
Case
No.
Actual
Age
Age per
Test
Dynamometer
Case
No.
Actual
Age
Age per
Test
Dynamometer
R.
L.
Total
R.
L.
Total
I
I45/12
10
34
26
60
15
I2Vi2
9
16
13
29
2
I310/12
II
26
25
SI
16
U8/12
10
IS
13
28
3
I34/12
12
25
23
48
17
98/i2
10
IS
12
27
4
I25/12
IO
25
23
48
18
I28/12
8
14
13
27
I76/12
12
24
23
47
19
I210/ll
ii
13
12
25
6
IS10/12
II
25
21
46
20
82/12
8
13
10
23
7
13
13
22
2O
42
21
8«/u
8
10
9
19
8
I35/!2
9
21
19
40
22
6»/ii
6
10
9
19
9
I3u/i2
ii
20
18
38
23
7Vl2
6
IO
9
19
10
I29/12
12
18
19
37
24
9Vl2
7
12
6
18
ii
I3Vl2
II
20
IS
35
25
64/12
6
8
8
16
12
I2Vl2
10
18
14
32
26
SVu
5
6
6
12
13
I59/12
9
16
16
32
27
73/12
86/12
ii
ii
22
14
I310/i2
ii
18
13
3i
In each of the above investigations the number of indi-
viduals examined is certainly small. The results are reported
in order to show the applicability of the tests to abnormal
children generally. Results with mentally deficient and with
normal children have been elsewhere reported.
THE PSYCHOLOGICAL CONCEPT OF CLEARNESS
BY E. B. TITCHENER
In 1913 C. A. Britz published a thesis for the Zurich
doctorate entitled Eine theoretische und experimentelle Unter-
suchung uber den psychologischen Begriff der Klarheit. Cir-
cumstances over which I have no control postponed my
first-hand acquaintance with the work to 1916. I have re-
gretted this the more because Britz deals in detail with two
psychological systems in which the notion of clearness holds
a prominent place, Wundt's and my own. Wundt he chose
for obvious reasons; myself, because my 'Standpunkt stellt
quasi ein Extrem dar.' Both of us receive a severe mauling:
which would be wholesome enough — since the criticism is
objective and based upon quotation — if only Britz had fol-
lowed a sound method. He has not. He struggles with the
Wundtian concepts of clearness and degree of consciousness
on the basis of the sixth edition of the ' Physiologische Psy-
chologic' and the seventh of the 'Grundriss'; and it never
occurs to him that the key to their understanding is a genetic
study of his author. He attacks my concept of sensory
clearness on the basis of my 'Text-book' and of Hillebrand's
review of my ' Feeling and Attention' in the Zeitschrift; he
has not referred, incredible as the thing appears, to the
Feeling and Attention itself.
External circumstances may be in part responsible. The
thesis was undertaken at Schumann's suggestion, and the
experimental portion was apparently completed under his
direction at Frankfurt. The author, however, returned for
his doctorate to his old university, and the thesis was accepted
by G. F. Lipps of Zurich. It is a fair assumption (is it not?)
that Schumann was chiefly interested in the experiments,
and that Lipps was generously disposed to a bit of work
originated and approved by a psychological colleague else-
where. In that case the critical chapters, about seventy per-
43
44 E. B. TITCHENER
cent of the whole paper, would have fallen, so to say, between
the two professorial chairs. Even so it is astonishing that
Britz should not have learned for himself the essentials of
scientific method.
I
I do not imagine that Wundt will find time to defend his
system against these latest charges, and I do not propose to
undertake the business for him. I shall, however, try to set
forth, in the light of a genetic study, the use and meaning of
the technical terms here in question. Such a study is doubly
instructive. It shows our modern psychology in the making;
the long series of Wundt's books reflects the recent history
of the science. It shows also the manner of Wundt's own
progress from logic to psychology, from activity to content.
His fundamental ideas have remained, for the most part,
unchanged; advance is made, habitually, by modification in
detail, by expansion and contraction, by redistribution of
topics and change of emphasis. Even when the system
suffers a decided innovation (we shall have a case presently),
there are always hints of the new departure, if we look closely
enough, in the previous work. Here, of course, I have space
only to give results. I hope to be able, nevertheless, to clear
up the difficulties which Britz and, perhaps, other readers
have found in the sixth edition of the ' Physiologische Psy-
chologic.'
We are to ask, accordingly, what Wundt means by con-
sciousness and degrees of consciousness, by clearness, by
degrees of apperception, and especially what is the relation
between degree of consciousness and degree of clearness. We
begin with consciousness.
The keynote of Wundt's psychological treatment of con-
sciousness is the notion of synthesis. Consciousness, the
condition of all inner experience (i-$Y or, more empirically,
the total contents of our immediate experience (6), cannot
be defined in psychological terms. We must be satisfied to
determine its conditions (i), that is, the phenomena that
xBy these numbers I indicate the editions in which the particular phrase occurs;
minor verbal changes must here be disregarded. References are given in later notes.
THE PSYCHOLOGICAL CONCEPT OF CLEARNESS 45
invariably accompany its manifestation in experience (2-5);
more exactly, we must be content to give the conditions under
which we observe such phenomena as we attribute to a con-
sciousness (6). There are, now, two psychological processes
which are bound up with consciousness and may be regarded
as its essential characters. The first of these is the formation
of ideas (and real feelings) from sensations (and simple
feelings). Our consciousness of ideas consists in the act of
synthesis whereby sensations are brought into temporal and
spatial form (i); in every act of ideation there is effected a
connection of elementary sensations (2-4). Ideas and real
feelings arise from a psychological synthesis of elements, and
this connection of elements is therefore one of the two charac-
teristics of consciousness (5, 6). The other is to be found in
the processes of reproduction and association of ideas (and
feelings). The connection of ideas takes place in conscious-
ness (i); it is only by way of reproduction and association
that consciousness can become aware of itself as persisting
without change through all the change of ideas (1-3); this
changing flow of ideas is itself aware of consciousness as a
synthetic activity connecting present ideas with those that
have gone before (1—3). Consciousness is empirically demon-
strable only on condition (unter der Foraussetzung) of a con-
nection of the ideas (and feelings) which follow one another
in time (4, 5); reproduction and association are therefore an
universal concomitant of consciousness (6). An orderly
connection of ideas (1-4) or, in more general phrase, a con-
nection of immediate experiences (5, 6) is, indeed, the con-
dition under which alone consciousness appears.
If, however, the primary thing about consciousness is
synthesis, then we must recognize the possibility of degrees
or grades of consciousness, since such connections as that of
sensations in the temporal or spatial idea may exist at various
levels (i). Self-observation reveals these degrees. "When-
ever we incorporate an impression but loosely in the context
of our ideas, or later remember it but imperfectly by reason
of this looseness of connection, we credit ourselves only with a
lower degree of consciousness at the time in question" (2-6).
46 E. B. TITCHENER
In these instances, capacity for the connection of ideas (or
psychical contents) is taken as measure of degree of con-
sciousness (2-5); or, as Wundt puts it in his final phrasing,
"the connection of psychical contents is a certain measure of
degree of consciousness" (6). Every connection of inner
states (or psychical elements) manifests some degree of
consciousness (2—6).
That is the Wundtian doctrine of consciousness and its
degrees, as set forth in the various editions of his great work.1
The intention of the discussion, from the very first, is psycho-
logical; even in 1874 Wundt is combating the logical tenden-
cies which showed their full force in the Vorlesungen of 1863. 2
His effort is not immediately successful; but, by degrees,
changing here a little and there a little, he moves away from
logic toward psychology, away from synthetic activity
toward observable connection. There is no reason to think
that he was, at every stage, fully aware of the significance of
the changes; he probably chose the wording that seemed, at
the time, best to express his thought — his original no less than
his present thought. In point of fact the corrections of the
early text serve in sum to change the whole atmosphere of the
discussion.
We find the same sort of progress in the treatment of
attention. The discussion of 1874 begins as follows: "In
the synthesis of sensations and in the association of ideas
consciousness apprehends itself as active. Thus arises that
expression of consciousness which we name attention. It
makes itself known in direct self-observation by the fact that
the interconnection of ideas, with which consciousness is
correlated, is by no means present to it at all times in the
same manner; consciousness is directed upon certain ideas in
higher measure than upon others." Later we have: "Beside
the coming and going of ideas we perceive within us not
infrequently (in varying fashion) and more or less plainly
an (inner) activity which we designate attention" (2-6).
lPhys. Psych., 1874, 707 f., 711 ff., 717; 1880, II., 195 f., 199, 201; 1887, II.,
225 f., 229, 231; 1893, II., 255 f., 259, 261; 1903, III., 320 f., 324 f.; 1911, III., 296, f.;
299 f.
2 See op. cit., 1874, 708 ff.
THE PSYCHOLOGICAL CONCEPT OF CLEARNESS 47
Degree of apperception is gauged by the subjective activity
with which consciousness turns to a particular sensory stimu-
lus (1-5). The simile of Blickpunkt and Blickfeld, the state-
ment that the Punkt is really a small Feld of varying extent
and that the main field darkens in proportion as the central
field brightens, and the distinction of perception and apper-
ception are present in all editions. Passive and active apper-
ception are distinguished in the second, the limen of conscious-
ness and the limen of attention only in the fourth edition.
All these things are familiar, and need not be dwelt upon.
What now of clearness, which ultimately becomes the sole
objective criterion of apperception?
The brightening of the Blickpunkt means, as we have
seen, that consciousness is directed upon certain ideas in
higher measure than upon others (1-4); certain contents
become more conscious than others (5); we observe in con-
sciousness different degrees of conscious status, variously and
varyingly distributed over its contents (6). If we consider
the apperceived contents themselves, we find the following
progression. We begin with a clearness of ideas, dependent
partly on the intensity of the ideas and partly on adaptation
of attention (1-3). Presently this clearness, dependent now
upon the intensity of the sensations composing the ideas and
upon adaptation of attention, is paired with distinctness;
clearness is predicable of an idea in its own right, distinctness
of an idea in its relation to other ideas. Feelings may be
distinct, but apparently can not be clear (4). Later still this
same clearness attaches to all complex conscious contents:
to ideas and feelings as wholes, and also to particular ele-
ments within ideas and feelings (5, 6). Clearness, which
originally belonged to ideas alone, thus remains to the end a
character of complex contents. "Clearness and distinctness
are exclusively characters of ideas, and may be transferred
to sensations only when these are considered as constituents
of ideas" (6).1
All this is fairly straightforward, though I must warn the
1 Op. dt., 1874, 717 f., 720, 722, 725, 729; 1880, II., 205 f., 208, 209, 212; 1887,
II., 235 f., 238, 239, 244; 1893, II., 266, 267, 269, 271, 272, 282; 1903, III., 331, 332 f.,
336, 337 f-, 339, 348, 349J 19", HI., 306, 307 f., 312 f., 314, 322, 323.
48 E. B. TITCHENER
reader that I have passed over certain passages which will
occupy us later. The pairing of distinctness with clearness
offers no difficulty; distinctness is always the subordinate
concept, and does not appear in the final summary of the
part-processes in an apperception.1 The irruption of the
new theory of feeling, in the fifth edition, does create a diffi-
culty— as I pointed out in Feeling and Attention — but it is
not one that directly concerns us here.
We are therefore ready to take up, in a preliminary way,
the relation of consciousness to attention and of degree of
consciousness to degree of clearness. The first of these
questions is easily answered. Consciousness, for Wundt, is
always wider than attention. In 1874 he wrote: "The
theory that consciousness and attention are identical is not
tenable." In 1911 he writes: "An impression that has sunk
below the limen of apperception does not therewith disappear
from consciousness;" and his whole treatment of the two
topics, from first to last, implies this distinction.2 Conscious-
ness is the total contents of our immediate experience; atten-
tion is the range of clear experience.
The second question may be answered, to begin with, by
the statement that degree of consciousness and degree of
clearness have, logically, no connection with each other.
Degree of consciousness is degree of organization of conscious
contents. Let us imagine (if we can) a consciousness with-
out attention. Such a consciousness would still show degrees
of consciousness, because the complex contents and the
groups of complex contents which make it up would differ
in closeness of connection or organization. There are pass-
ages in the first edition which seem to come very near to
such an imaginary consciousness;3 and the recurring phrase
"It is always association that puts ideas at the disposal of
apperception" at any rate suggests it.4 In reality, however,
1 See op. ciL, 1911, III., 316.
2 Op, ciL, 1874, 725; 1911, III., 314.
8 Op. cit.y 1874,795, 835-
4 Op. cit., 1880, II., 212; 1887, II. , 244; 1893, II., 279. I do not find the phrase
in the two last editions; and indeed it goes too far. See 1903, III., 524 f.; 1911, III.,
498 f.
THE PSYCHOLOGICAL CONCEPT OF CLEARNESS 49
consciousness comes to us in other guise. It makes its own
history from the very outset, and carries that history with it;
and the history is constantly interfering, so to say, with its
present course.1 Or, to put the same thing from another
point of view, it is organized, at whatever level and in what-
ever degree, as an attentive consciousness; associations are
formed in passive, apperceptive connections in active atten-
tion.2 While, then, degree of consciousness (or of organiza-
tion) and degree of clearness (or of conscious status) may be
distinguished logically, empirically they are bound together
in the most complicated fashion. One might suppose, per-
haps, that the difference between range of consciousness and
range of attention should be directly observable. Wundt
does not deny it, though he seems to think otherwise; it is a
subsequent apperception that ordinarily makes us aware of
the obscure fringe; and he does deny that the simultaneous
method is adequate to range of consciousness.3 One might
suppose, again, that degree of consciousness and degree of
clearness should run parallel; and in many cases, at many
moments of the history of consciousness, they doubtless do;
but we must remember that the one tends to be stable and
the other is essentially instable. Consider, indeed, any case
of active attention. The complex contents that now lie in
the obscurity of the Blickfeld, and that therefore have no
conscious status whatever, were once (probably, many times
over) given in passive apperception, in so far as they are
organized at all; and if their organization is high, as it may
be, they were given in active apperception. The complex
contents that occupy the Blickpunkt and therefore possess
various degrees of conscious status vary in degree of conscious-
ness, from moment to moment, according as apperception is
integrative or disruptive and their organization is corre-
spondingly strengthened or weakened. Or consider observa-
tion itself. Observation is always apperception; and we
cannot become aware of a low degree of consciousness unless
* Op. at., 1893, II., 284.
2 See W. B. Pillsbury, Amer. J. of PsychoL, 1897, 8, 329 ff.
3 Op. cit., 1911, III., 324, 330; cf. 1903, III., 351 if., and the stronger statements
of 1874, 726; 1880, II., 219; 1887, II., 261; 1893, II., 305.
50 E. B. TITCHENER
we give the poorly organized contents a high conscious
status.1 So the empirical relation of the two degrees (and I
have, of course, greatly oversimplified it in this brief account)
is complex in the extreme. Neither can exist without the
other; yet, since they do not run on parallel lines, their
separate treatment is a matter of practical convenience, if
not of necessity; a full account, historical and descriptive, of
any given consciousness implies constant reference to both.
If Wundt inclines, even in his latest writing, to make connec-
tion the fundamental character of consciousness and to
regard attention as an activity within consciousness2 — when
we might expect him to give the two factors equal rank —
the reasons are historical, and not least among them is his
reaction against unconscious ideas.
I believe that these answers to our two questions are
fair, and that they represent the essentials of Wundt's
doctrine; I confess that I have rounded off some rather
prickly passages.3 We have now to consider those divergent
statements to which I have already referred.
1 After a good deal of vacillation, Wundt settles down in the sixth edition to the
definite terminological distinction of Bewusstseinsstufe or Grad des Bewusstseins and
Grad der Bewusstheit: 1911, III., 299, 307. I have made this distinction throughout,
and have translated Bewusstheit by 'conscious status.'
2 Op. cit., 1911, III., 301.
8 Let me give an instance ! Wundt teaches that the contents at the Blickpunkt
(an area, be it remembered) are variously clear, and the contents in the outlying
Blickfeld obscure. In 1903, III., 353 (1911, III., 326) we are introduced to Grade der
Verdunkelung, degrees of obscurity. It looks, then, as if the contents below the
limen of attention might possess something more than degree of consciousness (which
is all that I have allowed them in the text), something that is, after all, very like
conscious status. I have, however, pointed out in 'Feeling and Attention' that there
is here a confusion of apperception with cognition, of attributive with cognitive clear-
ness, and that a recent worker in Wundt's own laboratory has called attention to it
(see 237 ff., 369, and cf. 230 f.). I have already remarked that Wundt's progress is
from activity to content: it is not till the fifth edition that 'consciousness' ceases to
be 'directed upon ideas,' and that 'contents become more conscious': cf. 1903, III.,
333 with 1893, II. , 267.
Again, in 1902, I., 323 (1908, I., 382) we are told that change of clearness, as
distinguished from change of intensity, alters the relation between contents; clearness
thus seems to be confused with distinctness. But the passage in which this state-
ment occurs is not represented in 1893; it harks back to 1887, I., 237, an edition in
which the distinction of clearness and distinctness had not yet been drawn. The
clearness of 1902 and 1908 (in these particular sentences) is therefore an undifferentiated
clearness and distinctness.
THE PSYCHOLOGICAL CONCEPT OF CLEARNESS 51
The discussion of the apperception-center in the fifth
edition surprises us by a reference to the clearness of sensa-
tion: surprises us all the more because the earlier editions
spoke in the same context only of the clearness of ideas and
impressions; because the same volume teaches that sensations
are constituted solely of intensity and quality; and because
the same edition, in a later volume, retains the orthodox
view that clearness and distinctness are exclusively characters
of ideas. This third volume adds, however, that clearness
may be predicated of sensations when they are considered
as constituents of ideas (the fundamental of a compound
tone, the color of a visual form); and so it seemed possible to
interpret the sensations which become clear in apperception
as sensations-in-ideas.1 That was a way out of the difficulty;
it was not the way the offending passage read. Those who
know the 'Physiologische Psychologic' historically know,
however, that its exposition is continually changing in detail,
and that the details are likely to prove important; I have
made the point earlier in this paper. So one hoped for more
light in a sixth edition; and the light came with a vengeance!
The reference to clearness of sensations in the discussion of
the apperception-center is now justified by entirely new
matter, which introduces the chapter on Intensity of Sensa-
tion. There are (we learn) intensive psychical magnitudes,
which accrue only to the simple elements of the mental life,
and there are extensive psychical magnitudes, which result
from the composition of elements. The three intensive mag-
nitudes are intensity, quality, and — clearness. And these
three characters are three coordinate dimensions of the
psychical elements; they are, that is to say, attributes of sensa-
tion. Compound contents show different degrees of clear-
ness in their different parts; degree of clearness is unequivocal
1 Op. cit., 1902, L, 322 f., 353; 1903, III., 338, 349; cf. 1880, 1., 218; 1887, L, 233;
1893, L, 228. These passages are to be sharply distinguished from the casual and
physiologically motived reference to the apperception of sensations (correlates of the
excitation of a sensory center) which occurs in 1902, L, 324, and which appears in all
editions from the second to the sixth.
52 E. B. TITCHENER
only in regard to the elements, to simple contents. Wundt's
statements are as definite as they could well be.1
Here, then, is a new kind of clearness, different from the
original clearness of ideas and real feelings, different also
from the clearness of the sensations-in-ideas. Wundt has
not made a clean sweep of things, however, even in this
volume; the old statement, running through all the editions,
that sensations are constituted solely of intensity and quality,2
is allowed to remain. A pure oversight, no doubt! The
few references to clearness in the second volume, of 1910, are
neutral. But then we come to the third volume, of 1911,
and there we are back again in the familiar atmosphere,
with the express assurance that clearness and distinctness
are exclusively characters of ideas! Could ever anything be
more bewildering?
Well! our bewilderment is at any rate less than Britz's.
For our genetic study proves that the two clearnesses do not
stand on a level. The clearness of the third volume, of 1911,
is the traditional clearness of the Wundtian system, deeply
rooted in nearly forty years of thought and expression; the
clearness of the first volume, of 1908, is a new phenomenon,
only casually foreshadowed in the corresponding volume of
1902. Something, it appears, was moving Wundt's ideas,
even at the earlier date, towards sensory clearness; and
something happened, between the fifth and sixth editions, to
precipitate and crystallize his ideas.3 Thereafter, in the
interval between 1908 and 1911, his interests turned away
from this something; he had forgotten all about his intensive
1 Op. cit., III., 1908, L, 539 ff. I myself urged as early as 1898 that clear-
ness should be recognized as a sensory attribute, but printed no extended discussion
of the question before this same year, 1908. See Phil. Rev., 8, 461 f.; 'Feeling and
Attention,' references to 'Clearness' in index.
2 Together (for a time) with feeling-tone. See 1874, 273; 1880, I., 272; 1887, I.,
290; 1893, L, 282; 1902, L, 353; 1908, L, 412.
8 1 am in this paper expounding Wundt, and neither criticizing his views nor
trying at all completely to trace their motivation; the first thing to do with an author
(and it is what Britz has failed to do) is to understand him. I think, however, that
it is safe to connect Wundt's new paragraphs with the revival of psychophysical
interest shown by the works of Miiller (1904), Lipps (1903, 1904, 1906), Titchener
(1905), Bruns (1906), Keller (1907) and others. In particular, G. F. Lipps was at
Leipsic from 1903 on.
THE PSYCHOLOGICAL CONCEPT OF CLEARNESS 53
magnitudes that accrue only to the elements; and so he
contented himself with the customary revision of the former
text. Natural enough, after all, in view of Wundt's age and
multifarious activities: the wonder is not so much that he
should have forgotten as that he should have had, in 1908,
the energy and the open-mindedness to attack once again
the whole problem of mental measurement, and in doing
this to effect a radical change in one of his most elaborate
systematic constructions. There is no possibility of recon-
ciliation of the two volumes; Britz's efforts are wasted labor.1
It would plainly be useless to reopen our questions of the
relation of consciousness to attention, and of degree of con-
sciousness to conscious status, in the light of Wundt's new
definitions. The clearness which is an intensive attribute of
sensation is at the same time degree of apperception (the
objective aspect of degree of attention) or of keenness of
apprehension.2 We may work out, if we will, what this
statement logically implies for the treatment of consciousness
and attention in the third volume; or we may wait patiently
for a seventh edition. At present the questions can be
answered intelligibly only if we ignore the intruding passages.
They can be answered, that is, only in the preliminary way
in which they have been answered above. This conclusion
seems to me to be a positive result, which justifies our re-
course to the genetic or historical method. It far outweighs,
just because it is the result of a sound method, any conclusion
reached by Britz.
II
I must now say something in reply to Britz's polemic
against my own doctrine of clearness. If only Britz knew
accurately what he is talking about! I give a few examples
to show that he does not.
(i) Britz finds five principal "criteria" of the sensory
attribute. I am said to rely only upon two, inseparability
1 But what was Klemm about, who read the proofs, that he did not call his chief's
attention to the discrepancy? See op. cit., 1908, L, X; 1911, III., V.
2 Op. cit., 1908, III., 541 ff. The italicized Apperzeptionsgrade of p. 541, 1. 16
should be Aufmerksamkeits grade.
54 E. B. TITCHENER
and independent variability; and I am further said to rely
mainly upon the second. Turn to 'Feeling and Attention'!
I there begin by criticizing the 'common definition' of an
attribute. From this definition I accept the criterion of
inseparability. As a mark of inseparability I instance the
reduction of the whole sensation to zero when a single attribute
reduces to zero. Britz, who raises this special case of in-
separability to the rank of a separate criterion, speculates as
to what I should make of the argument if I used it! From the
same definition I get the criterion of independent variability,
and point out that in fact "there are bound attributes as well
as free," so that "the test of independent variability, useful
enough for a preliminary survey, must be applied with caution
when we demand accuracy of detail." Yet this is 'Titch-
ener's Hauptmerkmal'! Hillebrand, from whom Britz de-
rives my two criteria, states the case correctly; so that Britz
has even misread Hillebrand.1
(2) Britz does not hesitate to criticize Bentley's experiment
with intensities of sound on the basis of the passing reference
in my 'Text-book.' What series were carried out, what
intensities of stimulus were employed, what precautions were
taken, of course he does not know.2
(3) I never use the phrase 'degree of consciousness';
from my point of view it is as nonsensical as 'degree of
matter' or 'degree of material existence' would be in physics.
Britz disregards my definition of consciousness, and his dis-
cussion of the place of Bewusstseinsgrad in my system is
consequently all in the air.3
(4) Britz credits me with 'the assumption that the num-
ber of degrees of clearness is not the same for all departments
of sense.' What I say is that we have 'to determine, intro-
spectively, how many degrees of clearness can be distinguished
[how many just noticeable differences of clearness there are]
1 Britz, op. cit., 14, 24; 'Feeling and Attention,' 8 ff.; F. Hillebrand, Z. /. Psych.,
1910, 58, 141.
2 Op. cit., 26; 'Text-book,' 1910, 280; 'Feeling and Attention,' 361 ff. I take this
opportunity to correct a misprint. In Table II., p. 364, the second rubric under
Height of Fall should be 74.4-89.6 cm.
8 Op. cit., 41 f.
THE PSYCHOLOGICAL CONCEPT OF CLEARNESS 55
in the various departments of sense'; I assume neither that
the number is the same nor that it is different.1
(5) I have just said that Britz misreads Hillebrand: here
is another case. " Hillebrand has shown in detail," he writes,
"to what absurdities we are led by the identification of
attention and clearness when we make clearness an attribute
of sensation." Hillebrand has shown no such thing. He
raises the question "whether every attribute has its own
clearness, as an attribute of the second order"; if it has, he
says, "that would lead to the absurd consequence that the
complete disregard of any one attribute brought with it, to
say the least, the disregard of all the others. . . . Here
Titchener seems to me to have overlooked obvious difficul-
ties." Perfectly fair criticism! but there is one absurdity
hinging on an 'if,' not an ausfuhrliche Darlegung of absurdities
in general.2
It is naturally disappointing, when one is made the part-sub-
ject of a doctorate thesis, to find one's views thus caricatured.
But enough has been said on that matter. Let us now see if
Britz makes any positive contribution to the discussion of
sensory clearness.
My thesis is that clearness or vividness (I am not yet sure
which is the better term, and there is historical warrant for
both) is one of the intensive attributes of sensation. Britz
complains that I say very little about its actual nature; and
in a sense that is true. You cannot say much about a thing
that you regard as ultimate to your science; any attempt at
a definition runs over, by force of circumstances, into what
Wundt would call a tautologische Umschreibung. I have had
recourse to a number of these periphrases; but I have tried,
above all, to exhibit the thing itself, to state conditions under
which it may be experienced and identified in experience.
Quality and intensity are here in the same box with clearness.
You can exhibit qualities or intensities, as you can exhibit
clearnesses; but when you attempt to define them, you find
yourself talking round them. If Britz had performed Geiss-
1 Op. cit., 41, 42; more correctly stated on p. 12; 'Feeling and Attention,' 277 f.;
* Text-book,' 295 f.
2 Op. cit., 42; Hillebrand, 146 f.
56 E. B. TITCHENER
ler's simple experiment with the two metronomes, equated
for quality and intensity of sound, he would have discovered
at first hand what I mean by sensory or attributive clearness.1
My thesis is, secondly, that sensory clearness is the ele-
mentary phenomenon in what is ordinarily called attention.
Just as sensory extension is the elementary phenomenon in
spatial perception, and sensory duration in temporal per-
ception, just so, mutatis mutandis, is sensory clearness the
unique thing, the psychologically ultimate thing, in attention.
Hence I remark in the 'Text-book' that "in the last resort,
and in its simplest terms, attention is identical with sensory
clearness." Analyze an attentive consciousness, and every-
thing is familiar to you but the one thing, which you finally
arrive at — this sensory clearness or vividness; that is new
and characteristic.
The importance of such a view for experimental psychology
is, I think, plain on the surface; a new road is opened, and a
road that by all analogy should take us an appreciable distance
to our goal, for an experimental attack upon attention. In
'Feeling and Attention' I speak accordingly of a 'simplified'
or 'elementary' psychology of attention; I suggest that we
start out, not from the gross facts of the attentive conscious-
ness, but from the 'rise' of the single sensation, the absolute
temporal limen, the carrying power of clearness under simple
conditions. "How far this elementary psychology of atten-
tion could be carried it is, evidently, impossible to predict,"
though there is no lack of specific problems; in any case, "the
results of experiment in these fields must be 'interpreted' by
a psychology of attention; the factors that make for clearness
must be separated from the other conditions involved, and
must if possible be separately estimated or 'weighted.'"
That is my view; and I am correspondingly surprised to
find Britz, who quotes correctly the sentence from the 'Text-
book' given above, asserting in several places that I identify
outright clearness with attention. If that were the case my
10p. cit., 40, 41, 44; 'Text-book,' 53, 279; 'Feeling and Attention,' 26, 183 ff.;
L. R. Geissler, in Amer. J. of Psych., 1909, 20, 510. Britz devotes a special section to
Geissler, as he does also to Wirth and Jaensch; all three will, I expect, find something
to say for themselves.
THE PSYCHOLOGICAL CONCEPT OF CLEARNESS 57
chapter on Attention would hardly have been written as it is.
But having made this identification, it is easy for my critic to
show that the introductory examples of attention — the shift
of interest due to the visit of a friend or to the receipt of a
telephone message — involve' more than sensory clearness, and
that I am therefore faithless to my theory before I have got it
formulated.1
I hold, thirdly, that clearness is not an attribute of the
simple feeling; and as clearness is an intensive attribute,
ranging from liminal obscurity to terminal clearness (just as
intensity ranges from the very weak to the very strong),
this means that feeling is for me neither clear nor obscure,
but only qualitative, intensive and durative. The traditional
* obscurity' of feeling rests, I believe, upon the customary
mixture of logic and psychology. I realize that the whole
psychology of feeling is debatable ground; but, after all, the
discussion in * Feeling and Attention' is seriously written and
deserves to be taken seriously. Britz gives a single sentence
to the matter. We ought, he says, to enquire carefully
whether the clearness of my two examples (the friend's visit
and the telephone message) is not applicable to feeling; anach
meinem Dafiirhalten kann er [der Begriff] angewendet
werden." But that clearness is evidently cognitive as well
as attributive; the distinction is clearly drawn, again, in
'Feeling and Attention.'2
I hold, lastly, that in cross-section the attentive con-
sciousness is arranged, for many and perhaps for most of us,
at two main levels, the upper of which certainly, and the
lower probably, are 'wrinkled' by minor differences of
sensory clearness. "A two-level type," Britz remarks,
"seems to me to be altogether beyond the range of psycho-
logical proof; it too obviously contradicts all and every
experience." Oddly enough, it seems to me to represent my
experience. Britz may very well belong to the multi-level
type, though he does not appear to have gone beyond casual
self-observation. He continues: "Within the apperceived
1 Op. cit.t 12, 40 ff.; 'Text-book,' 266 f.; 'Feeling and Attention,' 209 f., 251, 372.
2 Op. cit., 42; 'Feeling and Attention,' 237 ff.
58 E. B. TITCHENER
(beachteten) complex there are degrees of clearness which
may be lower than the highest upon the low level; therefore
it is not permissible to speak of a 'niveau,' * level,' plane or
surface." I do not know how he gets his data; but it is
surely plain that, if there are anywhere in a given conscious-
ness processes less clear than the clearest of the lower level,
these processes must be for me at the lower level. To say
that they are t within the apperceived complex' means, if it
means anything, that my critic is thinking of the unitary
object of attention; and that means that he has fallen into a
form of the stimulus-error.
Hillebrand's comment here is more to the point. He
asks why, if clearness is an intensive attribute of sensation,
there should be only two main levels of clearness at any
moment instead of an unbroken section of the attributive
continuum. I do not know, though I might if I knew more
physiology. I do not know either why the constant of atten-
tion is 6; our theories of attention are still nothing better
than more or less plausible hypotheses. I am trying only
to ascertain the psychological facts.1
We are not reaping an abundant harvest. Nor shall we
fare much better if we ask, as I now proceed to do, why Britz
objects to clearness as an attribute of sensation. He first
examines the criterion of inseparability, and finds that the
attribute of sensation is not separable by sensory attention
but is, of course, separable by abstraction. Then 'as regards
clearness' he adds: "not every phenomenon that is in-
separable by sensory attention is thereby given immediate
status as attribute of sensation." This statement, in the
absence of examples, is a little cryptic; Britz may be thinking
of some form of ' inseparable association.' We need not guess,
however, since the conclusion is simply a non liquet. He
asks, secondly, whether the sensation disappears as a whole
when clearness becomes zero, and replies that, for Wundt,
it does not; Wundt has two limens, the one of consciousness
and the other of attention. How this reply bears upon my
position I do not understand; nor, by his own admission, does
1 Op. cit., 53; Hillebrand, 148.
THE PSYCHOLOGICAL CONCEPT OF CLEARNESS 59
Britz. He examines, thirdly, the criterion of independent
variability, and finds that clearness is not an independent
variable. He forgets (though Hillebrand had told him) that
I recognize bound attributes as well as free. He asks, fourthly,
whether a reference to clearness is necessary to the complete
description of a sensation, and decides that it is not. "Can
we characterize a sensation completely without recourse to the
notion of clearness ? I must answer this question in the affirm-
ative." Yes, but he does not show us in a concrete case how
the thing may be done. Further: a supposed attribute may
prove to be analyzable into a number of really primary attri-
butes. Yes, and I have been on my guard; witness my treat-
ment of Aufdringlichkeit and of tonal quality. Britz quotes
no cases. Further : very few psychologists have regarded clear-
ness as an attribute of sensation. Yes, again: and how has
attention fared in the history of psychology ? Listen to Ebbing-
haus: "Attention is a real perplexity in psychology. Both in
the general run of English associationism and in certain com-
prehensive works down to the present day it is altogether ig-
nored. In other books it bears the strangest relation to the
systematic presentation of the whole subject, and sometimes
an author seems to be entirely helpless." And when psy-
chology came to deal with sensations? I quote Ebbinghaus
once more. "All statements of any exactness regarding
sensations, their attributes, their liminal values, etc., imply
from first to last — as everybody always understands without
being specially told — that a high degree of attention was given
to them." In other words: so long as psychology dealt with
the full attentive consciousness, we made no solid progress;
and when psychology acquired methods of precision, attention
was taken for granted. Now that a suggestion for the be-
ginnings of an exact psychology of attention are forthcoming,
one would think they were worth a trial. Very few psycholo-
gists have agreed, as a matter of fact, upon any general view
of attention.1
There remains, fifthly, the empirical side of an issue al-
I0p. cit., 23, 24, 30, 31; 'Text-book,' 54 f., 95; 'Feeling and Attention,' 26 f.,
326 f.; H. Ebbinghaus, 'Grundziige der Psychologic,' 1902, I., 585 f., 588.
60 E. B. TITCHENER
ready raised theoretically. Is clearness analyzable and not
simple, derivative and not primary? Britz replies, on the
ground of critical discussion and of experimental work, with
an emphatic Yes. Clearness (both Wundt's and mine, appar-
ently) is a very mixed concept, deriving partly from the meta-
physical philosophy of Leibniz, partly from popular psy-
chology (we talk of * clear' colors as we talk of 'pure' tones),
and partly from the properly psychological distinction of
degrees of consciousness; it may thus be very variously em-
ployed, under various empirical conditions, and its employ-
ment always implies a process which is of the nature of judg-
ment.
I have, now, said something of the value of Britz's critical
discussion, and I could say a good deal of the value of his
experiments. He worked with the tachistoscope (not know-
ing, of course, what I had said of that in 'Feeling and Atten-
tion,' though he had read Mittenzwey); he required his
observers to cognize and name (erkennen und benennen) the
colors exposed; and he employed a wissentliches Verjahren to
the extent, at any rate, that they knew the nature of his
problem. What the tachistoscopic analysis of the Erken-
nungsvorgang has to do, in any direct way, with the study of
attributive clearness, it is difficult to see.1 I shall not, how-
ever, enter into detailed criticism; experiment is best met by
experiment; and while a repetition of Britz's work will hardly
help us to a psychology of clearness, it may throw light upon
the psychology of Eindringlichkeit or insistence.
No higher honor can be paid a scientific theory than criti-
cal discussion based upon experiments which are conceived
and carried out expressly to test its validity. Here, however,
is a discussion that leaves out of account the original state-
ment of the theory, and relies wholly upon secondary sources;
and here are experiments that fall into line with the work
of Schumann and his school, but by the same token are
directed upon a complex process of assimilation. It is a
great disappointment.
1 Op. cit., 40, 42 f., 54 ff., 67 ff., 75; K. Mittenzwey, Psych. Stud., II., 1907, 386 ff.
("Im Begriff der Assimilation findet sich von einem Merkmal der Klarheit zunachst
gar nichts.")
THE PSYCHOLOGICAL CONCEPT OF CLEARNESS 61
Postscript. — Since writing this paper I have learned that
Dr. Britz is numbered among the victims of the war. It goes
against the grain to criticise thus sharply an author who can
no longer reply. Yet I am sure that Britz would have wished
his work to be seriously considered; and since the points really
at issue are not personal, but scientific, it is perhaps not too
much to hope that some other pupil of Schumann or Lipps
may carry further the study of clearness which Britz began.
COMPOUND SUBSTITUTION IN BEHAVIOR
BY S. BENT RUSSELL
St. Louis , Mo.
The average man does not seem to care much how his
brain works. There are many inquiring minds, however,
that are striving to add to the sum of human knowledge of
the brain and its operations which serve to adjust the man's
inner relations to his outer relations. The search for facts in
this field is certain to be guided more or less by theories that
are based on the results of preceding investigations.
These theories, of course, must be remodeled from time
to time. At the present day we find that there are certain
theories of nervous mechanisms that are given some accept-
ance in scientific circles. In brief outline these theories show
that each nervous impulse enters the nervous system at a
receptor that is either external or internal. It then passes
from neurone (nerve fiber) to neurone across the connecting
synapses or junctions. Following the course that is most
open to it, it tends to arrive at one or more muscles or other
effectors. The theories give a clear account of reflex action.
They also account in some degree for associative memory in
its elementary forms. They show for example an explanation
of substitution of one stimulus for another and an explana-
tion of memorizing a series of words.1
In this article it is proposed to extend these theories and
seek to explain more complex responses or rather responses to
more complex stimuli such as appear in first lessons in mental
arithmetic, for example. We will begin by noting some
effects of greater numbers of neurones. We will then discuss
briefly the formation of simple associations and the substitu-
tion of one stimulus for another through association and also
the linking of movements into a definite order or series. The
relation of such a movement series to what is known as
1 Watson, John B., 'Behavior,' New York, Holt, 1914, pp. 272, 274.
62
COMPOUND SUBSTITUTION IN BEHAVIOR 63
delayed reaction will then be taken up. With this prepara-
tion we will attack the main problem before us; i. e., responses
to complex stimuli.
To make a proper beginning we will start with a brief
definition of the more important terms we shall use.
Selective reaction is a form of simple association. It is
the type of behavior shown when a child points to an object
upon hearing its name pronounced. It is always the result
of training or experience.
Substitution will be used to express simple substitution
and is the type of behavior where one stimulus takes the
place of another in provoking a given movement. It is the
result of training as in the case of young chickens running up
at the sound, 'Here chick, chick,' without waiting to see the
food.
Delayed reaction is the term used for a case of selective
reaction or substitution when the response does not come
immediately after the determining or substitute stimulus.
Several seconds or several minutes may elapse between the
stimulus that determines the movement and the corre-
sponding response.
A memorized series is a series of movements made in a
definite order from previous training. When a dog is made
to 'fetch' he performs a memorized series that he has been
taught.
Compound substitution is the same in apparent form as
simple substitution but requires the cooperation of several
stimuli to provoke the response. Herbert Spencer in ex-
plaining instinct1 used the term compound reflex action to
describe automatic behavior in which complex stimuli pro-
duce complex movements. In compound substitution we
have a similar term. When you teach a boy to give a correct
answer to the question "Two and one make how many?"
you are developing a mechanism for compound substitution.
If we attempt to analyze this case after the habit is fixed, we
see that the sound of the word two cannot have a great
tendency to provoke the utterance of the word three. We see
1 'Principles of Psychology,' New York, Appleton, 1894, Vol. I, p. 432.
64 S. BENT RUSSELL
that the sound of the word one cannot have a great tendency
to provoke the word three. The same is true for the other
words of the question asked, when taken separately and yet
taken together, the words do provoke the response * Three.'
How may we account for this ?
We see at once that it is much harder to account for
compound substitution than it is for simple substitution. It
is not so very difficult to think of connecting nervous path-
ways that would allow one stimulus to excite a movement
which before training required a different stimulus. Neither
is it very difficult to think of connecting paths that would
allow a stimulus from one receptor to open up a passage for
a stimulus from another receptor, as in a case of selective
reaction. When we come to compound substitution, how-
ever, we find a different matter. We must now provide for
cooperation of stimuli, or, if you prefer it, for the resolution
of nervous impulses. It would seem that a new principle is
required. This is the problem which we are to discuss.
If the theories we now have covering association and sub-
stitution are worth demonstration, the extension of these
theories to cover compound substitution must be also worth
while.
Having this much knowledge of what we are seeking and
why, let us proceed with the search. Having noted what a
great gap there is between simple substitution and compound
substitution, let us now hunt for the missing links.
In the course of this discussion it will be shown that the
key to the mechanisms for a memorized series will prove to
be a missing link as it were.
Let us here briefly consider the effect of numbers of nerve
fibers in the system. The greater the number of nerve fibers
that are simultaneously aroused and tributary to a given
muscle the stronger will be the contraction if any occurs and
moreover the greater will be the tendency to contract actually.
The superiority of the human mind over the brute mind must
be largely due to the greatly superior number of the neurones
that control movement. A dog can be trained to respond to a
simple command but not to a long speech. A child can be
COMPOUND SUBSTITUTION IN BEHAVIOR 65
trained to respond to a simple command and by further
training, he comes to respond to very lengthy instructions.
The act of a thoughtful man is due to the joint excitation
of many neurones. To educate a child, we first train him
to respond to simple commands from the effect of association.
For each act thus learned, one or more association nerve
fibers must be developed. We may say that each step of
learning is the development of a particular association fiber.
Step by step, the child learns to talk. Step by step, he learns
to write, to read, and to count.
By way of further preparing a foundation for our dis-
cussion, let us agree that selective reaction and substitution
can be explained by the dual common path theory, which is
that for every unit association of which the nerve system is
capable, there is a common nerve path which is, we will say,
an association fiber. Moreover each association fiber has
two tributary fibers or private nerve paths coming from two
different receptors or sensory terminals.
Each impulse that follows the common path leaves a
resistance temporarily lessened for a subsequent impulse.
Each impulse from a private path goes by way of the associa-
tion fiber to a certain muscle or other effector. If the re-
sistance is low enough, a feeble impulse will reach the effector.
Let us call such an impulse a scout impulse when it only
serves to lower the resistance for the next impulse. Let us
also, on the other hand, call an impulse that follows a scout
impulse and is strong enough to aid in causing movement, a
worker impulse for convenience in discussion. When a volley
of worker impulses excite one of a pair of antagonistic muscles,
movement will take place unless there be equal excitation of
the opposed muscle.
By way of illustration, let us think of a child writing under
the direction of his teacher. A certain movement may be
prompted by a group of worker impulses from his eyes which
are guiding his hand together with a similar group from his
ear due to the teacher's words and another group of tactile
impulses from the feel of the penholder, etc., and still another
group from his hand and arm; i. e., kinaesthetic impulses from
66 S. BENT RUSSELL
the muscles and joints due to movements just made. All
these worker impulses reinforce each other and compel the
contraction of a certain muscle. This illustration shows how
some of the so-called voluntary movements are caused.
Such movements are determined by previous training as
each worker impulse comes by way of an association fiber
which has been developed by previous impulses. It is plain
that the formation of a single letter is due to a large number
of association nerve fibers working together like the instru-
ments of a great orchestra.
To apply the theory of the dual common path to a typi-
cal case of selective reaction, we may use the following illus-
tration. The sound of the word 'ball' spoken by the nurse
excites a short series of scout impulses in the boy's brain,
each opening up some common path. The sight of the ball
produces worker impulses that follow along the common
paths opened by the scout impulses. Of course these paths
were developed by the repetition of impulses in some previous
experience. The worker impulses now provoke the move-
ment of pointing at the ball. The result is due to recency
and frequency of stimuli in altering resistance to conduction.
In the case of substitution, the application is similar.
Let us take this illustration: Show a horse an ear of corn and
then call him, thus giving him a lesson. After several
lessons, he will come at your call without any corn. The
explanation is that the lessons opened up the dual common
paths so that the call after training produces a sufficient
volley of worker impulses to provoke a forward movement.
To express it in a different way the effect of the lessons is to
set up a * conditioned reflex' as it is called.
In his presidential address given in the March, 1916,
number of this journal Professor John B. Watson showed
how the conditioned reflex has been used in behavior experi-
ments and suggested its use for investigations of association
reaction, etc. Let us here note that in the cases he described
the conditioned reflex is the result of simple substitution and
appears when the substitute stimulus provokes a sufficient
volley of worker impulses to produce reaction. In the case
COMPOUND SUBSTITUTION IN BEHAVIOR 67
of compound substitution, however, the time element is a
factor. Hence, although we find a resemblance in the opera-
tions, the term reflex may be thought inappropriate.
In order to help the reader keep in mind the way impulses
tend to follow recent impulses from other receptors, the
following illustration is offered: Suppose that Mr. Brown
leaves his home in the suburbs in the morning after a snow-
storm, on the way to the railway station. He breaks a path
through the snow. A little later his neighbor Mr. Jones
leaves his home and soon strikes Brown's trail. His easiest
course is to follow Brown's track to the station. We may
say that Brown represents a scout impulse and Jones is like a
worker impulse that coming from a different source soon after
the scout impulse, tends to follow the same nerve path because
it offers less resistance.
Having now a conception of how elementary nerve
mechanisms are constituted, let us think how they may be
coordinated so as to account for a memorized series in which
the movements follow the same order in which they were
made before, in a similar situation. Each movement causes
afferent or kinsesthetic impulses, some of which may be
worker impulses that excite the next movement while others
are scout impulses that open up the paths for other move-
ments to follow. The kinsesthetic impulses have become
substituted for other stimuli that were received in early train-
ing before the habit was fixed. In this way the movements
become linked together so that it is only necessary to provoke
the first movement of the series and if the external conditions
are right, the other movements will follow automatically in
their proper order.
On consideration we see that the important factors for a
memorized series are the kinaesthetic impulses and the asso-
ciation nerve fibers leading to the muscles that must act. An
important thing for us to remember in this connection is that
the scout impulses sent through by one movement of the series
will facilitate the worker impulses for a movement that
follows after intervening movements and a considerable
interval of time.
68 S. BENT RUSSELL
Let us now pass on to the behavior known as delayed
reaction.
Let us suppose a small dog placed in a box so constructed
that he can see two openings in front of him, one to the right
and one to the left but cannot reach either until the bar in
front of him is removed. Over each opening is a green light
that can be switched on or off by the operator. One opening
always leads to food, the other does not. Sometimes one
opening and sometimes the other leads to food. To train
the dog the operator turns on the light over the opening
with food. After giving the dog time to notice the light he
turns it off and half a minute later releases him. After a
number of trials the dog becomes trained so that when re-
leased he goes at once to the opening where the light was
seen. We may explain the behavior in this way: The sight
of a light over the right-hand opening excites certain nerve
paths. That is, it causes impulses that follow certain nerve
paths. The performance of going to the right-hand opening
is a series of movements due to impulses following certain
nerve paths. Let us suppose that there are particular asso-
ciation nerve paths that are common to each group. That
is to say, there are particular association nerve fibers that are
aroused in executing the movement series that are also
aroused by the sight of the right-hand light.
The common paths will get two excitations whenever the
dog makes the proper turn at the proper time. The double
excitation may be assumed to have the useful effect of
opening these common paths; i. e., of lowering their resistance.
Each succeeding lesson will further facilitate the appropriate
behavior. When the common paths are sufficiently developed
we may explain the dog's actions by saying that the right-
hand light produces scout impulses that temporarily open
up further the common paths and the operation of releasing
the animal produces worker impulses that follow these com-
mon paths and provoke the movement of turning to the right.
It appears then that delayed reactions are explained by the
dual common path theory.
We may say that delayed reaction is by nature only a
COMPOUND SUBSTITUTION IN BEHAVIOR 69
special case of selective reaction. As a further aid to under-
standing the matter, we may note that a case of delayed
reaction is like a memorized series where some of the move-
ments are omitted so as to leave a gap in it. For in every
movement series that is habitually made, the movements in
the first part of the series are associated with each move-
ment in the last part of the series. In other words each
movement causes impulses that facilitate or insure any move-
ment that has a place further along in the series as we have
already observed.
We may say then that delayed reactions are special cases
of memorized series behavior. Or we may say with equal
propriety that a memorized series is a compound of delayed
reactions, meaning by this reactions that are more or less
delayed. For each movement of such a series, to some ex-
tent, is determined by impulses that preceded it by a con-
siderable interval of time.
Behavior students have measured the comparative in-
telligence of animals by the delayed reaction and by the
memorized series as shown in escaping from a maze or opening
a puzzle box. The more intelligent an animal is the quicker
it will master a given maze or puzzle box and the longer it
will remember what it has learned.
The principle of learning is no doubt the same with the
animal that learns in a few lessons as with the animal that
requires many. The difference must be in the nerve struc-
ture with which he begins the task. The animal of high
intelligence must have more association neurones to take
part in the proceedings.
From this brief study it would appear that the delayed
reaction can be explained in some measure by the same
mechanisms that are used for simpler forms of behavior.
We have considered four types of behavior, selective reac-
tion, substitution, the memorized series, and delayed reaction
and we find in all of them it is a matter of association nerve
fibers that register the frequency and recency of impulses
traversing them in the range of experience and then in turn
regulate the passage of impulses that provoke movement.
70 S. BENT RUSSELL
To put it another way association is the characteristic feature
of the four types of behavior.
This brings us to the main question, the explanation of
compound substitution.
Let us consider the case of a boy learning mental arith-
metic. In the beginning he is taught simple problems.
He is taught to make a certain response to a certain series of
ear stimulations. For example when he hears the question:
"Two and one make how many?" He must respond,
"Three." The reader will see that we have here a case of
compound substitution. As the boy advances in the study,
the series of excitations is increased in length so that there
are more factors to be kept in his mind as it were. In master-
ing each series he really develops certain association fibers;
i. e., his experience opens up certain nerve paths. The asso-
ciations formed in the earlier lessons take part in working
out the later problems.
But let us look into the matter more closely. Suppose
that the boy learns to add by counting on his fingers. At
the sound of "two" he holds up two fingers. At the sound
"one" he holds up one finger of the other hand. In response
to "How many?" he puts his hands together and counts
"one, two, three" and answers "Three." Here we have a
memorized series in which eye movements, finger movements
and ^ vocal movements are linked together by association.
Let us note that quite a large number of association neurones
are stimulated in the course of the performance. After the
boy has mastered this series, he can be trained to modify it
by counting in a whisper and uttering the final word three
aloud. Then he can be taught to suppress the finger move-
ments so that they are incipient only. The eye movements
also become incipient. It is fair to suppose that the number
of neurones stimulated though lessened, is now nearly as
great as when all the movements were actually made, so we
still have a large number of neurones taking part in the
response.
On consideration we see that a certain series of auditory
excitations provokes a memorized series composed of move-
COMPOUND SUBSTITUTION IN BEHAVIOR 71
ments or incipient movements that lead up to the desired
vocal response. In the case where the intermediate move-
ments are only incipient, we may say that the series of
excitations given by the question heard, provokes a series of
scout impulses and the effect upon the nerve muscle system
produces a group of worker impulses that excite the series of
actual movements included in the proper response. These
nervous operations by long practice become greatly abbrevi-
ated, but we may believe that the successive character of the
operations still remains.
This shows that a case of compound substitution may be
evolved from a memorized series, so we see how an elementary
case of compound substitution may be explained as a modi-
fied memorized series so to speak. The same reasoning may
be followed to build up cases of more and more complexity.
The behavior of a skillful lawyer engaged as referee in a legal
contest is largely determined by compound substitutions
which are due to his special training. His ability to keep in
mind an astonishing array of pros and cons has been built
up step by step. Each step in the case means more move-
ments linked in habit series and a greater number of asso-
ciation neurones stimulated. When the great lawyer finally
gives his decision for the plaintiff or for the defendant, the
deciding movement is prompted by the discharge of a large
number of association neurones which have been duly pre-
pared in the course of the proceedings.
A performing elephant will respond to his trainer's signal
by one trick or another depending upon the stage setting at
the time. In the same way, the great lawyer responds to his
stage setting; i. e., the pros and cons of the legal problem.
The lawyer's response, however, is due to a far greater num-
ber of association neurones and to a far greater number of
steps in the ladder of learning.
It may be noted that the above given explanation of com-
pound substitution rests to some extent upon an assumption
that all incipient movements can in some way cause afferent
72 ' S. BENT RUSSELL
impulses similar to the kinaesthetic impulses from actual
movements.1
If our view is the correct one, a case of compound substi-
tution may be regarded as a memorized series that has become
modified by suppressing the movements so that of the differ-
ent stimuli concerned, each takes part in determining the
resultant response and thus the stimuli do cooperate. It is
evident too that in the case of compound substitution of a
long series of stimuli, delayed reaction plays an important
part for the reason that it takes part in all memorized series
as we have seen.
It is hardly necessary to point out that language habits
are the basis of most compound substitutions in human
behavior. Writing movements also are an important aid in
forming the required associations.
From this demonstration we may conclude that a rational
theory can be formulated for nervous mechanisms for com-
pound substitutions in behavior if we may start with a very
large number of association nerve fibers to many of which
there are tributary afferent fibers coming from the muscles.
These fibers are developed by practice through the influence
of frequency and recency. As brooks come together to form
a river so do the worker impulses meet and provoke move-
ment. These nerve mechanisms are of great service in
adapting behavior to environments containing changeable
features. In the instinctive behavior of all highly organized
animals they have an important part. They are especially
important in child education. In fact we might almost say
that in any animal that has a cerebral cortex, behavior is
mainly determined by such mechanisms or at least modified
by them.
If our theory is the true one, we shall find that to have a
greater intelligence, the number of association fibers must
be increased in much greater ratio.
If we take a broad view of the subject, compound substi-
1 A discussion by the writer of 'The Function of Incipient Motor Processes' will
be found in the PSYCHOL. REV., 1915, 22, 163-166. The other main assumption
made in this article and termed the dual common path theory is discussed by the
writer in the PSYCHOL. REV., 1916, 23, 235.
COMPOUND SUBSTITUTION IN BEHAVIOR 73
tution in behavior may be defined as an increase of the
correspondence between the organized individual and its
environment in speciality and in complexity. Furthermore
it is the coordination and integration of correspondences, so
to speak. The subject is such an extensive one that the
demonstration given herein, to some readers, may appear
quite inadequate. On the other hand, to some readers, it
will seem a lengthy exposition of a simple proposition. It is
not practical in a short article to do more than offer a con-
tribution to the general study of this extensive subject with
sufficient elaboration to make the main idea plain to those
who are searching for mechanisms to explain mental growth.
THE DELAYED REACTION IN A CHILD
BY WALTER S. HUNTER
The University of Kansas
INTRODUCTION
The present paper derives its chief significance from the
fact that the child tested did not possess vocal language and
probably not gesture language either.
In the work with the delayed reaction on children previ-
ously published,1 five subjects were used: M., 8 years,
Hd., L., and H., each 6 years, and F., 2j years. All of these
children possessed vocal language. M. received 38 trials on
delays; Hd., 46 trials; H., 15 trials; L., 41 trials. All of them
succeeded with delays as great as 25 minutes. There was
some indication that the children first encountered particular
difficulty in the intervals from 4-6 sees. The child F. re-
ceived 507 trials on delays. Her maximal delay was 50 sees.
The periods of greatest difficulty were at 5 sees., 7 sees., 10
sees., 15 sees., 30 sees., and 40 sees. Although F. was given
45 trials on the one-minute delay and failed, it is possible
that more prolonged training would have enabled her to
master the interval.
When the present tests were begun, the writer hoped to
secure subjects who would fill in the great gap between the
child F. and the older children. Success has not attended
these efforts mainly because of a lack of convenient material.
One three-year-old child was tested, but the data are of so
little value that they will not be presented. The second and
fundamental purpose of this study was, however, to study the
delayed reaction in a child too young to possess vocal language.
This is of particular importance because of Watson's insistence
that the delayed reaction, if solved by internal factors other
than orientation, must be solved by vocal language.
1 Hunter, Walter S., 'The Delayed Reaction in Animals and Children,' Behav.
Monog., 1913, 2, No. I, pp. 52-62.
74
THE DELAYED REACTION IN A CHILD 75
The Subject Tested. — The subject of these tests was the
writer's daughter, Thayer, thirteen to sixteen months of age
from first to last of the experimentation. She was a normal,
healthy child physically — a little slow perhaps on the behavior
side, due to the lack of the constant attention that many
children receive. She learned to walk alone rather suddenly
at about 15 months; and by 16 months, she could indicate
with a little certainty her eyes, ears, nose and mouth. When
13 months old, she could l throw a kiss' and wave ' bye-bye.'
Even before this, as early as the tenth month, she swayed
and waved her arms to music. These observations are pre-
sented as a sample of her best accomplishments. She had
no vocal language. She made many sounds, some of which
were in response to definite stimuli; but in no case did she
use the sounds spontaneously and in no case did she use them
as symbols. Her equipment was not large in the first place,
and what there was was purely of a stimulus-response nature.
During the period covered by the experimentation, the
following vocal behavior was present: ' Daddy' she said
whenever a distant door was heard and I was away, or when
she heard me coming up the steps. 'Whitte' came in re-
sponse to the striking of the clock or the ringing of the door
bell. She could say 'boob-boob' for the dog; 'day-day' for
the duck; 'm-m-m' for the cow; a funny noise for the donkey;
and 'y-gob, y-gob' for the turkey. These were in response
to the specific questions, "What does the dog say?" etc.
Some of these she got mixed up and later forgot all but the
duck, the turkey, and the cow. In addition to these vocal
responses there occurred only the more conventional baby
noises.
One gesture might possibly be termed language, viz.,
raising her arms to be taken up. It is impossible to say,
however, that this was not in response to present stimuli.
It is also impossible to say with any certainty whether or
not the child used a vocal cry specifically "to attract atten-
tion" to her needs. I have no evidence to indicate that
such was the case. Experimentation was purposely stopped
before the first signs of language (in the conventional sense)
appeared.
76 WALTER S. HUNTER
Apparatus and Method. — The apparatus used is shown in
Fig. I. It consists of three boxes placed upon a stand whose
top is 6 inches above the floor. Each box is approximately
3 inches deep, 4 inches wide, and 5 inches long. Each is
covered by a hinged top. The apparatus was wired for
electric lights, but these were never used. Another type of
apparatus more nearly approximating that used with children
in the previous work was constructed. It, however, proved
unsuited to Thayer and was used only with the three-year-old
child.
The only features necessary to secure in devising an appar-
atus for the delayed reaction are these: (i) It must be adapted
to the size of the subject and to its mode of response — walking,
reaching, swimming or flying. (2) It must provide a means
for presenting a stimulus in one of several places. (3) These
stimulus positions must be equally accessible to the response.
And (4), the stimulus and the method employed should be
such as to present no differential cues to the subject during
the intervals of delay. These requirements, although rigid,
are simple and can be met for practically all organisms. I am
therefore unable to agree with Professor Yerkes when he
says1 that the multiple choice method of studying ideational
behavior is superior to all others in: (i) applicability to a
wide range of conditions; (2) susceptibility to standardiza-
tion; (3) quantitative nature of results; and (4) intelligibility
of data acquired. Nor can I acquiesce in the claim that
"It is already obvious that the method enables us to com-
pare, as has never before been possible, the responses to
certain standard situations, of human and infra-human, nor-
mal and abnormal, mature and immature subjects." Both
the multiple choice and the delayed reaction methods are
valuable for the study of human and animal behavior; but
they are applicable, I think, to very different problems.
In the experiments reported in this paper, the following
method was used. Thayer sat in front of the apparatus, as
shown in Fig. I, and the stimulus object was placed in her
1 Yerkes, Robt. M., 'Methods of Studying Ideational Behavior in Man and Other
Animals,' PSYCHOL. BULL., 1915, 12, 330-331.
THE DELAYED REACTION IN A CHILD 77
hand. A great variety of things were used for stimuli: dolls,
keys, rattles, shoe-buttoners, small books, etc. Every effort
was made to keep up the child's interest. Occasionally two
different stimulus objects were used in the course of the
day's work. Such methods are necessary if a child of this
age is even to approximate to the vigor with which a hungry
FIG. i. Thayer opening box c.
animal attacks its problem. The stimulus object was taken
away from the child almost as soon as she received it and
was placed in one of the boxes. The lid of the box was left
open and Thayer was pushed over (or permitted to lean over)
and made to look into the box. Often she tried to reach in
and get the stimulus; but in every case, her hand was with-
drawn and she was raised back to an upright position. The
lid of the box was now closed. Save for a few instances to
be mentioned in due time, the subject was distracted during
the interval of delay. Distraction took either one of several
forms: (i) I might place my hands over her eyes and rock
her body back and forth from right to left. (2) She might
be stood up, turned around with her back to the apparatus,
kept there awhile and then put down. (3) I might cause
her to turn her head by speaking to her. She would remain
in this position and imitate animals for me (as described
7S WALTER S. HUNTER
above under vocal habits) as long as any delays here used
required. About 2 sees, before the end of the delay period,
her body was straightened around and she sat facing the
middle box, &, entirely free from contact with me. (She
almost never looked back at me; and when she did, she paid
no further attention to the problem. I never spoke to her
during the delay. These facts together with the child's in-
ability to reach delays of a minute or more, indicate that she
was not deriving cues from the experimenter.) Thayer was
now left to her own devices until she opened the box con-
taining the stimulus object. In all but a few cases, she began
to hunt for the stimulus as soon as she was straightened
around. Time was taken with an ordinary watch and was
counted from the moment the box lid was closed until the
subject made some movement toward one of the boxes. In
the records particular attention was given to the orientation
at the time of response and to the behavior during the delay.
A reaction was counted wrong if the child opened any box
save that containing the stimulus.
EXPERIMENTAL RESULTS
In this experiment there was no period of learning the
association between the stimulus object and the three boxes.
Seeking for objects that had disappeared was already a
part of the subject's behavior equipment. I noticed as early
as her eleventh month that if I showed her a toy and then
hid it behind something, she would immediately reach or
creep toward the spot. I have no doubt, however, that this
type of reaction occurred earlier. The present test was
more complex than this in that the toy might be in either one
of three different places.
I quote the following from my diary records indicating
the results at the very beginning of the work. "Date, 10-
23-15. Trial I. Toys put in middle box and door shut.
(All this done by Thayer on her own initiative.) I now put
my hands over her eyes and shook her head and whole body
playfully but thoroughly. No orientation of upper part of
body retained. 13 sees, delay from time she straightened up
THE DELAYED REACTION IN A CHILD
79
after closing box until she reached toward b. She reacted
correctly, straight to middle box."
"Trial 2. Toys put in right box, a. 12 sees, delay with
distraction as above. Reacted correctly."
"Trial 3. Same as No. 2, 13 sees, delay. O. K."
"Trial 4. Thayer preferred right or middle box. I had
her put toys in left one, c. Distraction by standing her up
and turning her to me. 17 sees. O. K."
"Trial 5. Middle box used. Baby tired of test. 14 sees.
Distracted as in first trial. Reaction wrong."
In these tests made on the first day, Thayer missed one
of five or 20 percent. The delays ranged from 12-17 sees.
Distraction was always used. The same orientation at the
moment of release was held for all, viz., orientation to b.
These long delays were very startling and held out a promise
of very rapid development through training. This was the
last of October. A longer stage of delay was not success-
fully reached and consistently maintained until after Christ-
mas. This fact is shown in Table L
TABLE I
Delay in Sees.
Correct Trs.
Wrong Trs.
Percent. Correct
3
3
O
IOO
4
I
O
100
6
0
IOO
6
3
0
IOO
7
2
2
50
8
4
2
66
9
4
I
80
10
ii
4
73
ii
6
2
75
12
6
3
66
13
s
O
IOO
H
6
8
42
is
13
ii
54
16
5
5
50
17
4
3
57
18
2
i
66
19
I
i
50
20
9
IS
37
21
i
5
16
22
2
3
40
23
I
o
IOO
24
2
3
40
25
0
3
o
26
I
i
50
30
3
5
37
35
2
0
IOO
So
WALTER S. HUNTER
This table shows the size of the delays and the number of
correct and incorrect reactions made. All trials given the
subject through December 2 are included. The statement
is not chronological. The reason the intervals of delay
increase so gradually is that within a few seconds variation,
Thayer herself determined when she would begin the reac-
tion. All I could do was to place her facing the apparatus
and await results. Table II. groups the delays into five
TABLE II
Delays
3-7 sees.
8-12 "
13-17 "
18-22 "
Percent Correct
.72
•55
•37
•44
classes which may be called the 5 sec., 10 sec., 15 sec., 20 sec.,
and 25 sec. intervals. The 10 sec. interval may be regarded
as mastered, bat no higher interval. (This is understated,
as will be indicated below.) Table III. gives data gathered
TABLE III
Delay in Sees.
Right
Wronj
Percent Right
6
4
o
O
77
8
9
3
2
I
I
O
o
82
10
12
T r
IO
2
2C {
3
» 8
7c
20
2C
24
2
IO.
2
70
50
from January 2 through January 10. An interval of one
month had elapsed during which no tests were made. A com-
parison of Tables II. and III. indicates a marked gain in
ability to deal with the 15 sec. and 20 sec. intervals. In
Table III. the 15 sec. interval can be regarded as mastered
and the 20 sec. interval as practically perfected.
The tables just given are valuable in showing just what
the child actually did in the work as a whole. Her achieve-
THE DELAYED REACTION IN A CHILD 81
ments are much obscured, however, by such a presentation,
inasmuch as poor incentives and position factors frequently
dragged her total percentages down. Here, e. g., is the diary
record for the ten tests made November 12. (Data included
in Table I.) The delays were all above 20 sees., and 7 trials
of 10 were successful. Where the word 'distracted' is used,
TABLE IV
Box with Toy Delay Behavior
20 sees, distracted a
a
22 '
« ii
bed
c
21
«
. . . .abc
c
2O
«
. . . . c
b
d
35
22
«
stood her up . . .
. . . .b very slow and * careful.'
... .a
c
26
distracted
c
b
c
c . .
20
22
. .24.
«
stood her UD . .
....b
. . . .c
. .ac
Thayer's eyes were covered and her body was shaken back
and forth; or she was induced to look up at the ceiling and
listen to me count while I waved her arms about. The
letters in the last column indicate to which box the reactions
were made. In every case the subject was oriented, body and
face, to b at the moment of reaction. This day's record is
better than any that preceded it. Two weeks previously,
she had succeeded with intervals between n and 19 sees.
But during those two weeks that followed, she was largely
the victim of position habits. The day following the above
diary record, she again fell back into position habits. I kept
holding pretty well to 15 and 20 sec. delays with an occasional
one at 30 sees. The task was too difficult, however, and she
shifted from one position habit to another.
Work was discontinued for a month. By the end of this
time, the following changes had occurred in the child: (i) old
position habits were temporarily lost; (2) new interest was
taken in the problem; (3) greatly increased control of her
own body appeared — shown mostly in walking and balanc-
ing; and (4) a stronger aversion to being held during distrac-
tions had developed.
82
WALTER S. HUNTER
Table III. above summarizes the results for this period.
It also understates the subject's behavior. On January 5,
she made the entire day — 9 trials — at 15 sees, without error
and with no correlation between orientation and direction of
response. She now fell into a position habit, but recovered
and made 20 sees, delay correctly 5 times in succession, again
with no dependence upon orientation.
Thayer's best delays may be recorded as 20-24 secs- The
child F. used in the earlier work reached a delay of 50 secs.
F.'s record would probably have been higher had she been
tested with a method similar to the one here employed, i. e.,
a method where the satisfaction is derived from the stimulus
object and not from an associated food supply. The gap
between Thayer and F. would undoubtedly be bridged in a
gradual manner by a continuous increase in periods of delay.
Greatest interest now centers on children of less than one
year of age. How early ontogenetically does this ability to
react independently of orientation appear?
It remains to comment upon the position habits and
errors that appeared. The frequency with which these
stereotyped forms of response interfered with the work and
the j act that the child if permitted would watch the box containing
the toy during the interval of delay, indicate the great import-
ance of kinaesthesis in the response. Position habits occurred
with each of the three boxes so that during a particular posi-
tion habit period Thayer always chose a particular box first.
I made no tests where the choice lay between two boxes as
opposed to three. Time is limited both by the speed with
which a baby grows and even more by the necessity of staying
within the limits of the child's interest and patience. Table V
TABLE V
Total Reactions
Made
Order of response
cba
cab
CO,
cb
No made
17
11
8
28
66
Order of response ..........
abc
acb
ac
ab
No. made
6
3
2
II
22
Order of response
bac
bca
be
ba
No. made
7
3
7
9
26
THE DELAYED REACTION IN A CHILD 83
analyses all incorrect responses and gives the relative number
of times the subject followed the different possible orders.
Thus when an error was made, 17 times Thayer first opened c,
then b and then a. The table shows that three times more
errors were made beginning with box c than with any of
the others. When the subject opened c first, she opened b
next 45 times out of 66, or 68 percent of the time. When
she opened a first, she chose b next 15 times out of 20, or 75
percent of the time. When b was opened first, a was chosen
next 1 6 times out of 26, or 6 1 percent of the time. In other
words, when the reaction began at the end of the apparatus
the tendency was to take the boxes in order until the solution
was reached. Only six times in all did the subject go to the
same box twice in the same trial. These cases are distributed
throughout the entire period of experimentation. The fol-
lowing is a record of the order of the boxes chosen:
cccab
bcba
cacb
cacb
bcba
cacacab.
Of the 114 errors recorded in Table V., 32 (28 percent) oc-
curred when the box containing the toy on the last previous
trial was re-selected. Inasmuch, however, as such a mode of
response often led to success the percentage is very low.
This form of behavior as well as that of the six instances above
given is apparently far less current in the present subject
than in Hamilton's dog.1 The later study made by Hamilton2
reports the case of a child 26 months old. Out of 38 trials,
60.53 percent (34.21 plus 26.32) of the reactions involved the
type of behavior given just above as occurring but 6 times
during the present work, 264 trials. Since Thayer missed
120 trials (66 plus 22 plus 26 plus 6), her percentage is 5.
1 Hamilton, G. V. T., 'An Experimental Study of an Unusual Type of Reaction
in a Dog,' /. of Comp. Neur. PsychoL, 1907, 17, 329-341.
2 Hamilton, G. V. T., 'A Study of Trial and Error Reactions in Mammals,' /. of
Animal Behav., 1911, i, p. 51.
84
WALTER S. HUNTER
This extreme difference in behavior is undoubtedly due to
one or both of the following causes: (i) the guiding influence
of the absent stimulus in the delayed reaction tests; and (2)
the fact that only three boxes were used here as opposed to
Hamilton's four. It would be very interesting to determine
whether a variation in the number of boxes would result in a
corresponding variation in * reaction tendencies.' If this
were true, the possibility of phyletic correlations would be
pushed still farther back than appears in Hamilton's work.
It will be valuable to put beside this work, similar data
gathered on rats and raccoons in 1910-1912. The records
here given are representative and include only tests made
with three boxes on periods of delay. The following indicates
the maximal delays attained by the animals whose records
are used in this paper:
Rat No. 9, maximal delay 10 sees.
Rat No. 2, " " i "
Dog Blackie, " " 5 mins.
Raccoon Bob, " " 30-35 sees.1
Table VI. summarizes the errors made by these four
animals. It includes for comparative purposes the data for
Thayer. The raccoon's records include delays from I sec.
through 20 sees.; those for the dog, from i sec. through 7 sees.;
TABLE VI
Animal
No. of
Trs.
Total No. of
Errors A
3 Place
Errors B
Persistent
Errors C
Percent, of
A toC
Percent, of
.gto C
Thayer
264
1 2O
CA
6
c
II
Raccoon, Bob
Dog, Blackie
Rat No. 9
720
570
cyr
209
127
14-4.
78
75
4.2
29
25
j-j
13
19
37
33
•?o
Rat No. 2
345
152
69
47
32
60
those for rat No. 9, from the third stage of delay (turning
light off just as animal was released) through 7 sees.; and
those for rat No. 2, from the third stage of delay through
I sec. I have included the data represented by * percent of
B to C' because Hamilton's percentages are based only on
those reactions that included all the boxes of his apparatus.
1 These data are taken from my 'Delayed Reaction,' pp. 35-38.
THE DELAYED REACTION IN A CHILD 85
The column '3 place errors' includes the trials that involved
a testing by the animal of each of the three boxes. By
4 persistent errors/ I mean all errors that involve trying any
one box more than once each trial. These were all 3 place
errors. This column corresponds to reactions belonging to
Hamilton's types D and E.
The only one of Hamilton's human subjects whose per-
centage in D plus E rose above 6.45 percent was a 26-months-
old child whose grade was 60.53 percent. Of the animals
below man, the lowest grade (best record) was 22.58 percent
made by a dog. Hamilton's results and my own here pre-
sented indicate a marked difference between man and other
animals in reactive tendencies, i. e., in forms of kinaesthetic
habits. (There is, I think, no clear evidence as yet that the
tendencies are instinctive.) Whether this is caused by
phyletic factors, or by experimental and environmental con-
ditions is a matter undecided. My infra-human animals are
essentially on a par. And so I think are Hamilton's in that
practically all of them made their highest percentages in
what I here term ' persistent errors.' (His curves would be
quite different, naturally, if D and E were combined and if
B and C were combined.1) Much work is undoubtedly
needed to determine how minute a classification of reaction
tendencies can be and still be significant for animal ability.
THEORETICAL CONSIDERATIONS
There is very little in the way of interpretative comments
that I can add to what has already been said in other papers.2
The delayed reaction problem can be solved at least in two
ways:3 (i) by the maintenance of bodily orientation in whole
1 Hamilton, op. cit., 1911, p. 54.
2 'Delayed Reaction,' pp. 62-79 an^ Hunter, W. S., 'A Reply to Some Criticisms
of the Delayed Reaction,' /. of Phil. PsychoL, t3c.y 1915, 12, 38-41. See also Watson
J. B., 'Behavior,' 1914, pp. 224-227; and Ch. X.
8 If there is a present determining external stimulus, the reaction is not delayed.
A. C. Walton, 'The Influence of Diverting Stimuli during Delayed Reaction in
Dogs,' /. Animal Sehav., 1915, 5, 259-291, has shown that dogs can react successfully
to three boxes after delays of 30 sees, when they have been distracted during the
intervals. This is better than I had been able to show. Before deciding that the
dog belongs in a class with the raccoons — as perhaps he does — it will be necessary to
86 WALTER S. HUNTER
or in part during the interval of delay or by the chance
recovery of the proper orientation just at the moment of
release; and (2) by the use of some intra-organic factor which
is wow-observable by the experimenter. In the first method,
the animal always responds in accordance with orientation;
in the second, he does not. The cue used in the second
method may or may not be retained in the focus of neural
activity during the delay. It is highly improbable that such
retention occurs under conditions of distraction. What one
has, then, is a system of processes or cues which 'stand for'
certain differential responses as a result of association.
These cues are susceptible to selective re-arousal and subse-
quent successful functioning in initiating responses. This is
the condition which I have previously found in raccoons and
in one child, F., 2| years old. It is the condition here pre-
sented by Thayer, ages 13-16 months. This second method
of solution which I am describing may be mediated by any
type of intra-organic process which can be re-aroused with-
out the presence of the external stimulus, toys as used in
the present study. Inasmuch as kinaesthetic factors can be
so aroused,1 and inasmuch as they have been demonstrated
to have great importance not only in animal reactions in
general but in the delayed reaction in particular (position
habits and maintenance of orientation), it is most probable
that the intra-organic factors are kinsesthetic in nature.
(The genetic relations of sensation, image and imageless
thought are discussed in 'The Delayed Reaction/ pages cited.)
In certain cases this type of process has its locus in the vocal
organs with frequent resulting audible sounds. This we
term vocal language as it occurs in normal human adults and
in children of a certain development. In other cases the
observance of the sounds and their accompanying behavior
have data on the animals' orientations at the moment of response. This Walton
unfortunately does not give. The only facts that we have are that the animals did
not maintain their orientations during the delays. If it should appear that the animal
is able to recover the proper orientation in a large number of instances after a thorough
distraction and can then react correctly, this fact will itself be ofjjreat significance and
will require careful analysis.
1 See the all too brief comments in my review of Calkins' 'First Book in Psy-
chology,' PSYCHOL. BULL., 1915, 12, 189-190.
THE DELAYED REACTION IN A CHILD 87
does not indicate that the organism uses either the sound or
the parallel kinsesthesis as a substitute of the type above
described. This is the situation in all animals that indulge
in vocalization. But this intra-organic kinaesthetic factor
may arise elsewhere than from the throat. Some part of the
general bodily musculature may be the origin. Here, when
the behavior is overt, we speak of gesture language. When
it is not overt, the delayed reaction method has proved
serviceable in detecting it. Language is ideational in func-
tion. So also are the cues which function in many responses
of raccoons, of children and possibly of dogs (Walton). The
resulting conception of these cues is that they are kinaesthetic
sensory ideas. This line of reasoning leads one to conclude
that a true language non-vocal in character appears phylo-
genetically and ontogenetically prior to vocal language.
Such language, although undoubtedly of great service to the
individual in controlling his reactions, is of little social
significance.
VOL. XXIV. No. 2 March, 1917
THE PSYCHOLOGICAL REVIEW
THE LAWS OF RELATIVE FATIGUE1
BY RAYMOND DODGE
Wesleyan University
The temerity that ventures to speak of fatigue laws may
well arouse a critical attitude. But I shall not be quite so
indiscreet as the title might be misconstrued to imply. For
reasons that will presently appear in detail, I have no ex-
pectation that the laws of mental fatigue will be formulated
in the immediate future. Oeffner's so-called laws of fatigue
are obviously only empirical generalizations and summaries.
My subject is really much less pretentious. It concerns only
the relativity of fatigue. The laws of relative fatigue that
we shall discuss might with equal propriety have been called
the laws of fatigue relativity.
My excuse for selecting so threadbare a matter as fatigue
for the subject of the presidential address is largely personal.
As some of you know, I have been working on various phases
of mental fatigue experimentally for a number of years — too
long for self complacency. More than once it has seemed to
me that I was following a clear experimental path out of the
maze of fact, only to find myself back again at the starting
point, facing the same fundamental questions. But however
personal my interest in fatigue may be it certainly is not ex-
ceptional. I venture the guess that there is not a member of
this Association but has made fatigue the subject of direct,
indirect, or projected investigation. Certainly few psycho-
logical subjects have so widely interested investigators in the
allied sciences. Few seem to have at once such far-reaching
1 Address of the President before the American Psychological Association, New
York Meeting, December, 1916.
90 RAYMOND DODGE
bearings on psychological theory and the conduct of human
affairs. Few present such a bewildering literature, with
such an array of apparently mutually contradictory experi-
mental results. None is more confused with an equal pres-
sure for practical working rules. Confusion and eagerness
for practical results make a situation fraught with grave
peril to science. If anything could, they justify this attempt
to clarify and systematize the fundamental concept of mental
fatigue.
It would be an impracticable as well as an uncongenial
task for me to attempt a review of the literature of fatigue,
even if this were a fitting occasion. Our time limits, our
precedents, and my personal interests persuade me rather to
attempt what I hope may prove to be a more generally useful
undertaking, namely, a substantive analysis of the problem.
The first part of that task, as I apprehend it, is to clear the
problem of some misleading assumptions by which faulty
analogy and practical interests have confused the real issues.
Thus simplified we shall try to redefine the psycho-physical
problem on a scientific rather than on a practical basis.
Mental fatigue is one of those scientific problems that has
suffered from too much practical importance. In the enor-
mous number of investigations that have appeared since the
publication of Mosso's "epoch-making" book, just a quarter
of a century ago, educational, medical, and more recently
social and economic interests have given the dominant motifs.
It was indeed an alarming arraignment of the schools that
they ruined the health and impaired the eyes of pupils by their
excessive demands. An investigation of such charges was a
direct obligation on experimental pedagogy. Scarcely less
important than the school problems is a just determination
of the proper duration of an industrial day, with a fair con-
sideration for the welfare of the laborer and for the exigencies
of competition. None of us, moreover, is entirely free from
more personal practical difficulties in our desire to exploit
most effectively the time and energy at our disposal. Now
I would not for a moment be misunderstood to depreciate the
importance of these practical studies. My only contention
THE LAWS OF RELATIVE FATIGUE 91
with respect to them is that they all suffer more or less from
an inadequate scientific basis. But in spite of all the con-
fusion of alleged fact, all the premature and unverifiable
pronouncements, most of us still believe that an adequate
answer to such practical questions is both desirable and
possible.
Less insistent and obvious, but none the less real and
important, are the scientific problems of mental fatigue.
For the present at least it may be of some advantage to keep
separate the two lines of investigation, the practical and
the scientific. The former is, in the main, quite independent
of the latter. However the questions as to the nature and
laws of mental fatigue may finally be answered, careful
generalization from experience as to the expediency of certain
work and relaxation sequences will deserve and will receive
careful consideration in planning the day's work. So too
the most advisable length and distribution of recesses may
be settled purely empirically, entirely without reference to
any of the underlying bio-chemical processes. For all such
practical purposes the concept of fatigue is an accident. Its
function is not to recall the implications that it has in bio-
chemical science; but merely to serve as a vehicle for practical
maxims, a class name for all sorts of unanalyzed hindrances
to effective work. The hindrances would be just as real and
the practical maxims just as valuable even if it were proven
that fatigue had nothing to do with them. The scientific
problems as to the real nature and conditions of a supposi-
titious mental fatigue are quite independent of all such ques-
tions of practical expediency. Scientifically we must know
the differential characteristics by which mental fatigue can
be distinguished from the other limitations of the work curve;
as well as its elementary forms and their interrelations. We
must follow its implications as an indicator of the relationship
between mind and body; and correlate it with other bio-
chemical facts. It is these latter problems that appeal to
me personally with especial emphasis. Mental fatigue if it
exists in the physiological sense, must be connected in some
direct way with the energy transformations in nervous tissue,
92 RAYMOND DODGE
and the fundamental problems of inner psycho-physics. The
great problem whether our mental life conforms like the rest
of the organism to the underlying postulate of thermo-
dynamics, the conservation of energy, must be answered if at
all by the psycho-physics of work and fatigue. While I
sincerely hope that adequately equipped attempts to explore
these fundamental questions are not too far distant, there are
related problems that can be examined by simpler techniques.
Again mental fatigue if it exists ought to furnish us with an
instrument of dynamic analysis of the mental complexes,
reaching the inner mechanisms of our mental life. To
estimate its possible usefulness, one has only to think of the
analogous use of fatigue or adaptation in sense analysis; when
we adapt out one sense quality and note the effect of its loss
on the other qualities, or on perception in general. So, for
example, the relative composition of two purples might be
shown even if we had no other method, by adapting out
spectral blue, and comparing their resulting appearance.
It would seem that a similar process ought to be applicable
to mental experiments when we have no other means of ex-
perimentally eliminating the various, factors. One might
even outline the working postulates of such an analysis as
follows: I. Whenever in mental processes fatigue of one is
regularly accompanied by fatigue of another there must be
some dynamic factor common to both. II. Conversely,
whenever the fatigue of one mental process does not show as
fatigue of another, the two must depend on different dynamic
conditions. III. Whenever fatigue in one process is accom-
panied by the improvement of another process then the two
are probably related in the sense that the fatigued factor in
the former was inhibitory to the second. That such postu-
lates have borne little fruit hitherto, is not due to any inherent
logical unsoundness, but rather to our misapprehension of the
character of mental fatigue. At present their application to
the problems of analysis would be handicapped by the very
richness of the alleged correlations. It has proven embarrass-
ing to more than one of us to teach our students on one oc-
casion the very slight correlation between mental processes
THE LAWS OF RELATIVE FATIGUE 93
that seem very much alike; and on another occasion to teach
them how mental fatigue in general may be measured by the
pulse, the ergograph, addition, reaction time, the dermal
threshold, and other apparently disconnected events through
a long list of accredited extrinsic tests. To be sure the relia-
bility of these and other so-called tests is not universally
admitted. But the gross discrepancies between genetic and
dynamic correlations might well be taken a little more seri-
ously. Before any of these scientific tasks can be undertaken
with promise of success, we must know what mental fatigue
really is, if there is any such thing, and how it is conditioned.
THE CONCEPT OF MENTAL FATIGUE
The concept of mental fatigue is so familiar that a precise
analysis of its differentia has seldom seemed necessary.
Statements of its meaning, when they occur, regularly em-
phasize a diminution of some product of mental activity per
unit of time, incident to continued activity, and as Thorndike
insists, recoverable through rest. Actual recovery or the
capacity to recover through rest seems to me to be irrelevant.
On the one hand it excludes extreme fatigue; on the other
hand it fails to exclude all sorts of intercurrent disturbances.
But the diminution of production consequent to continuous
work with or without recoverability, is I believe an untenable
criterion of fatigue.
If the word fatigue has any scientific propriety in connec-
tion with our mental life, it seems to me that it should refer
to the metabolic conditions of mental action, not to any
predetermined characteristic of its consequences. This is
very much the same point that I made recently concerning
mental work. While that was not received with the una-
nimity that I had hoped, in the case of fatigue at least, failure
to realize the dynamic implications must lead to gross con-
fusion. Obviously psychology or pedagogy is entirely com-
petent to ignore the physiological concept of fatigue and to
develop its own empirical concept as decreased returns of
mental work. But if it ignores the metabolic implications
at the beginning, it may not assume at the end that physio-
94 RAYMOND DODGE
logical and pathological fatigue processes parallel the de-
creased returns. It is such gratuitous assumptions concerning
matters of fact that make psycho-physical parallelism a
dangerous working hypothesis. Moreover, such an inde-
pen.dent psychological concept would be scientifically de-
fensible only to the degree that work decrements, consequent
to work and eliminated by rest, prove to be homologous
processes with regular and definable antecedents. If on the
contrary work decrements show a large variety of types, or
follow any considerable variety of conditions, it would seem
to be good sense and sound science to enquire whether any
of the varieties of mental work decrement correspond to
physiological fatigue processes. These alone would then
seem to have a natural right to the name mental fatigue.
From this standpoint other decrements would be regarded as
pseudo-fatigues.
In order to conserve our time let me be quite direct and
frank. I regard it as improbable that any of the mental work
decrements so commonly treated as mental fatigue, are ever
simply conditioned by true fatigue processes in nervous tissue.
Conversely real fatigue may not appear as a decrement at all.
Some of the evidence for this position can only be indicated
here. Some of it must be given in more detail.
First one must note the physiological fact that nervous
tissue in situ has been found quite resistant to fatigue and
exhaustion under normal cirumstances. The axis cylinders
apparently never fatigue except under experimental conditions
when their environment is freed from oxygen, or when they
are narcotized so that they are unable to use the oxygen that
is present. Cell bodies are likewise resistant to fatigue under
normal circumstances. They can be exhausted in experi-
mental animals only under strychnine poisoning, after the
withdrawal of normal blood supply. Langfeld has shown that
in humans prolonged fasting produces no correlated decrease
of neural efficiency. Reflexes like the knee-jerk and the
protective lid reflex show no decrement after long series of
elicitation, if care be taken to prevent intercurrent general
depression of the nervous system. In those cases where
THE LAWS OF RELATIVE FATIGUE 95
fatigue decrement of the reflexes does occur, there is evidence
that neither the muscles, the nerves, nor the nervous centers
have lost their irritability except to the particular stimulus to
which they have become adapted. On the contrary hyper-
excitability is a common if not a regular phenomenon of
extreme so-called mental fatigue. At any rate it would seem
that the complete cessation of mental processes, like the
inability to recall an opposite, to complete a sentence, to
recite a series of nonsense syllables, or to multiply four-place
numbers by mental arithmetic cannot possibly mean the
fatigue of nervous tissue to the corresponding degree of com-
pleteness.
A second ground against the traditional differentiae of
fatigue is their failure to exclude normal psycho-physical
rhythms. In more than one respect it was an unfortunate
accident that the paradigm for the interpretation of the
phenomena of mental fatigue was the fatigue of a nerve-
muscle preparation. Undoubtedly there are many and im-
portant analogies between the action of lower spinal arcs and
cerebral processes. But after all the main task of physio-
logical psychology begins when it seeks to understand the
differences between the simpler processes and cerebral action.
Similarly, in connection with a supposititious mental fatigue
the regularly increasing work-paralysis of nerve-muscle prepa-
ration may be and in some respects must be a misleading
model. One of the great differences is that while the extir-
pated preparation changes only slowly under experimental
conditions when unstimulated, normal mental life precludes
unchanging neural conditions. In the complex intercon-
nections of human cortical processes the one statement that
can be made with completest conviction is that the experi-
mental subject never remains constant, quite apart from the
intended experimental changes. Even under the best possible
experimental conditions, the experimental change is only one
of the changes that we know to be occurring. The consti-
tution of these non-experimental changes in any given case
we know only in part. We believe that consciousness itself
is a process which involves more or less continuous inherent
96 RAYMOND DODGE
change. We know that there are also various intercurrent
physiological rhythms, cardiac, vascular, respiratory, in-
testinal, glandular, and muscular. Cortical action may also
initiate non-rhythmic changes in the glandular, circulatory,
and respiratory systems with far-reaching reactions of those
changes on the cortical action that originated them.
From the standpoint of the importance of the accompany-
ing mental changes, perhaps the most significant of the
rhythms is sleep. The fatigue-hunting enthusiasm that finds
in sleep the daily climax of fatigue is without physiological
justification. On the contrary, we have learned from experi-
mental investigations that for some persons evening may be
the time of most effective mental work. Moreover, it is
neurological commonplace that in serious extreme fatigue,
sleep may be impossible. Physiologists would welcome any
insight that we could give them into the causes of sleep. The
fatigue climax assumption simply is not tenable. Whatever
they may be, we know that the conditions of sleep are not
simple. Habit, the absence of stimuli, probably widespread
inhibitions, and possibly gland products and vaso-motor
changes cooperate in its production. Sleep may come from
restriction of activity quicker than from over-exertion.
Lecturers never go to sleep. The audience may. In view of
such complication of the conditions of continuous work dec-
rement the assumption that all diminished returns conse-
quent to work and eliminated by rest are fatigue seems to me
utterly untenable.
A third ground of suspicion against the true fatigue
character of most so-called mental fatigue is found in the
means that are commonly used to induce it. In nerve-muscle
fatigue experiments one isolates a specific tissue and stimulates
it successively in the same manner. In mental fatigue experi-
ments, on the contrary, repetition of the same stimulus is
systematically avoided. The more carefully one analyses
the assumptions of this anomalous technique, the more in-
congruous it appears. Let us take a concrete instance from
what Thorndike has taught us to regard as one of the purest
forms of mental work, mental arithmetic. If we strictly
THE LAWS OF RELATIVE FATIGUE 97
followed the analogy of physiological fatigue experiments some
association in mental arithmetic, say the multiplication of
two times two, should be repeated until work decrement or
paralysis indicated fatigue of the association process. As
far as I know that is never done. It seems absurd. The
experimental device of constantly changing the stimulus in
fatigue experiments is defensible only on the assumption
that all multiplication processes affect the same general group
of tissues, and that continuous multiplication of different
digits increases the sum of the fatigue of the whole. But
neurologically the assumption is certainly a strange one that
the nervous tissue which was involved in one association
fatigued more when a variety of different associations were
made than it did when all the burden fell on the same asso-
ciative elements, operating continuously or in rapid succession.
Moreover, there are no facts available to show that restriction
to a single field like multiplication will produce greater work
decrement than rapid change from one field to another. On
the contrary there is evidence that the greater the complexity
of the mental task the more pronounced is the decrement.
Such decrement, however, is more probably due to a confusion
between different paths of discharge than to fatigue of any
particular path. That confusion is real and a common ex-
perience every introspective account is evidence. The-
oretically it should be expected from the operation of the
known laws of association. Suppose, for example, that after
adding various digits to seven we come to the task of adding
four. The right associate is by hypothesis well known and
thoroughly practiced. But if other numbers have recently
appeared in the series they also tend to be reproduced on the
basis of recency. It is at least conceivable that the true
associate in such a case might be difficult to recall, not at all
from fatigue of the corresponding tissue but from effectual
inhibition because a more recent associate appears in its
stead. The necessity for inhibiting irrelevant and false asso-
ciates is certainly a common experience in the elementary
mathematics of some of us. But the tendency of recent
ideas to recur is not in any sense a fatigue or exhaustion
98 RAYMOND DODGE
process, but is probably a matter of residual excitation and
summation. Such work decrement then is not fatigue but
mere association rivalry.
A fourth ground against identifying work decrement and
fatigue may be found in the operation of incidental inhibitions.
Theoretically, every mental operation arouses more or less
widespread associated reverberations which manifest them-
selves in the sequences of actual associative recall, and may
on occasion, as we have just seen, operate to confuse the
regular sequence of work by a kind of associative rivalry.
Theoretically also, every actual association process involves
more or less widespread inhibitions of undesirable associations.
Now it is conceivable that these useful inhibitions of the irrele-
vant might operate to produce a pseudo-fatigue work decre-
ment in any extrinsic test. For example, I have published
experimental evidence that the most intense mental work of
an examination period commonly follows the first reading
of the examination questions. It is the period of adjustment
to the examination as a whole, when widespread association
systems are being organized. Such activities are not possible
in fullest degree without corresponding inhibitions. Ordi-
narily distracting stimuli pass unnoticed. Even physical
discomfort and pain may for a time be ignored. Now it is
conceivable that if at such a time the fatigue tester should
request the examinee to add digits for two minutes as rapidly
as possible, the response might show a degree of work decre-
ment that bordered on total incapacity. Or again, suppose
we would measure the fatigue of a Wall St. broker, hour by
hour, with the aesthesiometer test. And supposing as the
hour struck we should interrupt a selling campaign that was
taxing his skill as a broker by the request that he submit to
our compass point test. The chances are in favor of some
rather vigorous verbal defensive reactions with no discrimi-
nation at all. But if we were able to hold him to a promise
and actually start the test, is there any guarantee that gross
decrements in the measured function, all due to previous work
and remediable by rest, might not be due to his inability to
give his attention to our petty tests while his fortune was at
THE LAWS OF RELATIVE FATIGUE 99
stake on the floor? Of course the whole situation is absurd.
The most enthusiastic believer in space threshold tests would
hesitate to use such results as an indication of the broker's
general mental fatigue at that time.
We freely admit that these are extreme cases, and that
they break the most elementary rules for experimentation.
But have we any guarantee that similar discriminations
against some seemingly unimportant task might not occur
just after recess, or just before school lets out, when the
afternoon's escapades are in the making, or any time at the
interruption of seemingly important processes? Conversely
is there any guarantee that the interruption of annoying or
even fatiguing work by a few moments of trivial testing might
not be a joyous relief, giving results that might entirely hide
a supposititious real mental fatigue of the interrupted work?
I am not arguing that such inhibitions would not be very
much worth knowing; but merely that it confuses their real
bearings to call them all fatigue.
In addition to these specific inhibitory processes which are
commonly classed in psychology as phenomena of attention,
we are acquainted with secondary inhibitions through a dim-
inution of the supporting organic processes, glandular or
circulatory. Of the glandular changes I have no direct
knowledge. The initial increased pulse frequency, whenever
complete relaxation is interrupted by any mental activity, is
commonly followed by a gradually decreasing heart rate in
any prolonged experimental task. We may regard this as a
kind of adaptive process, an habituation to the task at hand.
It is difficult to conceive of it without reference to the gradual
elimination of extrinsic excitations, in which an initial general
excitation is followed by an inhibition which restricts the
excitement to selected processes. I have been able to dem-
onstrate that something of this sort occurs in every normal
reading pause. That continuous fixation of a trivial object
is inhibitory is shown by its familiar hypnagogic tendencies.
It is one of the methods of producing hypnosis. With some
probability we can predict a diminution in the organic condi-
tions for metabolism in all relatively unused neural centers
loo RAYMOND DODGE
during monotonous mental work. In extreme cases continu-
ous disuse leads to atrophy, muscular, neural, and glandular.
To regard work decrement which is due to more or less
complete atrophy of unused paths as fatigue would be a
manifestly absurd confusion of concepts. But work decre-
ment from secondary trophic deficiency, as in unused parts,
is just as surely not fatigue. Just as in periods of excitement
and important readjustment, there are undoubtedly vascular
and glandular changes which increase the activity of the
whole neural mechanism, reflexly reinforcing the processes
that initiated them; so it is probable that general depressions
of glandular or vascular origin accompany monotonous mental
work, in which even the centers that are most active finally
participate. But this again is not fatigue in any physio-
logical sense.
In as far as these various processes represent work decre-
ments or decreased returns that might be mistaken for
neural fatigue they may properly be called pseudo-fatigues.
We have described pseudo-fatigues of intercurrent rhythms,
of residual excitation and rivalry, and of specific and trophic
inhibition. The pathological evidence that Work decrement
is no true indicator of nervous fatigue is not new. Even to
summarize it would extend our paper too far. But I think
that without it, we have established the thesis that decreased
returns resulting from work and recoverable by rest if you
will, cannot be employed as simply and directly in the higher
neural systematizations as it can in simpler tissues. Ar-
bitrarily to define mental fatigue as work decrement is effec-
tual self-banishment from physiological tradition as well as
from clearly defined fields of investigation of the utmost
importance.
Having divested the mental fatigue concept of its irrele-
vant content as vehicle of the various work decrements, it is
now in order to inquire whether there is in our mental life a
real fatigue phenomenon. I believe that there is, but its
manifestations differ from the paradigm of nerve-muscle
fatigue in two important particulars. These are: first, the
inconstancy of the stimuli in mental work; and second, the
THE LAWS OF RELATIVE FATIGUE 101
interaction of competing paths. These two differences
combine to emphasize the relativity of all mental fatigue.
THE RELATIVITY OF FATIGUE
In the nerve-muscle fatigue experiment, the stimulus is
always simple, and usually constant in intensity, given at
regular time intervals. For a variety of reasons the stimulus
that is most used is the faradic current. It is capable of fine
adjustment, may be held at constant intensity over long
periods, and is exceedingly effective in quantities that do not
damage the tissue. No physiologist would start a fatigue
experiment with stimuli of unknown and variable intensity.
Unfortunately, that seems to be the only practicable method
at present in so-called mental fatigue experiments. Nobody
knows the relative stimulus value of two different mathe-
matical sums. But what is vastly more embarrassing, nobody
knows how to follow or to evaluate the ever-changing inner
factors in the total mental stimulus, such as the force of the
instructions, the personal interest of the subject in the scien-
tific aspect of his task, in its bearing on the particular exi-
gencies of his academic career, and so forth. It was one of
the great services of Kraepelin in his analysis of the work
curve to show how these inner stimuli may change during an
experimental period. The meaning of that analysis as I
apprehend it is not given in the precise variables or spurts
that he found, nor in the assumption that they are always
present, but rather in the demonstration that variables in
the inner stimuli may occur and must be reckoned with. It
would not take us long to add to his objectively defined list
many others taken from our experimental experience, such
as competition and personal pride, repetition of the instruc-
tions, encouragement and persuasion, the presence of the
instructor, rewards and penalties of various sorts, and the
unanalyzed mass of obligations.
I am not unaware that this matter of the inner stimuli
to mental work is packed with problems that we have no
adequate techniques to investigate. But that is no excuse
for ignoring them. It is our business as scientists to try to
102 RAYMOND DODGE
see things as they are, even if they are complex. There is at
least some ground for the suspicion that most if not all our
real mental fatigue of the work decrement type is really a
fatigue of the inner stimuli rather than of the capacity to
react. This at least would account for the extraordinary
correlations in the fatigue of the most diverse functions. In
many so-called mental fatigue experiments the only common
factor discernible to introspective analysis is the intent to
keep working as fast as possible to the neglect of competing
interests.
Now in the physiological experiment fatigue may be
shown in two ways, either by a rising threshold or by decreased
response to a constant supra-threshold stimulus. Only in the
latter case is there an obvious work decrement. The former
case implies a constant work output with a gradually increas-
ing stimulus intensity. In mental work we are often distinctly
aware of similar changes in the intensity of the inner stimuli
that keep us at a disagreeable or monotonous task. Mere
interest in the task may lose its force comparatively early.
Then the task is continued from stubbornness, the dislike to
fail, sense of obligation, honor, fear of ridicule, or hope of
reward, etc. All of these may operate in succession. In the
end all of them may lose their force and we say, " I do not care
what happens, I cannot go on with this thing any longer
to-night." There may have been no important work de-
crement until the break, as Yoakum calls it. But the process
is none the less a real fatigue if the continuation of work
depends on a change of the stimuli.
All of this emphasis on the importance of the neglected
factor of changing stimuli in the fatigue concept is probably
sufficient to justify the formal statement of a necessary
correction in the traditional definition of mental fatigue. We
may call it the first law of relative fatigue, neglected rather
than new. Without pretending to give it final formulation
we may express it as follows: Within physiological limits, all
fatigue decrement in the results of work is relative to the
intensity of the stimulus.
Education and society have a very practical interest in
THE LAWS OF RELATIVE FATIGUE 103
this phase of the fatigue problem.' They make use of a large
number of incentives in which as Thorndike wisely points out
the changes in satisfyingness may be a real cause of work
decrement. The adequate adjustment of stimuli to the
development of the individual and the needs of the case would
seem to be a very real problem in the training of backward
and gifted as well as normal children. It seems strange that
we have so little experimental knowledge of the relative value
of available reinforcements. Autogenic reinforcement is, I
believe, at least one factor in the underlying psycho-physics
of James's ' reservoirs of power' which may be quite as sig-
nificant for psychology as the action of adrenin to which
Cannon has introduced us. That continuous activity under
the reinforcement of emotion or even in the educational use
of play may be a source of serious fatigue we have been
warned by Kraepelin. Some other reinforcements are con-
spicuous for their insistence. Such a one is worry. It would
seem to be no accident that this is so closely connected with
exhaustion psychoses.
I believe that the relative value of the various inner stimuli
would repay the closest study. Just now it seems to be inter-
esting the abnormal rather than the normal psychologist.
Practical experience is full of rough approximations. Their
refinement by experimental techniques would not seem to be
an impossible task.
It is possible that we can study relative fatigue not merely
by the changes that occur during long series of repetitions but
more expeditiously in the relative refractory phase which the
genius of Verworn proved to be identical with the fatigue
process. Since the relative refractory phase is common to all
nervous tissue, I have asked the question whether we can
find in mental processes a similar phenomenon. This is
undoubtedly the case. In fact every mental process shows
something analogous. Repetitions of all sorts seem to be
avoided whenever practicable. The repetition of questions,
courses, lectures, phrases, and even words is possible enough,
but except for special reinforcing circumstances, it is post-
poned until the effect of the initial case is somewhat worn off.
104 RAYMOND DODGE
The routine is regularly less alluring than the unusual.
Mankind in general prefers new scenes, new plays, new walks,
new jokes, new styles, new investigations. Possibly the
decreased effectiveness of over-memorization is a case in point.
Possibly even the loss of attention to frequently repeated
processes, which is commonly regarded teleologically as a
freeing of consciousness for new adjustments, may be caused
by the longer refractory phase of the more complex systema-
tizations of attention, so that the rapidly repeated task is
dynamically excluded from conscious emphasis.
Works of art on the contrary are characteristically resistant
to the refractory phase. Possibly this results in some way
from their origin. Certainly one of the marks of good art is
the constancy of its appeals. The popular song, the clever
phrase, the good joke, soon finds us refractory to the point
of desperation, though it is notable that we become refractory
to their reception much quicker than to their execution. We
like to tell old jokes better than to hear them. But the great
classics in music and literature may be heard over and over
with increasing satisfaction. It is not impossible that Aris-
totle's catharsis by dramatic representation of suffering and
evil really operates by developing a refractory phase, and a
kind of relative fatigue. How far this principle operates in
habituation to environment, indifference to shocking condi-
tions of poverty and morals, to suffering, and to the horrors of
war, as well as to luxuries "when the novelty has worn off,"
I am not prepared to estimate.
It would seem that some of these or analogous phenomena
ought to yield data for a scientific study of the intensity of the
inner stimuli in connection with fatigue if we only knew how
to use them. But the very difficulties of technique emphasize
how far we are from a real knowledge of relative mental fatigue.
The simplicity of the nerve-muscle paradigm of mental
fatigue is further misleading in that it gives no indication of
certain important complications which are characteristic of
higher systematizations, and which Sherrington called their
competition. In a nerve-muscle preparation the impulse has
only one possible path. In the higher nervous system on the
THE LAWS OF RELATIVE FATIGUE 105
contrary any afferent impulse may theoretically activate
every efferent path. Just which motor process it finally
initiates, is determined by a kind of competition. Competi-
tion appears in the spinal reflexes though less conspicuously
than in cortically conditioned action, where it is the rule.
Unfortunately, however, just where it is most significant it
can seldom be followed objectively by our present means of
investigation. But there are clear evidences of its operation
in associative thinking, in attention, and in perception as well
as in conduct.
The relatively fixed tendencies of competition in the cord
are probably determined very simply by neural growth and
development. In higher systematizations the outcome of
competition tends to follow habitual patterns which have
originated in the varying life-history of the several competi-
tors. At any given moment in a developed system of this
sort, the outcome may be modified by a variety of reinforcing
and inhibiting accidents. Among the latter we must count
fatigue. In closely balanced competition the absolute de-
gree of fatigue need not be high to make it a deciding factor.
Indeed it is conceivable that if the balance of the other factors
is close enough, an infinitesimal fatigue, or the slightest
trace of a refractory phase may totally change the character
of the response, just as intrinsically trivial reinforcements or
accidental inhibitions may be the arbiters.
This relation of fatigue to balanced competition gives us
a second type of fatigue relativity. Fatigue is relative, not
only in the relation of apparent work decrement to stimulus,
as expressed in our first law; it is also relative in the sense of a
proportionate fatigue of the various factors in a competing
group. We may tentatively express this second type of
fatigue relativity in the form of a law which for want of a
better name we may call the Second Law of Relative Fatigue,
because it implies a higher systematization than the first law.
In any complex of competing tendencies the relatively greater
fatigue of one tendency will tend to eliminate it from the
competition in favor of the less fatigued tendencies.
Unfortunately the mechanism of competition cannot be
io6 RAYMOND DODGE
studied at all in simple preparations and only imperfectly in
the reflexes. The most characteristic systems are the least
accessible. In the search for accessible human systems of
greater complexity than the reflexes, it occurred to me, some-
thing over ten years ago, that the motor apparatus of the
eyes offered some unique advantages. There we may study
twelve intimately related and delicately adjusted final paths
which are directly connected with reactions of considerable
biological importance. Furthermore, every variation of their
interaction is capable of being recorded on a single plane,
without complicated mechanical devices, and without the
distortions incident to the moving of heavy masses, like the
limbs. Since that time the eye-movements have proved to
be unusually valuable indicators of neural conditions in some
forms of insanity and under the action of alcohol. In ex-
periments that are now in progress they give promise of being
the most consistent indicators of general neural conditions.
In the early hopes of using them for an analysis of fatigue
phenomena, I took a considerable number of binocular records
of rapid successions of eye-movements after the model of the
ergograph. Though reported on informally from time to
time these records have never been published before because
of my inability to account for some of their most conspicuous
peculiarities. As these difficulties have gradually been ex-
perimentally cleared, the records have been seen to illustrate
in a remarkable way some of the characteristic phenomena of
mental fatigue, and pseudo-fatigue. In particular they ad-
mirably schematize the second law of relative fatigue and
the "breaks" that it conditions.
Let me assume your familiarity with the technique of
photographically recording the eye-movements from the
corneal reflection. For the present records the eyes moved
horizontally through an arc of sixty degrees, fixating success-
ively two knitting needles which were situated thirty degrees
on either side of the primary position of the line of regard.
Each dot or dash on the records represents one phase of the
alternating current, and a time interval of about eight thou-
sandths of a second.
THE LAWS OF RELATIVE FATIGUE 107
sz
*".
io8 RAYMOND DODGE
The succession of eye-movements in the records that are here
reproduced was as rapid as practicable with subjectively
adequate successive fixation of the two fixation marks. Some
of the more characteristic fatigue phenomena which they
show are: (i) The speed of movement becomes less towards
the end of the series; (2) the fixations become less accurate;
(3) and finally the line of movement itself becomes more
irregular. Fig. 2 shows the climax of these processes in a
break. The gradual decrease in angle velocity corresponds to
the work decrement of extirpated muscle. But in this case,
in view of Sherrington's demonstration of the reciprocal in-
hibition of antagonistic eye-muscles, it doubtless involves
something more. The greatest angle velocity of eye-move-
ment could only occur when the relaxation of the antagonistic
was perfectly coordinated with the contraction, of the agon-
istic muscle. The pseudo-work-decrement in this case then
is not purely muscular but is in part a matter of defective
coordination. The increasing errors of coordination have a
similar origin. That is, the total elaboration of the con-
traction impulse and the corresponding relaxation of the
antagonistic becomes less exact in successive repetitions of
the act of fixation. But the coordination is not limited to the
internal and external recti as one might expect them to be in
horizontal movements of the eyes. All the records of 60"
eye-movements, which I have ever seen, show a vertical
factor. In all my records this vertical factor results in an
elevation of the line of regard. But it varies from movement
to movement. That these vertical components are not acci-
dents of purely muscular origin is shown by binocular records.
Since the disturbances are homologous for both eyes, their
origin must lie in the central nervous system. While occa-
sional gross disturbances occur early in the series of move-
ments, they become more and more conspicuous as the
series progresses. The vertical components represent the
intercurrent action of related and competing, but this is a
case of non-inhibiting systems. When they become extreme
they tend to interrupt for a moment the main rhythm of
horizontal movements. In some cases these various disturb-
THE LAWS OF RELATIVE FATIGUE 109
ances produce a moment of confusion and a break in the
process, which in ordinary mental fatigue experiments would
be interpreted as complete fatigue or exhaustion.
Our eye-movement schema for the relative fatigue of
competing systems is particularly free from complications
through extrinsic facilitations and inhibitions. Retinal fa-
tigue or adaptation is reduced to a minimum by the eye-
movement itself, and the consequent shift of the area of
stimulation. The homologous fixation marks, under constant
illumination present the same stimulus for each reaction in
the same direction. Cortical conditions of the successive
reactions, such as interest, attention and motives to continue
at work, cannot of course be guaranteed to remain constant.
But the experiment itself introduces no obvious distractions
like the physical discomfort of the ergograph. Moreover, all
our relative fatigue phenomena appear during short experi-
mental periods.
In order to protect our conclusions from the dangers in-
herent in a single line of experimental evidence, I sought other
similarly complicated coordination systems. While thus far
no other has been found with all the advantages of the hori-
zontal eye-movements, those movements of the index finger
which Bergstrom recommended for ergographic work show a
similar complication. Undoubtedly the strongest and best
practiced oscillatory movements of the index finger are the
flexion-extension movements. Considerably less easy for
most of us are movements of the finger sideways in the plane
of the hand. In any event, rapid oscillation of the finger in
this direction is always disturbed by intercurrent action of the
flexors and extensors. Their action prevents rectilinear
movement, decreases the angle velocity, and finally may so
confuse the process as to produce a break in the sequence of
oscillations, quite like the disturbances of the eye-movements.
It was the phenomena of these relatively accessible com-
plex systematizations that forced me to a re-analysis of the
mental fatigue concept. I believe that our eye-movement
paradigm gives us the clue not only for a more intelligent
experimental investigation of mental fatigue, but also for the
no RAYMOND DODGE
interpretation of previous investigations. The very irre-
gularity of the traditional results may be an expression of the
laws of relativity. But I hope that the time has passed when
an experimenter will be content to give us only the work dec-
rement as datum for the measure of fatigue. Certainly the
break can no longer be regarded as a temporary exhaustion of
a function. Perhaps the least expected change that the new
paradigm will make in our tradition is the place of the inter-
fering sensations of weariness. These may, after all, turn
out to be subjective indicators of real fatigue. Their effect
in apparent work decrement, however, will be determined
by their relative importance in the group of competing tenden-
cies. Under normal conditions at least I doubt if we should
call weariness a pseudo-fatigue.
It will be noted that the eye-movement paradigm is still
much too simple to apply directly to our mental processes.
In place of its anatomically restricted competition to the
nuclei of the third, fourth, and sixth cranial nerves, we have
reason to believe that cortical competitions are as indef-
initely complicated as the various active association ten-
dencies. That a variety of tendencies to associative repro-
duction are normally aroused as the effect of a mental stimulus
is indicated by the facts of the association experiment. This
normal spread of excitation, coupled with the effect of psycho-
physiological rhythms, and the complication of simultaneous
stimuli from the different receptor fields, gives the com-
petition in mental operations an almost chaotic complexity.
But in addition to all that, we must extend our notion of
competition and relative fatigue to those more slowly changing
inner determinants of action that we call motives, controls,
and the like. Indeed it seems probable that these inner
factors, in so far as they are the only continuously acting
factors in mental work, are more apt to be the locus of abso-
lute fatigue than the several discrete association tendencies
which are involved only occasionally in the mental task.
But aside from the obvious differences in complexity our
paradigm adequately represents the fundamental processes.
However long a mental process may be continued and how-
THE LAWS OF RELATIVE FATIGUE T 1 1
ever insignificant the decrement in returns, there comes a
moment when it stops. It may be interrupted by demands
for food, for sleep, or by some competing task. It may be
interrupted by the gradually increasing insistence of inhibiting
sensations like thirst, eye-strain, muscle pains, or pressure
pains from sitting still. In any case, the work decrement of
the consequent break can never be fully understood if we
regard it as a direct product of fatigue, but only in connection
with the intercurrent competing tendencies. Fatigue may
be a contributing factor, but the apparent decrement of the
break will bear no regular relation to the degree of absolute
fatigue in the tissues which performed the discontinued task.
This enables us to understand why in pathogenic nervous
exhaustion, the physician in search of a therapeutic measure
may seek to strengthen some competing interest. He may
even try to develop some fad, philanthropy, golf, the calcu-
lation of food calories, or what not, to compete with the old
system and its emotional, business, or religious reinforce-
ments. Most normal lives seem too full of competing ten-
dencies. In my own case I have been interested in observing
how every prolonged period of monotonous work like correct-
ing papers, for example, finds before its close some insistent
demand for interruption. If I successfully suppress one
demand, more insistent ones arise, until finally effective
voluntary reinforcement of the main task suddenly ends.
The voluntary reinforcements may have developed such sen-
sations of strain that the surrender to a competing impulse
brings great relief. I know that the interruption is not per-
manent. I consent to it to get the lesser matter off my mind,
expecting to return presently to the main task, freed from the
incubus of that particular competitor. In very much the
same way, after lying awake for a time on one side we turn
over, not because we could not lie on that side longer, not
because we expect any great improvement from the change,
certainly not because we expect to lie on the other side for-
ever. The displacement of the body mass is scarcely the
product of fatigue. But in the complex of competing ten-
dencies a little relative fatigue becomes the occasion for an
112 RAYMOND DODGE
entirely disproportionate result. Possibly social unrest fol-
lows a similar course. They seek a change in the govern-
ment, or the social and labor conditions, not because the
present is really unendurable, not because they expect a per-
manent betterment. In many cases at least, they act from
relative fatigue, to shift the pressure. I suppose all the
phenomena of restlessness and the corresponding attractive-
ness of change finally reduce to competition and the relative
refractory phase. They operate in work and play, in social
and economic activities, in politics and in religion. Without
their interference in our lives, unwelcome as it often is, we
must have continued indefinitely in the direction of our first
activity, with the consequent loss of that vital equilibrium
on which the organism as a unit of different parts depends for
its continued existence. Without their interference the initial
process must always work itself out to the final collapse of
complete exhaustion.
Relative fatigue, then, is not a mere limitation of human
efficiency. It is not exhaustion, but prevents it. It is a
conservator of organic equilibrium, as well as a condition of
organic development. The incapacity of the young child
for long-continued monotonous tasks may be a symptom of
an active, developing mind. Lack of competition would
result in mental deformity, or absolute exhaustion, just as
truly as the lack of stable reinforcing systems in the adult
would mean perpetual infantilism. Thus it seems to me
that the principles of relative fatigue have a direct bearing
on the practical problems of education which the traditional
doctrine of fatigue as apparent work decrement entirely
missed. The development of the capacity to sit still, to con-
tinue long at routine work, the adequate response to all formal
discipline demands more than the strengthening of the cor-
responding neural bonds. It demands the weakening or
elimination of competing tendencies. At least one of the
perils of routine education arises from this depression of spon-
taneity. But I have expressly disclaimed any right or
intent to discuss the practical side of the problem.
I cannot quite resist the temptation, however, in closing,
THE LAWS OF RELATIVE FATIGUE 1 13
to point a methodological moral. There has seemed to me
to be something almost humiliating in the eagerness with
which tests of mental fatigue have been sought, while there is
still so much that is uncertain in our knowledge of the funda-
mental nature of the process that we would test. If it is not
too great a strain on presidential license at a meeting like
this, when the program is so largely devoted to the matter
of tests, I would sound a note of warning that in my opinion
any tendency to supplant psychological investigation by
tests would contain a serious menace to the future of psy-
chology. Both have their proper place. But it can only
lead to confusion and work to the discredit of our science if
the search for practical tests blinds us to the necessity for
studying the dynamics of the processes that we hope to test.
We cannot afford to develop a new phrenology.
MORE CONCERNING THE TEMPORAL RELATIONS
OF MEANING AND IMAGERY
BY EDWARD CHACE TOLMAN
Northwestern University
The controversy between the imageless thought adherents
and their opponents has been lent a new aspect by the work
of Dr. T. V. Moore.1 He has ingeniously devised a simple
but most fruitful method for investigating the temporal rela-
tions of meaning and imagery. By the use of this method,
he has brought forward striking evidence which seems to
support the contentions of the imageless thought school.
The present investigation makes use of Dr. Moore's method,
but by applying it to a greater number of subjects has ob-
tained data which point to a qualification, if not a contra-
diction, of Dr. Moore's conclusions.
In the part of Dr. Moore's work which directly concerns
us, he presented to his subjects the names of easily visualizable
common objects, such as Zange, Fernglas, Pfeil, Messer,
Lampe, etc.,2 and asked them to react according to either
one of two instructions. One time he would instruct them
to react just as soon as they had obtained the meaning of the
word, another time just as soon as they had obtained a visual
image of the object which the word named. Nonsense words
were occasionally introduced as a check to make sure that
the subject was reacting to real meanings. By averaging up
the times separately for the two kinds of reaction, he found
whether, on the average, the subject could obtain meaning
or visual image in shorter time.
Of the 9 subjects to whom he applied the method, all
but one gave unambiguously shorter average reaction times
for meaning than for visual image. In the case of the one,
XT. V. Moore, 'The Temporal Relations of Meaning and Imagery,' PSYCHOL.
REV., 1915, 22, 177-225.
2 The work was performed at Prof. Kxilpe's laboratory in Germany.
114
TEMPORAL RELATIONS OF MEANING AND IMAGERY "5
the figures showed a tendency in the opposite direction. The
average reaction times were slightly shorter for visual image
than for meaning. The trustworthiness of these figures was,
however, called into question by the fact that this subject
was never cured of a habit of reacting to nonsense words as
readily as to real words. This fact led Moore to reject the
figures of this subject as inconclusive, and to draw all his
conclusions from the results of the 8 subjects who agreed.
Introspections were asked for after each reaction and it
was found that the introspection of these subjects bore out
the testimony of their objective reaction times. They all
agreed, that is, in reporting an awareness of simple meaning
which appeared in every case prior to the image. Further
introspection indicated that this awareness of meaning was a
totally different kind of content from image.
From these facts, combined with similar ones obtained
from experiments on the time relation of meaning and verbal
imagery, in which pictures of objects were shown the subject,
and he was sometimes instructed to react when he obtained
the meaning of the picture, and sometimes when he obtained
a verbal image of the name of the object, Dr. Moore con-
cludes that meaning as a psychological content is sui generis
and independent of imagery.
The present writer was led to question these conclusions
because of the conviction that his own consciousness of
meaning depended in no small part upon visual imagery.
With that conviction in mind, he attempted by a method
essentially similar to Dr. Moore's (to be described below)
to put the matter to test. Great was his surprise, however,
to discover that he himself, when the experimental conditions
were thus controlled, substantiated the results of Dr. Moore's
subjects, in that, on the average, he obtained meaning in less
time than he did visual image. In the course of more or less
haphazard experimenting, however, the writer, largely by
accident, discovered a subject who did fulfill the prediction
he had made for himself; a subject, namely, who obtained
visual image, on the average, in as short a time as she ob-
Ii6 EDWARD CHACE TOLMAN
tained meaning, and who declared that introspectively the
image was a part of, or essential to, the meaning.
This purely chance discovery suggested to the writer that,
if a large enough number of subjects were to be examined, a
small proportion might be found whose results would agree
with those of the subject just mentioned rather than with
those both of the writer himself and of Dr. Moore's 8 subjects.
In pursuance of such a possibility, an investigation was under-
taken of as many Northwestern University students as
possible who were at the time taking either the introductory
course in psychology or the laboratory training course.
The method employed was slightly different from that
used by Dr. Moore. Instead of presenting a purely chance
list of names, names of black or of white objects only were
presented. The subject was given two keys, one for the
right hand and one for the left, and instructed to react with
the right hand if the object were black, and with the left
hand if the object were white. The following typewritten
instructions, which were read by the subject before the be-
ginning of the experiment, will make the method clearer:
"You will either be shown the name of something which
is black or of something which is white. If it is black, you
are to press the right-hand key; if white, the left-hand one.
Sometimes you will be told beforehand that you are to react
(i. e.j press the appropriate key) the instant you know whether
the object is black or white, irrespective of how you know it.
Other times you will be told beforehand that you are to react
the instant you see from your visual image whether the object
is black or white. Introspection will be asked for from time
to time during the course of the experiment."
Each word was typewritten on a slip of paper which could
be fastened to a piece of black cardboard; the latter was cut
so as to slip into place directly behind a pair of shutters.
These were made to swing open towards the subject by means
of a camera-bulb. When opened, they exposed a black field
in the center of which appeared the slip of paper with the
typewritten word.
A Bergstrom pendulum chronoscope was used.1 This was
1 Described as Model No. 2 in PSYCHOL. REV., 1910, 17, 1-18.
TEMPORAL RELATIONS OF MEANING AND IMAGERY 117
arranged so that the opening of the shutters closed a circuit
which by means of a magnet released the pendulum. The
pressing of either reaction key closed another circuit, which
by means of another magnet stopped the pointer carried by
the pendulum. The scale over which the pointer passed was
calibrated to be read directly to thousandths of a second.
The complete swing of the pendulum lasted 2 seconds only; any
time longer than 2 seconds could not, therefore, be recorded.
The chronoscope was tested by means of a seconds pendu-
lum before the beginning of the investigation, and the strength
of the currents in the two magnet circuits were found such
that a complete swing on the seconds pendulum registered on
the chronoscope correctly to within 0.005 °f a second. These
strengths of current were noted and established throughout
the course of the experiments by means of rheostats which
were included in each of the two circuits.
Each subject was presented the same list of 24 words.
The reaction times of the first 4 were rejected. The remain-
ing 20, the times for which were counted, were the following:
coffee, lime, steam, snow, mud, coal, swan, iron, diamond,
crow, plaster, negro, cinders, raven, milk, print, teeth, jet,
cement, lard.
It will be observed that 10 of them are names of white
objects, and 10 names of black objects.
The programme was arranged so that for 5 of the 'white'
words the subject was instructed to react to meaning, and
for 5 instructed to react to image; the same held for the
* black' words. But different sets of 5 were used for meaning
and image, respectively, in presenting this same list to suc-
cessive subjects.
The testing of each subject took about 30 minutes, and
49 subjects in all were tested.1 In Table I. we present the
final results for all the subjects.
1 The writer wishes to express his indebtedness to Mr. Leslie B. Bunch and to
Mr. Wilbert C. Keiser, who helped as experimenters throughout this part of the
investigation, and to the latter also for his assistance in the preliminary experiments
and in setting up the apparatus.
If, for any reason, there was a slip on the part either of the experimenter or of
the subject, so that no reaction was obtained for one of these words, this was usually
rectified by making the subject react in the same way to another word of the proper
color at the end of list.
n8
EDWARD GRACE TOLMAN
TABLE I
Men
Women
Subject
M.
V. I.
Subject
M.
V.I.
Mr. Bark..
" Bu
•823
I.I99
'.617
.868
.796
.764
•7942
'.887
•843
.884
.792
.617
'ill
.8n
.972
.816
.893
i. 060
.888
1.136
1-385;
1.7841
1.014
.684
1.051
1.027
2.OOO3
1.289
•932
.918
1.047
1.044
1-457
I-4391
1.014
•757
.724
.802
.761
•975
Miss C
.742
1-378
778
Could not
visual
•834
•983
1.058
.892
1.003
.756
•654
.916
I-I93
.682
.870
.717
•784
•775
•561
.827
-736
.820
.862
•752
I.8881
1.056
obtain
images.
.942
.266
•145
.266
•247
.770
.841
-335
•9731
.891
.890
.889
.719
•743
•734
.528
.821
.698
•754
.912
" E
" Ca
" Go
" Gr
" Co
" De
" Gu
:: £:::::
" H
" Ed T
" M
" El T
" N .
" S. T
" Ka
" O
" Pa
" LI
' Mas
' Mi
" Ril
" Rit
" Si...
' Pa
" St
' Pe
« rp
' Po
" We
' Rea
" Bart*..
' S
" A*
" L.*
" B*
" Mi*
" Milln*
" Ki.*
" Kn *
" Li*
" Pe.*
" Wa*
" Mac*
" Rei *
" V*
1 Some of the individual reaction-times from which this mean was computed were
over 2 sees. But, since, as before mentioned (see above p. 117), the maximum range
of the chronoscope used was only 2 sees., we reckoned them at only 2 sees, each in
computing the mean. The direction of the results, it will be noted, however, were
not obscured by this method.
2 In the case of Mr. N., two of the individual reaction-times for " meaning "
exceeded 2 sees., but were reckoned at only 2 sees, in computing the mean.
3 In the case of Mr. N., all of the individual reaction-times for " visual image "
exceeded 2 sees., but were reckoned at 2 sees, in computing the mean.
The second column on each side gives the arithmetical
mean time for the subject's reaction to meaning; and the
third column the arithmetical mean time for his or her reac-
tion to visual image.
It will be observed that by far the greater majority of
the subjects (those whose names are not starred) showed a
decidedly shorter reaction time for meaning than for visual
image. And this holds in about equal proportions for men
TEMPORAL RELATIONS OF MEANING AND IMAGERY 119
and women. They belong, in short, to the type represented
both by Dr. Moore's 8 subjects and by the writer himself.
Their introspections also bear out this conclusion. They all
agree in reporting a meaning which appeared before visual
image.
Turning now, however, to the subjects whose names are
starred, we find a group for whom the results are quite differ-
ent. All but two of them, Mr. Bart, and Miss V., gave mean
reaction times which were actually longer for meaning than
for visual image, and these two gave introspections of a
character which, combined with the closeness of their reaction
time, suggested that a longer and more careful examination
of them might have resulted in their cases also, in a longer
reaction time for meaning than for visual image.
GROUP i
TABLE II
SUBJECT Miss B.
Meaning
T.
V.
Visual Image
T.
V.
Fork
I.O^O
.092
Square
I.4QO
.3-27
Turkey
.88c
•0^3
Ring
7^
.6q2
•jj/
.4.61
Banjo . .
.800
.138
Lion
I.O2C
.128
Rose
I.34.C
•4-O7
Candle
.co8
.CCC
Fly
872
.066
Steamer
.872
.281
Book
814.
.124.
Circle
I 362
2OO
Nest
.014
.61; I
•*T*
.287
Sofa
1. 17O
.217
Boot
.822
.116
Tree
•9 IS.
.238
Heart
.04. e
.OO7
Stocking
I.37O
.217
Tug. . .
.6c8
.280
Stairs
*o/~
.006
.24.7
Oven ....
.782
.IC6
Crown
.QQO
.163
Knife
.7C2
.186
Tower
I 23O
077
Cat
•O-*
.06 c
O27
Spoon
I COC
•7 C2
Cradle
6QO
.24.8
Cherry
I 3 CO
107
Rooster
I.C2C
.587
Brush
*-55«
I. OOo
.147
Mouse
.772
.166
Drum . .
.960
.193
Snake
OI2
O74.
i 308
ICC
Sled
.356
.4.18
Pear
I.IJ.O
.OI3
Mask
.12?
.187
Peacock
.078
.175
Flag.
.c8o
.1*8
Skull
I 708
.CCC
J. lag
Letter . .
•772
.166
Flask
I 620
•033
.467
Rabbit
.268
-5-7Q
Chain
868
.28c
Anchor
^oo
362
I 24.O
.087
782
•!>V4
I Co
.13U
Tntal
f
Tntal
20.503
5-75°
22.503
4.991
Mean =
Median =
•938
.847
.208
Median =
I.I52
I.I40
.250
I2O
EDWARD CHACE TOLMAN
TABLE III
Miss L.
Meaning
T.
V.
Visual Image
T.
V.
Turkey
.c6o
.044.
Square
7IO
1^6
Banjo ....
.C72
.1^4.
.OIO
Ring
;i8
.l^U
QCO
Rose
_rjT
ooc
Lion
•bl°
6cc
081
Candle
3QI
I2C
Steamer
•»55
60 c
Q-J T
Flv
6^?
117
Circle
723
I4.Q
Nest
.330
.186
Book
.C2O
.OC4.
Boot
.4.82
.O34
Sofa
.660
.086
Heart
-4-O4.
.112
Tree . .
.c8o
.QIC
TUR . .
.478
.o-?8
Stocking
c6o
OI4.
Oven .
.COC
.on
Stairs
•bvu
^60
.214.
Knife .
.4.QC
.III
Crown
.JWW
482
OQ2
Tower
•3QQ
126
Cat
568
OO6
Cradle
.CQO
.074
Spoon
.630
.<x6
Rooster
.q82
.066
Cherry
•VJW
.620
•^3^
.04.6
Hat
••y"
.60;
.080
Scissors
.TIC
.OCQ
Mouse
,C4.c
.020
Brush
C-72
.O4.2
Snake . .
OtO
.672
7
.ICO
Drum
cic
.OCQ
Sled
.66 c
14.0
Hammer
7OC
I-5I
Peacock
CIO
.i^y
OOO
Mask
c TO
064.
Ladder
.C2O
.OO4.
Flag
.rcc
.OIQ
Skull
•w»y
062
Total
.51^
.490
Total
, ,
PT/q
12.044
•47*
Median =
.510
.516
•°7S
Mean =
Median =
•574
.560
.070
To make sure of this result, however, we subjected these
14 subjects to further tests, and, in doing so, we decided to
copy Mr. Moore's method exactly, rather than to use the
preceding * discrimination between black and white' method.1
The words presented to the subjects were one and two
syllable names of common objects. The subjects were some-
times instructed to react as soon as they obtained the meaning
of the word, sometimes as soon as they obtained a visual image
of the object which the word named. The instructions 'react
to meaning,' and 'react to visual image' were in the case of
any one subject distributed irregularly, but with approxi-
mately equal frequency. A majority of the same words
1 The black and white method was adopted originally with the idea that it would
set a more definite check, in the case of reaction to meaning, by making sure that the
subject really understood and not merely recognized the word. The results, how-
ever, were so exactly similar to those obtained from Mr. Moore's method, that it was
thought as well to adopt the latter.
TEMPORAL RELATIONS OF MEANING AND IMAGERY 121
were used over again for the successive subjects in such a way
that the same word would with one subject call for the in-
struction 'react to meaning,' and with another that of 'react
to visual image.' Introspections were asked for after about
half the reactions only, owing to insufficient time, as each
subject could spare but a little over an hour, and it was
desired to obtain from each as long an objective series as
possible. Nonsense words, introduced occasionally, served
as checks to make sure that the subject was reacting fairly.
The results of the first few reactions in the case of each subject
were discarded. The writer served as sole experimenter
throughout this series.
Tables II.-XV. inclusive present the results for the
different subjects. In the first column are the words on
which the subjects received the instruction 'react to mean-
ing'; in the second column are the reaction times for these
words; and in the third column are the deviations of these
TABLE IV
MR. MILLN.
Meaning
T.
V.
Visual Image
T.
V.
Turkey
.96?
.OC2
Fork
.968
.022
Banjo
.74.8
«*3«
.161:
Square
1. 172
.226
Lion . .
.8t;6
.OC7
Ring
.04-2
.OO4.
Candle.
86c
•*o/
O4.8
Rose
OCX
O4. 1
Circle
93O
OI7
Flv
818
128
Book
QC2
O3Q
Steamer
QQC
O4.Q
Sofa
I.OAO
.127
Nest
.992
•U4y
.04.6
Tree
.7OI
.212
Boot . .
.872
.074.
Stocking
•/T*
.Q62
•O4.Q
Heart
.072
•w/*f
.026
Stairs
.QCQ
.O37
TUJJ
0.6 c
OI9
Crown ....
.017
OO4.
Oven
QCK
O4. 1
Knife
024.
OI I
Cat
I I O2
ic6
Tower
868
O4.C
I OCK
QCQ
Cradle
i 086
17-7
818
•*OjJ
128
Snake
1.28-?
.370
Rooster
.84.0
.106
Hammer
.700
.123
Scissors
"*TT
I.OIO
.070
Sled
.QCO
•O37
Hat
.814.
.132
Mask
^J
.6O2
.221
Brush
.873
.073
Peacock . .
,QJ.O
O27
Mouse
98 c
.O3O
Letter
.8l8
O7 c
Drum
O7O
O24,
Flask
•y4°
.027
Total
Total
IO 177
1'4"J
, ,
f.
Mean =
Median =
•913
•930
.091
Median =
.940
.967
.073
122
EDWARD GRACE TOLMAN
GROUP 2
TABLE V
MR. L.
Meaning
T.
V.
Visual Image
T.
V.
Banjo
6q8
O2 1
Ring
78 C
Q-2Q
680
O3Q
Fly
•/ •>
O4.O
IIC
Rose
•75°
.OTI
Circle
,8lO
.occ
Candle
.660
.CKQ
Book
.722
.0^3
Steamer
.760
.O4I
Nest
.Q2C
.170
Sofa
.027
.2O8
Boot
.718
037
Tree
768
O4.Q
Heart
I 378
623
Stocking
773
OCA
Ear
77C
O2O
TUR. . .
.720
.OOI
Stairs
•//!>
.64.0
.IIC
Oven
.7-7C
.Ol6
Crown
.812
•OC7
Hen
•/ J3
.6OO
.110
Snail
.760
.00 c
Basket
.Sio
.III
Eve
62O
.I-zc
Table
.713
.OO6
Bae
6c<?
IO2
Star
665
OC.4.
Horse
7 7O
OI C
Nose
.658
JT-
.OOI
Trunk
.620
.I'jr
Spade
.748
.O2Q
File
.co8
.IC7
Shoe
.7^0
.O3I
Picture
.628
.127
Balloon
•I,*
.611;
.IO4.
Scissors
822
' ~.'
.067
Egg
.CC.4.
~:T
.i6t;
Rooster
672
08 1
Cha iV '
•78?
nfi?
•7°^
.003
Total
1A "JAR
2 08 1
Total
I± ?86
*4o4°
14.300
A/r
Mean =
Median =
.719
.728
.063
Median =
•755
.722
times from the mean. Similarly, in the fourth, fifth, and
sixth columns are the words, times, and deviations for the
instruction 'react to visual image.'
Examining these tables we discover, first, a group of 3
subjects, Group I, Tables II., III., and IV., who appear to
belong after all to the type represented by Dr. Moore's 8 sub-
jects. For all 3 both the mean and the median times for
reaction to meaning are shorter than for reaction to visual
image. Next, we note 3 subjects, Group 2, Tables V., VI.,
and VII., whose figures begin to point in the opposite direction,
in that, whereas their median times are shorter for meaning,
their mean times are shorter for visual image. Finally, we
note 8 subjects, Group 3, Tables VIII. to XV. inclusive, who
substantiate our original hypothesis, in that for them both
median and mean times are shorter for visual image than for
meaning.
TEMPORAL RELATIONS OF MEANING AND IMAGERY 123
In order to be sure of these results, however, we must
compare the introspections. In parenthesis, be it noted that
our subjects were untrained,1 and hence the value of their
introspection is subject to qualification. The attempt was
made, however, to make their introspective task as definite
TABLE VI
Miss MAC.
Meaning
T.
V.
Visual Image
T.
V.
Square
.600
.on
Fork
,«8o
.Oil
Ring
;8o
O22
Turkey ...
CCQ
.043
Rose
C'? C
076
Banjo
.781
.188
Candle
.CIO
.O8l
Lion
.640
.047
Fly. .
•4-7Q
.IT2
Steamer
.^98
.ooc
Circle. . ,
7' *
.621
.OI2
Book
.489 •
.104.
Nest
.070
.-JCQ
Sofa
.861
.268
Boot
.COI
.no
Tree
.CA.2
.OCX
Heart
•i^A
£-4.6
06 c
Stairs
C42
.OCI
Stocking
-JQO
221
Oven
6^ <?
O4O
Jug
.C22
.089
Crown
.710
.117
Knife
.C^O
.O8l
Cat
.462
.I'll
Tower. . .
•4-4-1
.168
Spoon
.trie.
.078
Cradle
.778
.167
Cherry .
.4.4.7
.ICO
Scissors
,CJC
'
.O3O
Rooster
.CQO
.OO'?
Brush
760
I4.Q
Hat
4-4-2
.1^1
Mouse
.71-5
.122
Drum
.600
.007
Snake
.7'? 2
.121
Hammer
.478
.nt:
Sled
.740
.I2Q
Pear
.660
•"O
.067
Mask
.qiS
.OQ"?
Peacock .
rqg
.ooc
Flag
.qSo
.031
Ladder
ovu
,6cc
•WJ
.062
Skull . . .
•4-01
.I2O
Letter
.c6i
.O32
Rabbit
.717
.IO2
Flask
648
.occ
Anchor
.781;
.174.
Tiger
POO
OQ?
Apple
i, •>
.6-jc
0^4.
Hand
608
QIC
$rrK
Hen
.656
.04 c
Basket.
.633
.040
Snail
£
.60 c
•WT3
.OO6
Table
.696
.lO'?
Eve
«"P
c6o
•"-•ye
•5«J
.U^)l
Total
Total
Z.UJ4
i/.iiy
••7V7
_ .
Mean =
Median =
.611
.585
.100
Median =
•593
/S98
.075
and easy as possible. The method we chose was to ask them
such specific questions as: "Did you get a visual image?"
(in case of reaction to meaning). "Did you get a meaning?"
(in case of reaction to visual image); and "Which came
first?" Occasionally, in the case of a subject who seemed to
xThe great majority of them were but just finishing their first course in psy-
chology. See above, page 116.
I24
EDWARD CHACE TOLMAN
show native aptitude for introspection, he was asked to
describe, if he could, his consciousness of meaning; i. e.,
whether it seemed to him reducible to images or something
unique and not further analyzable. Descriptions of the
visual images obtained were also asked for from time to time.
Any other specific questions put to the subjects will be noted
along with the answers obtained.
TABLE VII
Miss REI.
Meaning
T.
v.
Visual Image
T.
V.
Fork
.808
.908
.790
.632
.663
.680
.685
I.OI2
.748
.678
I.OO2
•90S
&
.690
.630
•745
.900
.698
.682
•590
.061
.161
•043
•"5
.084
.067
.062
.265
.001
.069
.255
.158
.167
.087
.057
.117
.002
•153
.049
.06S
.157
Turkey
.910
I.I32
.704
.716
.620
.762
.620
.603
•538
.695
.872
•755
•772
•5"
.728
.920
.722
.601
.700
•232
•454
.026
.038
.058
.084
.058
•075
.140
.017
.194
.077
.094
.167
.050
.242
.044
.077
.022
Square
Banjo
Rine
Rose
Lion
Fly
Candle
Circle
Steamer . ...
Nest
Book
Boot . . .
Sofa
Tree
Heart
Stairs
Stocking
Oven
Tug. . .
Crown
Cat
Knife
Cherry
Spoon
Rooster .
Sled
Hat
Pear
Brush
Mask . .
MXDUSC
Peacock
Drum
Flag
Letter
Ladder
Flact
Rabbit
Total
12.881
2.149
Total
15.686
2.195
Mean =
Median =
.678
.716
•"3
Mean =
Median =
•747
.690
.105
Turning to the three subjects, Group I, whose results
agreed with those obtained by Dr. Moore, we find substantial
agreement between their introspection and that of his sub-
jects. When instructed to react to meaning, our three sub-
jects reported, almost without exception, that a consciousness
of meaning appeared first, and that a visual image did not
come until after this meaning. An interesting thing is,
however, that with these three subjects the visual image
almost always did come finally. Miss B. reported only
TEMPORAL RELATIONS OF MEANING AND IMAGERY 125
3 cases, Miss L. one case and Mr. Milln. no cases in which a
visual image did not follow the meaning. This shows what
ready visualizers the three subjects were, and how near they
were to crossing the border beyond which visual image comes
before, or at least simultaneously with, meaning. When
instructed to react to visual image, their introspections were
GROUP 3
TABLE VIII
Miss A.
Meaning
T.
v.
Visual Image
T.
V.
Spoon
ceo
.020
Tower
• 32C
.087
Cradle
SlS
O-J2
Cherry
• 37O
.042
Hat
rOC
02?
Brush
.4.22
.010
Mouse
•5QO
1 80
Snake
.681
.269
Drum
OV^
erg
OI2
Pear
.QOO
.488
Hammer. .
I.OOO
•43O
Mask
.392
.020
Sled . .
.4.84.
.086
Peacock
.4.CC
.043
Flag
.C7Q
.OOQ
Skull
• t»O4
.092
Flask
.COC
.wv/y
.06 c.
Letter
.-24.1:
2
.067
Rabbit
c6o
OIO
Chain
.C4.e
.Ml
Tiger
C72
OO2
Apple
.4.08
,OO4
Hen
•51*
cg2
OI2
Eve
.<?66
7:
.O4O
Hand
.=;8q
.029
Table
.-?oc
.107
Snail
'yy
.602
.122
Crab
.770
.O<? -?
Basket
•4-QI
.O7Q
Nose
.212
.2OO
Bae
.8cx
.23C.
Spade . . .
.-37C
•O37
Star
.uw^
.670
.IOO
Pig
.2J.C
->'
.167
Horse
4.17
.151
Balloon
.-2 JC
.OQ7
Trunk
4.0,8
O72
File
278
.134.
^hnp
Ao~
noo
Egg. .
.402
.560
.OIO
Total
7.822
2.076
Chair
•4*7
^bj
A ,
Total .
I2.C34.
I.Q24.
.412
.109
,. , ,.
Mean =
Median =
•570
•559
.087
•375
practically the same as those just discussed. They reported a
consciousness of meaning which came first, and a visual
image which came afterwards.
Below we give sample introspections for both kinds of
instructions.
126
EDWARD CHACE TOLMAN
Instruction: Meaning.
INTROSPECTION — GROUP i
Instruction: Visual image.
Subject Miss B.1
Cat: "Thought of cat at home and then Tower: "I thought of the meaning be-
saw it." fore I saw it."
Snake: "Thought of crawling thing,
visual image afterwards."
Spoon: "I got meaning before image.
I think I always do get meaning first."
Subject Miss L.
Banjo: "An idea of a banjo. After Steamer:
pressing key saw an image."
Saw the one I have so often
seen on the lake. Had a general idea
of a steamer first. I think the general
idea always comes first."
Rose: "No image until long afterwards."
Nest: "An idea of a nest. Then image
of a nest in a tree."
Scissors: "A general idea, then saw all
sorts of scissors, then selected one of
those, a pair of embroidery scissors."
TABLE IX
MR. Mi.
Meaning
T.
V.
Visual Image
T.
V.
Square
.662
.087
Fork
.460
.040
Banjo
.4.80
•OQC
Turkey
.370
.I3Q
CQ'J
.018
Ring
.C43
.O34.
Candle
CQ-7
.018
Rose
• 2QQ
.2IO
OVJ
4.18
137
Fly
.4.IO
.OQQ
Book
.C7O
.00 c
Circle
•f
.650
.141
Nest
.43O
.I4C
Sofa
.3IO
.199
Tree
7J
.OIO
•O^C
Boot
.41 C
.094
Stocking
.680
.IOC
Heart
/•
.632
.I23
Stairs
.576
.001
Tug
•wj-
.670
•*~j
.l6l
Crown
o/v
c6c
.OIO
Oven
.CQ2
.083
Cat
£ •>
64.2
.067
Knife
ov*
.OIO
.101
Hat
633
.058
Scissors
.562
.053
Mouse
.865
.290
Brush
.430
.079
Mask
•vvy
.6OO
.02 c
Sled
.468
.041
Flae
.C28
.047
Peacock
.682
.173
i: Ia°
Letter
.«8
.037
Skull
.C2I
.OI2
Rabbit
rcc
.020
Flask
•4-7$
.034
Tieer
O33
.162
.213
Chain
"rt J
.c6c
•~yt
.0^6
Total
IO.92O
1.413
Total
9.664
1.881
Mean =
Median =
•575
.576
.074
Mean =
Median =
.509
.521
.099
1 This subject and Mr. Milln. did report a few cases in which they seemed to get
the image first, but the majority of their introspections were like these quoted.
TEMPORAL RELATIONS OF MEANING AND IMAGERY 127
TABLE X
MR. PE.
Meaning
T.
v.
Meaning
T.
V.
Basin
.8qi
.078
Fork
.608
.130
Square
,7ir
.008
Banjo
.700
.0-58
Turkey
.814.
.001
Ring
.562
.176
Lion
OIO
OQ7
Candle
.948
.210
Rose
QIO
OQ7
Fly
•yv*
.620
.118
Steamer
688
I2C
Circle . .
.882
.144
Book
7OO
ll"i
Nest . .
.742
.004.
Sofa
.Q7O
.IC7
Boot
.678
•^1-
.060
Tree
.728
.08 c
Heart
.718
.020
«2C
OI2
Tus?
.84.0
.120
Stairs
.8oc
.008
Oven
.704
.034
Crown
671
.I4.O
Cat
.680
.oc8
Knife
•WJ
OQO
123
Spoon
.769
.031
Tower
QIC
IO2
Cherry
•74-1
.OO3
Cradle
7QO
O2"?
Scissors
.778
.O4.O
Rooster
Q2C
112
Drum
,773
.03 c
Hat . .
.088
.lye
Snake
.658
.080
Brush
.84.0
.027
Hammer
.970
.232
Mouse .
.84.8
.03 c
Pear
.829
.091
Sled
.708
.IOC
Peacock
.688
.oco
Mask
7Q8
.QIC
Ladder
.621
.117
Flag
.836
.023
Letter
.706
.032
Skull
•7^O
.063
Rabbit
.752
.014
Flack
Chain ....
.904
.708
.091
.IOC
Total
16.967
1.837
Total
2O 3 20
2.OIO
^Mean =
.738
.080
Mean =
Median =
.813
.814
.080
Median =
.718
TABLE XI
MR. BART.
Meaning
T.
V.
Visual Image
T.
v.
Basin . .
.800
.172
Fork
.603
.004
Square
.660
.032
Turkey
.882
• 27C
Banjo
.672
.04.4.
Ring
.522
.08 c
Lion
,-JIO
-u8
Candle
.768
f
.l6l
Rose
.702
.164.
Steamer
'
.62 c
.Ol8
Flv
•/y*
ci6
112
Book
.C-2O
.077
Circle
.680
.OC2
Sofa
•573
.034
Nest
.608
.O2O
Tree
•445
.162
Boot
.648
.O2O
Ear
.co8
.099
Heart . .
•~t*-'
.610
.Ol8
Tus?
.613
.006
Stocking
.622
.006
Oven . ...
.602
.005
Stairs
.623
.ooc
Crown
.615
.008
Total
7.S4-I
.963
Total
7.286
.934
Mean =
Median =
.628
.636
.080
Mean =
Median =
.607
.603
.078
128
EDWARD CHACE TOLMAN
TABLE XII
Miss Ki.
Meaning
T.
V.
Visual Image
T.
V.
CIO
.04. c
Finger
4. CO
.077
.^iu
4.6 C
OQO
Fork
4C2
07 c
Ring
.1QO
.165
Square
.CIO
•v/9
.008
Lion
.7CO
.IQC
Banjo
•j'-y
.612
.08 c
Fly
.4.08
.14-7
Rose
.087
Circle
.C7O
.OI C
Candle
460
.oc8
Boot
.<?4.C
.2IO
Steamer
6oc
.168
Tree
.CQ4.
.O^Q
Book
.C2O
.007
Ear
OV4-
480
O7C
Nest
.708
.181
Tue
•T-UW
08 c
4.3O
Bear
.c^o
OO'?
Crown
.6c8
.IO3
Heart.
.CQO
.063
Hen
•«3w
.6OQ
.OC4
Stocking
'J?~.
•4QO
.O^I
Hand
.730
.I7c
Stairs
.C73
.WJi
.046
Snail
.4.72
.083
Oven
•3/3
.c6/?
T^;:
.o^o
Basket
•W*
co6
.O4.O
Eve
AQC
.032
Crab
.V)U«J
.c6q
.014.
Cow
.698
.171
Glove
.C82
.027
Table
.482
.045
Star
.4-CO
.IOC
Leaf
.482
.04 c
Horse
.4.82
•O7'?
Bag
.2CO
.277
Trunk
.ceo
.00 c.
Nose
.Ci8
•OQQ
Total
11.105
2 OQO
Total
IO.C4.2
I.CO4
Mean =
Median =
•535
•530
.105
Mean =
Median =
•527
•519
•075
Subject Mr. Milln.1
Hammer: "Thought of hammer as some-
thing to drive nails, then saw a ham-
mer, no specific hammer."
Mask: "Thought right away of mask as
something to prevent detection, then
a visual image of black mask that
covered about half the face."
Mouse: "Thought of mouse as disagree-
able, then saw a little mouse on the
floor."
Hat: "Meaning that hat was to cover
head, then saw a felt hat."
We may conclude that these 3 subjects, both from their
objective results and their introspections, belong, generally
speaking, to the type represented by Mr. Moore's 8 subjects.
Let us turn now to the three subjects, Group 2, whose
results did not point definitely in either of the two directions.
In the cases of two of them, Mr. L. and Miss Mac., this
inconclusiveness of the objective results is supported by their
introspections. In the case of the third, Miss Rei., the
introspection does not bear out this inconclusiveness of
1 This subject and Miss B. did report a few cases in which they seemed to get the
image first, but the majority of their introspections were like those quoted.
TEMPORAL RELATIONS OF MEANING AND IMAGERY 129
TABLE XIII
Miss KN.
Meaning
T.
V.
Visual Image
T.
V.
Cradle
.690
.067
Cat
.732
.152
Rooster
.c88
.01 c
Tower
.690
.no
Scissors
.608
.07 c
Spoon . ....
.470
.no
Brush
6K
.008
Cherry
.4QO
.090
Blouse
.ui^
76 c
14.2
Hat
.472
.108
Snake
T\ C
112
Drum
.663
.083
Sled
SI
.00 c
Hammer
.748
.168
Pear
.860
.2^7
Flag
-653
.073
Mask
.62C
.OO2
Ladder
.568
.012
Peacock
C78
.04. c
Skull
.581
.001
Flask
CQO
•O1 3
Letter
.663
.083
Chain
c.22
.101
Rabbit
. ^j
.OC.O
.070
Tiger
C-7Q
OQ3
Anchor ...
.zoo
.080
Finger
78O
I C.7
Apple
.480
.100
06l
34O
Kev
.'ico
.230
Easel .
.7C.I
OT-'
.128
Fork
.708
.128
Banjo
.co'?
.I2O
Square
.780
.200
Rose
42O
.20"?
Turkey
.£,£,2
.028
Candle
74O
.117
Ring . .
.?2O
.060
Steamer
6QT
O7O
Lion
.q28
.OC.2
Book
CQO
127
Fly
.478
.IO2
Nest
•y~»u
.640
.OI/
Circle
.iJ./U
.670
.090
Bear
.616
.OO7
Sofa
.670
.090
Heart .
.cSo
.04. ^
Boot
.C7I
.OO9
Stocking
.JUW
.6^2
.OOO
Tree
.482
.098
Stairs
44Q
.174.
Ear
.662
.082
Oven
CQO
on
Jug
.c,82
.OO2
Eye
,c,oo
.I2/?
Crown
.525
.055
Table
.4.80
• 14"?
Hen
.645
.065
Crab
.eye
.048
Hand
.498
.082
Glove
.4.82
.IAI
Snail
.CIO
.O7O
Racket
O V"
/ifio
1 2O
Tntal
•4UU
ly.juo
^.951
Tntpl
n Q/-.7
\M
fai
lo-SS1
Median =
.023
.6lS
•°9S
Mean =
Median =
.580
•570
.088
objective results, but suggests rather that the subject belongs
with the last group, i. e., those for whom visual image defin-
itely comes before instead of after meaning. We will reserve
the introspection of this third subject to consider in that
connection, and will examine now the introspections of Mr. L.
and Miss Mac. only.
Mr. L.'s introspection is unique in that for him the
question seemed to be one not so much of meaning vs. visual
image as of some sorts of images vs. other sorts of images.
When instructed to react to meaning, he never obtained
130
EDWARD CHACE TOLMAN
TABLE XIV
Miss V.
Meaning
T.
V.
Visual Image
T.
V.
Finger , .
.7CQ
.120
K.ev
Q'l f
IO7
Basin .
I 14.2
.263
Fork
7O2
126
Square
760
110
Turkey
&I1
OI C
Ring
84.6
Q'l'l
Banjo
861
O'J'J
Lion
•7QO
•VJJ
.080
Candle
I.O4.C
.217
Rose
.913
.034.
Circle
.C.O2
*L
.•?2O
Fly..
*~*J
.q6o
.081
Nest.
.712
O^"
.Il6
Steamer
.Q1O
.OCI
Sofa
L*'"
.6C2
.176
Heart
.870
.000
Boot
.732
006
Scissors . .
1:4.0
•77Q
Tug
I O4.C.
217
Book
JT-
OO2
•joy
187
Hand
6qo
i^?8
Tower
7C.2
127
Bae
I O4.C
217
Crown
I.2QO
.4.11
Star. .
.8C2
.024.
Horse
.Ql8
.ocq
Nose
I.OOO
.172
Trunk
j^
Tjfi
Total
TT rRfi
Total
13.17O
^.057
A/r
0^0
Mean =
Median =
.879
.870
•137
Median =
.833
TABLE XV
MR. WA.
Meaning
T.
V.
Visual Image
T.
V.
Key. .
I.24.O
.24.7
Finger
84.0
06 1
Fork. . . .
I.2OO
.207
Basin
860
08 1
Square
.780
213
Turkey
c.84.
TQC
Ring
I OO8
QIC
Banjo
712
•*y5
067
Lion
I.I48
.ICC
Steamer
.QT.Q
.ici
Rose
.8OO
.IO"?
Circle
.8o<?
.024.
Candle
•0-8 ?
.OO8
Book
.UWJ
.6c,2
.127
Fly. .
•V^3
I.OoC
.072
Nest
.7J.C
QT.A
Boot
1.108
.lie
Sofa
808
O2Q
Tree
.Ql8
.07 c
Heart
800
02 1
Bear
.8qq
.004.
Stocking
71 C
064.
Tug
.uyy
.6Q2
.<2QI
Ear
80O
Ill
Oven
•975
.Ol8
Stairs. . . .
.788
.000
Total
12.908
I. 623
Total
10 127
.074.
Mean =
Median =
•993
.985
.125
Mean =
Median =
•779
.800
•075
anything imageless, but merely images of one kind or another;
sometimes visual, but often cutaneous or kinesthetic. When
instructed to react to visual images, he obtained visual image.
The following are typical introspections :
TEMPORAL RELATIONS OF MEANING AND IMAGERY 13*
INTROSPECTION — GROUP 2
Subject Mr. L.
Instruction: Meaning. Instruction: Visual image.
Steamer: "Visual image of white steamer Heart: "A feeling of tension or strain,
out on lake. Feeling of tension, i. f., then visual image. No consciousness
a kind of mental muddle before of meaning before the image."
image."
Stocking: " Kinaesthetic image, followed Horse: "Saw a dappled gray horse."
by visual image. They came almost
simultaneously. Meaning not really
made clear until the visual image."
Spade: "Combination of visual kinaes- Turkey: "A feeling of tension then a
thetic image. Visual came first." visual image."
Occasionally for * meaning' he did report the presence in
consciousness of the 'idea' of the object named. When
asked, however, to analyze this 'idea,' he invariably reduced
it to images; kinsesthetic, cutaneous, or even visual, but in
that case an image of something other than the immediate
object itself. The following are typical:
Instruction: Meaning.
Nose: "Idea of sniffing (seemed kinsesthetic and cutaneous)."
Star: "Idea of star as a source of light. (This idea a visual image of light.)" "Visual
image of a star after reacting."
We conclude that as far as his introspection goes meaning
for this subject was always image, of one sort or another.
If this be true, it suggests that the reason his objective time
for meaning averaged about the same as that for visual image
might be due to the fact that the various kinds of images
which he got for 'meaning' took on an average about the
same time to develop as the purely visual images which he
got when reacting to 'visual image.' In addition, however,
to suggesting an explanation of his objective reaction times,
this introspective testimony to the effect that meaning was
always image of one kind or another, would carry, of course,
a direct answer to our original problem. It would settle it
against the adherents of imageless thought. We cannot,
however, be certain of this decision until we are sure of the
absolute reliability of this subject's1 introspection. And for
1 This particular subject happened to be one of the better-trained (as well as one
of the most intelligent) students, since at the date when the test was performed he
had just completed a year's course in laboratory psychology, in addition to having
had two previous courses in psychology.
132 EDWARD CHACE TOLMAN
that we lack proof. We will not, therefore, attempt to make
any final statement, but simply draw attention to this sub-
ject's introspection and to the way in which it points.
Turning now to the introspection of Miss Mac., the
second subject for whom the objective time appeared approxi-
mately the same for both meaning and visual image, we find a
somewhat different state of affairs. This subject did not
succeed in analyzing her consciousness of meaning. The
presumption was, therefore, that it was unanalyzable because
imageless. The approximate equality of her times for mean-
ing and visual image, however, is explainable by the fact
that, introspectively, this awareness of meaning appeared
sometimes before the visual image and sometimes afterwards.
Out of 1 6 times in which she was asked to introspect upon
reaction to meaning, she reported 6 times in which meaning
came before the visual image, and 10 times in which image
came before meaning. When instructed to react to visual
image, she reported 8 times in which meaning came before
the visual image and 8 times in which the visual image came
before the meaning. Below are typical introspections:
INTROSPECTION — GROUP 2
Miss Mac.
Instruction: Meaning. Instruction: Visual image.
Snake: "Image first, then knew what it Drum: "Knew what it was, then saw
was, knowing a sort of memory of drum being carried in a parade."
snakes that I have seen."
Cradle: "Knew what it was when I saw Peacock: "Saw it, then meaning."
it."
Hen: "Visual image first; then recog- Ladder: "Knew what it was, then
nized the image. I remembered that image."
I had seen things before that looked
like that."
This subject's introspection, as far as it goes, substantiates
her objective results and we see why, in her case, reaction to
meaning and reaction to visual image required about equal
average times.
Combining her results with those of Mr. L., we note that
these two subjects taken together constitute a transition
group in that for them both reaction to meaning seems some-
TEMPORAL RELATIONS OF MEANING AND IMAGERY 133
times though not always to depend upon visual image.
In Mr. L.'s case meaning sometimes actually was visual
image; in Miss Mac.'s case it sometimes followed the visual
image and was dependent upon the latter.
We may turn now to the introspection of the third and
last group of subjects, the group for whom reactions to
meaning, as we shall see, did always depend upon a visual
image. Miss Rei., it will be remembered, belonged intro-
spectively to this group. In considering the introspection of
these 9 subjects (including Miss Rei.) we find it possible to
divide them, roughly speaking, into two subgroups: first, a
subgroup, A) composed of subjects who, as a rule, tend to
make a distinction between meaning and visual image; and
second, a subgroup, B, made up of subjects who, as a rule,
do not tend to make such distinction, but for whom meaning
is visual image.
Subgroup A comprises Miss A., Mr. Mi., Mr. Pe., and
Miss Rei.
Subgroup B comprises Mr. Bart., Miss Ki., Miss Kn.,
Miss V., and Mr. Wa.
The following are typical introspections for subgroup A.
GROUP 3 — SUBGROUP A
Instruction: Meaning. Instruction: Visual image.
Subject Miss A.
Rabbit: "Saw a rabbit hopping. Real- Letter: "I saw a letter lying open. No
ized it was a small animal, then re- particular letter."
acted."
Drum: "First saw a picture of a man File: "Saw both a letter file, and a file,
beating drum in orchestra, and then the tool."
thought, 'it is a musical instrument.' "
Subject Mr. Mi.1
Tree: "Saw a bunch of trees, then Scissors: "Image first; didn't think
thought what a tree really is." what it meant till after I pressed the
key."
Stairs: " Saw stairs, then got a meaning, Skull: "Image and that was all."
i. e., stairs something that you go up
on."
1 This subject reported some instances in which he thought he obtained a mean-
ing without a preceding image. His introspection tends, therefore, to class him to
some extent in group 2 with Miss Mac rather than here in group 3.
134 EDWARD CHACE TOLMAN
Subject Mr. Pe.
Cradle: " Saw a cradle first, and then the Scissors: " Saw scissors lying on a sewing-
idea to rock came on." table, that belonged to my mother."
Rooster: "Saw a big black rooster we Snake: "Visual image of a little green
used to own, and then thought of him garter snake running through the
as something to eat." grass."
Subject Miss Rei.
Cherry: " Saw a red cherry, then thought Mask: "Saw a black mask, nothing
of it as something to eat." besides the image."
Drum: "Saw a drum and then thought Ladder: "Saw a tall painters' ladder, no
of it as being a musical instrument." consciousness of meaning."
(In answer to question) "I think the Peacock: "Saw a peacock going up a
image is an aid to meaning." hill. It was at Lincoln Park."
Examining these introspections, we note that in the case
of Subgroup A visual image was always a precursor of mean-
ing, and that the meaning itself seemed to depend upon the
image. Turning now to Subgroup .5, the following are typi-
cal introspections:
GROUP 3— SUBGROUP B
Instruction: Meaning. Instruction: Visual image,
Subject Mr. Bart.
Lion: "A mixed image of the zoo, i. <?., Finger: "Image of a finger with ring
the line of cages in Lincoln Park. No upon it (no particular finger)."
other process in consciousness detect-
able."
Fly: "Visual image of fly paper with flies Ring: "Image of a key ring (no par-
stuck on it." ticular key ring)."
Subject Miss Ki.
Fly: "First saw a black fly, then some- Nest: "Image of a nest which I saw this
thing flying." morning and then of other nests of all
different kinds."
Basket: "Saw image of basket at same Bear: "Image of a bear in a cage."
time that I got meaning. Meaning
to me is image."
Crab: "I saw a single crab and the beach Book: "Image of a red book seen at an
behind him filled in." angle."
Subject Miss Kn.
Nest: "I said the word to myself. I Hen: "Image of a black and white hen."
realized it wasn't a nonsense syllable;
then I obtained a visual image of a Snail: "A greenish brown snail shell,
nest on a tree (not any particular nest); Nothing previous to this image,
then I reacted."
TEMPORAL RELATIONS OF MEANING AND IMAGERY 135
Oven: "Got an image first thing. As if Basket: "A market basket; nothing be-
looking into an open oven." fore the image."
Stocking: "An image of a new black
stocking the first thing in conscious-
ness. Nothing between image and
reacting."
Subject Miss V.
Eye: "Made up my mind beforehand Candle: "First I saw a candle, no par-
that this time I would get a meaning ticular candle. Then I saw a par-
before image, but got an image of an ticular candle, viz., the one I saw
'eye' almost before I saw the word." last."
Snail: (N. B. reaction to this word took Nest: "Image of the nest which I saw
over two seconds and was discarded last Sunday."
but the introspection is significant.)
"Could not get an image for a long Bag: "Image of travelling bag, also of
time. Had no meaning until I got paper bag."
the image. Felt just as if I were look-
ing at a nonsense syllable."
Subject Mr. Wa.
Ring: " Visual image of ring on my finger. Heart: " Visual image of a human heart."
Nothing previous to this image.
Afterwards an image of a circle on the
ground."
(In answer to specific question) "I feel that I go through the same process each
time no matter whether meaning or visual image is asked for."
In the case of Subgroup B the testimony seems to be
unanimous that the visual image was the meaning. No other
process which might play the part of meaning was ever
detected either before or after the image.
The common point in the introspection for both sub-
groups, it will be noted, is that visual image was the first
thing which came. The two subgroups disagree, however,
as to the extent to which this visual image was important for,
or a part of, the meaning. Subgroup B declared that it was
the whole of the meaning. Subgroup A reported merely
that it always came before the meaning but that meaning
itself was something more. This disagreement, together
with the fact that for the vast majority of subjects visual
image is entirely unneeded for meaning, sets us a problem.
Three solutions suggest themselves. We may try either
an out-and-out imageless position, an out-and-out image
136 EDWARD CHACE TOLMAN
position, or some sort of a compromise. The out-and-out
imageless position would have to contend that no matter
what were the objective reaction times to the contrary, nor
what the apparent introspective evidence, the visual image
can never have been really, in any true sense, a part of or
even necessary for the meaning. Such a position would have
to claim that the objectively shorter reaction times for
visual image in the case of both subgroups did not prove that
the visual image was necessary to the meaning, but that it
was merely an adventitious circumstance. It would have
to claim also that the introspective testimony from both
subgroups as to the subjective precedence in time of the
visual image was no additional proof of the prerequisiteness
of the image. The out-and-out imageless position would,
in short, have to deny every one of the evidences afforded
by the results of our third group of students. In support, it
would have only the results of the non-visual1 subjects, i. e.,
of Mr. Moore's 8 subjects and those of our investigation who
were like his. But since the method we have used is one
which traces the importance of the visual image only, the
support of the not extremely visual subjects carries little or
no weight. We feel justified in concluding, therefore, that
our results render the out-and-out imageless position unten-
able.
We turn now to the out-and-out 'image' position. Re-
sults which directly support it are the introspections of
Subgroup 5, that meaning is image, and image is meaning.
For it to be completely supported, however, two further
demands would have to be satisfied. The objective reaction
times for meaning and visual image in the case of Subgroup B
would have to be the same. This demand was not fulfilled;
the reaction times for meaning averaged longer than for
image, which implies that meaning for Subgroup B as for
Subgroup A involved something more than mere image.
Second, all cases of 'meaning' reported ought to be analyzable
by better introspection into images. This would demand
that the 'meaning' of Mr. Moore's 8 subjects and of the
1 Using 'non-visual ' in a merely relative sense.
TEMPORAL RELATIONS OF MEANING AND IMAGERY 13?
majority of our own original subjects, i. e., the * meaning'
which came before visual image, must really have been
imaginal. It must have been made up, that is, of verbal,
organic, kinsesthetic, or other images, which these subjects
failed to recognize. It would demand, similarly, that Miss
Mac.'s * meaning,' which came sometimes before and some-
times after the visual image, must have been made up of
images. And, finally, it would demand that the 'meaning'
of Subgroup A) which came after the visual image, and
which was often similar to an awareness of definition must
likewise have been made up of images. None of these
demands would be insurmountable, if we were strongly
prejudiced in favor of an out-and-out 'image' position. It
would be possible to assume, as has been done before, that
better introspection would eventually show images in proc-
esses where as yet nothing but imageless awarenesses have
been discovered. Such an assumption, however, has at the
present stage of psychology little but theoretical precon-
ceptions to support it. So that at present we consider it
safer to conclude that our results, while they do not com-
pletely contradict an out-and-out 'image' doctrine, do
nothing actively to support one.
Finally, we may turn to the consideration of an inter-
mediate doctrine which would both allow an essential im-
portance to the image, and yet admit an imageless component
as also necessary. Such a doctrine is directly suggested by
the results of Subgroup A. They, it will be remembered,
obtained first of all the visual image and then a 'meaning.'
It appeared that for them the visual image was a prerequisite
of the 'meaning,' but that the 'meaning' itself was something
different from image. It does not seem improbable that a
similar situation may have existed also in the cases of the
other groups. With the subjects of the Group I who obtained
'meaning' first, this 'meaning' may have come after kinaes-
thetic or organic images which were not identified. And in
the cases of Subgroup B the doctrine would explain the
longer reaction times obtained for meaning than for image.
We would simply have to assume that the 'meaning' which
138 EDWARD CHACE TOLMAN
followed the visual image was not recognized by this sub-
group, as it was by Subgroup A, as something distinct from
the image, but was confused with the visual image itself.
Such an assumption does not seem at all a difficult one.
Further experiments with subjects of the type of Subgroup B,
i. e.j those who introspectively declare meaning to be identi-
cal with visual image but who objectively require a longer
reaction time for meaning than for image, ought to throw
light upon the matter. We hope to be able in the near
future to report the results of such experiments.
We may sum up. The results of the y1 subjects for whom
both objectively and subjectively visual images came first,
render an out-and-out imageless position untenable. But
the fact, on the other hand, that for 2 of these 7 (Miss A.
and Mr. Pe., Subgroup A) meaning was distinct from the
visual image, as well as that for the great majority of all
subjects 'meaning' appears as something not analyzed into
images, gives no direct support to an out-and-out * image'
position. A compromise position, therefore, which assumes
that 'meaning' depends upon image but is itself distinct
from the latter, is the one most nearly suggested by our
results.
In conclusion, let us emphasize that the value of the
present investigation has lain not so much in the direction
of a positive proof of one or the other theory, as in showing
that, if a large enough sample of subjects be taken. Dr. Moore's
method in no way lends support to the out-and-out image-
less position.
1 This excludes Miss Rei. and Mr. Mi. The former's objective times and the
latter's introspection left it doubtful whether they really belonged in this third group
or in Group 2. See pp. 128-9, and footnote, p. 133.
EXPERIMENTS ON THE RELATIVE EFFICIENCY
OF MEN AND WOMEN IN MEMORY
AND REASONING1
BY ARTHUR I. GATES
The majority of psychologists and educators who have
expressed themselves on the subject are of the opinion that
women, as a rule, are considerably more efficient than men
in memory work and less efficient in applying the facts learned,
in self-expression, and in reasoning power. For example
one writer says:2 "Girls excel in learning and memorization
accepting studies on suggestion or authority, but are often
quite at sea when set to make tasks or experiments that give
individuality and a chance for self-expression, which is one
of the best things in boyhood."
Opinions similar to these seem to prevail generally among
psychologists, educators, and laymen. Many, moreover, are
of the opinion that women, in addition to having quicker and
more tenacious memories, are as a rule more diligent and
painstaking in their work; the boy may often be satisfied with
a fair knowledge of the general principles underlying a lesson,
while the girl seeks a more detailed and exact knowledge.
If such is the fact it should be taken into account in any
attempt to determine the sex-differences in memory, for obvi-
ously the differences in the time spent on the work might
easily account for the differences in the reproduction of the
ideas.
The experiments to be described presently were performed
first in 1913 and were repeated in 1914 and 1915, using as
subjects a class in elementary psychology consisting of from
158 to 275 students of both sexes of the sophomore, junior,
and senior years in the University of California.
1 From the Psychological Laboratory of the University of California.
2 Hall, G. S., "Youth: its Education, Regimen, and Hygiene." New York, 1912,
p. 287.
ARTHUR I. GATES
The data were obtained from the answers to three sets of
questions. Each set of two questions comprised the regular
weekly examination of the class. The first set called for a
somewhat detailed reproduction of facts presented in the
lectures of the week preceding. The second set called for the
application of facts or principles given in the lectures, the
purpose being to call into action a mental process as closely
as possible identical with that involved in reasoning. All the
questions were framed by Professor Stratton, who was in
charge of the classes, and who endeavored to make the
tests as nearly as possible adequate to the purpose of the
experiment. The papers in all cases were graded on a basis
of ten, but the averages below, for the sake of clearness, are
made on the basis of one hundred. All papers were corrected
by the regular * readers,' who were in no case aware that the
results were to be used for experimental purposes. It hap-
pened, moreover, that each of the nine sets of papers was
graded by a different 'reader.'
The following table shows the results in the case of memory
questions.
TABLE I
MEMORY
*9'3
1914
I9i3
No. of
Individuals
Grade
No. of
Individuals
Grade
No. of
Individuals
Grade
Women . •
QC
77
l62
89.2
1C A
86.4
Men
CO
73
IO2
85.0
Q8
81.0
Diff. in favor of women
4.0%
4-2%
5-4%
The women show a slight superiority in memory work,
amounting on the average to 4.5 percent. While the per-
centile difference is rather small, its reliability is indicated
by the fact that it appears in all cases, although different
questions were given at different times to three entirely
different groups of individuals.
Table II, shows the average grades obtained by men and
women to questions that involved reasoning.
The evidence indicates a slight superiority of the men
in this sort of work. The average difference is approximately
EFFICIENCY IN MEMORY AND REASONING
141
TABLE II
REASON
19*3
1914
1915
No. of
Individuals
Grade
No. of
Individuals
Grade
No. of
Individuals
Grade
Women
QO
77. c
153
83.3
154
88.4
^/fen
58
79-5
IO-?
86.4
99
89.2
Diff. in favor of men
2.0%
3-1%
0.8%
2 percent, a difference which is so small as to have but little
significance were it not for the fact that it is repeated by the
three separate groups.
Table III. shows the results of tests in which the subjects
were given free choice between a memory and a reason ques-
tion. The two questions, constituting the regular weekly
examination as before, were presented and the students were
permitted to take their choice.
TABLE III
ONE MEMORY AND ONE REASON QUESTION
1913
Memory Question
Grade
Reason Question
Grade
No. of
Individuals
Percent of
Individuals of
That Sex
No. of
Individuals
Percent of
Individuals of
That Sex
Women
Men
Diff
60
19
72.3
28.8
43-5
85
82
3
23
47
27.7
71.2
43-5
86
87
i
1914
Women . .
I2Q
84.8
7O.4.
23
1C 2
77
Men
80
78.4.
64.. c
22
21 6
80
Diff
70.4.
6.4
5-9
6.4
3
1915
Women
Men
144
85
91.8
74.2
88.4
87.0
\l
8.2
15.8
87.2
89.8
Diff
17.6
1.4
ij.^
7.6
2.6
Although both sexes show a distinct preference for the
memory question, the preference is much greater in the case
of women. The men show more willingness than do the
women to take the reason questions, although the actual
number of either sex that take these questions is small. In
142 ARTHUR I. GATES
1913 and 1915 twice as great a ratio of men, and in 1914 a
ratio one third greater of men than of women chose the
reason question. The grades received in the memory tests
confirm the earlier finding that the women excel in this kind
of work. The women excel in every case, although in two
(1913 and 1915) the differences in their favor are very small.
The grades received on the reason questions also confirm
the earlier finding that the men excel slightly in this type of
work. Although the superiority of the men is small it
appears in every case.
Our general conclusions from the experiment thus far are
as follows:
1. The women excel the men in memory work.
2. The men excel the women, but to a less degree, in
reason work.
3. Both sexes prefer memory work but more men show a
willingness to do reason work in lieu of memory work.
To let the experiment remain as it stands and to accept
without further question the conclusions just enumerated
would be hazardous and would certainly not take into
account all of the factors which have an influence here.
There is at least one possibility which if proven to be a reality
would force us to modify the conclusions at which we have
just arrived. It is possible that the apparent superiority of
the women in reproduction from memory is due merely to a
greater amount of study and not to an innate superiority of
memory.
To take into account this possibility the following test
was employed. The news item given below was read to the
class at the beginning of the lecture hour, the students being
warned to pay particular attention to the contents, without
being informed, however, of the purpose of the test. The
item follows:
THREE HOUSES BuRNED1
Boston, September 5. A serious fire last night destroyed three houses in the
center of the city. Seventeen families are without a home. The loss exceeds fifty
thousand dollars. In rescuing a child, one of the firemen was badly burned about
the hands and arms.
1 See Whipple, G. M., 'Manual of Physical and Mental Tests,' Baltimore, 1910,
p. 504.
EFFICIENCY IN MEMORY AND REASONING
'43
3-
4-
5-
6.
7-
The students were first requested to write down all the
facts they could recall from the article. Following the free
account, they were asked to answer the following questions:
1. In what city did the fire occur?
2. What was the date of the item? J*^
When did the fire break out? Kv»^
How many houses were destroyed?
In what part of the city were these houses ?
How many families were left homeless ?
What was the total loss (in dollars) ?
8. Who was burned?
9. On what part or parts of the body was this individual "•'
burned?
10. What was this individual doing when the burns were
received ?
The data used were obtained from the answers to the ten
definite questions, for it was found that the additions or
alterations of these answers from the free accounts were so
slight as to be negligible. The papers were graded on a basis
of ten, one unit being allowed for the correct answer to each
question. Table IV. gives the average results.
TABLE IV
Percentage
Reported
Percentage
Correct
Percentage
Positive Errors
Ratio of Pos.
Errors to Amt.
Reported
Ratio of Pos.
Errors to Pos.
Errors Plus
Amt. Not
Reported
1913
Women
97-4-
84.4
11. O
.111
.833
^£en
QO.O
8O.O
IO.O
.IOO
•COO
1914
W^omen
98.4.
86.4.
12. 0
.121
.888
^len
04.. 2
81. i
II O
.116
.6co
I9IS
Women
Q.1. 1
82 c
II 6
121
.662
Men
880
«^o
76 4.
10 6
1 2O
.44-1
Average of above.
Women
96.6
84.4
12.2
.I2C
.782
ft^en
90.7
79.8
IO.C
.112
.qiQ
Diff
c.o
/y.o
4.6
1.7
.OI 1
.263
The women in every case report a greater amount of the
content of the item, as well as a greater amount of it cor-
rectly. On the average the women report 96.6 percent of the
144 ARTHUR I. GATES
item and 84.4 percent of it correctly while the men report
but 90.7 percent with 79.8 percent correct. The men, how-
ever, make fewer mistakes. The actual number of errors
made by the women is greater in every case, although the
differences between the sexes is small. The ratios of the
number of errors to the total amount reported show even
smaller differences because of the fact that the women in all
cases report a larger amount. But the ratios of the amount
of positive errors to the total amount of positive errors plus
the amount not reported — i. e., to the field in which suggestion
and kindred forces could operate because the ideas were not
correctly remembered — were much larger for the women.
That is to say, the women, much more than the men, were
likely to make erroneous statements rather than mere omis-
sions. This ratio is, on the average, about one third larger
for the women.
The general conclusion from this test is that the women
in immediate memory tests can correctly reproduce more of
the detail of a given group of facts but at the same time make
more mistakes.
A question, however, may be raised with regard to the
application of the results gained by this method to the deter-
mination of the relative ability shown by men and women
in the tests first considered, because the present method
tests immediate memory, or immediate reproduction, rather
than delayed memory which is the function operative in the
examinations.
Accordingly, the same students were requested, one week
or five weeks after the immediate-memory test, to write,
without previous warning, all that they could remember of
the news item given above. The same set of ten questions
as before was used. Table V. gives the results.
In delayed as well as in immediate memory the women
have a greater range of report, a greater number of details
are reported correctly, and more positive errors are made.
The amount by which the sexes differ is about the same in
both types of memory.
The experiments with the news item justify the following
conclusions:
EFFICIENCY IN MEMORY AND REASONING
'45
TABLE V
DELAYED MEMORY
1913. After 5 Weeks
Ratio of Positive
Percentage
Reported
Percentage
Correct
Percentage
Positive Errors
Ratio of Posi-
tive Errors to
Amount
Errors to Posi-
tive Errors Plus
Amount Not
Reported
Reported
W^omen
82 o
64. o
18.0
.210
.COO
Men
72.0
S7-o
15.0
.208
•349
After 5 Weeks
Women
Men
89.0
79.0
68.0
60.0
2I.O
I9.O
.236
.240
.655
•475
1915. After I Week
Women
Men
944
89.8
80.7
75-8
13-7
12.0
.145
.137
.650
•495
Average of Above
Women
88.5
80.3
8.2
70.9
6ti
17.9
M-|
2.6
.200
.195
.005
.602
.440
.162
Men
Diff
1. The women excel the men in tests of immediate or
delayed memory, at least in so far as the amount of material
correctly reproduced is concerned.
2. The women, however, make more positive errors in
reporting.
The results obtained by other investigators are for the
most part in harmony with the present findings. A summary
of such experiments will be found in Whipple1 who concludes :
" Sex differences in this test [memory for ideas], as in the rote
memory test, are in favor of girls."2
A final consideration is the possibility that the women
employed in these experiments constitute a more select group
than the men. It is possible that these women are on the
whole more capable, or that their previous training has
better adapted them to the particular subject of psychology.
There is no obvious reason why this should be the case, but
in order to throw some light upon it the average grade in
the course has been computed for each sex.
1 Whipple, G. M., 'Manual of Mental and Physical Tests,' Part II., 17-43, 149"
223.
2 Op. cit., p. 213.
146 ARTHUR I. GATES
TABLE VI
i9J3
1914
1915
The women received an average grade of
77 o
yr r
74. O
The men received an average grade of
75-5
75-0
72.0
The women have slightly the higher grade. The mass of
experimental evidence from other investigations, however,
indicates that in groups of men and women of equal endow-
ment and training, the women usually excel in memory work.
We have found that in the three groups just considered, the
women excel in memory. It seems that the small amount
by which the women excel the men in the grades received in
the course may be accounted for by the great predominance
of memory work in the weekly and final examinations on
which the grades are based. The women who apparently
excel in memory work should in a long series of tests of that
nature, come out with a somewhat better average.
The three main conclusions that the investigation seems
to justify are as follows:
1. The women excel the men noticeably in either immedi-
ate or delayed memory work.
2. The men excel the women, but to a less degree, in
reason work.
3. Both sexes prefer memory work, but a greater relative
number of men show a willingness to do reason work in lieu
of memory work.
Results obtained by other investigators and the supple-
mentary tests for possible sources of error have brought
forth no evidence contradictory to the conclusions we have
reached.
INDIVIDUAL DIFFERENCES IN JUDGMENTS OF
THE BEAUTY OF SIMPLE FORMS
BY EDWARD L. THORNDIKE
Teachers College, Columbia University
Students of esthetic appreciation have commonly been
especially interested in the general drift or average tendency
toward this or that preference and have perhaps given an
impression of greater uniformity than exists. The diversity
of the judgments whose average favors the golden section,
for example, is really very great. It seems worth while
therefore to report certain rather extensive measurements of
esthetic preference which I have made.
The subjects of the experiment were college juniors and,
with few exceptions, of the female sex. The judgments
made were of the order of esthetic merit (the question being,
"Which rectangle do you like the looks of most? Next
most? etc.") of (A) rectangles 22-33, (&) triangles 41-52,
(C) Crosses 61-66 and 81-86, (D) designs A-L and (E)
147
148
EDWARD L. THORNDIKE
the 24 unnumbered designs. Each set was shown as here
save that the dimensions were in each case double those
here (quadrupling the areas).
(C)
I give the facts for from 100 to 250 individuals who made
the judgments, in the form of the percent of them assigning
a given form to a given position.
INDIVIDUAL DIFFERENCES IN JUDGMENTS OF BEAUTY 149
In the case of the rectangles it will be observed that 27,
28, and 29, those most liked, still have some ratings in the
lowest position of all; and that 33, the one least liked, still
has ratings in the highest position. In only 3 cases out of
144 do over 25 percent of the ratings give a rectangle the
same position.
II
II II
(D)
(D)
In the case of the triangles there is a pronounced drift of
opinion against the tall triangles, but even so almost every
position has votes in the case of each. This is still more the
case with the crosses.
In the case of the designs where the sequence by propor-
tions is more hidden, the variability becomes enormous.
EDWARD L. THORNDIKE
Although any one person may feel very decided preferences,
these are never shared by enough of his fellows to make
TABLE I
FREQUENCIES OF EACH POSITION FOR EACH RECTANGLE COMPUTED FROM ORDERS OF
MERIT REPORTED BY 200 INDIVIDUALS: IN PERCENTS
Rectangles
Positions
22
23
24
25
26
27
28
29
3°
31
32
33
I
4-5
5
3-5
8.5
6.5
IS
10.5
IS
16
6-5
4-5
4-5
2
i-5
7
6
7-5
H
7
16
17
II
8
3-S
i-S
3
4
2-5
8
ii
9-5
15
IS
ii
8.5
7
5-5
3
4
2-5
3-5
ii
12
10
16.5
I3-S
7-5
II
7
4-5
2
5
2
4
5-5
12
18
iS-S
8-5
12
8
7
4
3
6
2
4-5
9
14-5
22
9
12.5
10
7-5
5-5
3
i-5
7
5-5
7-5
12
15
4
14-5
9-5
6.5
9
ii
4-5
i-S
8
6
9
IS-S
6
12.5
2-5
5-5
ii
8
ii-S
10
i-S
9
9-5
15
12
10
i-5
2
2
5
IS
ii-S
ii
5-5
10
ii-S
18.5
H
i-5
i-5
2-5
I
i-5
4-5
22
H
7
ii
IS
21
3
2
3
2
i-5
3
35
14-5
12
36
2-5
0-5
o-S
0-5
3
i-5
o-S
54-5
anything like universal agreement. In the series of 12
designs, not one has 25 percent of ratings in any one position.
INDIVIDUAL DIFFERENCES IN JUDGMENTS OF BEAUTY I51
In the series of 24 designs, in only about one case out of
thirty are there 10 percent or more of ratings in any one
position.
TABLE II
FREQUENCIES OF EACH POSITION FOR EACH TRIANGLE COMPUTED FROM ORDERS OF
MERIT REPORTED BY 250 INDIVIDUALS: IN PERCENTS
Positions
Triangles
4*
42
43
44
45
46
47
48
49
So
51
52
I
26.0
14.0
21.2
14.0
7.2
4.8
3-6
2.O
1.2
3-2
.8
1.6
2
10.8
31.2
16.8
12.0
7.6
6.8
7-2
2.8
2.8
1.6
.8
3
80
.0
14.8
30.4
l6.4
12.0
3-6
3-6
3-6
3-2
2.O
1.2
.8
4
9.6
7.2
6.8
33-2
14.0
II. 2
8.0
4.0
1.6
3.6
.8
4
5
6.4
S.6
7-2
7.2
36.0
12.0
9-2
8.0
4.0
2.4
1.6
.8
6
7.2
3-2
4-4
6.4
6.4
41.2
12.0
7-6
7-2
4
2.4
1.6
7
44
1-2
3-2
24
4.8
6.8
4S-6
12.4
6.8
2.8
2.4
2.8
8
2.4
4.8
1.2
1.2
6.8
4.8
4.8
49-2
I2.O
8.0
2.4
1.6
9
•5.6
4.8
2.4
4.8
2.4
2.8
1.6
S-6
.IJ2.0
IO.O
4.0
3-6
10
5.2
2.0
44
1.2
1.2
3.6
2.0
2.0
4-8
61.2
9.2
4.0
ii
6.8
4.8
1.2
4
.8
2.0
1.6
.8
3-2
4.8
67.6
44
12
5.6
?8
0.8
4
.8
4
.8
1.2
16
5.6
76.8
?
I 2
4
.8
No great value attaches to the general drift of the con-
sensus, since the responses to the objects displayed as they
were and with criteria of symmetry so strongly suggested
TABLE III
FREQUENCIES OF EACH POSITION FOR EACH CROSS COMPUTED FROM ORDERS OF
MERIT REPORTED BY 140 INDIVIDUALS: IN PERCENTS
Crosses
Positions
61
62
63
64
65
66
81
82
83
84
8s
86
I
2.1
•7
IO.O
15.0
24-3
7-i
64
19-3
5-7
2.1
2.1
S-o
2
2.1
3-6
IO.O
22.1
!6.4
4-3
S-o
14-3
12.9
1.4
7-i
1.4
3
64
8.6
15-7
IO.O
11.4
4-3
18.6
8.6
10.7
1.4
4-3
4
3.6
3.6
IO.O
15.0
9-3
5.0
iS-o
13.6
15.0
2.1
64
1.4
5
7-9
15.7
7-i
10.7
6.4
13.6
iS-o
15.0
4-3
3-6
1.4
6
' V-6
8.6
11.4
9-3
5-7
8.6
13.6
7-9
12. 1
11.4
3-6
4-3
7
1.4
5-o
11.4
3-6
11.4
12.9
12. 1
4-3
15-7
10.7
5-7
S-o
8
S-o
IO.O
3.6
2.1
6.4
17.9
10.7
3.6
2.9
24-3
9-3
5-7
9
2.1
11.4
S-o
64
2-9
9-3
7-i
2.1
6.4
17.1
24-3
4-3
10
II.4
IO.O
13.6
2.9
5-0
57
•7
4-3
8.6
17.1
20.7
ii
IS.O
32.1
•7
2.1
5-7
S-o
•7
1.4
3-6
17.1
iS-7
12
53-6
0.7
•7
•7
6-4
1.4
3.6
2.1
30.7
may be different from the responses to the same objects in
isolation or in different surroundings. However, it may be
of interest to some to record that: The most liked rectangles
152
EDWARD L. THORNDIKE
TABLE IV
FREQUENCIES OF EACH POSITION FOR EACH LETTERED DESIGN COMPUTED FROM
ORDERS OF MERIT REPORTED BY 100 INDIVIDUALS: IN PERCENTS
Positions
Designs
A
B
C
D
E
F
G
H
I
J
K
L
I
14
6
12
9
2
S
6
2
IS
13
7
10
2
12
8
12
8
2
2
i
9
13
16
S
II
3
16
2
12
IO
IO
5
3
5
12
H
4
9
4
16
II
II
II
9
10
3
2
5
9
5
9
5
H
9
12
II
10
7
3
ij
7
6
4
12
6
8
6
13
i<;
6
17
5
6
5
9
i
IO
7
II
13
7
6
7
IO
ii
7
6
6
7
8
8
5
9
ii
13
8
8
7
IO
4
ii
6
7
9
i
6
6
7
19
12
13
6
7
10
12
10
2
7
2
5
12
16
7
19
10
ii
3
5
ii
I
10
c
II
C
17
22
1C
-j
7
12
13
2
4
3
24
7
I
2
41
2
TABLE V
FREQUENCIES OF EACH POSITION FOR EACH UNNUMBERED DESIGN OF THE FIRST
Two Rows COMPUTED FROM ORDERS OF MERIT REPORTED BY 250
INDIVIDUALS: IN PERCENTS. THE RESULTS FOR THE OTHER
Two Rows SHOW THE SAME VARIABILITY
Positions
Designs
Row
i
I
i
i
4
I
i
2
2
2
2
2
2
Number
i
2
3
5
6
I
2
3
4
5
6
I
2.8
10.0
12.8
2.8
1.2
4.4
2.O
.8
•4
7.2
2.4
1.2
2
4.0
9.6
8.8
2.0
.8
5-2
6-4
4.4
3-2
4.0
4-4
1.2
3
3-6
9-2
8.8
3-6
6.8
4.4
2.8
2.8
4.8
2.8
•4
4
5-6
IO.2
7.2
2.4
' i '.6 '
3-6
4-4
3-2
3.6
3-6
4.0
3.6
6.0
8.0
4.0
2.4
•4
8.0
3-6
6.8
5-6
3-2
2.O
1.6
6
4.8
8.4
3-6
2.4
3-6
8.4
9.2
7-2
7-2
2.8
1.6
2.4
8
9
8.0
g
4-4
3-2
4-4
7-2
2.0
1.6
3-6
2.8
3.6
4.8
4.4
4.8
4-4
8.4
7-6
7-6
7.2
4.8
4-8
6.0
4.0
4-8
6.0
4.8
4.8
it
2.8
4.8
6.0
3.6
2.8
3-2
10
3-2
3-6
2.8
4.0
6.0
5-2
4.8
6.8
10.0
4.0
5-6
3.6
ii
5-2
2.4
2.8
4.8
5.2
4.8
4.8
4-4
7.2
6.0
4.0
44
12
3-2
4.4
4.0
2.8
6.0
3-6
5-2
4.4
6.4
7.2
6.4
5-2
13
5-2
4.8
2.8
6.4
4.0
4-4
4.8
3-6
2.8
3-2
6.0
6.0
H
4.8
3.6
3-2
2.8
4.0
3-2
5-2
2.8
3-6
n o
O.O
5-6
6.4
IS
4.0
1.6
2.8
3-2
4.0
3-6
3.6
5-2
4.0
4-8
IO.O
4-4
16
17
4.4
2.8
1.6
2.0
1.2
3-2
4.0
2.4
4.8
4.8
Ii
6-4
4-4
4.8
3-2
4.0
5-6
I8
6.0
6.0
3-6
44
2.4
18
2.0
1.6
3-2
5.6
6.8
1.6
2.0
6.0
2.8
1.2
5-2
44
19
3-2
2.0
4.0
4.8
8.4
4.0
1.2
3-2
3-2
4.0
3.6
4.0
20
2.O
1.6
3-2
8.4
5.6
1.6
1.6
4-4
3-2
1.6
4.0
8.4
21
2.8
1.2
3.6
10.8
6.8
2.0
2.8
5-2
2.0
2.4
2.4
IO.O
22
3.6
1.2
II. 2
3-2
1.6
2.0
4-4
.8
1.2
3.6
IO.O
23
1.6
.8
3-2
1.6
3-6
2.4
.8
1.2
1.2
.8
2.4
2.4
24
4.0
.8
1.2
1.2
4.4
.8
.8
•4
4-4
•4
•4
4.0
INDIVIDUAL DIFFERENCES IN JUDGMENTS OF BEAUTY 153
TABLE VI
ORDER OF MERIT ASSIGNED BY THE CONSENSUS OF COLLEGE STUDENTS
Unnumbered De-
signs. The Num-
Rectangles
Triangles
Crosses
Lettered Designs
bers Here Follow
the Order of
Printing*
29
43,44
64, 82, 65
28
42
81, 83, 66
A
14
27 and 30
4i,4S
84,63
J
2
26
46
62, 85
CDIL
3
25 and 31
47
61, 86
BF
6 15
24
48
E
i 7, 13
23 and 32
49
G.H.
8 9, 10, 20
22 and 33
50
K.
II 17, 22, 24
Si
4 5, 12, 18
52
16 23
19 21
* That is, the first design in the second row is 7, the next is 8; the first design in
the third row is 13, the next is 14, etc.
had, as the ratio of altitude to base, 1.83 to I. The most
liked triangles had, as similar ratios, 1.6 to I and 1.7 to I
(43 and 44 being equally well liked). The most liked of
the crosses had a bar half of the length of the upright and
such a bar is best liked when it cuts the upright so as to leave
one fourth above and three fourths below. A bar two fifths
of the length of the upright is nearly as well liked. The
most liked of the unnumbered designs is the second one of
the third row. The first and third of the fourth row are the
most disliked. In the lettered designs the space relations
may vary widely so long as the design remains obvious,
and so long as neither bareness nor crowdedness is suggested.
A and / are liked about equally; G, H and K are disliked
about equally.
The order of merit of the consensus is given for each
group of designs in Table VI.
PRELIMINARY REPORT ON THE RELATIVE IN-
TENSITY OF SUCCESSIVE, SIMULTANEOUS,
ASCENDING, AND DESCENDING TONES
BY A. P. WEISS
Ohio State University
The attribute of tone intensity has been relatively neg-
lected in experiments in audition because of the technical
difficulty in producing pure tones which may be varied in
their loudness or intensity in a definite and measurable
manner.
The apparatus with which the experiments in this paper
were performed was developed at the University of Missouri
under the guidance and suggestions of Dr. M. F. Meyer.
The extended report showing the details of the apparatus
construction and the manner in which the experiments were
conducted is being published as a PSYCHOLOGICAL REVIEW
MONOGRAPH.
The apparatus makes it possible to produce tuning-fork
tones which meet the following conditions :
1. The tones are pure in the sense that no lower or upper
harmonics can be detected.
2. The tones can be quickly varied from weak to strong
in any number of steps and each degree of intensity can be
repeated as often as necessary.
3. The tones 'come in' and 'go out' at their full intensity
without disturbing, starting, or stopping noises.
4. The phase relations of the tuning forks is under control.
The nature of the experiments may be understood from
the following illustrations.
I. Relative intensity of successive and simultaneous tones:
Suppose we have the tone 200 which, during a given trial, is
always sounded at a medium and constant intensity. An-
other tone 250 can be easily varied from weak to strong
RELATIVE INTENSITY OF TONES '55
(ascending order)1 or from strong to weak (descending order).
Suppose we sound 200 and 250 alternately and vary the
intensity of 250 (either ascending or descending) until it
seems to have the same intensity as 200; the question now
arises, if 200 and 250 are sounded simultaneously, are they
still of the same intensity?
2. Relative intensity of ascending and descending tones:
Suppose 200 is kept constant in intensity and 250 is varied
in descending order, will the point at which 250 is considered
equal to 200 be the same as when 250 is varied in ascending
order?
The tones used in this experiment were the four tones
150, 200, 250, 300. Each tone was compared with each of
the other three tones in four ways.
1. Successively, with the comparison tone varying in
ascending order.
2. Successively, with the comparison tone varying in
descending order.
3. Simultaneously, with the comparison tone varying in
ascending order.
4. Simultaneously, with the comparison tone varying in
descending order.
Both the lower and the higher tones were used as standard
in each pair. Each pair of tones was further compared for
ten degrees of intensity ranging from a weak tone which was
nevertheless clearly heard, to a strong tone which was not,
however, so loud that it became disagreeable. That is, the
range of conveniently obtainable intensities was divided into
ten steps and the various tone combinations were compared
for each of these steps.
The method of making the judgments was that of " Selbst-
einstellung." One tone (the standard) was kept at constant
intensity while the observer varied the intensity of the com-
parison tone until it seemed equal in intensity to the tone
which was being used as the standard.
xThe numbers 200 and 250 refer to the vibration rates. Ascending order or
ascending tones refer to tones which are varied from silence to weak to strong. De-
scending order or descending tones refer to tones whose intensity is varied from strong
to weak.
150 A. P. WEISS
The tones were produced by resonators suspended over
silently vibrating tuning forks of constant amplitude and the
objective intensity of the tone was measured by the distance
of the mouth of the resonator from the prongs of the tuning
fork.
FIG. i.
The above diagram shows the results obtained. The
number in the right arm of each brace indicates the intensity
relations between the series connected by the brace. Thus
the number .86 in the brace connecting the circles of the
successive-ascending and successive-descending series, means
that when a tone of constant intensity was compared suc-
cessively, first with an ascending tone, and second with a
descending tone, the descending tone was made .86 step
stronger than the ascending tone. Subjectively this implies
that the descending tone was actually heard weaker than the
ascending tone, since if it had been set at the intensity of
the ascending tone it would have been judged weaker than
the standard tone.
A negative sign in the diagram means that the series was
made weaker (objectively) than the companion series. In
subjective terms this means that the tone was actually heard
stronger. This opposition between "made stronger objec-
tively" and "heard weaker subjectively" is rather confusing
and it was thought worth while to add the subjective impli-
cations parenthetically in the statement of the conclusion.
Each of the conclusions which follow are based on at
least 4,800 judgments or reactions. The term 'step' refers
to one tenth the total range of intensities used in the experi-
ments, or one tenth of the range of conveniently obtainable
intensities.
RELATIVE INTENSITY OF TONES 157
1. When compared successively with a tone of constant
intensity, descending tones are made .86 step stronger (or
heard .86 step weaker) than ascending tones.
2. When compared simultaneously with a tone of constant
intensity, descending tones are made .92 step stronger (or
heard .92 step weaker) than ascending tones.
3. When compared with a tone of constant intensity,
ascending tones, when compared simultaneously, are made .52
step weaker (or heard .52 step stronger) than when compared
successively.
4. When compared with a tone of constant intensity,
descending tones, when compared simultaneously, are made
.55 step weaker (or heard .55 step stronger) than when com-
pared successively.
5. When compared with a tone of constant intensity in
mixed simultaneous and successive order, descending tones
are made .86 step stronger (or heard .86 step weaker) than
ascending tones.
6. When compared with a tone of constant intensity in
mixed ascending and descending order, simultaneous tones
are made .34 step weaker (or heard .34 step stronger) than
successive tones.
The deviations of one half of the intensity judgments
above or below the objective intensity was .74 step. This
seems to indicate that within the range of conveniently ob-
tainable intensities used in this experiment 15 steps might
have been discriminated. Taking very weak and very loud
tones it seems that 25 steps should be possible.
The value .74 is also an indication of the reliability with
which intensity judgments can be made. This is about the
same for all the intensities used, being somewhat less for the
medium intensities than for the extremes, as might have been
expected.
The variability of the intensity judgments is not influenced
as much by difference in vibration rates as was expected.
The greatest difference between any of the tones of this ex-
periment was an octave (150 vibrations) and the comparisons
158 A. P. WEISS
between these two tones were no more variable than where
the difference was 50 vibrations. This seems to show that
even between tones whose vibration difference is considerable,
the intensity judgments can be made with a degree of accuracy
which promises well for an experimental analysis of the sound
intensity reaction.
DISCUSSION
A NEW METHOD OF HETEROCHROMATIC PHOTOM-
ETRY—A REPLY TO DR. JOHNSON
In the September number of this journal appears a discussion
entitled 'A Note on Ferree and Rand's Method of Photometry,' by
Dr. H. M. Johnson, of the Nela Park Laboratory. This discussion,
we may perhaps be pardoned for noting, is remarkable chiefly for
its numerous mistakes and incorrect or misleading representations,
a few of which we take opportunity here to rectify. The net service
of the discussion is to call the authors' attention to the omission of a
decimal point in the original article, for which they duly acknowledge
their debt.
1. In his opening paragraph Dr. Johnson says: "The authors
claim for their method that with respect both to sensitivity and
reproducibility it surpasses the equality of brightness method, even
when the photometer head used is of the best Lummer-Brodhun
type." In regard to this statement we beg to point out that Dr.
Johnson has omitted from what was actually said all that makes a
difference between a reasonable and an absurd claim. We had
claimed in our paper greater reproducibility of setting for the method
in question as compared with the equality of brightness method
only in case of hetero chromatic photometry, in which respect as is
well known the equality of brightness method is notably deficient.
The possibility of a service to heterochromatic photometry alone
is the reason given in the paper for applying to the rating of artificial
lights a principle formerly used by us for an entirely different pur-
pose. Also the special reference to heterochromatic photometry
was featured in the title.
2. Dr. Johnson next says: "The authors assumed that the two
elements making up the photometer screen ' received equal amounts
of light from the source to be measured.' Even if the elements
were equidistant from the lamp . . . the truth of this assumption
does not follow from the data given, In some of the work the
results of which are presented in the authors' table, the angular
separation of the compared elements was 14° to 15° at the source.
Now the radiation from a carbon or tungsten lamp is not equal in
159
160 c. E. FERREE AND GERTRUDE RAND
all directions as is that from an ideal point source. In fact, for
lamps of such types, differences of several per cent, in different
directions normal to the long axis of the lamp are the rule, and a
considerable difference might occur in a range of 15°."
With reference to the above statements we wish to note in the
first place that it was never assumed by us that there were only two
elements in the photometer screen. This erroneous interpretation
of the principle on which the method is based can be attributed to
Dr. Johnson only. Secondly, that when the angular separation of
the elements referred to (the stimulus patch and the measuring
disc), is correctly computed from the data contained in the original
article it is found to vary between 4.5° and II0,1 and not to have a
range of 15°. And thirdly, that when the question of the influence
of the distribution curve on the general applicability of the method
to working practice was raised by us in a paper presented to the
Philadelphia Section of the Illuminating Engineering Society in
February, I9I4,2 it was the consensus of opinion in the discussion
that followed that the possibility of error from this source is of neg-
ligible consequence in a field presenting so many difficulties as
heterochromatic photometry, and that the effective check on these
and many other points which were raised by us at that time — in
addition to those now raised by Dr. Johnson — must come in a
comparison of the results with those obtained by the equality of
brightness method. Because of this confirmatory opinion of a group
of specialists fully familiar with all the technical and working details
of photometry and because of the check experiments we had run on
the point to convince ourselves of the negligible influence of the
factor for the conditions under which we worked (see this paper,
p. 165), we had not considered it necessary to raise the discussion
in the preliminary exposition of the principles on which the pro-
posed method is based, contained in the article in question. How-
ever, since the point has been raised by Dr. Johnson, the following
comments may not be out of place.
1 It is assumed here of course that Dr. Johnson referred to the angle for the colorless
light. There can have been no reasonable doubt in his mind that the colored light was
not obtained from the naked carbon or tungsten lamps to which his comments on
distribution refer. (See footnote, original article p. 9).
2 With reference to the foregoing point and to others taken up in this discussion it
is scarcely needful to state that principles and descriptions of conditions of a technical
nature were taken up in a fuller and more detailed way when a statement of the method
was presented to auditors technically interested in photometry than was done in the
article criticized by Dr. Johnson.
HETEROCHROMATIC PHOTOMETRY
161
(a) A general statement of the type which Dr. Johnson has
made about the inequality of distribution of carbon and tungsten
lamps is incomplete to the point of being somewhat misleading.
As is well known, the distribution curve of an incandescent fila-
ment lamp depends upon the shape of the filament. While, for
example, the single oval filament of the ordinary carbon lamp
gives considerable unevenness of distribution, if wide enough angles
are considered, the single loop tungsten filament of the Mazda
lamp, series type, gives a curve which deviates so little from
FIG. I. Showing the distribution curve in the horizontal plane of a 5O-watt
carbon lamp, single oval filament — readings taken at 5 ft. radius; lamp operated at
6.8 horizontal cp.; watts per horizontal cp., 2.97.
a circle as to be scarcely detectable with the exception of a very
small region in the plane of the filament. The curves for these
lamps are appended in Figs. I and 2. In Fig. 3 is given also the
curve for the ordinary type B Mazda lamp.1 This curve shows
more variation than the series lamp but it is so nearly uniform as
to be considered circular for practical purposes. However, neither
this nor the single oval filament carbon lamp have ever been used
1 The determinations represented in these curves were made by the photometric
laboratory of the General Electric Co., Schenectady, N. Y.
1 62 c. E. FERREE AND GERTRUDE RAND
by us in connection with the method of photometry in question
without some device to secure greater uniformity of distribution of
light. In case a naked lamp were used at all it has always been of
the series type, single-loop tip-anchored filament, and care has been
taken to have the lamp set on the bar so that the light was taken at
right angles to the plane of the filament or from the most uniform
part of the curve. But even were a carbon lamp used and the arrow
FIG. 2. Showing the distribution curve of a 60 cp. series Mazda lamp (clear),
single loop tip anchored filament, 6.6 amps. — readings taken at 5 ft. radius; lamp
operated at 60 horizontal cp.; watts per horizontal cp., 1.18.
or 'fiducial' mark scratched in a plane at right angles to the plane
of the filament, the distribution would fall off so evenly on either
side (see Fig. i)1 that the difference in the illumination of the
stimulus patch and measuring disc, not exceeding 5.5° on either
side, should be negligible.
1 It should be noted that in making the cuts for the curves in Figs. I and 3 the
true deviations from uniformity have been exaggerated by small but considerable
amounts.
HETEROCHROMATIC PHOTOMETRY 163
(b) So far as the question of uniformity of angular distribution of
light is concerned, stress seems to be laid in the criticism on the
equality of illumination of the stimulus patch and the measuring
disc alone from the lights to be photometered. This is not at all
in keeping with a correct interpretation of the method, for the
photometric balance does not consist in the judgments of the actual
amounts of light falling on the stimulus patch and measuring disc.
FIG. 3. Showing the distribution curve in the horizontal plane of a 4O-watt G. E.
Mazda lamp (clear), regular type small bulb, no volts — reading taken at 5 ft. radius;
lamp operated at 32.5 horizontal cp.; watts per horizontal cp., 1.23.
The apparent brightness of the stimulus patch is, for example, the
result of three factors: the actual amount of light falling on the
stimulus patch, the amount falling on the surrounding screen (rather
in both cases the amount reflected to the eye), and the physiological
induction caused by the difference in the brightnesses of these two
surfaces. Dr. Johnson, however, as indicated above, in considering
the question of the distribution of the illumination and its probable
effect on the results of the method, seems throughout his discussion
164 C. E. FERREE AND GERTRUDE RAND
to take into account only the relative amounts of light received by
the stimulus patch and the measuring disc, and in so doing shows a
fundamental misunderstanding of the principle on which the method
is based. The illumination of the field surrounding the stimulus
patch is just as important as the illumination of the stimulus patch
itself, for it is an equal factor in producing the induction and is, so far
as any one knows, effective for induction up to the measuring disc;
and there is, it is scarcely needful to point out, not an angular
separation of if between this screen and the measuring disc. The
important point is rather that there shall be no effective difference
in the collective situation influencing the induction and its measure-
ment for the standard and the comparison lamp. That is, although
the two surfaces are compared in each judgment, the comparison
of the two light sources is based on the results of two judgments,
and if there is no difference in the collective situation influencing the
two judgments, no injustice is done to the lights compared. If,
therefore, we were considering with Dr. Johnson the relative illu-
mination of stimulus patch and measuring disc to the exclusion
of other factors, and to what degree this relative illumination is
influenced by the distribution curve of the light source, the impor-
tant item is not that there is an angular separation between them
of a given number of degrees and a possible difference of illumination
in consequence, but how much this varies for the position of the
standard and comparison lamps on the photometer bar. For the
nearest position of the standard lamp, the difference in the angular
separation for the two lamps was 11°; for the farthest position for
the distances as given in the table, the difference would have been
4.5°. However, for the greater distances that would have been
required for the standard lamp from the screen to establish a balance
with the less intense colored lights, a part of the reduction was
produced by sectored discs, because in the form and set-up of ap-
paratus employed for that work, distances of light from screen of
134-160 cm. (Table I., original article, p. 9) could not conveniently
be attained owing to the angle of the shadow cast by the observer's
head. This reduction was converted into terms of the law of
squares to make the results comparable in the table with those
obtained by the equality of brightness method. The actual setting
of the lamp on the photometer bar for the greatest of these distances
was 104 instead of 160 cm. The difference between the angular
separation of stimulus patch and measuring disc was in this case,
therefore, 7°. The actual range of variation of angular separation
HETEROCHROMATIC PHOTOMETRY
165
of stimulus patch and measuring disc was thus only from 7° to 11°.
There is, it is obvious, considerable difference between these values
and the 15° with which Dr. Johnson confronts us. Furthermore, in
the course of the original work we ran a series of check experiments
to determine whether this difference in angular separation in case
of the standard and comparison lights produced any significant
error. That is, in these check experiments both lights were kept
in the same position and the light for the more intense, the standard,
was reduced by means of sectored discs very accurately cut from
sheet aluminum, the open sectors of which were measured with a
protractor provided with a Vernier scale reading to minutes. The
results of these experiments are given in Table I.
TABLE I.
SHOWING A COMPARISON OF THE RESULTS OBTAINED FOR THE LIGHTS REPRESENTED
IN THE ORIGINAL TABLE WHEN THE PHOTOMETRIC BALANCE WAS MADE (a) BY
CHANGING THE SETTING OF THE LIGHTS ON THE PHOTOMETER BAR; AND (b) BY THE
USE OF THE SECTORED Disc
Value of
Dis-
Open Sec-
tance of
tor Giving
Source of Colored Light
Color
White
Light
Giving
Equal-
ity of
Illumi-
Ratio of
Candle-
power.
Color:
White
Equality
of Illumi-
nation with
Distance
of White
and Col-
Ratio of
Candle-
power.
Color;
White
Differ-
ence in
Ratio
Differ-
ence in
Per
Cent.
Candle-
power
nation
ored
Cm.'
Lights
Equal
87 cp. 41 cm. distant
Red
66.6
0-379
137-5°
0.382
0.0030
0.785
from photometric
screen
Blue-green
ro.r
O.4.74.8
I72,O
o 4.778
O.OO3O
0.628
52 cp. 38 cm. distant
Red
37 0
82.2
T"/TU
0.2137
A / **W
77-75
\J.£^j 1 <J
0.2160
O.OO23
1. 06
from photometric
screen . .
Blue-green
7O C
O 2QO£
IOC C
O 2Q3 1
O.OO26
0.887
13 cp. 38 cm. distant
Red
i ->
160.0
w.^yw.j
0.0564
•*OO
20.5
^-^yj L
0.05694
O.OOO54
0.948
from photometric
screen
Blue-green
134-9
0.0793
28.85
0.08014
O.OOO79
0.985
Moreover, so far as inequalities of illumination of stimulus patch
and measuring disc are concerned, we may point out that a naked
lamp was not even used in the experiments the results of which are
given in the original table. Partly because the colored light was
secured by means of colored filters, and partly as a precaution
against unevenness of illumination of stimulus patch, measuring
disc, and surrounding field for a height and breadth sufficient for
the purpose of the experiment, the light was placed in a lamp-house
166 c. E. FERREE AND GERTRUDE RAND
(see original article, footnote p. 9).1 This lamp-house was 24 cm.
high, 14 cm. wide and 14 cm. deep. At the lower end of the lamp-
house was an opening 5 cm. square through which the light passed
to the screen. The lamp-house was lined with mat white paper so
shaped as to round off the edges and corners and to give as much as
possible in the lower part of the enclosure the effect of the segment
of a sphere. No light passed directly from the lamp to the screen
as the tip of the lamp was for the different lamps used from 2 to 9
cm. above the opening for the emission of the light. Owing to the
high absorption of the Wrattan and Wainwright filters the light
from the lamp used to establish a balance with that transmitted
from the filters had to be greatly reduced at the opening of the lamp-
house by means of colorless absorbing screens, which served further
to diffuse the light. To determine whether or not any serious dif-
ference in the distribution of light to measuring disc and stimulus
patch was present in case of this device, the light was photometered
at stimulus patch and at measuring disc for each position of the
lights on the bar. No difference could be detected for these two
positions by the equality of brightness method. Also the distribu-
tion curve for the light coming through this opening was found to be
circular through an angle greater than the n° in question. Our
statement in the original article then was correct that the equality
of distance of the measuring disc and stimulus patch on either side
of the photometer bar guaranteed that they receive equal illumina-
tion from the light source employed. It would also be true within
any reasonable margin of error for the single loop filament series
lamp set as described above (without a lamp house), and even for
the single oval carbon filament within a margin of error quite ac-
ceptable for work in heterochromatic photometry.
3. In a later paragraph (p. 394) Dr. Johnson conveys the im-
pression that we claim an agreement between the results of the
new method and those of the equality of brightness method within
1 The lamp-house is not shown in the photograph of apparatus given in the original
article. The photograph was a part of the general description of the method and the
apparatus that might be used with it. In making this photograph the apparatus
was regrouped, the object being merely to show the type of bar used, the screen and
the measuring disc. In this photograph it will also be noted that the apparatus was
not even shown in the position in which it is used in making the determinations. The
use of a lamp-house is mentioned in another part of the article, namely the part treating
of the results that were given as a sample of what might be obtained with the method.
In the first photographs that were made the lamp-house was included, but its size and
position in the foreground made it appear so disproportionately large that it was decided
to omit it and to give the photograph the general character mentioned above.
HETEROCHROMATIC PHOTOMETRY 167
a fraction of one per cent.1 Of this we have to say that no numerical
value whatever was assigned to the agreement in the original article
nor was any general statement made that would warrant the infer-
ence that we claimed an agreement within so small a margin. All
that appears in the article in this connection is a very brief table of
results containing no reference whatever to the point in question
accompanied by an 8-line paragraph stating that the table is ap-
pended as a sample of the results obtained, that the results are
averages from 25 determinations, etc. It is the custom in photom-
etry when a numerical expression is made of agreements, mean
deviations, etc., to give these in per cent, illumination or per cent,
candlepower. When this is done for the table in question, the
agreement shown by the data given falls within 1.5 per cent, instead
of 'within a fraction of I per cent.' as is stated by Dr. Johnson.
And this it will be remembered, is an agreement in the average.
When the individual determinations are compared, the deviations
reach values of the order of + 10 and— 12 per cent. Some idea of
this may be had from an inspection of the per cent, mean variations
appearing in the table for the results obtained by the equality
of brightness method. Thus it will be seen that the actual
closeness of agreement of results is not surprising. It has been
made to appear so only by our critic's method of presentation.
1 On p. 393 Dr. Johnson says: "The authors do not describe their mode of pro-
cedure in making their measurements by the method of direct comparison. I assume,
therefore, .... Under these conditions and working with the lamps beyond certain mini-
mal distances from the photometer head, the luminous intensities of the compared
sources would be inversely [italics ours] as the squares of their distances from the photom-
eter screen at valid settings for equality of brightness on the two halves of the photometer
field." We did not suppose that in an article on photometry it was necessary to give
a description of the equality of brightness method over 100 years after its principles
were laid down for all time (Pierre Bouguer, 1760, and Sir Benjamin Thompson,
Count of Rumford, 1793). However, we do wish to say now that Dr. Johnson has
raised the question that we conformed to all that is essential in his very elementary
directions with the exception that we chose rather to follow the custom to which we
know of no exception either in practice or recommendation, of calculating the luminous
intensities of the light sources on the basis of the direct squares of the distances of these
light sources from the photometer head, instead of the inverse squares. In replying
to an advanced criticism on photometric method, one should not have to point out
that the law of inverse squares applies to the intensity of illumination at different
distances from a given source; while the converse of this relation, namely, the direct
squares, applies to the comparative intensities of two sources which produce equal
illumination on a given screen or photometer head. That is, the former is used in the
computations of intensity of illumination: foot-candles, meter-candles, etc.; and the
latter in the computation of the relative intensities of light sources : candlepower, lam-
berts, millilamberts, etc.
1 68 c. E. FERREE AND GERTRUDE RAND
4. Also on p. 394 Dr. Johnson presents a table in which it is
represented that the measuring disc in the work for which our table
of results was submitted was 3 cm. nearer to the observer than the
plane of the screen containing the stimulus patch. Applying the
law of inverse squares he demonstrates that the illumination of the
stimulus patch and measuring disc was in case of each light source
unequal. Since the colored lights were all nearer the screen and
measuring disc than the standard white light in proportions varying
from 41 /59 to 387 160 (actually 417 59 to 387 106 because, as stated
earlier, a sectored disc was used for the lights requiring the greater
distance of setting from the screen), the 3 cm. caused a greater
difference between the illumination of the measuring disc than of
the stimulus patch for the colored lights than for the white light by
percentages ranging from 5.4 to 15.5. From the showing of this
table without further inquiry into causes, it was concluded that
'the authors' procedure in making the settings was faulty,' the
* method is insensitive' and that the evidence of agreement of the
two methods is * spurious,' for the explanation of which latter point
there seems to have been no hypothesis worthy of mention but that
the settings of one method were biased by a knowledge ofthe settings
of the other — a smashing and uncompromising arraignment truly!
However, we beg in passing to say a word of this table ourselves.
In the first place, as a matter of only minor consequence to the
present discussion, we wish to point out that in all of the compu-
tations given by Dr. Johnson of the deviations in per cent, from
proportionality of illumination of stimulus patch and measuring disc,
errors have been made, and that in 5 out of a total of 6 cases ap-
pearing in his table these errors have ranged from 1.8 to 1 1 per cent,
of the correct value, with a leaning in some of the most important
cases towards the advantage of the critic. This, we may be par-
doned for noting, is under the circumstances somewhat surprising,
and is of value perhaps chiefly in demonstrating that it is possible
for mistakes to occur even in a critique levelled at the accuracy of
the work of others without furnishing a justification for the im-
pugning of motives and integrities. And secondly we wish to state
that, as might have been suspected by our critic himself,1 the 3 cm.
xThe above statement is made for the following reasons, (a) It is obvious on
a priori grounds to one having even the least rudimentary knowledge of the principles
on which photometry is based, that a just balance could not be established between the
colored and white lights involving so wide a difference in setting on the bar if the
measuring disc was 3 cm. in front of the photometer screen. And (b) even an approxi-
mate set-up of the apparatus with the lights in position demonstrates at a glance that
HETEROCHROMATIC PHOTOMETRY 169
was a typographical error. In the original data still in our posses-
sion, the distance of the measuring disc from the screen is given as
.3 cm.1 When the law of inverse squares is applied to this, the dis-
crepancy of illumination of stimulus patch and measuring disc for
the distances used by Dr. Johnson in his computations ranges from
.464 to 1.22 per cent., and for the actual distances used, from .464
to 1.03 percent. — an amount which the experienced photometrist
will, we think, grant is relatively negligible among the much greater
sources of error present in heterochromatic photometry.
We have, however, been sufficiently curious to know what results
would be obtained with the measuring disc placed 3 cm. in front of
the screen to repeat the work represented in the original table for
the four highest intensities with this change in the set-up. Dif-
ferences from the results quoted in the original tables — also, as it
happens for the cases tested, the amount of deviation from agree-
ment with the equality of brightness results — ranged from 13.5 to
25 per cent, when the determination was begun with the weaker
light, and from 18.6 to 29 per cent, when the determination was
begun with the stronger light.2 These figures indicate that rather
than being remarkable for its insensitivity, as is charged by Dr.
Johnson on the basis of too narrow a consideration of possibilities
and apparently no first-hand knowledge whatever of the facts in
question, the method shows by still another test a very high degree
of sensitivity.
5. The error in our critic's final conclusion (pp. 395-6) should by
this time be so obvious as to need no comment. We will, therefore,
rest our case so far as we recognize that a case has existed, until
space can be had for a further presentation of results. In this
regard it is hardly necessary to mention that we do not consider,
the conditions produced are not compatible with the principles on which the method
of making the balance is based. For example, when illuminated directly from the
lamp on the bar a sharp shadow is cast by the disc on the screen, which is plainly
in the view of the observer at the angle at which the observation is made. This is the
equivalent of surrounding the disc with a black band which varies in width as the
position of the lamp on the bar is changed. This is obviously not permissible. In
fact the error is of a kind which is usually handled in a note of inquiry to the authors.
1 Also there are, we might mention, a number of witnesses to the set-up of the
apparatus used by us in the work on heterochromatic photometry.
2 On account of the limited space allowed, an explanation of why such excessive
deviations are obtained with this incorrect set-up will have to be deferred until later
work; also the very obvious explanation of why a greater distance of measuring disc
from screen was permissible, in fact pf advantage, in the work in which the method
was used to detect changes in the diffuse illumination of an optics-room (PsvcHOL.
BULL., 1913, 10, p. 371) than when it was applied to the rating of lights on a bar.
170 C. E. FERREE AND GERTRUDE RAND
as our critic seems to have thought, that a place has been won for
our method among those hoary and worn with service on the basis
of a single sample table appended to a preliminary description of
method and apparatus and representing the results of only one
observer for two colors and only six of the possible settings on the
photometer bar..
NOTE. — Dr. Johnson mentioned the use of a rotator toequalize the light radiation in
different directions; also the deviations found by Wright from Lambert's law of re-
flection for mat surfaces. Since neither of these points was raised in the original article,
it might be inferred that they were not known and taken into account by the authors.
It will probably not be prejudicial to either side of the case to mention here that one of
the writers supervised the construction of his first lamp rotator for work in photometry
in 1901 while a teacher of physics, and is well acquainted with the uses and need of a
rotator. Also in 1903 while a graduate student of physics he was assigned a study of
the reflection from mat surfaces as a problem for investigation, the object being to
continue along the lines mapped out by Wright. Both from his reading and in-
struction with regard to the work of Wright and others, however, he is totally unable
to concur in a single comment that Dr. Johnson has made on the subject of diffuse re-
flection in the footnote on p. 392. Dr. Johnson says: "Another source of error which
the authors appear not to have taken into account may be worthy of mention. The
angles at which the light was diffusely reflected into the eye from the stimulus patch
and the disc at the fixation point were not the same. The percentage of incident light
reflected into the eye would have been different, therefore, even if the two surfaces had
been of the same material. Furthermore, the difference in percentage of incident light
reflected in the direction of the eye is not constant for any two positions of the source.
Cf. Wright, H. R., 'Photometry of the Diffuse Reflection of Light on Matt Surfaces,'
Philos. Trans., 1900, 49, Ser. 5, pp. 199-216." Of the sentences quoted the second is
the only one that can be said to be true. The angle of emission e from the stimulus
patch in relation to the eye was approximately o°; while for the measuring disc it was
25°. The reflection, therefore, in the direction of the eye from a given point or unit
surface in the area fixated of the measuring disc was less than that from the stimulus
patch by an amount equal to the cosine of 25°. Dr. Johnson, however, neglects to take
into account in considering the case presented by our method that the observation is
not confined to a single point or unit of area and that the area of surface viewed increases
as the secant (the reciprocal of the cosine) of the angle at which the surface is
viewed measured from the normal. That is, the increase of the area viewed just
compensates for the lessened amount of reflection from unit area. Nutting, for ex-
ample, says: "A red hot metal plate is of the same brightness viewed at any angle since
the foreshortening of the area just compensates for the variation in the radiation from
a given area. Lambert's law holds for mat surfaces for both emitted and reflected
radiation." Even the author referred to by our critic, in discussing the two possible
methods of making the photometric determination in his investigation of the reflection
from mat surfaces, says in effect the same thing (cf. Wright, p. 205), so without exception
does every other author after whom we have read. Therefore when two mat surfaces
are observed whose areas are not limited, the apparent brightness of these surfaces is
the same for different angles of observation provided that the angle of incidence and
amount of incident lignt are the same for both surfaces as was the case for the stimulus
patch and measuring disc in our work for any one setting of the light on the bar; for
HETEROCHROMAT1C PHOTOMETRY 1 7 l
although the reflection from unit area decreases as the cosine of the angle of reflection,
the area from which the eye receives its light increases as the secant of the same angle;
from which it follows that the amount of light entering or reflected in the direction of
the eye is independent of the angle at which the surface is viewed.
It is obvious, then, that Dr. Johnson's statement that the percentage of incident
light reflected in the direction of the eye would have been different, even if the two
surfaces had been of the same material, is not true. From this it is equally obvious
that his next statement also is not true, namely, that the difference in the percentage
of incident light reflected in the direction of the eye is not constant for any two posi-
tions of the source, for as shown above there is no difference in the percentage of
incident light reflected to the eye from the two surfaces for any given setting of the
light on the bar. In other words, the possible bearing of Lambert's law and Wright's
results with regard to this law, is not what Dr. Johnson has stated it to be. Just
what this bearing is will be discussed further on in this note. What we wish to do
at this point is to show that even if it were true that the percentage of incident light
reflected to the eye were different for any one setting of the light on the photometer
bar, this would make no difference whatever in the results obtained by our method.
That is, if less light were reflected to the eye from the measuring disc than from the
stimulus patch for the first light set upon the bar, it would mean merely that the coef-
ficient of reflection of the measuring disc would have to be reduced by a corresponding
amount to obtain the match. Then when the comparison light was placed on the
bar and its distance adjusted until as much light was given to the screen as was re-
ceived from the first light, the stimulus patch and measuring disc would again match,
for neither the difference in angle of reflection to the eye nor the reflection coefficients
would have been changed. Dr. Johnson's point, granting its verity, would have appli-
cation only if the stimulus patch were illuminated alone by one of the lights and the
measuring disc by the other and the method of balancing consisted in bringing these
two surfaces to equality — then it would be necessary that each reflect to the eye the
same percentage of the light received by it; but the point is clearly quite irrelevant to the
method described by us in which the two surfaces are illuminated for each judgment by
only one of the lights, and the balance consists in so adjusting the distance of the two
lights in the successive judgments that the match for the one based on the amount of
induction produced at the stimulus patch holds also for the other. In this case it is
important only that the physical situation and other factors be kept the same for both
judgments — not that they be equal each to each for the single judgment — for the balance
is based on the principle that if all the factors are kept constant the amount of induction
at the stimulus patch will always be the same when the same amounts of light are re-
ceived on the screen. It is obvious also that the same considerations are true with
regard to the materials forming the stimulus patch and measuring disc. Moreover,
with reference to this point, it may also be said that there was, as a matter of fact, very
little difference in the materials forming the two surfaces; for one sector of the measuring
disc was identical with the stimulus patch and the other sector was a darker gray of the
same series of papers (Hering's series of standard grays).
In concluding our comments on this footnote which has revealed so much of our
critic's point of view, we will indicate briefly and only in a general way the relation of
Lambert's law of reflection from mat surfaces and Wright's findings with regard to this
law to the practical working of our method. As already shown, Dr. Johnson's criticism
was based both on an erroneous understanding of this law as applied to the making of
the photometric judgment by any method whatsoever and on a wrong conception of
172 C. E. FERREE AND GERTRUDE RAND
the principles of the method criticized. Our actual chance of error in terms of Lam-
bert's law is that the angle of incidence (Johnson's 'difference in angle of reflection'
has nothing whatever to do with photometry from mat surfaces) on the stimulus patch
and its surrounding field is different for the light from the standard and comparison
lamps when they are of different intensities and a different setting on the bar is required
to establish the photometric balance. That is, according to Lambert's law the intensity
of the illumination of the stimulus patch and its surrounding field is proportional to
the cosine of the angle of incidence (the cosine i ).
Now considering for the sake of simplicity the stimulus patch alone, the variation
in the cosine of the angle of the incident light for the entire range covered in the work
criticized from the least to the greatest distance of the source of light from the screen,
falls within I per cent. While this would mean only a comparatively slight difference
in the induction situation from the lights compared, we have from the beginning in
our own thinking frankly faced it as a small source of error in case the reductions of the
light on the screen are produced by changing the position of the lamps on the bar.
However, it would not enter in at all, as will be readily seen, if the reduction of light is
produced by means of a sectored disc or any device: absorbing screen, Nicol's prism,
grating, etc., which does not change the distance of the source of light from the screen
and, therefore, the angle of incidence of the light on the stimulus patch. In this
regard it should be remembered too that our photometer is no more at fault in physical
principle than the equality of brightness photometer after Rumford as ordinarily con-
structed, in which also the angle of incidence is changed with a change of the position of
the light on the bar — not so much at fault perhaps, for compensating factors operate
in our method of getting the balance which are not present in the Rumford method.
The relation of Wright's results to the situation described here is that he found that
there are certain small deviations from the law of the cosine i as the angle of incidence
is changed. Now just how great the chance of error is in our method from the law of
the cosine i considered in relation with the results of Wright it is utterly impossible to
estimate with any acceptable degree of precision from the principles involved for the
following reasons: (a) The surrounding field as well as the stimulus patch must be
taken into account in applying the law of cosines. The difference in the angle of in-
cidence for the different points in this field vary for any two positions of the light on the
bar — towards zero as a limit, for example, for the points between the stimulus patch
and the end of the bar, and differently in other directions. (£) The effect is not direct
but operates through induction, the quantitative relations of which are not definitely
known. And (c) Wright apparently considered it worth while to make no change of
angle of incidence smaller than 20°, while the entire range of variation of this angle in
our work from greatest to least distance of lamp from screen was for the colorless
light 2° and for both the colored and colorless lights 5°.
Rather, therefore, than indulge in bootless speculation in regard to the possibilities
of error from these sources, it is obviously much more to the point to get some empirical
measure of their effective importance. The effective importance of this factor along
with others not mentioned by Dr. Johnson may be checked up (a) by a comparison of
results in the average with those obtained by the equality of brightness method (see
table in original article, p. 9); and (b) still more definitely and directly by comparing
the results obtained by the method when the reductions of the light on the screen are
produced by changing the distances of the sources from the screen and when the dis-
tance of the source and, therefore, the angle of incidence of the light is kept constant
and the reductions are made by means of a sectored disc (see Table I. of this discussion).
HE TEROCHROMA TIC PHO TOME TRY 1 73
Even had these comparisons not been made, the probable relative unimportance of
these sources of error as compared with the high variable error obtained for one or any
small number of determinations by the equality of brightness method, should, we
think, be obvious to all who have a working familiarity with the latter method in hetero-
chromatic photometry. On the point of sureness of principle, moreover, it is instructive
to compare the agreements of the induction and equality of brightness methods shown
in the tables referred to above with those obtained for the equality of brightness and
flicker methods, for example, for lights presenting the same amount of color difference.
BRYN MAWR COLLEGE,
C. E. FERREE,
GERTRUDE RAND.
[The above discussion, which exceeds our usual limits, has been
accepted by the Editors in order that the authors might have ample
opportunity to clear up the points raised in Dr. Johnson's NOTE.
The questions at issue are so specialized and technical that we be-
lieve it unprofitable to continue the discussion in the pages of the
REVIEW. A committee of experts acceptable to both parties may
be suggested as the best means of settling any differences which re-
main between the writers and their critic. — THE EDITORS.]
THE STANFORD (1915) AND THE VINELAND (1911)
REVISIONS OF THE BINET SCALE
A brief analysis of the Stanford and Vineland revisions is here
attempted in order to indicate a few of the chief points of dif-
ference.
A hasty review of the Stanford revised scale gives one the
impression that it is much more difficult throughout.1 The exten-
sion of the scale to age 19% (superior adult)is certainly a commend-
able advance.
The scale begins at age three and each age contains six tests,
in addition to from one to three alternate tests for each year up to
age ten. The placing of six tests in each year permits assigning a
two months' value to each test. There are no tests for age n, the
only other ages listed being 12, 14, 16, and 18. Since there are no
tests for age n, the eight tests in age 12 are each made to count
toward three months mental age, yielding a total value of two
years. By this means tests for 14, 16 and 18, six per year, are
made to cover without break the range of mental development
from 12 to 19^.
The Vineland revision consists of 49 tests for the ages between
3 and 12. Covering the same period in the Stanford revision they
number 56, with 13 additional questions which may be used as
alternates.
In the following tables the evolution through which the Vine-
land revision passes is indicated test for test. Tests not in the
Vineland (1911) are printed in italics.
VINELAND (1911) STANFORD (1915)
Age— Test
Age III. Test I — Pointg. eyes, etc.. . .remains, .becomes III-i
" 2 — Rpts. 6 syll remains, .becomes III-6
" 3 — Rpts. 2 nos omitted
" 4 — Enumer. of pic remains . .becomes III-3
" 5 — Knows name remains . . becomes IH~5
Vineld.IV-i becomes.... III-4
" IV-2 " ....III-2
" IV-3 " ....III-A. i
1 See foot-note, p. 179.
'74
REVISIONS OF THE BINET SCALE 175
VINELAND (1911) STANFORD (1915)
Age IV. Test I — Knows sex. . shifted to III-4
" 2— Recog. key,
etc " " III-2
" 3— Rpts.snos.. " "III-A.i
" 4 — Comprs. lines remains . . becomes IV-i
Klmns. Form Discr.. . . = IV-2
Vineld.V-4 = IV-3
" V-2 =IV-4
Binet Compr. I deg = IV~S
Stanfd. 4 digits = IV-6
Vineld. V-3 = IV-A. I
Age V. Test i — Compares wts remains . . becomes V-i
" 2 — Copies sq.. . .shifted to IV~4
" 3— Rpts. ii syll. " " IV-A. i
" 4— Counts 4c... " " IV-3
" 5 — "Patience" remains, .becomes V-g
Vineld. VII-s = V-2
" VI-s = V-3
" VI-2 = V-4
" VI-3 = V-6
Binet, age = V-A. i
Age VI. Test I — A. M.-P. M remains . . becomes VI-A. i
" 2 — Definit'ns, use . . shifted to V~4
" 3— 3 direct'ns " "V-6
" 4 — R. hand, L. ear remains . . becomes VI-I
" 5 — Aesthet. Comp.. shifted to V-3
Vineld. VII-3 = VI-2
" VII-i = VI-3
" X-4(istser.). = VI-4
" X-i(part).... -VI-S
Stanfd. 16-18 syll = VI-6
Age VII. Test i— Counts I3c. . .shifted to VI~3
! 2 — Descr. pic remains . . becomes = VII-2
" 3 — Unfin. pic shifted to VI-2
" 4 — Copies Diamd remains . .becomes = VII-6
" s— Colors shifted to V-2
Binet, fingers = VII-I
Vineld. VIII-5 = VII-3
Stanfd. bow-knot = VII-4
Vineld. VIII-i = VII-S
" VIII-3 = VII-A. i
Stanfd. rpt. 3 no. bkwd. = VII-A. 2
1 76 SAMUEL C. KOHS
VINELAND (1911) STANFORD (1915)
Age VIII. Test I— Differences . shifted to VII-S
" 2 — 20-1, bkwds remains, .becomes VIII-2
" 3 — Rpts. days,
shifted to VII-A. i
" 4 — Counts stamps,
shifted to IX-A. 2
" 5 — Rpts. 5 nos.,
shifted to VII-3
Stanfd. ball-fid = VIII-i
Vineld. X-4 (2d. sr.) . = VIII-3
Binet, similar = VIII-4
Vineld. IX-2 = VIII-5
Stanfd. vocab. 20 = VIII-6
Vineld. X-i (part). . . . = VIII-A. i
Binet, dictation = VIII-A. 2
Age IX. Test I — Change 20-4 remains . . becomes IX~3
" 2 — Superior def . . shifted to VIII-5
1 3 — Date remains . . becomes IX-i
" 4— Months " .. " IX-A.i
" s— Arrange wts " .. " IX-2
Stan. rpt. 4 no. bkwd. . = IX~4
Vineld. X-$ = IX-5
" XI-4 =IX-6
" VIII-4 = IX-A. 2
Age X. Test I — Money (part) . . shifted to VI~5
(part)
shifted to VIII-A. i
(part) omitted
c 2 — Design remains . . becomes = X~3
" 3— Rpts. 6 nos " . . " = X-A. i
' 4 — Comprh. 1st. ser.,
shifted to VI~4
pt. 2nd. ser.,
shifted to VIII-3
pt. 2nd. ser remains . . becomes = X— 5
" 5 — Sentence shifted to IX-$
Stanfd. vocab. 30 = X-i
Vineld. XI-i = X-2
Binet, 8 memor = X~4
Vineld. XI-3 = X-6
" XII-3 = X-A. 2
Healy Constr. Puz = X-A. 3
Age XL Test I — Absurdity — shifted to X-2
" 2 — Simple sent. . " " IX~5
« 3_6o wds " " X-6 No XI
" 4— Rhymes " " IX-6
" 5 — Dissected sent. " " XII-4
RE V 1 SI 'ON S OF THE BINET SCALE
177
VINELAND (1911) STANFORD (1915)
Age XII. Test I— Rpts. 7 nos.,
shifted to XIV-A. i
: 2 — Abstract def remains, .becomes XII-2
" 3— Rpts. 23 syll.,
shifted to X-A. 2
' 4 — Line suggest omitted
" 5— Problems. . . shifted to XIV-4
Stanfd. vocab. 40 = XII-i
" ball-fid, sup. . . = XII-3
Vineld.XI-5 = XII-4
Stanfd. fables = XII-s
" rpt.sno.bk... = XII-6
Vineld.XV-i =XII-7
Stanfd. sim. 3 thgs = XII-8
Age XIV.
No TESTS FOR XIV
Stanfd. vocab. 50 = XIV-i
" induct, test = XIV-2
Vineld. Adult-4 = XIV-3
" XII-s = XIV-4
Stanfd. arith. prob = XIV~5
Vineld. XV-2 = XIV-6
" XII-i = XIV-A. i
Age XV. Test i — Interp. pic. . shifted to XII-7
" 2— Clock hands " " XIV-6
" 3— Code " " XVI-6
" 4 — Opposites omitted
Age XVI. (Average Adult)
Adult-
Test I— Ctg. paper. . . shifted to XVIII-2
" 2 — Reversed triang omitted
3— Diff. abstr. wds. shifted toXVI~3
4— Diff. pres. kg. shifted to XIV-3
5— Sense select. . " " XVIII-4
No TESTS FOR XV
Stanfd. vocab. 65 = XVI-i
" interp. fables.. = XVI-2
Vineld. Adult-3 = XVI-3
Stanfd. enclsd. boxs.. . . = XVI-4
" rep. 6 no. bk... = XVI-5
Vineld. XV-3 = XVI-6
Stanfd. rep. 28 syll.. = XVI-A. i
" comp.phys.rel. — XVI-A. 2
Age XVIII. (Superior Adult) Stanfd. vocab. 75 = XVIII-i
Vineld. Adult-i = XVIII-2
Stanfd. rept. 8 no = XVIII-3
Vineld. Adult-s = XVIII-4
Stanfd. rep. 7 no. bk... = XVIII-5
" ingen. test = XVIII-6
Summarizing the above tables:
In age 3, four tests remain and one is omitted.
In age 4, one test remains and three are shifted to an earlier age,
being too easy for four-year-olds.
i78
SAMUEL C. KOHS
In age 5, two tests remain and three are shifted to age 4, being too
easy for children of five.
In age 6, two tests remain and three are shifted to age 5.
In age 7, two tests remain, two are shifted to age 6, and one to
age 5.
In age 8, one test remains, three are shifted to age 7, and one to
age 9.
In age 9, four tests remain and one is shifted to age 8.
In age 10, two tests and a portion of a third remains, one test is
shifted to age 9, one part each of two tests is shifted to age 6,
and one part each of two tests is shifted to age 8.
In age n, all the tests are shifted, there being no corresponding
n-year group in the Stanford revision: two are shifted to age 9,
two to age 10, and one to age 12.
In age 12, one test remains, one is omitted, and three are shifted:
one to age 10, and two to age 14.
In age 15, one test is omitted and the other three are shifted as
follows: one to age 12, one to age 14, and one to age 16.
Of the "Adult" tests, one is omitted, one becomes a test for age
14, one a test for age 16, and two are tests in age 18.
The above changes are indicated in the following table:
TESTS
Shifted
Age
No Change
Omitted
Earlier
Later
Total
i
2
3
4
i
2
3
4
I
4
i
3
s
2
3
6
2
3
7
2
2
I
8
I
3
I
9
4
i
10
|
i
1
1
ii
—
2
2
I
12
i
I
I
2
IS
—
I
I
I
I
Adult1
3
I
1
Total
22i
41
19
5l
I
I
3
2
58
Percent. . . .
38.5
7-5
32.8
9-8
i-7
I.I
5-2
3-4
IOO.O
1(16-18)!
It will be observed that between 3 and 7 years twelve tests
have been removed to earlier years, and no tests to later years.
REVISIONS OF THE BINET SCALE 179
These changes will dispose of the criticism that the lower end of
the scale is too easy. But between 8 and 12 years 143 tests have
been removed to earlier years, and only five tests to later years,
thus making the scale still more difficult at its upper end.1
38.5 percent of the scale remains unchanged. 7.5 percent of
the tests in the Vineland revision are omitted. 45.4 percent of the
tests are found too easy for their respective ages and are shifted to
earlier years, 32.8 percent being placed in the next earlier age.
7.6 percent of the tests are found too difficult for their respective
ages and are consequently placed in later years.
No test is placed more than four years below its original position.
No test is placed more than two years above its original position.
It might be well to keep in mind that the tests appearing only
in the Stanford revision have not been considered in the above
tabulation and in the summarization of the test data. For that
reason some of the statements made need not be regarded as
seriously critical.
Using the Stanford revision, Terman and his collaborators
found that (a) by using the intelligence quotient one can transform
the 'age grade scale' into a * point scale' automatically, should one
prefer expressing the development of intelligence in that manner.
"As such it would seem to be greatly superior to the Yerkes-
Bridges scale, for it includes a much larger number of tests and its
points have definite meaning and equal value." (b) Sex-differ-
ences are found to be so small as to be negligible for practical
purposes, (c) The younger the children the greater the influence
of social status on intelligence.
The Stanford revision is to be welcomed in its effort toward a
scale free from those objections which are still being quixotically
hurled against it. SAMUEL C. KOHS
STANFORD UNIVERSITY
1 The following article, however, " L. M. Terman and H. E. Knollin: Some Problems
Relating to the Detection of Borderline Cases of Mental Deficiency" /. Psycho-
Asthen. 1915, 20: 1-15, coming to the notice of the writer after the above was in
type, shows the reverse to be true. Tabulating the reactions of borderline subjects
(mental ages by the Stanford Revision between 12 and 14, — 104 adults) they found
that by the Vineland Revision, weighted for tests above 12, the median age for these
subjects was reduced as much as one and one-half years, and with the tests unweighted
the reduction was greater, namely two years. It ought also be mentioned, in this
connection, that the procedure and scoring of quite a number of tests have been
changed in the Stanford Revision. Consequently a strict analysis of test displacement
must take these facts into consideration. Change of procedure or scoring may so
modify the statistical data obtained for a test as to warrant its transfer to some lower
year without necessarily increasing the difficulty of the scale at that particular point.
VOL. XXIV. No. 3 May, 1917
THE PSYCHOLOGICAL REVIEW
THE NATURE OF MENTAL PROCESS
BY HARVEY CARR
University of Chicago
This paper proposes the somewhat unorthodox view that
the mental functions with which psychology concerns itself
are in reality psychophysical, and at times neural, activities^
and that psychology shall study and attempt to comprehend
these functions in their entirety. The author adopted this
conception of the nature of mental process several years ago
and is convinced from his teaching experience that such a
mode of treatment possesses certain distinct advantages.
The conception may be contrasted with the more ortho-
dox * subjective' view which postulates psychophysical paral-
lelism but which confines its efforts exclusively to a compre-
hension of the conscious or subjective aspect of these psycho-
physical events. Such a psychology studies color, sound,
taste, and pain as experiential results but contends that the
neural correlates of these sensory experiences belong to the
domain of physiology. An emotion as a psychological phe-
nomenon is described and defined in terms of sensational and
affective processes subjectively regarded; the neural events
involved in an emotion are relegated to the domain of physi-
ology. The acts of memory, imagination, reasoning and will
in so far as psychology is concerned with them consist merely
of those aspects which can be immediately experienced; the
neural events involved in the acts are quite important but
their consideration involves a trespass into domains right-
fully belonging to another science. Psychology thus deals
exclusively with the purely conscious or psychic as opposed'
to the nervous and material- Psychology is differentiated
181
1 82 HARVEY CARR
from the material sciences in virtue of a peculiar subject
matter and a peculiar method of apprehending its data.
Our conception also accepts psychophysical parallelism as
a working hypothesis; it contends that psychology shall study
psychophysical processes in their entirety, and that it shall
include within its domain activities which lie outside the field
of consciousness. Psychology will study emotions, and acts
of reasoning, memory and will, but it will define and envisage
these acts as psychophysical processes, and attempt to com-
prehend the neural events involved as well as those aspects
immediately experienced.
The conception allows a division of the field of organic
functions between the sciences of biology, physiology and
psychology along natural lines of cleavage based upon differ-
ences of interest, training and technical procedure. Irre-
spective of definitions, psychology has been concerned with
an ultimate comprehension of those operations by which an
organism in virtue of its previous experience is enabled to
adapt itself to a complex and variable environment. Phys-
iology and biology have been interested in other types of
functional activity — different in character, evolutionary his-
tory, and biological significance. The three sciences represent
distinctions which appeal to radically different types of human
motive and scientific interest, and which necessitate different
sorts of training and technical equipment. This paper is not
concerned with the formulation of an exact definition de-
limiting the boundaries of the three sciences.
The conception is unorthodox only in relation to prevailing
definitions of psychology. To my mind it is essentially in
harmony with the dominant point of view of the science, and
it is not wholly inconsistent with much of current practice.
Sciences have a way of developing and outgrowing their def-
initions. Practice and attainment often fail to square with
theory and definition. The subjective conception of the
nature of mental process (mental as opposed to material)
originated from philosophical interests at a time when a dual-
istic conception of the human organism prevailed. At
present the prevailing point of view in the science is biological,
THE NATURE OF MENTAL PROCESS i§3
— a view which emphasizes the essentially unitary character
of the human organism. The science is interested in certain
modes of adjustment, and any adequate understanding of
these processes of adjustment must necessitate a compre-
hension of the acts in their entirety. A division of any act of
adjustment into its material and conscious aspects with the
consequent treatment of but one component certainly gives a
very inadequate comprehension of the phenomenon in ques-
tion, and introduces a distinction which is not only without
value, but which is likely to involve the student in many dis-
tracting perplexities. Our proposition, however, will neces-
sitate no radical changes in much of current modes of pro-
cedure. As a matter of fact, many psychologies treat mental
operations as psychophysical processes with only occasional
lapses into a consistency with their subjective definitions of
the science. It is our definitions that need revision, a
revision in harmony with current tendencies and ideals.
The conception allows of a matter of fact treatment of the
cause and effect relations in mental activity. One can assert
that behavior is influenced by previous acts of memory or will
and mean exactly what we say and what the unsophisticated
mind understands by the statement. Psychologists will not
be compelled to add for the benefit of the sophisticated quali-
fying phrases to the effect that although they asserted a
causal influence of a mental act, yet they really did not
mean it, but were forced to employ such statements by certain
inadequacies of language. Any conception which allows of a I
natural and matter of fact treatment of the causal category in I
mental operations is at least deserving of respect. The new'
definition of the mental will permit a restatement and a
solution of the mind-body problem more in accordance with
common sense. Interactionism is logically possible; in
accordance with popular belief we may say that our mind is
influenced by bodily conditions and that our mind is also an
effective influence upon bodily activities, for mind and body
have been so conceived and defined in relation to each other
that such statements are no longer logically or factually im-
possible. With our definition the distinction of mind and
body is merely a distinction of two systems of organic function.
184 HARVEY CARR
Our definition will include within the domain of psychology
I the non-conscious components of mental life. I refer to such
phenomena as retention, memory disintegration, conflict of
impulses, Aufgabe, unconscious motivation, the concept of
habit, and the wealth of subterranean activities brought to
notice by the investigations of abnormal psychology. The
subjective psychologies have assumed several attitudes toward
these phenomena. Some are logically consistent with their
presuppositions and attempt to ignore such intruders. Others
admit the significance of these data for their purposes, but
consistently remind their readers that after all these phe-
nomena really belong to the domain of physiology. Others
include these events within the domain of their science, but
feel compelled by motives of consistency to impose upon these
processes some sort of a * conscious label.' Witness such
terms as unconscious, subconscious, co-conscious, psychical
dispositions, etc. These are not merely negative terms,
equivalent to neural; they have a positive significance.
Since these activities have a mental significance and since the
mental must also be conscious, these acts must be conceived in
such a way as to possess certain positive characteristics of
conscious process. For my part, I prefer the remaining
possible mode of procedure, viz., to revise my definition of the
nature of mental process in such a way as to include such data.
The subjective conception of mental process as something
immaterial constitutes an inadequate tool for the physician
in his attempt to comprehend the nature of the mental, or
functional disorders. Watson, in a recent article,1 has as-
serted his inability to understand the medical concept of
mental disease. He cites a case which was diagnosed as
being ( purely mental,' and which was described and defined
wholly in conscious terms. Watson gives the impression
that the physician was of the opinion that this disorder could
not in any manner be stated in neural terms, that it was a
disorder exclusively on the conscious plane without neural
counterpart. Such a conception of the nature of the so-called
mental diseases is of course foreign to current psychological
1 'Behavior and the Concept of Mental Disease,' /. of Phil., Psychol., &c.t 1916,
13, 589-
THE NATURE OF MENTAL PROCESS i§5
doctrines, and I doubt very much that such a view is uni-
versally prevalent in the medical profession. I am willing
to admit that this and similar crude and preposterous con-
ceptions are to be met with, but, unlike Watson, I am inclined
to place the blame for this unfortunate state of affairs upon
psychology rather than upon medicine. Medicine has merely
adopted current conceptions. Psychology must be held re-
sponsible forthe factthat betterand more adequate conceptions
were not available. Given a conception of mind as something
non-nervous and non-material, crude notions of the nature of
a mental disease must result. The mental will be distin-
guished from other diseases in terms of immaterial vs. ma-
terial instead of functional vs. organic. The conception
needlessly introduces into the discussion of mental disease
and its treatment the old philosophical question of the
relation of mind and body stated again in terms of the con-
scious vs. the physiological, or neural. Certainly the in-
jection of such philosophical questions in discussions of the
treatment of mental disease adds nothing of any positive value,
introduces a perplexing distraction, and to my mind is wholly
unnecessary. The old philosophical problem vanishes at
once if we start out with the assumption that the disordered ^-
mental functions are in reality psychophysical events.
The psychophysical conception of mental process offers a \
mediating point of contact for the two extremes of subjectiv- \
ism and behaviorism. Such a view permits the widest
latitude as to methods of approach; it permits mental processes
to be studied from the standpoint of immediate experience, of
objective observation, or of clinical data. This suggestion
of a common ground for the subjectivists and the objectivists,
I am well aware, will evoke no approval from either of the
warring camps. This phase of the argument is designed
exclusively for the benefit of the neutrals. Our program will
differ from that of the subjectivists in allowing an objective
mode of approach to the problems of psychology; it will differ
from behaviorism in two respects: it admits that the study of
conscious data has given us much useful knowledge of the
nature of mental operations and that further progress is
1 86 HARVEY CARR
possible in the future. Behaviorism as defined logically
includes the whole field of organic function. Psychology
should be content with a more modest program and make a
more reasonable division of the field of functional activity
among the sciences of psychology, biology, and physiology.
The conception will modify our attitude toward the pur-
poses and methods of comparative psychology. The sub-
jectivist to be consistent must define the object of comparative
psychology as a reconstruction of the inner life of an animal.
Such a program appeals to some minds; to other minds it is
repellent. The latter are impressed with the insuperable
difficulties and limitations of the interpretative process, and
are inclined to regard any results achieved as somewhat futile
even if logically valid. Our conception makes possible for
those who prefer it a purely objective or behavioristic science
of comparative psychology, and yet permits others to recon-
struct the conscious life of organisms if they so desire.
An exclusively subjective psychology is prone to meet
certain needless difficulties in its presentation. The psycho-
logical discussion of a mental -event is usually followed by an
explanation in terms of neural mechanisms which are labelled
physiological. The psychological process viewed by itself
impresses the mind of the student as a peculiarly inert and
unimportant thing. When the neural mechanism is now
added, the process takes on life and significance; mental
processes can now do things and be effective instruments of
organic adjustment. Moreover, the neural chain of events is,
by hypothesis, more complete than the conscious aspect for
not all neural events are represented in consciousness. It is
small wonder that many students decide that they must
look to physiology for any complete and real explanation of
mental life. This sceptical attitude of the student toward
psychology is not wholly unjustified, nor is it confined entirely
to immature students; as a matter of fact this attitude toward
our science is quite prevalent among the physiologists. I do
not contend that it is impossible to meet such a position in a
satisfactory manner; the significant point is that this atti-
tude of scepticism is wholly unnecessary and exceedingly
THE NATURE OF MENTAL PROCESS 187
detrimental. No science can afford to be placed in a de-
fensive and apologetic position. The dual presentation
further raises such distracting questions as the relative values
of the neural and conscious components in an act of adjust-
ment, their causal interrelations, and the necessity for a
double treatment of mental events. All these are valid
problems but they are philosophical in character and have
no place in an empirical, matter of fact, science. Psychology ,
generally recognizes the philosophical character of these
questions and their tendency to distract the mind of the
student from the main task at hand, and the usual method of
attempting to ignore them is by the adoption of parallelism
as a working principle. To my mind this method fails utterly
to achieve its purpose. Such philosophical questions must
necessarily obtrude with a dualistic mode of presentation; in
fact this method was developed at a time when psychology
was studied primarily as an introduction to philosophical
problems, and a better method for this purpose would be hard
to devise. These difficulties are at least minimized if not
eliminated by adopting the conception of this paper. Paral-
lelism is adopted as a working hypothesis in a matter of fact
way without calling the student's attention to it. The total
activity is made the object of study; the dichotomy involved
is not one of process, but one of method of approach or appre-
hension.
In conclusion we may say that there are no fixed and im-
mutable boundary lines between sciences. Any science in-
cludes within its domain whatever is pertinent to its primary
interest. If mental acts are means of organic adjustment,
then these acts must be conceived and studied as such devices.
If neural events are essential parts of the act, the conception
of the mental must be broadened so as to include them. *
There is no a priori necessity for defining the mental in non-
material terms, except habit and the force of tradition. The j
concept of the mental must be adapted to further the primary j
task of psychology as it is now conceived.
A REFORMULATION OF THE LAW OF
ASSOCIATION
BY WALTER S. HUNTER
University of Kansas
I
The reformulation that this paper undertakes is to render
explicit certain facts and points of view that the writer finds
implicit in current psychological thinking. The 'law of asso-
ciation' as formulated and discussed by British psychologists
concerned itself with ideas and sequences of thoughts and
not with sequences of conscious states in general (which
would have included sequences of sensory data). Such was
the case with Aristotle also and with the intervening writers.
Observation was warped primarily by the prevailing interests
in logic and epistemology and by the tendency to overestimate
the importance of vision in consciousness.
Aristotle says, e. g.: "Hence, when we are recollecting we
keep stimulating certain earlier experiences until we have
stimulated one which the one in question is wont to succeed.
And just so we hunt through the sequence, thinking along
from the present or some other [thought], and from similar or
contrasted or contiguous."1
In Berkeley's writings much use is made of association, or
suggestion as he terms it. And it is particularly to be noted
that sequences of sensations as opposed to sequences of ideas
are produced by the Deity. In David Hartley there is a
thoroughgoing physiological as opposed to a logical associ-
ationism. Hartley's classical formulation is as follows: "Any
sensations A, B, C, etc., by being associated with one another
a sufficient number of times, get such a power over the cor-
responding ideas a, b, c, etc., that any one of the sensations A^
1 Quoted from H. C. Warren, 'Mental Association from Plato to Hume,' PSYCHOL.
REV., 1916, 23, p. 210.
188
THE LAW OF ASSOCIATION 189
when impressed alone, shall be able to excite in the mind
by r, etc., the ideas of the rest."1 Hartley's physiological
interests led him to apply the law specifically and in detail
to habit formation. Obviously here the cases described —
particularly the speech habit (see below), — must furnish data
in harmony with our present contention; but Hartley never
realized it. Had he or other writers done so, they must have
recast their general law.
The point of view of these writers and even their formu-
lations of the law have persisted into current writing in spite
of the fact that a constantly increasing growth in the knowl-
edge of habit and thought has been making the conventional
law inadequate. The following quotations from current
textbooks will indicate the present general attitude.
In Calkins2 we find : " Successive association is the sequence
of an imagination on a perception (or another imagination), a
sequence which is attributed (in after-reflection) to the
previous occurrence, simultaneously or in swift succession,
of the two experiences." This is repeated in the table on
page 359-
Titchener3 says: "We then find this: that whenever a
sensory or imaginal process occurs in consciousness, there are
likely to appear with it (of course in imaginal terms)4 all those
sensory and imaginal processes which occurred together with
it in the earlier conscious present. This we may term the
law of association."
Angell5 formulates the matter as follows: "The law of
association asserts that whenever two images, or ideas, have
been at any time juxtaposed in the mind, there is a tendency,
if the first of them recurs, for the other to come with it."
Judd and Thorndike state the law in a very general form which
could be interpreted to cover our present point; but the con-
text in each case indicates clearly that it is the old view of the
1 Priestly, Jos., 'Hartley's Theory of the Human Mind.' London, 1775, P- I4>
Prop. 5.
2 Calkins, M. W., 'First Book in Psychology/ 4th Ed., 1914, p. 116.
•Titchener, E. B., 'Textbook of Psychology,' 1910, p. 378.
4 Italics mine.
6 Angell, Jas. R., 'Psychology,' 4th ed., 1908, p. 206.
190 WALTER S. HUNTER
nature of the second term of the association which is in their
minds. The present thesis receives no consideration.
Since the Greeks it has been recognized that the principle
of association is the principle of habit formation. But the
development and interrelations of the two thoughts has ended
there because of the interests above mentioned. At the
present, behavior studies and the critical literature on thought
afford us the ground for and even require a recasting of the
doctrine of association. The revised version needs, I believe,
but to be stated in order to carry conviction in many minds:
// A and B are experienced together in space or time and if later
one is experienced either in sensory or in imaginal form, it
tends to arouse the other either in sensory or in imaginal form.
In other words the second member of an association may be
and often, if not usually, is a sensory process. By the term
sensory process I would include both sensation and perception.
The essential point is that the conscious state which forms
the second term of the association is often conditioned by a
present and on-going peripheral (sensory) activity. It is
passing strange that this has not been explicitly stated and
systematically incorporated before now. Ever since Eb-
binghaus's tests on nonsense syllables * associations' have
been studied which have involved the reproduction or recall
of sensory material (auditory- vocal-motor). Even sub-lim-
inal associations have been tested by the saving method and
otherwise with the constantly present purpose of making
possible the recall in sensory form of certain material. The
present formulation does not attempt to explain the fact of
association, nor does it attempt to analyze the factors in-
volved in the formation of associations (habits). It proceeds
on the assumption that fundamentally 'associationism'
cannot be outgrown and left behind any more than neural
habits and the setting of synaptic connections can be so
treated. It is therefore important that the 'law' be made
adequate and general.
II
When psychologists from Hartley to the present time have
attempted to sketch the neural side of the 'law of association'
THE LAW OF ASSOCIATION I91
they have talked in brain terms and not in terms of the nervous
system as a whole. This has followed by virtue of the as-
sumption that the second term of an association must be an
image (or a centrally aroused process). Hear James1 on
the neural side: "I shall try to show, in the pages which
immediately follow, that there is no other elementary causal
law of association than the law of neural habit. All the
materials of our thought are due to the way in which one ele-
mentary process of the cerebral hemispheres tends to excite
whatever other elementary process it may have excited at
some former time. The number of elementary processes
at work, however, and the nature of those which at any time
are fully effective in rousing the others, determine the char-
acter of the total brain action, and, as a consequence of this,
they determine the object thought of at the time. According
as this resultant object is one thing or another, we call it a
product of association by contiguity or of association by
similarity, or contrast, or whatever other sorts we may have
recognized as ultimate." "Let us then assume as the basis
of all our subsequent reasoning this law: When two elementary
brain-processes have been active together, or in immediate suc-
cession, one of them, on reoccurring, tends to propagate its
excitement into the other." In the last sentence James indi-
cates that the 'law of neural habit,' essential for association
to his mind, is one dealing with central processes.
All that this does is to read over into neural terms the
faulty formulation of the law as indicated above. Intro-
spection can at the best only note the sequences of conscious
states. If these can be traced to a prior contiguity, they are
cases of association. On the basis of the introspective report
on the nature of the associated contents, one can then proceed
to formulate the neural changes involved. The following
illustrations make clear that such neural changes involve far
more than the cortex, — that we are dealing with peripherally
and not with centrally aroused processes: I see (A) and
touch (B} an object which results in my drawing back (C),
trembling (D), screaming (E), or swearing (F). Later I see
1 James, Wm., 'Principles of Psychology, 1890, vol. I., p. 566.
192 WALTER S. HUNTER
(A) or touch (B) the object (or do both) and at once C, Z), Ey
or F appear. And they may and do appear in sensory form.
Here belong the so-called cases of emotional and affective
memory which were so much discussed some years ago. I
see (A) an object and it gets bound up with unpleasantness
(B). Later when I have A or its image ay the unpleasantness
returns as B in the concrete with all its organic and motor
changes. Or again I see a word and hear it at the same time,
later I see or think of the word and the sound is again pro-
duced by my vocal organs. It is hardly necessary to point
out how in these cases the nervous impulses shuttle back and
forth between the brain and the periphery of the body. Cen-
tral arousal and images do not monopolize these situations.
On the conscious side we have had a succession of sensations
whose connection reflection can trace to a prior contiguity.
This is what we mean by "mental" association and this is
where the law starts or gets its basis. Neurally we have had
widespread activity unconfined to the hemispheres.
Current writing is emphasizing the relation of language and
the vocal processes to thought. It seems clearly established
that thinking can go on without detectable imagery. The
'meanings' to most psychologists (and I agree) are not 'pure
thoughts,' but accompany, among other things, the kinsesthe-
tic sensations (sensory processes) of the vocal organs. These
language cases are functionally the most important instances
where the second member or succeeding members of an as-
sociative train are sensations (as I should say), or peripher-
ally initiated processes (as others would say).
It should not be objected that with the increasing auto-
matism these peripheral processes become short-circuited and
drop from consciousness. The same could be said of the
imaginal members of the association possibly even more
truly. It is doubtlessly the case that with both sorts of
material the meaning is usually the clearest and most promi-
nent feature; but that again is a point for any formulation of
the law of association and not alone for the present one.
The essential thing for us here is that the meaning can be
carried, and is carried, as well by sensory as by imaginal proc-
THE LAW OF ASSOCIATION 193
esses and that by virtue of the internal control of their
stimuli these sensory processes may constitute the second
terms of an associative sequence.
Nor should it be objected that the present account fails
to consider simultaneous association or persistent association,
where one part of an associated complex persists over into
the succeeding moment. The explanation of such cases
need be no easier with images than with sensory processes.
The play of muscles from moment to moment offers perhaps
the clearest illustration of this type of event. Muscles i-io,
let us say, are active at a given moment and by their activity
lead to the contraction of muscles 8-15. A common element,
muscles 8-10, is found in each activity. If we speak from the
conscious side, we have a core of kinaesthetic (and cutaneous)
sensory experiences persisting through two successive mo-
ments of time. Practically this situation is always present
and gives us our feeling of bodily presence and of the con-
tinuity of the self. Here again our account could be much
more complete if we could formulate the causes of the setting
of associations (synaptic connections). Such incompleteness,
however, does not affect our essential point, nor is it a dif-
ficulty peculiar to the present thesis.
Ill
Sensory processes have their functional significances in the
responses that they initiate. And so, stated the other way
round, what adaptive movements shall take place depends
very largely upon the sensory processes available. Certain
of these sensory processes are practically beyond the organ-
ism's control. Certain others on the contrary are very much
within the control of certain organisms and to a variable
degree within the control of all. An animal's control of its
own movements or responses is to a large extent dependent
upon the sensory processes that it can itself initiate. Self-
initiated stimuli in most animals are those of hearing, cuta-
neous sensitivity (pain excluded?), and kinaesthetic, organic
and static sensitivity. Those that are not self-initiated are
taste, smell, and vision. The last statement, of course, needs
194 WALTER S. HUNTER
much qualification. Many animals produce odors and others
have phosphorescent organs which emit light. The same may
be true also of taste, particularly with aquatic animals. In
man the qualification that I feel it most necessary to urge is
that of visual sensations produced by intention and probably
initiated by the efferent optic fibers.
Wherever we find sequences of habits, there we should
expect to find stimuli under the organism's control. Other-
wise in order to have a habit sequence such as speech, running
a maze, or solving the delayed reaction, it would be necessary
that some non-controlable stimulus initiate each link in the
response. But the contraction of one set of muscles stimu-
lates receptors which arouse the succeeding muscular activity.
If this sequence is essentially constant and if the receptors
condition consciousness, we have a typical case where an
associative sequence is sensory in its entirety. Writers on
association at the time the law was fixed upon psychology
were too fascinated by vision (and logic) — and vision is a
sense whose stimulus (light) is practically non-producible by
man's own activity. What more natural then than that in
formulating a rule of sequences, they should have said that
the second terms of associations could not be sensory but
must be imaginal. Indeed the last sentence is misleading.
The question was not clearly enough considered for them to
use the terms 'could not be' this or 'must be' that. But
this error, I believe, proceeded from tacit assumptions as I
have indicated. And the same assumptions, tacitly made,
still persist and lead to our conventional formulations, al-
though such formulations are as wide of the mark as I have
tried to indicate. It is true that where writers come to con-
sider kinsesthetic habits, e. g., they give essentially the account
that I have sketched above. But at the same time they forget
that so far as the conscious side is concerned they are dealing
with genuine cases of the law of association; and when for-
mulating that law, they forget the other situation. Nowhere
do they bring the two into working union. Suggestive state-
ments are alone to be found. I cite three instances:
First, Hartley's account1 of the growth of the speech habit:
1 Priestley, op. cit., pp. 33-34.
THE LAW OF ASSOCIATION 195
"About the same time . . . the muscles of speech act occa-
sionally in various combinations, according to the associations
of the motory vibratiuncles with each other. Suppose now
the muscles of speech to act in these combinations at the same
time that sound is produced from some agreeable impression,
a mere sensation, or a slight associated cause, which must be
supposed to be often the case, since it is so observable that
young children, when in a state of health and pleasure, exert
a variety of actions at the same time. It is evident, that an
articulate sound, or one approaching thereto, will sometimes
be produced by this conjoint action of the trunk, larynx,
tongue, and lips; and that both these articulate sounds, and
inarticulate ones, will often recur, from the recurrence of the
same accidental causes. After they have recurred a sufficient
number of times, the impression which these sounds, articu-
late and inarticulate, make upon the ear, will become an
associated circumstance (for the child always hears himself
speak, at the same time that he exerts the action) sufficient
to produce a repetition of them. And thus it is that children
repeat the same sounds over and over again, for many suc-
cessions, the impression of the last sound upon the ear exciting
a fresh one, and so on, till the organs be tired."
Miss Calkins also comes near making the essential point
when she points out that any state of consciousness plus a
feeling of generality is a concept. (". . . when I think of
fear, the consciousness of generality accompanies a genuine,
though certainly faint, experience of that emotion."1) Now
since thinking consists in a sequence of concepts, or their
physiological equivalents, she should have added that it is
possible to have a sequence of internally controlled but peri-
pherally initiated states of consciousness. And this would
have forced her to re-state the law of association whose
phrasing I have quoted above.
James, too, came perilously near discovering the point
that I am making. There is one sentence in the very chapter
on Association that gives a beautiful illustration of the fact
that the second term of an association may be sensory.
1 op. dt., p. 147.
196 WALTER S. HUNTER
". . . probably every one who bathes himself in a certain
fixed manner is familiar with the fact that each part of his
body over which the water is squeezed from the sponge
awakens a premonitory tingling consciousness in that portion
of skin which is habitually next to be deluged."1 Surely only
the momentum of historical usage and the absence of the
new psychology of thought could have prevented James from
correlating his discussions of Habit, Association and Will with
reference to this point!
I would once more call attention to the semi-behavioristic
background of the present discussion. Studies of the delayed
reaction in young children and in raccoons (possibly also in
dogs2) have indicated the functional presence of internal
factors which can initiate responses in the absence of the
accustomed external stimulus. These internal factors may
theoretically be either centrally or peripherally aroused
processes which are called up by the sensory stimulation due
to operating the release box of the apparatus. Elsewhere3
I have given reasons for deciding in favor of the peripheral
processes. The importance of the appearance in the animal
series of this ability to control adaptive muscular response by
internal stimuli which need not be constantly present can
hardly be overestimated. Man's language sequences are
but the development of the animal form of sensory associa-
tions— a development which is centralized more and more in
the vocal apparatus.
1 James, op. cit.t p. 555.
2 Walton, A. C, 'The Influence of Distracting Stimuli during Delayed Reaction in
Dogs.' /. of Animal Behav., 1915, 5, 259-291.
3 Hunter, W. S., 'Delayed Reaction in Animals and Children,' Behav. Mon.,
1913, 2, No. i.
THE SCIENTIFIC PRODUCTIVITY OF AMER-
ICAN PROFESSIONAL PSYCHOLOGISTS
BY SHEPHERD IVORY FRANZ
Within the past few years there have appeared reviews of
the progress of psychology for different periods of time.1
That general progress has been made is evident. That in
America progress has been made is shown by the inauguration
and rapid increase in the membership of the American Psy-
chological Association, by the foundation of journals devoted
to the publication of psychological discussions and researches,
by the appearance in numbers of psychological books and
articles, by the great popular interest which has been aroused,
by the birth and growth of numerous laboratories for teaching
and investigation, by the creation of special university chairs
apart from those for allied disciplines (especially philosophy
and education), and by the establishment of special research
laboratories for applied psychology. It is also pertinent to
remark that during this developing period of psychology there
have been begun and continued several lines of investigation
which, although not entirely American, have continued in
mass and in importance to be due to workers in this country.
Although there is abundant evidence of advance, those
historians who have recounted the progress have dealt with
their theme in an impersonal way. They have reported the
number and the character of the published investigations,
the establishment of independent departments, and the
number of conferred doctorates. They have not dealt with
an equally important subject which is germane to the one
regarding which they have written. We have not been
informed by whom the psychological advances have been
made, or whether or not in view of the increasing number of
1 See C. A. Ruckmich, 'The Last Decade of Psychology in Review,' PSYCHOL.
BULL., 1916, 13, 109-120. This contains references to previous reviews of like char-
acter.
197
198 SHEPHERD IVORY FRANZ
professional psychologists there has been a corresponding
increase in the number or in the value of the published in-
vestigations. In other words, although it is admitted that
advance has been made, we are as far from knowing whether
or not the advance has been satisfactory and corresponds with
the number of psychologists.
The estimation of the value of an individual's contributions
has been attempted from time to time and in a variety of
ways. There is the well-known attempt to grade psychol-
ogists by votes of a few selected individuals. The general
results of this grading have been reported.1 There has also
been a negative grading of psychologists in general in the
elections to the National Academy of Sciences. Of the five
so-called psychologists who have been members, three at the
time of their election were professors of philosophy and
primarily interested and concerned in the teaching and pub-
lication of philosophical (ontological, epistemological, and
logical) doctrines. There has also been a continuing selection
(and promotion) of men for professorships and other academic
positions, which selection acts as a grading of a less obvious
kind. The last kind of selection brings about two adverse
conditions, one of them being that an individual selected for
a position because of time-serving or personal recommendation
has automatically acquired a claim to a grade to which his
previous position and work did not apparently entitle him,
the other being that because of this there has been a consider-
able amount of time-serving and inbreeding in some of our
institutions.2
All of these methods of estimating the relative ranks of
individuals have obvious and with respect to even their tem-
porary value perhaps vital defects. Much the same may be
1 'American Men of Science.' Ed. by J. McK. Cattell. 2d ed. 1910. See es-
pecially pp. 537-596. The results have been given in only a general way; the names
of the judges are not mentioned, and we must depend upon the editor's word that they
were competent.
2 While it may not apply to psychologists it is not unknown in scientific circles
that some individuals, because of personal charm or characteristics such as self-appre-
ciation and a hypercritical attitude toward others, have been able to attract students
and have been able to persuade others of their importance and to get a self-enhanced
reputation spread.
PRODUCTIVITY OF AMERICAN PSYCHOLOGISTS *99
said regarding methods now in use for the determination of
the value of a published research or other contribution. If
the different methods were used in combination and subjected
to safeguards respecting individual interests they would
probably give better estimates than have hitherto been made.
To take individuals at a given time and settle upon values or
grades may be necessary, but it should be recognized that the
grades shift at different times. The same holds true with
regard to estimated grades or values of performed work. The
supposed values of the nineteenth century do not hold for the
twentieth, nor do those of today hold for tomorrow. It is,
however, possible to determine whether or not a given in-
dividual or a group has carried out some of its functions, and
to determine whether or not certain individuals have fallen
short or have done more than their normal or expected share.
We may assume as proven that in psychology there has been
satisfactory advance in general, both in character and in
quantity of the work, and we may inquire: "By whom have
the advances been made, and in what manner have individuals
or groups contributed to the advance?"
It has already been said that if we wish to deal with ab-
solute value it is not possible to make satisfactory judgments,
and answers to the questions could not be given. No one
psychologist has sufficient knowledge of methods and results
in all branches of psychology to be considered a sufficiently
expert judge. Nor has any one the confidence of all or of a
majority of psychologists. Each judge of values is influenced
in making his judgments by considerations of training, of
knowledge, and perhaps of special personal interest. There
is no absolute impartiality. The introspectionist does not
hesitate to say that behavior studies are not truly psycho-
logical. The behaviorist may reply that there is no such
thing as the introspection that is talked about so much. He
who does not hesitate to teach and write about the functions
of the parts of the nervous system may never have performed
or seen any of the classical cerebral experiments; by some
psychologists the realm of the abnormal has never been visited
and by them it is known only like uncharted parts in a geog-
200 SHEPHERD IVORY FRANZ
raphy; and the remainder of applied psychology is for many
like a prostitute whose acquaintance is not desired and who is
to be shunned because of fear of infection.
Notwithstanding the limitations of value of individual
opinion, much might be gained by taking a consensus of
opinion of those of divergent special interests and of those
who are admitted to be fair jurymen.1 But we can also do
something very definite by determining that a certain in-
dividual has or has not made any published contribution
towards psychological advance.2 This is a comparatively
easy method giving positive results. It admits of little or
no discussion of a judge's partiality, it rests solely upon the
admission of published material as the facts with which to deal.
We may also determine that one has contributed by books,
or by the publications of investigations, or by any other means
which we may select. And there is also the possibility of
answering the question, "Has the progress, as measured by the
number of publications, corresponded with the number of
individuals who have become professional psychologists?"3
1 Not related academically to the individuals to be judged, supplied with all the
data respecting the individuals (not with only a part as was done in Cattell's 'American
Mec of Science' classification), and of sufficiently diverse individual interests to form
a jury representative of all branches of psychology.
2 There is little possibility of determining the value to psychology of the indi-
vidual's college or university teaching. Some indication might be obtained by the de-
termination of the academic antecedents of psychologists and of those who have taken
advanced courses in psychology but who have not become professional psychologists.
This does not mean the determination of the number of conferred doctorates. It is
often forgotten that the interest of the individual has been created before he has known
of the possibility of advanced work, and the doctorates give in the main only a clue
to the institutions which are recommended by those who have created the interest, or
which are selected because of material advantages. Another method of determining
the value of the individual teacher would be that of estimating the average value and
the amount of published work which his students put forth after leaving him. Not all
of their subsequent work is due to his influence but an estimate could be made.
3 The consideration of these matters has been somewhat forced upon me in con-
nection with editorial duties during the past few years, since it was necessary to know
what lines of work were being investigated and by whom. During a longer period of
time I have not infrequently been asked to offer suggestions regarding possible candi-
dates for positions, and having had very few students fitted for such positions, I have
taken the opportunity to recommend those who have exhibited some accomplishment
(publication) rather than those who were known to me as individuals of 'promise' in
their advanced work. The results of some of these casual inquiries regarding accom-
plishment were so different from what I had expected that it led to the consideration
of a larger group.
PRODUCTIVITY OF AMERICAN PSYCHOLOGISTS 201
We have available for the purpose well-known yearly bib-
liographies, and the present article has been made possible
because of them. The membership list of the American
Psychological Association was consulted to obtain a list of
our * professional' psychologists. Since membership in the
Association is not limited to those who are instructors or
professors of psychology, only those whose official positions
consisted solely in relation to psychology were considered to
be the professional psychologists.1 Since those few who hold
only research positions can not be dealt with in the same man-
ner as teachers, I decided to omit their names from the list
to be investigated. This left 87 names, of which three were
omitted because of the lack of certain data which I thought
essential. The 84 individuals represent 48 institutions, 45 as
professors, 26 as assistant, associate or adjunct professors,
and 13 as instructors. The academic titles mean little but
they are mentioned only to indicate the range or the number
of departments of psychology. Five individuals were con-
nected with one institution, four with another, there were
three in each of nine institutions, there were two in each of
eleven, and one in each of twenty-six institutions. Half of the
men had received their doctorate during the decade as follows :
1906, 2; 1908, 6; 1909, 6; 1910, 4; 1911, 2; 1912, 10; 1913, 7;
1914, 4; 1915, i.
Since my immediate interests have been connected with
recent publications I selected the past ten years, 1906 to
1915, inclusive, and have tabulated the contributions of each
of the 84 individuals for each year. The contributions which
have been listed were found in the PSYCHOLOGICAL INDEX.2
xThe 1916 membership list was used. This contains the titles of positions as
supplied by the members, and is presumably correct. I am convinced that some of the
titles are incorrect, but the mistakes are relatively unimportant. I have not gone
beyond the official returns. Some professional psychologists who are not members of
the Association are not included. Some of them are known to the writer, but it seemed
unfair to include them, since they are not members of the national body of professional
psychologists.
2 This bibliography does not give references to all of the publications of psychol-
ogists, some contained in inaccessible periodicals are not given, some which are not
psychological are not noted, but the failure to list all the psychological publications of
any one individual rests solely with that individual, since the INDEX asks that omissions
202 SHEPHERD IVORY FRANZ
The different kinds of contributions have been listed under
the following titles: Monographs, Original Articles, Discus-
sions, Books, General Reviews, Reports of Meetings. A few
translations of American books into foreign languages were
omitted, on the ground that the original authors had already
made the contribution and the work of translation was that of
another. Translations of foreign books into English were
also omitted from consideration. In some cases it is difficult
to decide whether a certain publication is a discussion or an
original article, in other cases whether a publication be an
original article or a monograph, or a monograph or a book.
Some periodicals publish long articles which in other series
would be published as monographs, but they have been dealt
with as they stood. When doubt arose as to the character of
the publication the contributor was given the benefit of the
doubt.1 The inclusion of general reviews and reports of
meetings may need justification. They have appeared to me
to be legitimate methods of advancing psychology because
they may create interest to investigate certain matters or give
facts to others which would not ordinarily be discovered, etc.
Their value, however, must be considerably less than those of
publications containing the results of personal investigations.
Somewhat similarly with books, since they are largely com-
pilations of the work of others.
Table I. gives the results of the examination of the INDEX
for the individuals and the decade under consideration.
Here there are shown the numbers of each kind of publication
for each year and for the ten-year period. It will be seen that
with the exception of a few lean years (1907, 1914) the number
and corrections be supplied. A few mistakes were noted, when they were obvious
they were taken account of in the tabulations. The tabulations by years are not always
exact on account of the inability of the INDEX compilers to secure all the references for a
particular year at the time of publication, but titles omitted in one year are found in
the next number.
1 That is, as will be noted below, an artificial value was assigned to each kind of
publication, and when doubt arose the higher value was assigned. At the same time a
'joint' article, etc., was assigned to both individuals, and given in each case its full
value. I was at first inclined to reduce the article value for 'minor contributions,'
but I concluded to let the above consideration hold. The recent publication of the
Proc. Nat. Acad. of Sciences was another difficult matter. The ' articles ' in that journal
are in reality abstracts or summaries (Auto-refer ate).
PRODUCTIVITY OF AMERICAN PSYCHOLOGISTS
203
TABLE I
KINDS OF PUBLISHED CONTRIBUTIONS OF PROFESSIONAL PSYCHOLOGISTS FOR THE
DECADE, 1906-1915, INCLUSIVE
1906
1907
1908
1909
1910
1911
1912
i9»3
1914
1915
Totals
Monographs
I
•i
2
2
C
C
5
IO
6
6
45
Articles
16
36
4-1
49
42
44
59
58
39
76
492
Discussions
7
4'
c
7
2
8
9
12
9
3
66
Books
2
i
IO
6
•2
6
8
5
9
5
55
Reviews ...
4"
i
4"
•j
4.
29
12
25
76
33
161
Reports
I
o
4"
•7
4,
6
7
2
5
32
Totals
61
45
68
70
6O
98
116
112
93
128
851
of research articles and monographs has not differed very
greatly from year to year and there is a fairly gradual increase.
There is a slight increase in 1912 and 1913 over the preceding
years and a greater increase in 1915 over 1913 and 1912.
The notable decrease in 1914 is not explained.1 It may be
that the increase of interest in the subject of tests is re-
sponsible, those who had taken up this line of work being
occupied in 1913 and 1914 with the accumulation of facts
which were published in 191 £.2 When we look at the yearly
totals we note that the increase began in 1911, and this was
coincident with the inauguration of the series of general
reviews in the PSYCHOLOGICAL BULLETIN.
Table II. shows the distribution of the contributors over
the ten-year period. It is a remarkable fact that the first
five years are almost constant in number of contributors and
that the increase in the number of contributors has taken
1 A writer in the New York Times book review supplement has mentioned that in
general literature 1914 was a lean year. The reason is not obvious in that case or in
the case of our psychological publications, since the European war could not have its
effect on production until very late in the year. In connection with psychological
journals this effect should have been more noticeable in 1915 if it existed because of
the war.
2 Ruckmich reports over 800 original articles for the decade 1905-1914. Probably
the number in 1905 was nearer that for 1906 than that for 1915 of the present table.
On that assumption we would have about 500 titles of original articles and monographs
for that decade by professional psychologists. Not all of these have appeared in the
magazines listed by Ruckmich, and in comparison with the 'over 800' noted by
Ruckmich it is apparent that fully 3/8 of the grand total was contributed by the
non-professional psychologists, or by those whose loyalty was divided between psy-
chology and philosophy, education, etc.
204
SHEPHERD IVORY FRANZ
TABLE II
GROUPING OF PROFESSIONAL PSYCHOLOGISTS ACCORDING TO YEARLY NUMBER OF
PUBLICATIONS, FOR THE DECADE, 1906-1915, INCLUSIVE
1906
1907
1908
1909
1910
1911
1912
19I3
1914
i9IS
I contribution
1C
TlS
12
II
17
2O
17
19
?R
2 contributions
3 contributions
IO
•J
7
-z
8
c
8
6
8
•z
9
7
13
c
13
c
13
12
6
4 contributions
o
o
I
2
HP
o
2
c
2
6
C contributions ....
2
o
I
I
2
6
I
•j
2
i
6 contributions
o
I
o
2
o
o
2
2
I
2
7 or more contributions
I
0
2
0
0
2
4
I
0
2
Totals
31
27
29
30
32
38
47
46
45
57
place during the past five years. The greatest number of
contributors is found in 1915, the year of the greatest number
of contributions. This table also shows what may be termed
the scientific activities of psychologists for it gives the number
of publications of groups of individuals. For the past five
years about 30 per cent, of those who contributed published
three or more articles, etc., each year. This is, of course,
not to be taken to mean that the same individual did this
from year to year, although it may be mentioned that the
tabulation of the material indicates that the man who does
it one year is more apt to repeat with more than a single con-
tribution for the following years.
Since not all of the individuals on our list have been, nor
could they be expected to be, active in publication during the
whole of the decade it is of interest to compare the figures
which are given in Table II. with the examination of the
years of the doctorate or other higher degree which is held.
There were 42 individuals who had obtained their higher
degree anterior to I9O6.1 The other 42 could not be expected
to publish work before their doctorate, although some did so.
Counting the date of the doctorate as the date when publi-
cation might reasonably be expected, the numbers of indi-
viduals expected to publish were obtained by adding the new
doctors to the original 42 at the beginning of the decade. It
has already been mentioned that some published before the
doctorate, the number being 14 (out of a total of 42), and a
few published more than one article.
1 Two individuals on the list hold the Master's degree only.
PRODUCTIVITY OF AMERICAN PSYCHOLOGISTS
205
When now we compare the number of expected contri-
butors with the actual number of contributors we find an
interesting condition. Table III. gives the data. Although
the number of expected contributors increased during the
first five years the yearly number of actual contributors
did not increase. The increase in contributors came
during the second half of the decade, but the percentage of
expected contributors who published is the same (68 per cent.)
at the beginning and at the end, there being a decrease in the
percentage up to the year 1910 and an increase thereafter.
If we separate out those contributions which are intended to
convey new facts or new interpretations,1 which would limit
us mainly to articles and monographs, we find an even more
TABLE III
THE PUBLICATION ACTIVITIES OF PROFESSIONAL PSYCHOLOGISTS FOR THE DECADE,
1906-1915, INCLUSIVE: THE NUMBERS OF EXPECTED CONTRIBUTORS, OF ACTUAL
CONTRIBUTORS, OF THOSE CONTRIBUTING ARTICLES AND MONOGRAPHS, AND
THOSE CONTRIBUTING PREVIOUS TO DOCTORATE
1906
1907
1908
1909
1910
60
29
28
3
1911
1912
1913
1914
83
45
35
o
1915
Expected contributors
44
30
28
I
44
26
24
i
S?
26
23
3
56
28
26
2
62
37
27
i
72
43
30
4
79
45
34
i
84
57
40
0
Actual contributors
Contributors of articles and monographs
Contributors previous to doctorate (addi-
tional)
interesting comparison. At the beginning of the decade
there was a total of 44 individuals expected to contribute.
Of this number 28 made contributions of articles and mono-
graphs (64 per cent.). The percentage (actual contributors
in relation to expected contributors) decreased in the follow-
ing years as follows: 55, 46, 46, 47, 44, 42, 43, 42, 48. The
contributions anterior to the doctorate by those who con-
tributed previous to their doctorate are not included in
these calculations. It is to be observed, therefore, that the
percentage of original contributors has decreased and the
percentage of total contributors at first decreased and later
reached its original figure. The great differences in the per-
centages from 1911 to 1915 inclusive are to be understood
1 It is not intended to say that books, discussions, reports or reviews do not
contain new facts and new explanations, but that they are less apt to do so.
206 SHEPHERD IVORY FRANZ
primarily as the result of the publication of general reviews
and an apparent satisfaction of the authors of these reviews
in their accomplishment.
The subject of books deserves a separate paragraph.
These were mainly the work of the group of older men, for of
the 55 books which have been published in the decade only
five have been written by those whose doctorates were granted
in 1906 and subsequently. Two men were responsible for
one book each, two for two books each, four for three books
each, one for four books, and one for fifteen books. The years
of publication of the 55 books are shown in Table I.
Mention has already been made of the increase in the
expected contributors owing to the granting of the doctorate.
The original 42 names have had added to them an equal
number. Of these additional 42, 14 published previous to
the doctorate. The first publication after the doctorate, in
most cases dissertations or parts of dissertations, of 18 was
made in the year of the doctorate, of 10 the first publication
was in the year following the doctorate, of 9 in the second
year, of i in the third year, of 2 in the fourth year, and of I
in the fifth year after the granting of the doctorate. No
publication by one who received the doctorate in 1913 had
been made up to and including 1915. These figures show that
either our means of publication are insufficient, or that the
dissertations presented in partial satisfaction of the doctorate
are not nearly ready for (or are not worthy of) publication,
or both.1
The 42 individuals who had received the doctorate prior
to I9062 might each have contributed something in each of the
ten years under consideration, but only seven did so. Seven
others contributed in 9 of the ten years; 4 in 8; 4 in 7; 6 in 6;
I in 5; 2 in 4; 3 in 3; 2 in 2; 2 in only one of the years; and 4
had no publications. Those who received the doctorate in
1906 and subsequently can not be dealt with in the same man-
1 It might be well for university authorities to deal a little more strictly with the
matter of publication. The publication of a dissertation in part or as a whole is the
only evidence to the world outside of the particular university that the individual has
shown a capacity for investigation, one of the main doctorate requirements in all in-
stitutions of which I have knowledge.
2 Including the two who have not taken the doctorate.
PRODUCTIVITY OF AMERICAN PSYCHOLOGISTS 207
ner, but the groups in accordance with the percentage of years
in which contributions were made, counting the total of
years since the doctorate as the expected total, are as follows :
o, i; from I to 10 per cent., 7; n to 20 per cent, 5; 21 to 30, 4;
31 to 40, 5; 41 to 50, 13; 51 to 60, I ; 61 to 70, i; 71 to 80, 3;
100 per cent., 8. The number of older men who averaged at
least one contribution for every two years, or more often, is
double that of the younger men. The younger men had
more than two thirds of their number who did not publish
as much as one contribution for every two years. It should
further be stated that of the younger men placed in the most
regular class (100 per cent.) one contributed original work in
only one of 8 years, a second in only two of 6 years, a third in
six of 7 years, and a fourth in two of 3 years. The other 4
contributed an article or monograph in each of the expected
years. Of the older men in the 100 per cent, class, one failed
to report original work in only one of the ten years, one did
not report such work for two of the years, and two for three
of the years. Three contributed at least one article or mono-
graph in each of the ten years.
Comparing the two groups we find that of the 420 ex-
pected individual years of publication of the older group (10
years each for 42 individuals) there were only 257 individual
years of publication, a percentage of 61; of the younger group
there were 214 expected years and an actual total of 109, a
percentage of 51. This difference is entirely accounted for by
the long delay in publication after the doctorate, for if the
delay periods be subtracted from the total there is a percentage
of 6 1 for the younger men. Although there is not a sufficient
number of years to make the conclusion certain, the figures
would lead to the belief that when the younger men start
there is not a great difference in total productivity between
them and the older men. My impression previous to tabu-
lation was the reverse, probably because of some notable
examples of productivity of the older men. The latter is
counterbalanced by the fact that there are five of our pro-
fessional psychologists, four of the older group and one of the
younger group (the latter case mentioned above), who have
208
SHEPHERD IVORY FRANZ,
not published anything worthy of citation in the PSYCHO-
LOGICAL INDEX in ten years.
The individual differences which have been mentioned are
better shown when we compare the totals and yearly averages
for the individuals of the groups. It would not be expedient
to mention names of individuals, or to designate them in
recognizable terms, so that we must fall back upon generalities
of individual differences. Of the older group there are four
who did not make a scientific contribution of such a character
or in such a journal as to be deemed worthy of mention in
the INDEX in the decade; there is only one of the younger
group. In addition the contributions of one of the older
TABLE IV
DISTRIBUTION OF PROFESSIONAL PSYCHOLOGISTS ACCORDING TO THE NUMBERS OF
CONTRIBUTIONS IN THE DECADE, 1906-1915, INCLUSIVE
Number of Publications
Total Publications
Articles and Monographs
Older
Younger
Older
Younger
o
4
14
12
5
7
I
35
4
i
i
,1
16
4
i
I
38
2
I
O
I tO IO
II tO 2O
21 tO 30
3 1 and over
TABLE V
COMPARISON OF TOTALS OF PSYCHOLOGICAL CONTRIBUTIONS BY THE OLDER AND
YOUNGER GROUPS OF PROFESSIONAL PSYCHOLOGISTS FOR THE DECADE,
1906-1915, INCLUSIVE
Articles
Mono-
graphs
Books
Discus-
sions
General
Reviews
Reports
Totals
Older
Younger
369
123
19
26
50
5
55
II
109
52
18
14
620
231
group consisted exclusively of the class of general reviews and
reports. The accompanying tables show the distribution
of the men in accordance with their scientific-literary pro-
ductivity. In Table IV. there are shown the numbers of
individuals in each group in relation to the total number of
publications and in relation to the publications of articles
and monographs. In Table V. there are shown the different
PRODUCTIVITY OF AMERICAN PSYCHOLOGISTS 209
kinds of publications (articles, monographs, books, etc.)
according to the groups. The average total number of con-
tributions for the decade by the older men is 14.8, for the
younger men it is 5.5; the average of articles and monographs
by the older men is 9.2, and by the younger men only 3.5.
These figures may be somewhat misleading if taken as they
stand, on account of the difference in the total number of
years that might be expected for scientific publication. The
total years for the older group has already been said to be
420 (42 individuals for 10 years), and 214 for the younger
group. When the comparison is made of total publications
and of original (monograph and article) publications of the
older group and of the younger group in relation to the ex-
pected number of years it is also found that the older group
outranks the younger. Thus the individual yearly average
for total contributions for the older group is 1.5, and only I.I
for the younger group; and the individual yearly averages for
articles and monographs are respectively .92 and .69 for the
older and the younger men.
It is interesting to speculate on the reasons for these dif-
ferences. Doubtless in most institutions the younger men
are employed a greater part of the time in preparation of
materials for the laboratory work of students and in the
grading of themes, etc. In the smaller and less well endowed
institutions there is less aid for the prosecution of investiga-
tions, and if aid can be obtained it takes an exorbitant
amount of time to get the administrative machinery in
running order. At the same time the younger group has to
take considerable time in the preparation of material for their
courses of instruction, and they are more frequently called
upon to act as subjects or assistants for other research workers.
On the whole the older group has the advantage of long estab-
lished policies, of equipment, and of professional and me-
chanical assistance. At the same time the labor of teaching
is correspondingly less on account of the previous experiences.
Perhaps if we should compare the work of the older group, or
of as many of them as held positions in the decade of 1891
to 1900, with that of the present younger group we should
not find as great differences as now exist.
210 SHEPHERD IVORY FRANZ
On account of the differences in time opportunity it is to
be expected that the younger group would publish less in total
and that the individual total would be less than many of
those of the older group. Thus we find that nearly one half
(19) of the older group reached or exceeded the average of 1.5
publications per year, and that only one third (12) of the
younger group equalled or exceeded the yearly average for
that group (.5 publications per year). Table IV. gives the
results of the groupings. This shows that more than one
half of the older men averaged one publication per year,
seven having more than three per year The figures for the
younger group must be read in the light of the total yearly
expectation (214 instead of 420). Each individual should be
considered in relation to the total of his expected years of
publication. Thus the six individuals of the younger group
who exceeded ten publications had a total publication ex-
pectation of 44 years. The total of publications was 124,
which gives a yearly average less than 3.
The data with regard to the character of the publications
of the two groups are given in Table V. The difference in
the number of books has already been mentioned. Mono-
graphs are apparently the prerogative of the younger man,
the newly created doctor, for he has in proportion to the
expected years two and one half times as many as his older
colleague. The older men publish 33 per cent, more articles,
as has been mentioned, they take part in more controversies
as judged by the number of discussions, and they contribute
an equal share of the general reviews. With regard to the
last the remark may be necessary that the general review is
most frequently due to ' request' and not infrequently a request
to the elder is declined in favor of a younger colleague.
Since the different kinds of contributions are so varied no
direct comparison may be made of individuals except in terms
such as have already been used (total number of contribu-
tions, number of monographs and articles, and the relation of
these to the expected years of publication). An indirect
comparison may, however, be made if we assign to the differ-
ent classes of publications an arbitrary numerical value. This
PRODUCTIVITY OF AMERICAN PSYCHOLOGISTS 2 1 1
must be very arbitrary on account of the impossibility of
making good comparisons of values, as has already been ex-
plained. It is also to be kept in mind that no arbitrary value
for a class can be defended in particular cases, for if we con-
sider two articles we may immediately note that one deals
with an investigation in which some new methods have been
used and the analysis of factors appears to be well wrought
out, and we may also find that the second consists of a con-
firmation of previous work by the use of the same methods
which had previously been used. Even in two articles which
contain new facts or demonstrate new methods we may find
similarities and divergencies of completeness or of apparent
originality. It favors the majority of low grade (if we may use
such a term) publications if we assign the same arbitrary value
to all of one class, and this I haveidone. The values which I have
selected for the different classes of publications are as follows :
Reports of Meetings, i; General Reviews, 2; Discussions, 3;
Books, 6; Articles, 6; Monographs, 9. No justification of
these arbitrary values will be attempted, but the following
were in mind when the values were assigned. A discussion is
often of no observable value in adding to our knowledge, and
scientifically is worse than useless when it takes on the char-
acter of a personal attack. On the other hand, it may tend
to clear up doubtful points, bring up new ways of viewing a
situation, and at the same time by pointing out gaps in our
knowledge indicate lines of investigation. In so far as a
discussion does any of the latter things it appeared that it
has a real value beyond that of a report of a meeting, and
since a discussion also tends at times to inter in a suitable
manner some supposed facts which very generally, but er-
roneously, have been accepted it was thought worthy of a
greater value than that of a general review. It is assumed
that all of our psychological discussions are of the good char-
acter mentioned. Differences exist in general reviews. Some
are summaries of a few contributions of others. Some give a
fairly complete account of current work with an evaluation
of the material and thus help others who are not specializing
in the subject to obtain a better view than would be ob-
2 1 2 SHEPHERD IVOR Y FRANZ
tained by looking over the mass of details which the original
sources contain. Some of the general reviews which we are
considering here may be worthy of a higher value than some
of the discussions, but many might bear a reduction of the
general value figure. Here again it became necessary to
decide for the majority rather than for the few. Much the
same may be said with regard to books. Although the writer
does not pretend to have the specialist's critical ability in
every branch of psychology he is satisfied that many books
published during the period are nothing more than general
reviews, and at times poor ones. Some of the books have
brought out new facts and explanations, they have added
considerably to our psychological advance. But because of
their general character it is thought that a value double
that of a discussion and triple that of a general review would
be an ample average value. The original, or research article
(whether experimental or otherwise) was assigned a value
equal to that of a book, and the monograph (but only on
account of its length and supposed completeness) was assigned
a value fifty per cent, higher. The remark previously made
regarding monographs, that some monographs would be
articles in other publication series, is a point against such a
valuation, and the valuation is not insisted upon. The
increase in value is on the side of the younger men, who
most need the extra count.
Having selected the arbitrary values which have been
mentioned it becomes possible to make comparisons of the
output of different individuals without making special refer-
ence of an identifying character. At the same time it
becomes possible to make comparisons of the groups since the
heterogeneity of the different kinds of publications has been
translated into a homogeneity.
The calculation of the individual values for the ten-year
period shows that the range is from zero to 244. This means
that some have contributed nothing and that others range
from the zero point up to a valuation of 24.4 points per year.
It is necessary to translate all the total valuations of the
younger group into * expected' values, or average yearly
PRODUCTIVITY OF AMERICAN PSYCHOLOGISTS
213
values in accordance with the number of expected years of
publication. When this is done for all individuals we find
that there is a general average of 6.7 points per year for the
group as a whole, and a median of 4.2. When the five in-
dividuals who have not contributed even as much as a report
of a meeting during the decade are omitted the average is
7.2 and the median is 4.5. The distribution of the 84 in-
dividuals in respect to average yearly values of publications
is given in Table VI. Eighteen of the older group and
twenty-five of the younger group are below the median of
all values; 22 of the older and 30 of the younger groups are
below the average of the total. Arranging all in sequence of
average yearly values and dividing into four equal parts, each
containing 21 men, we find the following distribution of the
older and younger respectively in the groups from lowest to
highest: older, 10, 7, n, 14; younger, II, 14, 10, 7.
TABLE VI
GROUPING OF PROFESSIONAL PSYCHOLOGISTS ACCORDING TO AVERAGE YEARLY
"VALUES" OF CONTRIBUTIONS FOR THE DECADE, 1906-1915, INCLUSIVE
I.O
Values
and
I.I—
2.1-
3-1-
4.1-
5-1-
6.1-
7.1-
8.1-
9.1-
IO.I-
15.1-
Under
2.O
3-°
4.0
5-0
6.0
7.0
8.0
9.0
10.0
15-0
*5-o
Number of men
Q
14.
Q
Q
Q
2
o
c
c
4.
The older group on account of time, material equipment,
academic relations, and other conditions has advantages which
make it of special interest. All of the individuals had attained
their higher degrees previous to 1906, about three quarters
are heads of departments, and about the same number are
connected with well-equipped and long-established labor-
atories. This group furnished four individuals who did not
make a published contribution of any kind to psychological
advancement in the decade. Six of the group contributed
one original article or monograph during the ten-year period;
two contributed two original contributions; two contributed
three; and three contributed four. We have in this group,
therefore, forty per cent, who have not averaged an original
contribution once in two years. Some of these seventeen
214 SHEPHERD IVORY FRANZ
individuals did contribute in other directions besides mono-
graphs and articles, for they published 5 books, 33 general
reviews, 2 discussions, and wrote one report of a meeting.
Besides the four who did not make any kind of a contribution
there were four additional who made no contribution beyond
the original articles and monographs published. On the
whole the younger group, while not as productive as the
older, show better results with respect to the publication of
individuals. Only one (duration three years) has not pub-
lished, as is mentioned above. Ten others have not published
as frequently as once in three years; there is a total of sixteen
who have not published as frequently as once in two years.
Since the older group had the opportunity to publish for
ten years we need not deal with averages entirely, but may
consider totals as well, on account of the homogeneity of the
series. It is of interest to know that seven of the group
(17 per cent.) contributed a total of 159 articles and mono-
graphs (41 per cent.); the highest half of the group (21)
contributed eighty-seven per cent, of all the articles and
monographs of the group (337 articles and monographs).
This leaves for fifty per cent, of the older men only thirteen
per cent, of the articles and monographs published by the
group.
The majority of older men who have contributed little in
the way of articles and monographs, and also to the total,
have held their present positions for many years, and have
apparently * grown up' in their present locations, they are
located in some of our better endowed institutions, and in
those with good laboratory facilities, they have colleagues
teaching in the same lines, and they occupy what may be
properly called positions of prominence in their respective
institutions. They have not the apparent disadvantages of
isolation, or of having to carry the burden of the psycho-
logical world upon their shoulders since there are colleagues
to help in teaching and perhaps in research. But these are
some of the men who are representing psychology as a science
in their respective university niches. Those of the younger
group who are least productive, and have given little evidence
PRODUCTIVITY OF AMERICAN PSYCHOLOGISTS 2 1 5
of interest in psychological advance by publication, are
mostly located in the smaller institutions, where there are no
colleagues of sufficient training or productiveness to be elected
members of the American Psychological Association. There
are in the younger group exceptions to this, several notable
cases being evident when the list is inspected. Without
going into the figures for total or original publications, since
the number of men at the different institutions is small, it
may be stated that those younger men who have the oppor-
tunity to remain at a long-established department have done
better than those in the more recently created departments.
But this does not hold for particular cases, since there are
surprising exceptions of individuals with apparently all the
advantages which can be obtained in the better endowed
institutions doing little or nothing which is prepared for the
edification of their scientific colleagues.
It should not be assumed, and it is here stated to the
contrary in order that there may be no misunderstanding,
that these men are doing nothing for psychological advance.
Some may have editorial duties, some may conceal themselves
in the work of their students, and some (like Herbert Spencer)
may be reserving their energies for some magna opera which
will be given to the world in due time. It seems unlikely,
however, that as many as 40 per cent, of the older group are
engaged in the accumulation of material for the development
of a cosmology, or of a system of psychology, or of an ex-
haustive history of the science, or of other large projects which
should not be laid aside in favor of the minor contributions
such as articles and monographs. It is apparent that a few
of those who hold chairs of psychology are contenting them-
selves with teaching and the carrying out of the social obli-
gations which fall to the lot of every scientific man, who must
meet his colleagues and take part in the life of the university
with which he is connected. Whether or not there are more
psychologists who are doing these things than scientific men
in other lines can not now be determined. And, the other
view may be expressed that those who are contributing much
do so without proper scientific care and because of reportorial
2 1 6 SHEPHERD IVOR Y FRANZ
tendencies. The character of the work of many of our most
productive men shows that they have not done their work at
the expense of care. But, the writer feels that some of the
so-called 'professional' psychologists should be classed with
dilettantes; they are not scientific professionals in the sense
that they are forging ahead and that they are succeeding
because of their efforts in scientific work.
Much has been written recently about university positions and university control;
the professor always being the oppressed and the university as represented by the
president being the oppressor. I have a feeling that part of the dissatisfaction may
be due to the 'great promise' for which candidates have been recommended never
showing up in practice. A president of one of our leading institutions has been criti-
cized for saying that every man on the faculty may be expected to publish at least one
article every two years. It may be that he had had experience with those of 'promise'
and not of performance. Much has also been written about academic tenure, as if
that was a sacred right (or rite), but the critic might well imagine a better state of
affairs to follow in some of our universities if there was a power of recall for those
members of the faculty who do not measure up to the expected or to the average amount
of performance. If the present methods of filling positions, from the grade of professor
down to that of instructor, be continued we shall always have some of little or no ac-
complishment (the dilettantes) filling important chairs, especially if there be the added
conditions that promotions be made as vacancies occur. It is the writer's belief that
much of the inconsistencies would be eliminated if every vacancy were advertised as
open to application, stating that applicants were expected to send in the accounts of
their academic careers, copies of publications, etc., and that these would be graded by
a non-interested committee of specialists (advisory board) who would report to the
faculty or to trustees the results of their findings. We should probably escape the
absurdity of having in one year two young men recommended from one university as
'geniuses of the first water.' This has happened, and the writer is not certain which is
to be the more pitied, the university which permitted its 'geniuses' to go to other in-
stitutions, or the universities which would accept men who were recommended as such.
Another matter may be worth considering briefly. This
is the relation of productivity to the institutions from which
the higher degrees were received. In pointing to psycho-
logical advances in institutions, we have hitherto been
content to mention the number of doctorates which have
been granted. But it is obvious that numbers count for
very little in progress, unless one is going to play a mass game
or overrun a weak nation. What should be considered is the
question: What have the doctors done after leaving an in-
stitution? And we may also ask: Has the training which is
supposed to fit the man for research been effective in stimu-
lating him to efforts in that direction? Part of these ques-
PRODUCTIVITY OF AMERICAN PSYCHOLOGISTS
217
tions may be answered by considering the performance of
those granted higher degrees by different institutions. Of the
84 men on the present list 17 were scattering with respect to
institutions, but the remainder were distributed over seven
institutions. Only one institution is mentioned by name
(Leipzig), the other individual institutions (American) are
represented by letters. The 17 scattering cases are grouped
together as 'other foreign' and 'other American.' Table VII.
TABLE VII
COMPARISON OF PUBLICATIONS OF ALL KINDS AND OF ARTICLES AND MONOGRAPHS,
BY GROUPS OF PROFESSIONAL PSYCHOLOGISTS, ARRANGED ACCORDING TO IN-
STITUTIONS CONFERRING DOCTORATES. (Two INDIVIDUALS WHO HAVE NOT
RECEIVED THE DOCTORATE ARE INCLUDED UNDER THE INSTITUTIONS WHICH
CONFERRED THEIR HIGHER DEGREES)
Institutions
Leipzig
A
B
c
D
E
F
Other
Foreign
Other
American
Numbers of men
8
78
859
1 1.0
3-6-
18.6
603
7-7
II
79
674
8.8
1-5-
19-3
Sio
6-5
7
54
421
7.8
i.i-
23.0
360
6.7
15
132
884
6.7
o.o-
24.4
708
54
II
74
38i
5-i
1.4-
14.0
312
4.2
8
58
226
3-9
0.6-
"•3
198
3-4
7
42
103
2-5
o.o-
7.5
96
2.3
6
57
422
7-4
0.6-
9-4
348
6.1
II
60
269
4-5
o.o-
10.2
222
3-7
Total expected years
All publications:
Total credits
Credits -T- years
Range of average yearly
credits of individuals
Articles and monographs:
Credits
Credits -f- years
gives the numbers of men from different institutions, the total
years under consideration (the total * expected' years of all
in the special group), the total credits, the relation of credits
to the total expected years of publication, and the minimum
and maximum average yearly credits by the individuals in the
groups.
Probably the total expected years, rather than number
of individuals, is a better means of comparison with the totals
of publications, and the relation of these two are shown in the
fifth line. Here it will be observed that the 8 men from Leip-
zig far surpass the groups from the other universities. They
have the highest minimum, showing that each is publishing
fairly regularly, and although they are surpassed in maxima
by three other institutions this is largely due to a few very
2 1 8 SHEPHERD I FOR Y FRANZ
productive men. The men from Leipzig have mostly been
of the older group, as indicated by the average of expected
years (9.8), but the average is not much greater than that
(9.5 years) for the group from Other Foreign Institutions.
Leipzig also closely approaches the total of C, although the
number of men is only slightly over half, and the total number
of expected years is about 60 per cent, of this American
university. The order of institutions in total credits divided
by expected years is as follows: Leipzig, A, B, Other Foreign,
C, D, Other American, E, and F. When now we compare
the performance of the doctors from the different institutions
in relation to publication of articles and monographs we find
the results shown in the last two lines of Table VII. Here
again Leipzig shows a great superiority. The order of in-
stitutions in credits divided by years has not materially
changed, there being a reversal of the order for B and A. The
remainder of the table speaks for itself.
One matter remains to be briefly considered. This is the
relation of administrative work to the carrying on of scientific
production. Our group contains five individuals who have
given their occupation as partly that of dean or president.
Two of these individuals notwithstanding their arduous
administrative duties have managed to carry on investiga-
tions and to publish the results of them. At the same time
they have both for total and for original publications managed
to be in the more productive class. The other three probably
have their administrative duties as reasons for their relative
non-performance. In the group there are also eleven who
have greater or less editorial duties in connection with the
publication of periodicals. Of this group two have been
below the average in performance. Both of these extra-
scientific duties (administration and editing) have not inter-
fered with the production of an average amount of scientific
articles by two thirds of the men thus engaged, and it may
be that these added functions have been beneficial rather than
the reverse. In the estimation of productivity of individuals
editorial functions have been omitted and had these been
added to the totals in some numerical way it is quite likely
PRODUCTIVITY OF AMERICAN PSYCHOLOGISTS 219
that more of the group of editors would have approached the
top. As the arrangement now stands, counting only articles,
books, reviews, etc., we find that two thirds are in the higher
half of the distribution, and of these most are in the higher
quarter.
It is pertinent to remark that psychology appears to be
getting from those whose chief interests are not in its de-
velopment (from the non-professional psychologists, in other
words) as much as, if not more than, from many of its own
men.1 From some of its own psychology is receiving much
less than should be expected, and there is, perhaps, some
reason for the assertion that "were it not for an academic title
it would be difficult to discover the reason why certain in-
dividuals are called psychologists." In conclusion the at-
tention of the reader is called to the consideration of the
wisdom of the action of certain scientific societies which
require that a member shall retain membership in them only
as long as he continues to show an active interest in the
advancement of his science by publication, provided con-
tinued ill-health or other disabling conditions (old age and
the like) do not prevent.2
1 1 expect to deal with the psychological contributions of these non-professional
members of the Association in another article.
2 A constitutional amendment of this character was recommended by its council
to the American Psychological Association at one of its meetings, and at the subsequent
meeting the council reversed its recommendation (see the Proceedings of the Associ-
ation for the years 1906 and 1907).
THE PSYCHOLOGY OF THINKING IN THE CASE
OF READING
BY EDWARD L. THORNDIKE
Teachers College, Columbia University
If a person is presented with a paragraph to read and
questions about it, his responses provide useful material for
studying some of the facts and laws of thinking. Consider,
for example, the following task and the following responses
made to the first question by pupils of grades 5 and 6:
/
Read this and then write the answers to I, 2, 3, 4, and 5. Read it again as
often as you need to.
Nearly fifteen thousand of the city's workers joined in the parade on
September seventh, and passed before two hundred thousand cheering spec-
tators. There were workers of both sexes in the parade, though the men far
out-numbered the women.
1. What is said about the number of persons who marched in the parade?
2. Which sex was in the majority?
3. What did the people who looked at the parade do when it passed by?
4. How many people saw the parade?
5. On what date did the event described in the paragraph occur?
Two hundred people About two thousand
Three thousand Maybe No. 12
Thousand About 2700
Eighteen thousand Two hundred thousand spectators workers
Two thousand in the parade
Five thousand Two hundred thousand spectators
Ninety thousand Two hundred cheering
Twenty-five thousand Nearly 115000 on Sept. of people
About thirty-five thousand Nearly sixteen thousand
Nearly twenty thousand Hundred thousand spectators
More than ten thousand It is said about the number or group of
There were about 25000 people
200,000 It is said that they are great
It was 200,000 A very great deal
220
THE PSYCHOLOGY OF THINKING IN THE CASE OF READING 221
A lot of people
Congregation
There were a great lot of men
The men outnumbered the women
The men were more than the women
There were more men
They outnumbered the women
There the par on number the
The men were far ahead of the women
Men and woman
Citizens
They were workers
There were workers of both sexes
Workmen in the parade
Of all the working men
That the city workers joined the parade
Workers joined the parade
That they rejoin in the parade
They were joined
A number of workers joined the parade
Joined the parade
Workers join
They joined
They pass two hundred spectators
Before the spectators
Passed before two hundred thousand spec-
tators
They two hundred thousand cheering
spectators
Passed before 200000 and 15000
They passed nearly 5000
Passed before two hundred spectators
They marched before cheering spectators
Three thousand cheering them
People of both sexes cheering them
They are cheered
Parade before two hundred spectators
Parade spectators
They marched nice
They marched very nice
They kept in step
They marched very straight
They did good or bad
They look so nice
They clap their hands when they see the
American flag
They keep their step and many others
There character
Honorable and good
The people said the parade large
Most of them were old
They are soldiers and marched
They say halt
The captain says march
There was a lot of floats
The people are killed by the war
The meddles
September seventh
Irish
The variety of responses to this one fairly unambiguous
question is a challenge. There is a challenge also in the
relative frequency of the different responses.
I shall report here some general facts which are displayed
by some hundreds of responses to each of a dozen or more
sets of questions upon a paragraph, which I have examined.
The first is that:
When a question on a paragraph is answered, any one
word may be over-potent in determining the response. As a
limit we have the case where a word produces a response due
to that word alone irrespective of all else in the situation. Or,
more generally, any element in a situation may be over-potent
to any degree.
The evidence supporting this claim is the existence of
222 EDWARD L. THORNDIKE
answers or elements in answers which could come as probable
results or over-potent action of single words, but whose oc-
currence otherwise is highly improbable. For example, in
the case of the words of the paragraph such influence is seen of:
thousand, in this answer to 3: "The people cheered thousands.'*
city, in this answer to 2: "City workers."
workers, in these answers to 2: "Workers," "Workers of sex";
and in this answer to 3: "There were workers."
In what follows a number in parentheses preceding a
response designates the question as an answer to which the
response was given.
joined, in (i) "They joined," (i) "They were joined," (3) "They
joined in," (3) "They were joined in the parade," (3) "They
cheered then joined in."
parade, in (2) "in the parade," (2) "Both in parade," (2) "Sexes in
the parade."
September, in (2) "There were workers of the September."
seventh, in (2) "Seventh," (3) "September seventh," (3) "Seventh
Avenue."
Lest the bizarre nature of some of these errors lead the
reader to fancy that they are fragments, or answers misplaced
a line too high or too low, I may note here that every quota-
tion that has been or will be given in this article is, unless
specially noted at the time, a complete answer, as given by
some pupils, and undoubtedly intended for the question whose
number it bears. Quotations are exact except that the first
word is capitalized whether or not this was done by the pupil.
passed, in (3) "Two hundred thousand cheering spectators passed."
spectators in (i) "Parade spectators," (i) "Cheering spectators,"
(1) "Two hundred thousand spectators workers in the parade,"
(2) "Spectators and working," (3) "The people who looked
cheered the spectators," (3) "The people looked and cheered
the spectators," (2) The sex spectators.
cheering, in (2) "Cheering."
both, in (2) "There were both workers," (3) "There were workers in
both spectators."
sexes, in (3) "Six in the parade," (3) "Sixes," (3) "Cheered the sex
in the parade."
THE PSYCHOLOGY OF THINKING IN THE CASE OF READING 223
though, in (2) "Though the men marched in the parade," (2)
"Though the men far outjoined parade."
men, in (3) "They cheered the men in the parade."
far, in (2) "Far out women."
out, in (3) "They counted out how many women."
numbered, in (3) "The men numbered women," (3) "Numbered the
women," (3) "They numbered the people."
out-numbered, in (2) "Sex out numbered of women," (2) "Out-
numbered."
women, in (2) "Sexes of women," (2) "Men and women," (3) "The
women."
Any phrase or other part of a sentence tends in a similar
manner to produce a response due to that phrase per se.
Thus, fifteen thousand appears as an answer to questions 2,
3 and 4 and (in "Passed before 200,000 and 15,000") as an
answer to question I. The following are responses probably
explainable by the independent action of phrases or other
groups of words.
the city's workers, (2) " City workers."
joined the parade, (i) "Joined the parade," and (2) "They joined
the parade."
September seventh, (i) "September seventh."
passed before two hundred, (3) "Passed before two hundred."
two hundred, (i) "Passed before 200 spectators," (2) "Two hundred
spectators," and (4) "Two hundred."
hundred thousand, (4) "About a hundred thousand."
two thousand, (i) "Two thousand," and (4) "Two thousand."
two hundred thousand, (i) "Two hundred thousand," (2) "Two
hundred thousand," and (3) "Two hundred thousand."
There were workers, (2) "There were workers."
of both sexes, (i) "People of both sexes cheering them," and (5)
"Sept. yth both sexes."
There were workers of both sexes, (i) "There were workers of both
sexes," and (3) "There were workers of both sexes in the
parade."
though the men, (3) "Though the men far out joined parade," and
(5) "Thought the man fat out."
outnumbered the women, (i) "They outnumbered the women," and
(2) "Sex outnumbered of women."
In the same way it can be shown that every word and every
224 EDWARD L. THORNDIKE
word group in a question tends to produce a response due to
that word or word-group per se. Consider for example the
responses from pupils in grades 7 and 8 to Test M.
M
However certain it may seem to be that men work only because they
must, and would avoid labor except for the food, clothing and luxuries that
are its rewards, the facts may well be to the contrary. It can hardly be the
case that men dislike work because they wish to be utterly idle. For mere rest,
mere inactivity, is not commonly enjoyed. To have nothing to do is not what
men seek. Were that so, we should envy the prisoner shut up in his cell. If
men had to choose between a life spent at eight hours of work daily in a factory
and a life spent at eight hours of sitting on a throne without moving hand or
foot, many of them would, after trying both, choose the former. Activity of
body or mind, at which a man can succeed, is, in and of itself, rather enjoyed
than disliked.
1. What is it that this paragraph says may seem sure, but probably is false?
2. In what respect is a prisoner in his cell like a man with a million dollars?
3. If the absence of any activity were what we wished for, what would be our
attitude toward a prisoner in his cell?
4. What is stated in the paragraph to be really liked and not objected to?
5. What choice is described in the paragraph as an argument that work,
merely as such, is not always avoided?
In question I the influence of Paragraph is seen in such
responses as:
"Work,"
"Labor,"
"Idle men and working men,"
"Idleness and wealth," and
"Men that do not work."
The influence of Paragraph says is seen in:
"It says that men should not avoid labor,"
"It says that some men work and some of them would be idle," and
"Some men envy prisoners in a cell rather than work."
The influence of Sure is seen in:
"For mere rest mere inactivity is not enjoyed."
Similar evidence in the case of questions 2, 3, 4 and 5 is as
follows :
THE PSYCHOLOGY OF THINKING IN THE CASE OF READING 225
QUESTION 2
prisoner in his cell — "Because he is shut up in a cell," and "A man
in prison is sitting 8 hours daily with chains."
QUESTION 3
activity — "Is rather enjoyed than disliked," "For mere rest, mere
inactivity is not enjoyed," and "A man succeed or rather
enjoyed than disliked."
absence of activity — "Sluggish," "To have nothing to do is what
men seek," and "Idleness."
prisoner in his cell — "He would not be there unless do something
wrong," and " Because it is not a good thing to be a prisoner."
QUESTION 4
is stated — "Men who work," and "That the American man wants
to work."
really liked — "Good clothing and luxuries," and "A hard work man
should be liked."
objected to — "Mere rest mere inactivity is not commonly enjoyed,"
"Inactivity," and "To have nothing to do."
QUESTION 5
choice — "The man who works and the one who does not."
described — "Why men work."
avoided — "They would avoid labor if it was for food."
work avoided — "Idle and stealing."
Consider also the following responses to questions I, 2, 3
and 4 on paragraph /.
/
In Franklin, attendance upon school is required of every child between
the ages of seven and fourteen on every day when school is in session unless
the child is so ill as to be unable to go to school, or some person in his house
is ill with a contagious disease, or the roads are impassable.
1. What is the general topic of the paragraph?
2. On what day would a ten-year-old girl not be expected to attend school?
3. Between what years is attendance upon school compulsory in Franklin?
4. How many causes are stated which make absence excusable?
226 EDWARD L. THORNDIKE
5. What kind of illness may permit a boy to stay away from school, even
though he is not sick himself?
6. What condition in a pupil would justify his non-attendance?
7. At what age may a boy leave school to go to work in Franklin?
1. "Every paragraph must have a period "") ,
"A group of complete sentences" } due tO
2. "Monday" 1
"Wednesday" ^due to day.
"Friday" j
"The ten year old girl will be 5a," due to ten-year-old girl.
3. "It was a great inventor" \, r LJ.
,,n . . ,. r due to / ranklin.
"Because its a great invention J
"Because it is a small city," due to Franklin.
4. Twenty five ^
about ninety L due to How many.
2000 J
A boy should bring a note"! ,
Tr , . }• due to absence excusable.
If you bring a note J
In the illustrations given so far the action of the element
has been accompanied usually by some vague action of the
situation as a whole, but cases may be found where this
reduces to about as near zero as is possible, provided the pupil
writes any answer whatever.
Consider, for example, the following response to questions
I, 2, and 5 on paragraph / (on page 220).
QUESTION i
"Most of them were old," which shows almost no effect of anything
save the "persons."
"There were a lot of floats," which shows almost no influence save
of "parade."
"Irish," which could fit any paragraph or any question provided
parade or persons or both occurred therein, almost as well as it
fits this.
QUESTION 2
" The chief commander of all " and " Captain and lieutenant" seem to
show no influence save the major of majority plus a faint effect
of parade.
THE PSYCHOLOGY OF THINKING IN THE CASE OF READING 227
QUESTION 5
" 1492" and " 1776" seem the products of date in total neglect of all
else in the question and paragraph.
Some of the cases above, if taken alone, are perhaps as
explainable by other causes as by the tendency of each word
to act irrespective of the total of which it is an element, but
no one, I think, will assume that the collection can be well
explained save by supposing that the single words do have
that tendency. We must then think of the pupil who ex-
amines the paragraph and the questions as beset by a tendency
to answer each question by each word in the paragraph and
have each word in the questions evoke a response that is
bound to it alone. Most of these tendencies are of so nearly
zero strength that they almost never compete successfully
with other tendencies; and most of the resulting thoughts are
so absurd that if they come to mind they are promptly dis-
missed. But a sound theory must accept their existence,
traces of which we have illustrated in the case of Paragraphs 7,
/ and M. Every element in a situation tends to arouse the
response which is connected with it.
Just as any element of the situation may be, relatively to
others, far too potent in determining the response, so also it
may be not nearly potent enough. Calling Pa the potency
of element a and Pb the potency of element b, Pa/Pb may
vary between o and oo as limits.
The following are some cases of under-potency in the case
of / (page 220) :
QUESTION i
nearly — (Failure to include this in the response to I is of course
very, very common.)
fifteen— "Thousand"
what is said about the — "Honorable and good," "They march very
nice," "They marched very straight," "They did good or bad,"
and many similar responses.
number of persons — "They were workers," "Men and women,"
"That they rejoin in the parade," "Passed before cheering
spectators."
228 EDWARD L. THORNDIKE
who marched in the parade — The many responses of "200000,"
"They cheered them," etc.
All of question I except parade is under-potent — "There were a lot
of floats."
QUESTION 2
the "out" of outnumbered — "A number of women."
which — "Both sexes was in the parade," "There were both sexes
there," "Workers of both sexes," "Men and women," "Two
sexes," "Two of them."
sex— "City workers," "City workers of N. Y.," "The chief com-
mander of all," "Working," "The front ones," "Spectators,"
"Cheering," "Fifteen thousand."
in the majority — "Women," "Sex outnumbered of women."
which . . . was in the majority — "The sex spectators," "Sexes,"
"In the parade," "Sexes in the parade," "There were men of
other sex in the parade."
QUESTION 3
what did the . . . do — "They were cheered by the people," "Two
hundred."
people— "Tip his hat."
people who looked at — "Passed before two hundred," "Passed before
a number of cheering spectators."
when it passed by — "Two hundred thousand cheering spectators
passed."
it— "They saluted them," "They cheered them."
cheering — "Inspected the parade," "They were glad to see it,"
"They talked about it" (and many others).
All save parade underpotent — "September seventh," "Seventh
Avenue."
QUESTION 4
saw the parade — "The men outnumbered the women," "Far out-
numbered."
two — "About 100000," "One hundred thousand," "Three hundred
thousand."
hundred— "Two thousand."
thousand — "Two hundred."
two hundred [thousand] — "Fifteen thousand," "Nearly fifteen
thousand," "Over 25000," "Over five hundred," "About
10000," "About 5000," "About 1000."
THE PSYCHOLOGY OF THINKING IN THE CASE OF READING 229
QUESTION 5
what date — "There were workers of both sexes in the parade,"
"Thought the man fat out/' "Described," "Sexes of the pa-
rade," "The parade," "And outnumbered women."
event described in paragraph — "March 4, 1915," "March 17,"
"April 23, 1903," "November^" "December^" "On Friday,"
"March 17," "March 18," "St. Patrick's day," "On the
twenty-second of February," "St. Pattac," "1492," "1776,"
"1820."
Seventh — "September seventeenth" (a common error, often due to
misperception or memory, probably) "September."
In any situation we may distinguish in a rough way
between words meaning things, qualities and events on the
one hand and words meaning relations between them on the
other. Thus we may think of question 2 below as:
[(Use of a gas range) instead of (use of a coal range)]
effect upon
temperature of the kitchen
H
You need a coal range in winter for kitchen warmth and for continuous
hot-water supply, but in summer when you want a cool kitchen and less hot
water, a gas range is better. The xyz ovens are safe. In the end-ovens there
is an extra set of burners for broiling.
1. What two varieties of stoves does the paragraph mention?
2. What effect has the use of a gas range instead of a coal range upon the
temperature of the kitchen?
3. What is needed to provide a supply of hot water all day long?
4. For what purpose is the extra set of burners?
5. In what part of the stove are they situated?
6. During what season of the year is a gas range preferable?
These relating elements are subject to over-potency and
under-potency like any others, but especially to under-
potency. Thus in the question quoted, we have as re-
sponses due to under-potency of the instead of:
230 EDWARD L. THORNDIKE
"The stove makes heat" and "You need a coal range in winter and
a gas range in summer."
As responses due to under-potency of effect upon, we find:
"You can cook better," "Heats quicker," "Because it is summer,"
and "Because we need no coal."
As responses showing under-potency of both instead of and
effect upon, we have:
"Gas to cook and coal to get warm," and "Gas range in summer."
Elements may act in substantially correct potency but out
of their proper relations. Responses to arithmetical problems
will illustrate this richly. In reading we have such cases as
"A cool kitchen is used for a gas range" in response to What
effect has the use of a gas range instead of a coal range upon the
temperature of the kitchen?
The connections leading from any element or group of
elements may be wrong whether the element is under-potent,
of correct potency, or over-potent. They may be wrong in
toto in the sense of leading to unserviceable responses to that
element for any purpose, or wrong for any defined set of
purposes, or wrong for the test's purpose in the sense of leading
to responses unserviceable for the particular need or problem.
Thus "A complete sentence" is wrong for almost all purposes
as a response to paragraph; "Commas and periods" is perhaps
even more universally futile; "A group of sentences" is
wrong for the particular purpose of answering the question
of /i.
Inadequacy may be considered as a special case of wrong-
ness by lack or insufficient connections, and failure of response
as the limiting case of inadequacy.
Incorrect thinking due to wrong or inadequate bonds lead-
ing from one or more elements of the situation is a simple
consequence of the general facts of connection-forming that
does not need demonstration here. We know a priori that
every element tends to call up what has followed or accom-
panied it.
I will report only three illustrations. The paragraph being
John had two brothers who were both tall. Their names were Will and
Fred. John's sister, who was short, was named Mary. John liked Fred
THE PSYCHOLOGY OF THINKING IN THE CASE OF READING 231
better than either of the others. All of these children except Will had red
hair. He had brown hair.
and the questions being
5. Who had brown hair?,
6. Who had red hair?..
there are with pupils of all grades from the third to the eighth
a large per cent, number of responses of "John" and "Fred"
to "Who had brown hair?" and 30 per cent, of responses of
"Will" to Who had red hair? Under-potency of All the
children except caused the formation of the connection of Will
with red hair. This wrong connection not only produced the
large crop of errors to question 6, but also worked back to
prevent the correct answer to question 5.
To question I on paragraph / we find the following an-
swers, all being more or less wrong connections leading from
the element paragraph.
'The sentence," "Subject and predicate,"
'A sentence that made sense," "A letter,"
' Period," " In every paragraph must have a period,"
'Capital," "Capital letter in the first letter,"
'A capital letter," "Commas and periods,"
'To begin with a capital," "The paragraph man was marsh."
To the question, on paragraph 7", What do you think
61 heaven's azure" means? we find the following answers from
college freshmen, in whom right bonds leading from azure
are either absent or so weak as to be suppressed by the wrong
(for the purpose) bonds leading from heaven or from the general
sense of glorified moralizing which the passage establishes.
Only about one college freshman in three thinks of so plain
a thing as the blue sky!
T
But it is to you, ye Workers, who do already work, and are as grown men,
noble and honorable in a sort, that the whole world calls for new work and
nobleness. Subdue mutiny, discord, widespread despair, by manfulness,
justice, mercy and wisdom. Chaos is dark, deep as Hell; let light be, and
there is instead a green flowery world. Oh, it is great, and there is no other
greatness. To make some nook of God's creation a little fruitfuller, better
more worthy of God; to make some human hearts a little wiser, manfuller,
happier, — more blessed, less accursed! It is work for a God. Sooty Hell of
mutiny and savagery and despair can, by man's energy, be made a kind of
Heaven; cleared of its soot, of its mutiny, of its need to mutiny; the ever-
lasting arch of Heaven's azure overspanning it too, and its cunning mechan-
232 EDWARD L. THORNDIKE
isms and tall chimney-steeples, as a birth of Heaven; God and all men looking
on it well pleased.
"The everlasting rainbow," "God's love,"
'"Peace," "God's influence,"
"Peace, universal," "God's approval of work,"
'Peace progress and justice one to an- "God's creation,"
other," "A world in which +,"
"Peace and the world purged of strife," "A place where men +,"
"Peace and happiness love and manli- "The light of the world,"
ness," "The light of a new method — of freedom
"The peace and love of man's fellow men," and happiness,"
"The peacefulness one sees in a God- "Man's outlook on life,"
fearing community," "High ideal +,"
"Harmony in life," "The clearing of mutiny,"
"Themillenium," "Perfect condition of things with every
"Appreciation of the beauty of nature," one happy and working for the best,"
"Wisdom mercy," "The work that man has accomplished,"
"Happiness," "Eternity,"
"Enjoyment," "Eternal life,"
"The happiness and contentment that "Eternal blessing,"
comes-}- " (-}- here and later means "Love your neighbor as yourself -}-,"
that more was contained in the re- "God is love. Do unto others as ye
sponse; but nothing correct), would have others do unto you."
We must then think of a pupil who answers the questions
concerning a paragraph as beset by tendencies of each word
and word group in the question to assume undue potency, to
become dislocated from its proper relations, and to call up
its past accompaniments and sequents. For him to answer
rightly means that an elaborate hierarchy of bonds is active
and that an intricate set of forces maintains a balance of
power. One may become directly aware of at least a part of
this complex coordination and subordination of tendencies if
he will note just what it implies to respond correctly to this
question on paragraph T: What does the author refer to as a
" Sooty Hell of mutiny and savagery and despair"? One may
infer it less directly but more adequately by surveying the
hundreds of different tendencies to respond which a paragraph
and question evokes in a thousand pupils and realizing that
almost or quite all of these tendencies were present as truly
in any one successful pupil, but were prevented from deter-
mining final response by some organization within them-
selves or by some guiding tendencies from without.
Sometimes the correct balance or organization reduces to
THE PSYCHOLOGY OF THINKING IN THE CASE OF READING 233
the simple case of letting one element be as potent as it may
and reducing the potency of all other elements to negligible
amounts. We then have the case of reasoning of which
James has given the classic description. But the task of
thought is, it seems, not usually to choose only one element in
the situation for potency, or to accept one only of the facts
evoked by that element. Usually there are many elements
to be let work together and many evoked facts to be used for
the purpose at hand. In our illustration, the "workers,"
"sooty," "mechanisms" and "tall chimney" all need to be
given potency to secure the response of "a region of fac-
tories," "a manufacturing community," "a factory town dis-
turbed by labor troubles," or the like, and the "God,"
"chaos," "Hell" and "Heaven" of Carlyle's grandiloquence
all need to be somewhat tempered in their tendencies to call
up the world, human nature or other things of vast scope and
great moral importance.
Elaborate as are the compositions of forces which give
thought its final motion and direction, the forces themselves
are of simple nature, being elements in situations and con-
nections leading from these elements to responses which use
and (in my opinion) satisfying accompaniments have yoked
to each element.
Three simple mechanisms — under-potency and over-po-
tency of elements, dislocation or disrelation of elements, and
wrongness or inadequacy of connections — seem to be all that
are needed to explain errors in thinking. Conversely, proper
balance and organization of elements and right bonds there-
with seem to explain correct thinking, no matter how elaborate
or subtle. Thinking and reasoning do not seem to be in any
useful sense opposites of automatism, custom, or habit, but
simply the action of habits in cases where the elements of the
situation compete and cooperate notably.
It is of course the case that, along with the balanced
action of elements, there goes an inspection and validation of
them and the ideas or acts they evoke, whereby each suc-
ceeding situation is often amended by increasing or reducing
the potency of certain of its elements, and whereby certain
234 EDWARD L. THORN DIKE
futile ideas may be cast away entirely. These welcomings
and rejectings, retainings and letting go, are however them-
selves nothing more than situation-response bonds, where the
response is attending to or turning from, cherishing, repeating,
saying no or yes to, or the like. It is also the case that the
"set" or adjustment of the organism plays a more striking
role in reasoning than it does in mere day-dreaming or routine
habit-action, but not a different sort of a role.
I conclude therefore that the general laws of human be-
havior which explain why a pupil puts his clothes on or off and
eats or leaves uneaten his breakfast explain why he succeeds
or fails in making geometrical demonstrations or scientific
researches, and that there exists no fundamental physiological
contrast between fixed habits and reasoning.
THE SIMILARITY OF BROTHERS AND SISTERS IN
MENTAL TRAITS1
BY DANIEL STARCH
University of Wisconsin
In this study, information was sought on two questions:
(i) To what extent are adult siblings (children of the same
parents) alike in mental characteristics and (2) is the similar-
ity greater in those mental traits which are directly affected
by training in school work than it is in those traits which are
not directly affected by school work?
Some experimental work has been done on these problems
with siblings and twins of various ages below adult life.2 The
purpose of the present study was to investigate the mental
resemblance of adult siblings, and with these problems in
mind two series of tests were carried out.
I. For measuring mental functions which are directly
affected by school work, the following tests were used : Speed
and comprehension of reading ability, size of reading voca-
bulary, speed and quality of hand-writing, and ability in
spelling. These tests were made in exact accord with the
methods described elsewhere3 and hence no further details
will be given here. Abilities in the four fundamental oper-
ations in arithmetic were measured by the Courtis tests,
Series B.* Ability in arithmetical reasoning was measured
by the writer's scale (' Educational Measurements,' p. 114).
II. For measuring mental functions which are not directly
affected by school work, the following tests were selected:
Two perception tests were used: (a) The well-known A-test
1 This investigation was planned by the writer. The tests themselves and the
computation of the results were made very carefully by Clara Fuller Taylor.
2 Thorndike, E. L. 'Measurements of Twins,' Archiv. of Phil, PsychoL, t3c., No. I.
Starch, D. 'The Inheritance of Abilities in School Studies,' School and Soc., 2,
608-610.
3 Starch, D., 'Educational Measurements' (Macmillan), pp. 20, 38, 60, and 89.
4 Courtis, S. A., 'Manual of Instructions' (Detroit), p. 58.
235
236 DANIEL STARCH
which determines the number of A's cancelled in one minute,
and (b) a geometrical form test which determines the number
of a certain geometrical figure cancelled in one minute on a
page of similar forms. Memory was measured by the
number of words that could be recalled after hearing a series
of ten monosyllabic nouns read at the rate of one word per
second. Motor capacity was tested by the tapping test in
which the task consisted in tapping with the right hand, as
rapidly as possible for thirty seconds.
Each test was carried out twice on two different days with
a group of eighteen pairs of adult siblings who were students
in the University of Wisconsin. The ages of these persons
ranged from nineteen to thirty-two. The tests were repeated
so as to obtain more reliable measurements of the capacities
involved than a single test would yield.
The results may best be presented in the accompanying
table of coefficients of correlation computed according to the
formula
6 sum d2
r = *^r-—r
In arranging the pairs of siblings, the older ones of the dif-
ferent pairs were placed on one side and the corresponding
younger members were arranged on the other. The scores
made by each person in each test were then tabulated opposite
each person's name. The older members of the various
pairs were then ranked by themselves in each test and the
younger ones were likewise ranked by themselves. On the
basis of these ranks the coefficients of correlation were com-
puted and found to be as follows:
Reading — speed 51
Reading — comprehension 64
Writing — speed 72
Writing — quality 46
Size of reading vocabulary 07
Spelling 05
Arithmetical reasoning 38
Addition — attempts 71
Addition- -rights 44
Subtraction — attempts 43
Subtraction — rights 29
THE SIMILARITY OF BROTHERS AND SISTERS *37
Multiplication — attempts 37
Multiplication — rights 25
Division — attempts 46
Division — rights 56
Average 42
Memory 31
A-Test 50
Geometrical form test 07
Tapping 65
Average 38
Coefficients based on ranks in all rests combined 73
Several interesting results appear in this table, (i) The
resemblance of siblings is apparently no greater in those
mental traits which are directly affected by school work than
in those which are not so affected. The average correlation
in the former group of tests is .42 and in the latter .38. This
seems to indicate that the mental similarities of children of
the same parents are due primarily to heredity rather than
to similarity of environment since the resemblance is no
greater in those traits which are more directly affected by
environment.
(2) The resemblance of siblings is approximately as great
in mental traits as in physical traits. Pearson found the
correlation between brother and brother in height to be .50
and in cephalic index (ratio of length to width of head) .49.
These correlations for physical traits are a little larger than
the ones found here for mental traits taken separately. The
correlation, however, calculated on the basis of a combined
rank for each person in all mental tests together was found to
be .73. This greater correlation for all tests combined as
compared with the correlations for single traits is due partly
to the variation of the correlations among the single traits
and partly to the imperfections in the separate tests, which are
counterbalanced to some extent in a combined ranking.
(3) Apparently the resemblance is greater in some traits
than in others. Spelling ability, range of vocabulary, and
perception of geometrical forms seem to be correlated very
slightly; whereas speed in writing, speed in tapping, and speed
in addition seem to be correlated very closely. To what
238 DANIEL STARCH
extent different mental traits are correlated by greater or less
amounts can not be stated with confidence on the basis of
the present tests. Further measurements are necessary.
The chief significance of the present results consists in
further corroborating the notion that the mental make-up of
human beings is as much a matter of heredity as their physical
make-up and that environment plays a relatively small part
in producing the resemblance of closely related individuals.
A METHOD OF RECORDING ERRORS IN FORM
BOARD TESTS
BY E. K. STRONG, JR. AND EDWARD P. GILCHRIST
George Peabody College for Teachers
The length of time required to perform a Form Board
Test has been, in most cases, the principal element in measur-
ing the performance. Various efforts have been made to
utilize the error-element also. But most of these have not
led to anything of real value except to the originator of the
scheme, himself. Yet all who have used form board tests
have come to realize that they learned as much, if not more,
from the way the test was performed as from the time it
required. The reason the errors have not been used more
has been due to the difficulty of recording them in the few
seconds one has at his disposal.
Miss Schmitt1 has recently very vigorously protested
against using time as a measure of efficiency and has substi-
tuted in its place an analysis of the ' method of attack.' She
reduces the child's method to one of three types: 'planned,'
'trial and error,' and 'chance.' The writers believe that
this analysis should lead to some very. desirable additions to
our present methods of evaluating a child's ability, but they
are not prepared as yet, to accept Miss Schmitt's rather
arbitrary method of determining what is a 'planned' or 'trial
and error' performance. However, her position raises again
the great need of devising a method by which one can keep
track of the various movements by which a child solves the
form board test. The following method has been found use-
ful in this connection and is recorded here as a possible help
to others who have been confronted by the same problem.
Suppose we are working with the Healy-Fernald Con-
struction Test A. This test can be solved by making five
1 Schmitt, C., "Standardization of Tests for Defective Children," PSYCHOL.
MONOG., No. 83, 1915.
239
240
E. K. STRONG AND EDWARD P. GILCHRIST
moves and in three seconds (our best record): or it may be
solved after a great number of moves and after a long period
of time. Ordinarily five or ten minutes is considered suf-
ficient time in which to test the child's capacities. Now, if
one uses coordinate paper, illustrated in the plate, he may
draw a diagonal line upwards and to the right for every
* placement'; a horizontal line to the right for every 're-
arrangement'; a diagonal line downward and to the right for
every 'removal.' The record illustrated in the plate tells
this following story. Three blocks were picked up, one after
the other, and placed in the rectangle: two rearrangement
moves followed; one block was removed from the frame;
333"*
FIG. i. Showing Method of Recording Movements in Solving Healy-Fernald
Construction Test A.1
two rearrangements of the two remaining blocks followed;
another block was placed in the rectangle; three rearrange-
ments followed; the last two blocks were placed in position.
1 The lower cut should have been so prepared that the two numerals " 2 " in the
center of the cut and the three uumerals " 3 " near the right of the cut should be
placed on their respective horizontal lines instead of above them. If this had been
done the distinction between a numeral on a diagonal and one on a horizontal line
would appear more clearly to the eye.
A METHOD OF RECORDING ERRORS IN FORM BOARD TESTS 241
This record can be summarized as follows: Placements — 6;
Removals — i; Rearrangements — 7; Total — 14.
If one cares to go into further detail in his analysis of
the movements, he may write a number from one to five
instead of drawing the horizontal or diagonal lines; each
number standing for a block. In this way, just what block
was involved in the movement may be recorded. Such a
record will appear as in the lower part of the figure.
Note: The blocks are numbered in accordance with their
length; blocks four and five being, of course, identical. From
the record it is clear that block I was correctly placed; block 2
was not correctly placed; block 3 was also not correctly
placed; it was moved twice and then removed from the
rectangle; block 2 was then moved twice, the second time
being placed correctly; then block 3 was replaced incorrectly
and moved three times, when it was finally correctly placed;
then blocks 4 and 5 were placed correctly.
A slight amount of practice enables one to think the
correct number when a block is touched; also to write the
numbers uphill when they are being placed, downhill when
being removed, and on the line when they are being moved
about. And all of this can be done while keeping the eyes
for the most part upon the subject.
DISCUSSION
INTROSPECTION VERSUS THE SUBCONSCIOUS
Do not the following experiments show that an assumption lies
at the base of the conclusions drawn from certain introspective
data regarding the mental processes involved:
1. An experienced observer is given a number, say 6, to multiply
by 3, and he answers 18, having had no visual nor auditory images
nor indeed any thing, so far as he could determine, to give to pro-
tocal. But if he is previously told to observe the steps of the process
9
in multiplying 9 by 4 he often reports a visual image 4 and perhaps
36
also auditory and other images in connection with the solution of
the problem. That is, in conducting the experiment in the second
way the steps of the process appear.
2. Something similar is also observed often when an experienced
observer is instructed to reverse the hands of a clock, that is, he
has nothing to report except the correct answer until he has been
told to observe the steps of the process with greater attention.
3. The same thing came out in a striking way when a student
investigating recognition once complained to me that an experienced
reagent with whom he was working was evidently a very careless
observer as he had little or nothing to give to protocal. After the
reagent had been told of the complaint and instructed to observe
more carefully the experimenter informed me that this reagent's
introspections had become more detailed, as regards the steps of the
process.
4. In reaction time experiments the introspections are usually
much fuller after the observer has been told to observe what
occurred in the Vorperiode, etc., that is, when instructions have been
given similar to those given by Ach in his reaction time experi-
ments. That is to say, in the above experiments, the steps of the
process were observable only after the reagent had been directed
to note them, that is, after they had been drawn from under the
threshold of consciousness through the method employed.
As to whether the steps of the process as they appear in the
242
INTROSPECTION VERSUS THE SUBCONSCIOUS 243
second mode of conducting the experiments just mentioned are the
same as in the first mode, one may suppose, (i) they are or (2) they
are not or (3) may have no definite opinion regarding the matter.
If one supposes the processes are the same, he is assuming, is he
not, that the processes above and below the threshold of conscious-
ness are the same, that is, that the conscious and the subconscious
are fundamentally one; if, on the other hand, he rejects this idea
altogether he holds the opposite view regarding the relation between
the conscious and the subconscious; if he does not know what con-
clusion to draw from the data obtained by the two different methods
he has no definite opinion regarding the relation of the conscious to
the subconscious. It would seem, would it not, from the data given
that the opinion one holds regarding the relation between the con-
scious and subconscious has a somewhat greater significance in intro-
spective work than has been generally supposed? In short, that
our opinion of the value of some introspective data depends upon
our view of the relation between the conscious and the subconscious,
in fact, of the subconscious itself.
LILLIEN J. MARTIN
LEIAND STANFORD JR. UNIVERSITY
THE MNEMONIC FEAT OF THE 'SHASS POLLAK'
Some years ago, through the kindness of my friend Professor
Hollander, of the Johns Hopkins University, my attention was
directed to a special achievement in memorizing which I venture to
report; since, so far as I know, it has remained unnoticed by
psychologists, and yet should be stored among the data long and
still richly gathering for the study of extraordinary feats of memory.
The facts of the case I can hardly do better than to allow the
witnesses themselves to state. And first the Reverend Dr. David
Philipson, of Cincinnati, to whom I was first referred by Professor
Hollander.
"The Babylonian Talmud" he has been good enough to write
me, "consist of twelve large folio volumes comprising thousands of
pages. All the printed editions of the Talmud have exactly the
same number of pages and the same words on each page. This
must be borne in mind in order to understand the remarkable feat
of memory about to be described. There have been, as there
undoubtedly still are, men who know the whole text of the Talmud
by heart. Some years ago one of these men, a native of Poland,
was in this country. I witnessed his remarkable feats of memory.
Thus, one of us would throw open one of the volumes of the Talmud,
say the tractate Berakhot, at page 10; a pin would be placed on a
word, let us say, the fourth word in line eight; the memory sharp
would then be asked what word is in this same spot on page thirty-
eight or page fifty or any other page; the pin would be pressed
through the volume until it reached page thirty eight or page fifty
or any other page designated; the memory sharp would then mention
the word and it was found invariably correct. He had visualized
in his brain the whole Talmud; in other words, the pages of the
Talmud were photographed on his brain. It was one of the most
stupendous feats of memory I have ever witnessed and there was
no fake about it. In the company gathered about the table were a
number of Talmudic experts who would readily have discovered
fraud had there been any. The technical name which was used by
the Jews of aforet;mes to designate these memory experts was
Shass Pollak; Shass is the abbreviation for the Hebrew terms for
the Talmud, and Pollak is Pole; nearly all these memory experts
244
MNEMONIC FEAT OF THE 'SHASS POLL A K' 245
came from Poland; a Shass Pollak then is a Pole who has memorized
the entire contents of the Talmud and is able to give exhibitions of
his mnemonic powers like those mentioned above."
And next let me quote from Judge Mayer Sulzberger, of Phila-
delphia, who in answer to my inquiry, wrote as follows:
"I have met but one 'Shass Pollak' in my life. He was brought
into my library one evening by a friend. I conversed with him and
experimented upon him.
"After he had been introduced as the expert in question I
expressed some curiosity with perhaps a mien of incredulity. He
was eager for the fray.
"You are of course aware that all (or nearly all) modern editions
of the Talmud are paged alike and printed alike, each page beginning
and ending with the same word in all the editions.
"I went to the case and took out a volume of the first edition
which has its own paging not followed by the other editions. He
made an automatic dive for a word in a particular part of the page,
and lo! it was not there.
"Confounded by this unexpected event, he thought at first
that this was not a Talmud I was showing him; and when convinced
finally that it was, seemed to bear it some resentment for its im-
proper behavior.
"I then brought out the corresponding volume of an ordinary
edition and he undoubtedly made good.
"He would take a pencil and merely glancing at the page put
it down anywhere and without looking told the word on which his
pencil had lighted. This he did over and over again. There is
no reasonable ground for the suspicion that he saw the words. I
watched him closely and am convinced that he did not. He had,
I feel sure, a perfect image of the page and the position of every
word on it in his 'head.'"
Finally, let me give the testimony of Dr. Schechter, of New York,
the late President of the Jewish Theological Seminary of America —
testimony the more interesting in that while it depends upon the
recollection of an experience many years ago, yet it is an independent
account of the same kind of testing which Dr. Philipson reports
— namely, by pricking through the pages — and consequently con-
firms the opinion of Judge Sulzberger that the success of the 'Shass
Pollak' who was tested merely by pencil was not due to a sly catch-
ing of the word by eye.
President Schechter stated to me by letter that once he had
246 GEORGE M. STRATTON
come across a 'Shass Pollak' but that it was too long a time ago
to give an account of him with definiteness. "It is at least forty-
five years since the incident occurred," he wrote. "What I re-
member was that he could tell you the contents of every page of
the Talmud by heart. I remember also that the people amused
themselves by prying a needle into any volume of the Talmud,
and he could tell exactly the word on which the needle touched. But
I also recollect distinctly that it was nothing more than a verbal or
rather local memory, the students all maintaining that he knew
very little about the meaning of the contents, their interpretation
and application. I heard afterwards of many similar ' Shass Pollaks,'
but it is a fact that none of them ever attained to any prominence
in the scholarly world."
This absence of any scholarly grasp of the contents thus memor-
ized, of which President Schechter speaks, also appears in the
judgment of Dr. Philipson. "I looked upon his achievement at
the time I witnessed it as purely mechanical," he writes. "It is
quite likely that he could not interpret the Talmud though he knew
its contents by heart." And Judge Sulzberger, when proposing to
his 'Shass Pollak' that he use his knowledge to some scientific or
literary end, was listened to with respect, but nevertheless received
the impression that such proposals were deemed by his man to be
nonsensical.
All of which confirms the oft-repeated observation, that such
extraordinary powers of memory may exist in a kind of intellectual
disproportion where there is no corresponding development of
other powers — where, indeed, there may be an actual stunting of
other powers and interests; as though the mind had 'run' to mem-
ory, and been enlarged here at the expense of other functions.
As to the more precise amount of matter that was memorized,
it should be noted that a page of the Babylonian Talmud consists,
as my colleague Dr. Popper, has pointed out to me1, of the text
proper, called the Gemarah, and printed as a more central portion
on the page, and of a commentary printed below and around this
text. Upon special inquiry whether the mnemonic feat applied
only to the Gemarah or included also the Commentary, Dr. Philipson
1 Professor Popper has also referred me to the articles "Talmud" and "Mnemo-
nics" in The Jewish Encyclopedia for evidence that at one period the Talmud was
handed down solely by memory. The feat of the Poles here recounted may therefore
be regarded perhaps as the survival of a custom among early Jewish students in many
and widely-separated communities. The work of Briill, Die Mnemotechnik des Tal-
muds, Vienna 1864, should also be cited.
MNEMONIC FEAT OF THE 'SHASS POLL A K' *47
states that the test which he witnessed was upon the Gemarah only;
and Judge Sulzberger is of the opinion that this was also true in
the case that came under his observation. Even so, the task must
have been a stupendous one; the amount of reading-matter upon
each page is still great, and the number of pages is enormous.
In closing may I express my thanks, in which other students of
psychology will certainly unite with me, to the gentlemen who have
so generously given the facts above recorded.
GEORGE M. STRATTON
UNIVERSITY OF CALIFORNIA
VOL. XXIV. No. 4 July, 1917
THE PSYCHOLOGICAL REVIEW
VARIETIES OF PSYCHOLOGICAL EXPERIENCE1
BY JOSEPH JASTROW
University of Wisconsin
A commemorative occasion justifies a retrospective, though
hardly a reminiscent mood. The personal justification lies
in the fact that I speak as one of a small group — in this and
all countries — who have held a monogamous professorship
of psychology for a quarter century. The contrast of then
and now stands forth partly as a shift in intellectual temper-
ament, partly as a diverging succession of interests. Both
are responsible for the historical moving-picture, which to
our near vision still flickers by reason of imperfect fusion.
The dominant interest under which I began to profess
psychology was clearly the experimental one; it set a novel
and a positive programme. Equally assertive was the
physiological plank. The two stamped one's alliance, in a
sense made one a partisan. The emblem of the one was the
laboratory, of the other an evolutionary faith and a sense of
the reality of the body in the affairs of mind. They might
have been emblazoned as a Hipp chronoscope rampant, and a
copy of Darwin couchant. The bearer of this coat-of-arms
was in many quarters under suspicion. He was more than a
radical, less than a renegade. By implication he was chal-
lenging the accredited "mental science" of the colleges,
which was a branch — in some cases a stunted twig — of
philosophy, and a perquisite of the president. At educational
gatherings — even more inconsequential and vaporous then
than now — he was asked to defend the superiority of experi-
1 Address given on the occasion of the celebration of the twenty-fifth anniversary
of the American Psychological Association, New York, December 28, 1916.
249
250 JOSEPH JASTROW
mental over rational psychology; also to indicate what was
the fate of the soul under the new regime. To the philoso-
phers the psychologists seemed needlessly young and irre-
sponsible; to the men of science they seemed out of focus or
tangential. Now that the philosopher and the biologist have
become our ready allies, the attitude toward us as militant
invaders is almost forgotten, and may well remain so.
The credentials upon which psychology added a star to
the united sciences were significant ones; it advanced from
territorial governorship to statehood by patent of ancient
right reestablished under successful pioneering. It was not
— as in the case of sociology — a squatter settlement in un-
occupied land where the adjoining sciences failed to meet, but
the declaration of independence of a domain quite too large
and distinctive for colonial status. The first insistence was
naturally upon technique. The novel sight, which to some
was amusing, was not the philosopher descended to earth
from his mythical habitat in the clouds, but actually donning
overalls and using his hands. The garb seemed strange and
lowly, even grimy. But rapidly enough the new psychology
— like the new woman — became a more familiar and less
forbidding Erscheinung, even revealing the eternally human
traits. Problems followed the eclectic clue of technique.
The early Wundt was compilational rather than systematic,
though the Teutonic Grundlichkeit extended to architectural
plan as well as building material. Wundt set the interests of
the first group of American students of modern psychology,
in which ancient and honorable body I may claim a place.
The rallying point was the Johns Hopkins University — itself
as new as psychology — under the leadership of the author of
"Aspects of German Culture," G. Stanley Hall. He, how-
ever, found a small group of students already waiting; he
quickly attracted others, and then (after five years) trans-
ferred his influence to Clark University. Stanley Hall's
doctorate was the first given in psychology; Harvard Uni-
versity, 1878. My own was the second and the first given at
Johns Hopkins specifically in psychology.
In so far as American psychology was a native product, it
VARIETIES OF PSYCHOLOGICAL EXPERIENCE 251
reflected the pioneering spirit, the spirit of William James.
No speaker on a commemorative occasion would forego the
privilege of placing a wreath upon his monument. He was
the first among us to bear the title professor of psychology;
he holds that position not by priority but by preeminence.
He stands as the exponent of the value of varieties of psycho-
logical experience. He made clear in his own person the
intimate dependence of pursuit upon temperament. It was
peculiarly fortunate for the recognition of the new psychology
that its academic status was assured by the acknowledged
leadership of so commanding an academician, so distinguished
a scholar, so great a man. It was equally fortunate for the
career of psychology in America that the stamp of the
Jamesian genius pervaded its progress and directed its un-
foldment to the desirability of seeing psychology steadily and
seeing it whole.
The development of a new discipline in a new country
encounters directly the social situation. Of first consequence
was the accredited recognition by the universities and colleges ;
which meant the establishment of courses that could play a
worthy part in the curriculum. In the conflict of classics and
science, which in the days I am recalling was reaching its
declining and conciliatory stage, psychology occupied an
advantageous position, though it was exposed to the danger of
falling down between two stools: the unrecognized stool of the
laboratory, the exalted throne of philosophy. Yet the trend
of opinion was pacifist and not militant. In the readjustment
of the curriculum psychology may more frequently have been
offered as a compromise than as a solution. But all this was
helpful; the method of intelligence would have been better;
but the method of trial and error was acceptable. Prag-
matically the situation offered positions to those who would
undertake the new training. In the words of the college song:
"One day the summons came out of the West;
'Get Ph.D's and come,' rang the request."
I may readily indicate the status of teaching at that time
by mentioning that when I responded there was no elementary
252 JOSEPH J AST ROW
text available. Carpenter's ' Mental Physiology' both in
title and content was about the only widely read book that
reflected the new data and the new interpretation. The
pioneer contribution was the invaluable work of Professor
Ladd; and this celebrated its quarter century a few years ago
by the appearance of the Ladd-Woodworth edition. The
text implied a more generous course and a better type of
preparation than the college curriculum afforded. To the
teacher of psychology this monumental work was a daily
support, and to the ablest students, often of nearly the same
age as the teacher, a guide, philosopher and friend. The
work of James appeared three years later. With the labor-
atory and seminar the lecture and research and the training
course, and then the American Journal of Psychology, at
Johns Hopkins; with the large personal influence and the
.systematic work at Yale, the decisive prominence of psy-
chology in the graduate work in philosophy at Harvard,
American psychology was launched. The auspices could not
have been more favorable. The varieties of psychological
experience represented by these creative personalities assured
the development of psychology in this country a breadth of
outlook, a soundness of technique, and a vital contact with
the dominant intellectual interests, which have been most
important assets. They may well stand central in our com-
memorative respect, a tribute to Hall, Ladd and James.
For a time the development of courses, the preparation of
texts, the establishment of academic positions kept the small
band of psychologists busy; also many of them could secure
their positions only by a willingness to share the responsibility
for allied teaching, mainly in philosophy, logic and ethics.
Indeed these disciplines as commonly represented the major
preparation, to which psychology was added. But the out-
ward and visible sign of the psychologist was his laboratory;
and the growth of these both for instruction and research
represents the most distinctive American contribution.
There is no parallel to it elsewhere. In a direct sense the
laboratory is responsible for the largest increase in varieties of
psychological experience. It set the experimental attitude
VARIETIES OF PSYCHOLOGICAL EXPERIENCE 253
in inquiry, and this of all the several independent factors un-
mistakably shaped the career which we are passing in review.
It has always been true in the history of science that the kind
of questions men ask is determined by the facilities which they
have acquired in answering them; though it is equally true
that the sense of limitation of facilities has been the motive
force in extending them to the inclusion of larger interests
and newer problems.
I propose to touch only upon the larger varieties of psy-
chological experience thus furthered. They may be sum-
marized under five overlapping waves. The first is the direct
exploration of the intimate structure of mental process, the
direct analytical interest of the man of the laboratory. It
immerses him in technique and device and the envisagement
of problems; it immerses him unduly unless it carries with it
the poise conferred by interpretation.
The second is the comparative interest, which in recent
years has almost entirely reconstructed the view of animal
behavior, and has yielded its interpretative product in the
"behaviorist" position. Its high-power study of the begin-
nings of mind gives it the apect of psychological histology;
but its far-reaching conclusions extend to the entire evolu-
tionary programme of the mental life. It forces upon
psychology the problem of the distinctive nature of the
human endowment, and the stages of differentiation in the
onward series.
One may place third and thus in the central position the
growth, almost the overgrowth, of applied psychology. This
reflects the practical stress of the environment and the prag-
matic temper of the American Weltanschauung. It was born
in the laboratory. The term l mental test' is so distinctive of
English, which in this case is American usage, that it has been
adopted in German, French and Italian literature. It arose
from the consideration that analysis is not only of the factors
of a process but of their place in the individual psychology.
The work of Galton should be recalled on this occasion for the
reason that while the biological aspects of his versatile studies
attracted most attention in England, the psychological sig-
254 JOSEPH J AST ROW
nificance of his methods and results was more influential on
this side. He emphasized the application of psychological
tests in combination with anthropological traits. He es-
tablished the first laboratory for such purpose; and it may be
recorded that the first installation on this side was in connec-
tion with the World's Columbian exposition in Chicago, 1893,
where also the visible embodiment of experimental psychology
was shown to the public.
In applied psychology two aspects of psychological ex-
perience were involved: the first emphasized that the training
of mind could proceed wisely only upon a knowledge of mind;
and the teacher was referred, not without misgivings, to
psychology. The second aspect was focused upon the ability
of the tests to reveal individual capacity. Thus were laid the
foundations of educational psychology which in the present
outlook looms so momentously large that it heralds the divid-
ing line of further specialization. It is making for a competi-
tive share in professional status; the future division of function
is indicated, though localization is somewhat uncertain.
With the present expansion of departments one exponent is
likely to assume responsibility for the analytical, theoretical
disciplines, and another for their applications. Upon the
warmly disputed question whether pedagogy is a science or an
imposition, and its pursuit a profession or a misfortune, we
may maintain an impeccably neutral position. We cannot
overlook the fact that the psychology of the schooling proc-
esses, by sheer force of practical importance, is entitled to a
commanding place in the training of teachers; while collater-
ally the investigations thus resulting give promise of rounding
out the analyses of the learning processes to the great benefit
of general problems of primary import. Here lie varieties of
psychological experience of sterling value. To some the
field has the appearance of extensions of city plots in unborn
suburbs, neatly staked out in building-lots, with cement
sidewalks but no habitations. The promoter in psychology is
not unknown; but the new settlement seems less speculative
to one who has a retrospective standard.
A distinctly settled section is that of tests of capacity, the
VARIETIES OF PSYCHOLOGICAL EXPERIENCE 255
Binet-Simon addition. Applied individually the test invites
a diagnostic use. The vocational pressure is intense; it should
be encouraged and only its premature and cock-sure decisions
resisted. It is characteristic that the sins of commission and
omission alike are far more common among the practical
cultivator of trees than among the more theoretical con-
servator of forests. Get-wise-quick methods offer the most
lucrative rewards to the needy psychologist. At the moment
we are besieged by requests to enlighten men of affairs how to
choose employees, how to detect capacities, and by individuals
how to increase mental efficiency. Only one who is blind to
the lessons of history can fail to see the dangers in this great
white way of psychology. One cannot look upon phre-
nology, physiognomy and the reading of character as merely
the slums of psychology. Whatever the rating of the fakir,
his customers come from all shades and grades of education.
Here may be gathered interesting varieties of psychological
experience; and a candidatus philosophicus in psychology may
find a promising thesis by illuminating the psychology of
fraud. Yet psychology will be shirking its social responsi-
bilities if it declines to cross an unsavory threshold. It is not
a sign of virtue to fear to tread, just because frauds rush in.
Too proud to investigate is not a proper attitude. The prob-
lem of vocational and individual fitness is a wholly legitimate
and, for the cautious and modest psychologist, an engaging
pursuit. The reputation of the psychologist will depend
upon the restraint with which he exercises authority, and
pronounces judgments.
The practical varieties of psychological experience are
worthy of respect in their own right. We all know that the
road from theory to practice is the more indirect the more
complex the situation. Qualities are far more generic than
their applications. We must insist upon the legitimacy of
the psychological perspective and decline to assume while
yet we respect that of the practical inquirer with a narrower
interest. A man will become a persuasive salesman, or a
shrewd employer or a good teacher far more regularly upon
the basis of a general equipment than of correspondence
256 JOSEPH JASTROW
courses. At all events it is in the interpretation of the under-
lying qualities of men that psychology, pragmatically dis-
posed, finds its metier, however ready to utilize the trends of
employment, and to direct inquiry to practically significant
relations. The two perspectives must differ. The variety
of experience is valuable; the forms of experience imposed by
modern conditions acquire a peculiar importance; but none of
these interests should distort the far more significant varieties
of an historically larger and intrinsically deeper experience.
In precisely the same sense in which the sociologist, fixing his
attention upon modern conditions, will be handicapped by a
narrow vision if he forgets that what he is studying with a
specialized interest is in reality a transferred biological situ-
ation to be interpreted under the principles of biological re-
lations, will the applied psychologist become a mere trained
craftsman if his sense of design is unilluminated by the inter-
pretative insight conferred by long immersion in the principles
of psychology.
I have chosen deliberately to enlarge upon the practical
varieties of psychological experience, for the reason that in
this vista the retrospective view directs the enterprise which
will plan the highways of the future. The attitude of Ameri-
can psychologists toward the possible and desirable applica-
tions of their pursuits, even the mode of capitalization of their
personal value for public consumption, seems to me so pe-
culiarly important that I have chosen to project an "insert"
on a larger scale in the moving picture which I am unreeling.
The psychologist, I repeat, must insist upon complete author-
ity as an architect of his science; what consideration shall he
give to the expressed needs and wishes of a possible clientele?
The practical and the theoretical perspectives are distinct;
how shall they be made to converge, and yet retain that
singleness of vision which is indispensable to a solid, realistic,
stereoscopic effect?
It requires no prophetic but only a presbyopic vision to
foresee that the insistent demands of practice will form a
league to enforce attention. The psychological practitioner
is coming; upon us rests the responsibility that, when he comes
VARIETIES OF PSYCHOLOGICAL EXPERIENCE 257
he shall be in no measure a quack or an opportunist, neither
papal nor encyclopedic, nor pretend to be all things to all
men. The training of psychologists cannot undertake the
development of geniuses, who as a rule have found the
academic environment unstimulating. A scientifically minded
sense of proportion is the central equipment; with it must
be combined a clinical sense for the recognition of vari-
eties of experience when encountered. I resent the implica-
tion that because a large amount of money is spent in ad-
vertising, the psychology of advertising thereby gains in sig-
nificance or importance. I welcome the fact that the actual
interest in advertising supplies a variety of psychological
experience which we may utilize to the full. Problems that
loom large in application may have but slight illumination in
principle, and be rather barren in enlightenment. The
analysis of a practical occupation into its underlying factors
is the legitimate work of the psychological practitioner;
advice will gain in value as it strikes root in a soil enriched by
scientific cultivation. The tendency which most I deplore
is the neglect of the wider and the truer interpretation, because
in some of its aspects it is less amenable to the rigid technique
of the quantitative method. Falling in love with technique
may be a pleasant but is hardly a rational indulgence. It
forms one of the temptations, the idols of the practical mind.
You see it pitiably at work in the pursuit of efficiency, with
the result that there is more attention paid to cost-accounting
than to the value of the article when produced. You see its
menacing shadow thrown across the academic portals in the
impertinent attempt to measure service by unit-hours and
neglect quality and all the finer values incommensurate with
the crude and irrelevant yardstick. If the method is con-
tinued, education will have but one purpose, ambitious and
charitable at once: to make people efficient though incom-
petent. My plea is for the recognition, selection and culti-
vation of a psychological competency for practitioner and
theorist alike; which is a plea, in an expert sense, for the vari-
eties of psychological experience.
I have referred to the danger of mistaking a quantitative
258 JOSEPH JASTROW
result for an important one, of supposing that what is measur-
able is significant; likewise to the danger of supposing that
what is practically demanded is by that right entitled to a
large place in a scientific perspective. With these is combined
the danger of proceeding to far-flung battle lines of conclusions
upon a slender campaign of experimental results. None of
these dangers operates simply; they combine subtly and
intrude subconsciously, as is the manner of fallacy and her
tribe. I can point the moral most quickly by using a tale for
adornment, though I run the risk of stepping upon toes, the
owners of which I respect. The extent of my hardihood will
be clear when I say that I shall illustrate in terms of that
vexed question of the mental differences of men and women.
On the basis of well-designed experiments one observer con-
cludes that women are less disposed than men to be affected
by argument. In deciding whether one aggregation of
markings contains more or fewer individual components than
another, the women proved more tenacious of their original
opinions (whether this is consistency or obstinacy is an unwise
question) than did the men. Ergo women are less desirable
than men as members of a jury. In citing this bit of evidence
I am fortunate in that I agree with the significance of the
findings; but I remain wholly unconvinced by the conclusion.
The second illustration finds me in the reverse attitude.
On the basis of a painstaking and well-devised series of sen-
sory and mental tests, the convergence of results measuring
specific capacities proves to be far more striking than the
divergences of men and women. Upon these data is based the
conclusion that intellectual distinctions among the sexes do
not exist; they are either the result of imposition of masculine
dominance upon femine complacency, or the prejudiced views
of tradition. The comprehensive evidence of the varieties
of psychological experience embodied in the history of culture
is cavalierly disregarded. Everything is ruled out of court
except the findings of the laboratory in parallel columns of
figures. The contributions to the subject are disposed of as
belonging either to the literature of fact or the literature of
opinion; which to my view is at once a specious and destruc-
VARIETIES OF PSYCHOLOGICAL EXPERIENCE 259
tive distinction. Your quantitative fact, however exactly
determined, must pass the judgment-seat of interpretation,
personified in the venerable and wordly-wise figure of ex-
perience; then only will its contributive value appear. On
the other hand opinion may be as valueless as gossip, and as
important as any brand of truth accessible to the present
generation. More particularly, the best of opinion is founded
upon exactly the same appreciation of precise investigation,
is imbued with precisely the same scientific method, reflects
a parallel training and allegiance to accredited principles, as
that which guides the experimental devotees. To disregard
such contributions and to ignore historical experience on the
strength of data authentic in their own province but only
modestly significant outside of it, is to violate conspicuously
the appreciation of varieties of psychological experience,
which constitutes the most important equipment for the
loyal psychologist.
Not to court misunderstanding, let me explain that I have
properly selected instances of unquestioned authority, and
also that the questionings which I have raised are directed
with unbiased neutrality to point out the prejudice that is
invited by generic interpretation, and by the neglect thereof.
If we enlarge small findings to large conclusions we exceed
our warrant. If the experimentalist insists upon the supreme
value of his experience above all other varieties, proposes to
disregard scientifically rigorous thinking expended in other
problems than his, is convinced that the quantitative pattern
is the only authentic one, that amenability to measurement is
the indispensable passport for psychological citizenship, the
future of psychology faces an undesirable and unnecessary
impoverishment. Moreover it is just because the promising
growth of the applied field favors so largely the sharply defined
and technically interesting varieties of psychological ex-
perience, that the pure experimentalist should safeguard the
comprehensive and broader aspects of his function.
With pardonable overemphasis I have cited the logical pro-
cedures leading to an agreement with Mme. de Stael: les dmes
n'ont pas des sexes. According to this view minds have no sex;
260 JOSEPH J AST ROW
according to Freud they have little else. And thus we reach
a further variety of experience in the abnormal. The as-
cendency of the Freudian movement occupies the head-line
in this section of the revue; whether by reputation or by no-
toriety critics cannot agree. Either view pointedly illustrates
the complex significance of varieties of psychological experi-
ence and of the attitudes that lead to their favor or disfavor.
Freudian psychology must be saved even more persistently
from its friends than from its detractors. As I grasp its
bearing, it forms an important and essentially true contribu-
tion made by the wrong men; its germinal ideas are sound
despite the loosely woven evidence. To some the Freudian
orchestra makes unseemly noise and nothing else; others hail
it as the music of the future. Personally I am convinced that
the acceptable Freudian sonata remains to be written; it will
be composed by one imbued with the spirit of its method and
possessed of a rare sense of the value of phrasing. The sex
motif will be less insistent and strident, not silenced or
ignored but sublimated. I have every sympathy with those
who are nauseated by its seemingly cherished obscenity, and
irritated by its seemingly malicious slander of the human
mind. But my logical conscience rebukes me by a reminder
that any such attitude is irrelevant. The Freudian principles
and the Freudian mechanisms must be considered for their
value as varieties of psychological experience, quite apart from
the bad form and bad taste and bad logic of their support
and supporters; the merit of the cause or the campaign and of
the tactics or munitions must be judged separately. The
Freudian reconstruction — for a time obscurely and disparag-
ingly received, and only recently advancing to a conspicuous
place — has vitalized a large realm of observation bearing
upon the abnormal but equally valid for the normal experi-
ence. In every future retrospective view its place is secure.
For the moment it may impress us as a capricious, disorderly
vers libre, a libel upon the fair name of poetry or psychology;
when the exotic and chaotic and neurotic elements have
given way to fairer expression its contributions will be more
fairly judged and seen.
VARIETIES OF PSYCHOLOGICAL EXPERIENCE 261
My concern with the abnormal is limited to its extension
of the varieties of experience, and their interpretation for
normal use. This the Freudian doctrine attempts, ifi
attempts it practically in the method of psychoanalysis, but
confines the procedure to the disclosure of entanglements and
disqualifications that impede and distress normal functioning.
Yet psychoanalysis is a broader practical procedure, related
to an applied psychology and the allied interest of diagnosis
in all difficult situations. It embraces the experimental
methods in its use of the association-material. It brings to
the fore the province of the subconscious which constitutes
one of the most significant enlargements of psychological
experience, but one readily misinterpreted in favor of unas-
similated hypothesis. The inclusion of the domain of the sub-
conscious stands as one of the most significant annexations
in the retrospect — a distinct settlement of territory wrested
by a difficult exploration in an uncertain jungle.
Going back twenty-five years, one comes upon the thick of
a fray that almost threatened a world-war in psychology.
For the first five or ten years of my professional experience the
popular view of a psychologist was a ghost-hunter. Psychical
research held the field of popular favor; that a laboratory was
anything else than a seance-chamber seemed incredible. The
psychologist had to listen to inane stories of coincidences, that
spoiled many a promising dinner-party. When he couldn't
explain a lame and distorted tale, his inconsequence was made
shamelessly and publicly evident. Scepticism was considered
a mask for ignorance, and an unpleasant substitute for the
lie direct. But the most remarkable aspect of the movement
was the hold it gained upon men eminent in science — but not
in psychology — through whose advocacy psychical research
attained a prestige far beyond that accorded to psychology
as we know it. Indeed in many quarters the interest in psy-
chology was prompted by a hope that it would solve questions
foremost in the minds of psychic researchers — such as telepa-
thy and the survival of personality. That in some sense
significant varieties of psychological experience pervaded
that racially old field of belief, that persistent recurrence of
262 JOSEPH J AST ROW
reversal of orderly mental sequence, was a position variously
defended. All this has in these days the flavor of dried me-
mentos. Yet it was only a half a dozen years ago, in these
scientifically dedicated precincts, that a small group of us
found it necessary to allay popular and learned unrest in
regard to the specific mechanism by which a homemade bass-
wood table defied the law of physics in the presence of a
much heralded and sponsored Neapolitan and of a company
sufficiently impressed to pay handsomely for each levitation
of a skillful foot. The experience was at once humiliating
and enlightening. Fraud-hunting as an indoor sport may
have some zest if not value; or it may become a social obliga-
tion. It has helped to dissipate the surviving myth that the
man of the academy is wholly lost in the mazes of a wicked
world. Those of us who in the older days protested against
the profanation of psychology, and were often snubbed for our
insolence, may properly indulge in a moment of congratu-
lation to the younger psychologists who have not this role
to play. Yet the true interpretation of the phenomena that
psychical research attempted to invite to unlawful secession
remains a part of the gain of abnormal psychology.
The concluding phase of the composite psychological wave
is represented by the psychology of social relations. Its
pursuit obviously requires a capacity for broad interpretation,
for the analysis of the deepest motives of human conduct, for
the envisagement of underlying similarity of situation despite
complexity of circumstance, for catching psychology on the
hoof, and sensing its living products and pulsating throbs;
my cental plea for catholicity of experience receives here its
most direct justification. The laboratory with its simplified
and scheduled analyses finds its corrective in the intricate
worldly composite of conflicting forces. Social psychology
represents the most elaborate phases of an applied science.
It is applied in the tests of life. It quickens every interest
that is entrusted to the psychologist, and justifies as it am-
plifies his problems. For it makes clear in how many direc-
tions psychology has a voice with other concerns, makes it
clear that to be a psychologist implies a capacity to seize the
VARIETIES OF PSYCHOLOGICAL EXPERIENCE 263
psychological value of experience which to other interests
presents different aspects and appeals, — to single out the psy-
chological instruments in the orchestration of life. All life
is of one living; the psychologist, far less than the devotee of
other specialties can afford to diagram or artificialize his Fach.
He should be in the world and of it, in the best sense an inter-
preter of human values, capacities, enterprises; for his is the
duty of directing and safeguarding the precious mechanisms
upon which all living proceeds. Such must be his insight and
his training that he may be rewarded by the respect of his
fellows in science, his cooperators in practical affairs.
I look to the increasing study of social psychology for a
redemption from too restricted specialization, for a balance to
the dangers of absorption in technique, for a compensation
for the limitations of the quantitative method. A sense of
the reality, of the richness and fullness, the complexity and
conflict, the growths, changes and transformations of the
mind's products permeates this field as no other. In a measure
it projects the culmination of psychological experience; it
projects culture and the vaster problems of political and social
striving as a psychological evolution, testifying to the great-
ness of the forces of mind. And we of to-day are witnessing
the largest and most appalling issues of estrangement in
ideals, sentiments, allegiances, that the world has faced; the
psychology of war must be considered in the establishment of
enduring human relations. The world is going to be wisely
ruled, the endeavors of organized men more sanely directed,
the errors of the past less disastrously repeated, if a body of
men find participation in the direction of affairs possessed of a
psychological discernment; for this insight is as indispensable
to modern conditions in certain relations as is an economical,
a political, or a business sense in others.
Lying close to this domain and making a parallel appeal
is a body of knowledge, engaging in itself and vital in its
applications, which I select for special consideration. I do
so because it sets forth so amply and so pointedly the quali-
fications of the psychologist; also because it presents manifold
relations to all the several divisions of the present review,
264 JOSEPH J AST ROW
and thereby points the moral of my thesis. Dessoir in his
"History of Psychology" recognizes it as one of the three
great interests in the science of mind, with ancient antecedents
and constant influence; he calls it psychognosis. In modern
relations it may be viewed as an oblique or irregular section
through the entirety of interests that form the composite of
psychology; the specific disciplines consider the several fields
more directly and disinterestedly. The nature of individual
differences, the sources of these in human trends, their ex-
pression and emphasis in historical circumstance, their
liabilities and possibilities in cultivation and decay, their
contributions to human institutions — are all involved in
psychognosis. The directions of temperament, the founda-
tions of character, the total determinations of capacity and
career form its subject matter. The temper of its pursuit is
practical, but always with that wider and deeper foundation
in varieties of human experience, which sends it back for
analysis and authority to one and another discipline of
psychology. It develops congenially under the type of
interest that supports social psychology, has its strongest
affiliations there; but it utilizes the entire range of psycho-
logical science and requires a catholic interpretation of the
psychologist's function. Its restatement under the tremen-
dous enrichment of the last twenty-five years of psychological
investigation is an urgent desideratum of the present; and
this obligation will do much to invite psychologists to that
attitude toward their province which shall not be provincial
but in the worthiest sense cosmopolitan.
The prospective and the prophetic venture is more en-
gaging than the retrospective; commemorative occasions
invite a Janus-faced, judicial attitude. The westward course
of empire returning upon itself sets the gaze upon the east
once more; with the rounding of the circle the conquests of
the future turn to an inward advancement, to the perfection
of the human equipment that comes from a comprehension
of its nature, origin and history. In this vast reconstruction
the psychologist comes to his own. He expresses and confers
upon his disciples an interpretative sense, which in the be-
VARIETIES OF PSYCHOLOGICAL EXPERIENCE 265
wilderment of change traces the orderly relations of law.
Whatever the progress of the future, he recognizes in Kipling's
words, that
We can bring no more to living,
Than the powers we bring to life.
THE NEED FOR SOCIAL PSYCHOLOGY1
BY JOHN DEWEY
Columbia University
On the surface it is just coincidence that the foundation
of this association and the publication of the 'Principles of
Psychology' of William James were so nearly contemporan-
eous, their respective dates being, as you know, 1891 and
1890. In view, however, of the depth and breadth of the in-
fluence of James, we who are celebrating to-day our twenty-
fifth anniversary are at liberty, I think, to consider the coin-
cidence as more than chronological, and to date back by one
year the gestation of our association. At all events, it would
be ungrateful to engage in any discussion of the past and
future of social psychology without recalling the few rich
pages of the 'Principles' which are devoted to the social self,
and, in the discussion of instincts, to the native reactions of
human beings in the presence of one another. Big books
have been written since which are hardly more than an am-
plification of suggestions found in these few pages. When,
for example, a few years later, the Socius became the hero of a
psychological drama, not many recalled that he had already
been introduced under that very name in the pages of James.
Again it is outwardly a mere coincidence that the work of
Tarde on the 'Laws of Imitation' was published in the year
in which the 'Principles' saw the light of day, and that
practically all of Tarde's work fell within the decade lying
between 1890 and 1900. But behind the pure coincidence
there was the recognition of the need for social ends of a more
scientific treatment of collective human nature, and the
important role of psychology in building up the new social
science. While James confined himself to pregnant sugges-
tions concerning the new forms which human experience and
1 Address given on the occasion of the celebration of the twenty-fifth anniversary
of the American Psychological Association, New York, December 28, 1916.
266
THE NEED FOR SOCIAL PSYCHOLOGY 267
selfhood take on because of the presence of other human
selves, Tarde attempted an ambitious interpretation of almost
all facts of social organization, progress and degeneration in
terms of certain rubrics to which he gave a psychological
quality. For more than a decade his work and that of his
followers in France and in the United States — among whom
we may cite in diverse directions Baldwin and Ross — domi-
nated social psychology and almost sociology. I shall not
rehearse the old discussions about Imitation as a psychological
fact and a social force. I shall assume with most of contem-
porary psychological critics that as a descriptive and explana-
tory conception it misplaced emphasis and tended to distort
facts. But nevertheless we cannot minimize the immense
power of this stage of social science in popularizing the idea of
social psychology, and in bringing into recognition many
facts, such as the importance of prestige, fashion, sensitive-
ness to the beliefs of others, the difficulties which innovation,
no matter how reasonable, has to meet, etc., facts which are
permanently imbedded in social science. Tarde himself was
certainly one of the most stimulating and varied of writers,
and I do not think we shall ever outgrow some of his contri-
butions, although to my mind they are found rather in logic
than in psychology — such as the necessity for reducing the
gross phenomena of social life into minuter events which may
then be analyzed one by one. The most fruitful of his psy-
chological conceptions was ahead of his time and went almost
unnoted. It was that all psychological phenomena can be
divided into the physiological and the social, and that when
we have relegated elementary sensation and appetite to the
former head, all that is left of our mental life, our beliefs,
ideas and desires, falls within the scope of social psychology.
I hope I may find general agreement in pointing to the
work of McDqugall and Thorndike respectively as indicative
of the next great force in social psychology, together with such
writings as those, upon the social side, of Graham Wallas.
Aside from valuable contributions in detail, the significance
of these contributions lies, to my mind, in ^recalling social
psychology from the wrong track in which the Imitation and
268 JOHN DEWEY
Suggestibility schools had set it going. For those schools
gave the dawning science a wrong twist in carrying over into
science the old popular and practical antithesis of the indi-
vidual and the social, and thus setting up two independent
and even contrary sciences — individual and social psychology.
As a concrete illustration of the absurd^results to which this
antithesis led, it is perhaps sufficient to refer to those bizarre
writings on the psychology of the crowd in which it was
assumed that the psychology of the individual left to himself
is reflective and rational, while man's emotional obsessions
and irrationalities are to be accounted for by the psychology
of association with others. From the root of all such aber-
rations we were recalled the very moment the problem was
presented not as one of the relationship of a mythical
psychology of an isolated individual mind to the even more
mythical psychology of a mass or crowd or public mind, but
as the problem of the relationship of original or native ac-
tivities to acquired capacities and habits. Henceforth our
social psychology is placed upon the sure ground of observa-
tion of instinctive behavior; it can develop upon the basis of
fact undistorted by the requirement of meeting preconceived
notions imported from without. The whole question of
imitation, for example, reduces itself to one of fact: Is imi-
tativeness one of the original tendencies of human nature.
If so, what is its intensity and mode of working in conjunction
with the other unlearned activities?
The popularizers of science will doubtless remain half a
generation behind this as well as other scientific advances, but
for those who have learned the lesson of recourse to funda-
mental responses, the way is opened for emancipation from
the greatest foe with which social science has had to contend
— which I shall take the liberty of calling the monistic. How
often have we been invited to build up our social, political, and
ethical explanations in terms of some single and supposedly
dominant mental constituent! How often discussions and
disputes have been, at bottom, only a question as to which
of rival single claimants we shall yield allegiance. Instincts
to power, to control of others, fear of authority, sex, love of
THE NEED FOR SOCIAL PSYCHOLOGY 269
pleasure, of ease, all have been appealed to, and explanations
constructed in terms of one or another exclusively. Hence-
forth it is, I submit, pure wilfulness if any one pretending to
a scientific treatment starts from any other than a pluralistic
basis: the complexity and specific variety of the factors of
human nature, each operating in response to its own highly
specific stimulus, and each subject to almost infinite shadings
and modulations as it enters into combination and competition
with others. The conception of social psychology resulting
from this mode of approach becomes essentially one with
that set forth by Professor W. I. Thomas in his paper on the
province of social psychology at the St. Louis Congress of
Arts and Science in 1904, On the one hand our problem is
to know the modifications wrought in the native constitution
of man by the fact that the elements of his endowment operate
in this or that social medium; on the other hand, we want to
know how control of the environment may be better secured
by means of the operation of this or that native capacity.
Under these general heads are summed up the infinity of
special and difficult problems relating to education on the
one hand and to constructive modification of our social in-
stitutions on the other. To form a mind out of certain
native instincts by selecting an environment which evokes
them and directs their course; to re-form social institutions
by breaking up habits and giving peculiar intensity and
scope to some impulse is the problem of social control in its
two phases. To describe how such changes take place is the
task of social psychology stated in generalized terms.
I hope I do not need to disclaim an attempt to give in even
the barest summary the history of social psychology during
the past twenty-five years. My object has been quite other.
I have only wanted to refer to some salients in the intellectual
fortifications constructed during this period for the sake of
pointing out, in equally general terms, something of what now
confronts us, waiting, nay demanding, to be done. Before
passing on to this point, I feel I must avert possible misunder-
standing by mentioning two allied factors which have also
influenced the development of which I have spoken. One is
270 JOHN DEWEY
the application of statistical methods to psychological re-
search; the other, the behavioristic movement. Neither was
devised primarily in the interests of social psychology. The
requirements of education have, however, been a powerful
agency in promoting the former, while education presents,
of course, one phase of the problem of social control. Speak-
ing more broadly, social phenomena are of a kind which demand
statistical mathematics rather than the type of mathematics
which has been evolved especially for use in dealing with
physical facts. Condorcet's great essay on 'The Progress of
the Human Mind' forecasts a future in which human arrange-
ments would be regulated by science. In dealing with the
influence of mathematical science he points to the newly
developing theory of probabilities as that branch of mathe-
matics which is fraught with infinite potentiality for control
of social progress. I think it is only fair to see in statistical
psychology a step forward, short and halting though it be
for the immediate present, in the realization of Condorcet's
prophecy.
The behavioristic movement inevitably tends to confirm
the tendency of which I have already spoken in connection
with the writings of James, McDougall, and Thorndike. It
transfers attention from vague generalities regarding social
consciousness and social mind to the specific processes of
interaction which take place among human beings, and to the
details of group-behavior. It emphasizes the importance of
knowledge of the primary activities of human nature, and of
the modifications and reorganizations they undergo in asso-
ciation with the activities of others. It radically simplifies
the whole problem by making it clear that social institutions
and arrangements, including the whole apparatus of tradition
and transmission, represent simply the acquired transforma-
tions of original human endowments.
This provides the possibility of a positive method for
analyzing social phenomena. I shall avoid engaging in passing
in the disputed question of the value of an introspective
psychology. But it seems almost self-evident that even if
introspection were a valid method in individual psychology,
THE NEED FOR SOCIAL PSYCHOLOGY 271
so called, it could not be of use in the investigation of social
facts, even though those facts be labelled social mind or
consciousness. Yet one has only to look at the writings of
the Austrian and German school of "folk-psychologists"
(say of Wundt, obviously the most important) to see how this
treatment has been affected by an assumed need of making
the method and results of social psychology conform to the
received categories of introspective psychology. From such
deforming of facts the behavioristic outlook immediately
redeems us; it represents not an improvement in detail but a
different mode of attack. It is not as yet possible to estimate
the significance of this alteration. In my opinion, however,
the chief cause of the backwardness of social psychology has
resided in the artificiality of the endeavor to adapt the rubrics
of introspective psychology to the facts of objective associated
life. The opening of another road of approach may therefore
be expected to emancipate inquiry.
I thus come to the explicit statement of the purpose of my
reminiscent sketch. The aim was to justify the presentation
of the conviction that the quarter century in which this
Association has existed marks just the emancipation of social
psychology from influences which prevented its development
on its own feet and its own merits, while the work done on
lines which (as it seems to me) must be abandoned, have
nevertheless done the great service of enforcing the vast
field open to a social psychology, and the great need it has to
serve. I turn accordingly from the past to the future, or if
you will from prophecy taking the guise of history to prophecy
frankly avowing itself as such.
I foresee, then, a great reflex wave from social psychology
back into general psychology. An important conclusion in
the psychology of native activities does not seem to have been
drawn as yet by those who would base a scientific psychology
upon this foundation. The conclusion seems inevitable that
since 'mind' does not appear in the original list of instincts,
it represents something acquired. It represents a reorgan-
ization of original activities through their operation in a
given environment. It is a formation, not a datum; a pro-
272 JOHN DEWEY
duct, and a cause only after it has been produced. Now
theoretically it is possible that the reorganization of native
activities which constitute mind may occur through their exer-
cise within a purely physical medium. Empirically, however,
this is highly improbable. A consideration of the dependence
in infancy of the organization of the native activities into
intelligence upon the presence of others, upon sharing in joint
activities and upon language, make it obvious that the sort
of mind capable of development through the operation of
native endowment in a non-social environment is of the moron
order, and is practically, if not theoretically, negligible.
The net outcome of the newer type of psychological
method is thus an unexpected confirmation of the insight of
Tarde that what we call 'mind' means essentially the working
of certain beliefs and desires; and that these in the concrete —
in the only sense in which mind may be said to exist — are
functions of associated behavior, varying with the structure
and operation of social groups. Speaking in general terms,
there is no more a problem of the origin of society than there
is of the origin of chemical reactions; things are made that
way. But a certain kind of associated or joint life when
brought into being has an unexpected by-product — the for-
mation of those peculiar acquired dispositions, sets, attitudes,
which are termed mind. This by-product continually gains
in relative importance. It increasingly becomes the signi-
ficant acquisition among all the varied reorganizations of
native tendencies. That anything which may properly be
called mind or intelligence is not an original possession but is
a consequence of the manifestation of instincts under the
conditions supplied by associated life in the family, the
school, the market place and the forum, is no remote inference
from a speculative reconstruction of the mind of primitive
man; it is a conclusion confirmed by the development of
specific beliefs, ideas and purposes in the life of every infant
now observable.
On the face of it, this conclusion has implications only for
the theory of psychology. But slight scrutiny makes obvious
its consequences for the struggle to gain control of the forces
THE NEED FOR SOCIAL PSYCHOLOGY 273
forming society. The ultimate refuge of the standpatter in
every field, education, religion, politics, industrial and
domestic life, has been the notion of an alleged fixed structure
of mind. As long as mind is conceived as an antecedent and
ready-made thing, institutions and customs may be regarded
as its offspring. By its own nature the ready-made mind
works to produce them as they have existed and now exist.
There is no use in kicking against necessity. The most
powerful apologetics for any arrangement or institution is the
conception that it is an inevitable result of fixed conditions of
human nature. Consequently, in one disguise or another,
directly or by extreme and elaborate indirection, we find the
assumed constitution of an antecedently given mind appealed
to in justification of the established order as to the family,
the school, the government, industry, commerce and every
other institution. Our increased knowledge of the past of
man has, indeed, given this complacent assumption a certain
shock, but it has not as yet seriously modified it. Evolution
in the sense of a progressive unfolding of original potencies
latent in a ready-made mind has been used to reconcile the
conception of mind as an original datum with the historic
facts of social change which can no longer be ignored. The
effect on the effort at deliberate social control and construction
remained the same. All man could do was to wait and watch
the panorama of a ready-formed mind unroll. The French
school of imitation, and its present successor, the Durkheim
school of collective mind, has practically the same outcome
as the German school of Volk-geist in this respect. All are
engaged in explaining the past and present, and (if they predict
at all) in predicting the future on the basis of the past. The
new point of view treats social facts as the material of an
experimental science, where the problem is that of modifying
belief and desire — that is to say mind — by enacting specific
changes in the social environment. Until this experimental
attitude is established, the historical method, in spite of all
the proof of past change which it adduces, will remain in
effect a bulwark of conservatism. For, I repeat, it reduces
the role of mind to that of beholding and recording the oper-
274 JOHN DEWEY
ations of man after they have happened. The historic
method may give emotional inspiration or consolation in
arousing the belief that a lot more changes are still to happen,
but it does not show man how his mind is to take part in
giving these changes one direction rather than another.
The advent of a type of psychology which builds frankly
on the original activities of man and asks how these are
altered, requalified and reorganized in consequence of their
exercise in specifically different environments brings with it-
self the experimental attitude, and thereby substitutes the
interest in control for the interest in merely recording and
what is called 'explaining.' If mind, in any definitely con-
crete sense of that word, is an offspring of the life of associ-
ation, intercourse, transmission, and accumulation rather
than a ready-made antecedent cause of these things, then the
attitude of polite aloofness or condescending justification as
to social institutions has its nerve cut, and with this the in-
tellectual resources of sanctified conservatism disappear.
Instincts become mind when they are organized and directed
with reference to the ends of attention, esteem, and endeavor
which are supplied by the shared life of the place and time.
The kind of mind they become depends upon the kind of
objects of attention and affection which the specific social
conditions supply. The task of unravelling the arrangements
which exist into elements of native instinct and past acquisi-
tions is indeed an infinitely complex and difficult one; not the
less hard and extensive is the job of showing how this and that
association with other persons develops this and that intel-
lectual and emotional disposition — or mind — in this and that
individual having his own peculiar original endowment. But
if the history of human achievement in knowledge proves
anything, it is that the all-decisive discovery is that of an
effective and fruitful method. When men once hit, after
endless awkwardness, upon the right road, the rest takes care
of itself. Scientific movement becomes orderly and cumu-
lative in the very process of occurring. Social and mental
phenomena become intelligible because they come within the
scope of the experimental method of attack. And again the
THE NEED FOR SOCIAL PSYCHOLOGY 275
history of science testifies to a conclusion which may also be
arrived at theoretically — the introduction of the experi-
mental method is all one with interest in control — in modi-
fication of the future.
There is a genuine modesty, and there is a stupid simu-
lation of modesty which is only a mask for lazy complacency.
No science has so much cause to be humble about its actual
achievements as has social science, including social psy-
chology. But in prospect, in possibility, social science seems
to me to stand about where physical science stood three
centures ago in the early years of the seventeenth century.
There is the same halting and obstructed tendency to move
from the attitude of the outside spectator, classifier and
justifier of things as they are outwardly given to that of the
active participant and modifier, from that of wholesale or-
ganization to that of retail reorganization. The experimental
method in physical matters brought with it a technique
of control — a technique of invention and construction.
Specific desired ends can be formulated in specifically ana-
lyzed terms; the conditions of their attainment stated; these
conditions subdivided into known and unknown factors, and
some definite estimate made as to the praticability, at the
given time, of attacking the problem. That we are without
any such technique in social matters is self-evident. That
the attainment within reasonable time of a similar technique
stands and falls with the possibility of developing a human
psychology which shall be experimentally applicable to the
understanding of social affairs is not, however, self-evident,
and is my excuse for reiteration.
I venture accordingly to repeat a thought which I had the
honor of presenting before this association some years ago.
The need of social control is, of course, as old as associated
life itself. But the need of that control at the present time
is tremendously accentuated by the enormous lack of balance
between existing methods of physical and social direction.
The utilization of physical energies made possible by the
advance of physics and chemistry has enormously complicated
the industrial and political problem. The question of the
276 JOHN DEWEY
distribution of economic resources, of the relationships of rich
and poor was never so acute nor so portentous as it is now; and
this state of affairs is as much the result of progress in physical
science as is the recognition of the Copernician astronomy.
The present war is too vast and too tragic to permit one
lightly to summon it as evidence for any merely theoretical
thesis. But is it not, I ask, a demonstration made to order of
those old words of Thomas Hobbes? "The utility of moral and
civil philosophy is to be estimated, not so much by the com-
modities we have from knowing those sciences as from the
calamities we receive from not knowing them." Such a
conception is not fashionable just now; it is easier to place
blame upon fate or upon the innate wickedness of human
nature as seen in this or in that set of human beings. But
the ultimate fate is the fatality of ignorance, and the ultimate
wickedness is lack of faith in the possibilities of intelligence
applied inventively and constructively.
Physical science has got to the point of bringing even the
ends of the earth into physical, forceful relations with one
another, and to the point of mobilizing all its resources for a
contest in aggression and endurance. We are overwhelmed
by the consequences of the very sciences into which have gone
our best thought and energy for these last few hundred years.
We apparently do not control them; they control us and wreak
their vengeance upon us. Yet how infantile and pusil-
lanimous are those who talk about the bankruptcy of science
and who blame the increase of knowledge for our situation.
Physical knowledge, and the consequent technique of control
of physical forces, has far out-run social knowledge and its
technique. The recourse of a courageous humanity is to
press forward in the latter until we have a control of human
nature comparable to our control of physical nature.
From the point of view of the psychology of behavior all
psychology is either biological or social psychology. And if it
still be true that man is not only an animal but a social
animal, the two cannot be dissevered when we deal with man.
Hence it is that subsequent years have enabled me to find
added meaning in words which I spoke before this association
THE NEED FOR SOCIAL PSYCHOLOGY 277
years ago, and which in conclusion I venture to repeat. " We
are not called upon to be either boasters or sentimentalists
regarding the possibilities of our science. . . . But we are
entitled in our daily work to be sustained by the conviction
that we are not working in indifference to or at cross purposes
with the practical strivings of a common humanity. The
psychologist in his most remote and technical occupation
with mechanism may be contributing his bit to that ordered
knowledge which alone enables mankind to secure a larger
and to direct a more equal flow of the values of life."
THE CASE OF SELF AGAINST SOUL1
BY MARY WHITON CALKINS
Wellesley College
Most contemporary psychology tacitly ignores that con-
cretely real being, the self or I. Psychologists experiment on
mental processes of varying sorts, percepts, images, and
emotions; they dispute about the attributes of sensation, the
occurrence of imageless thought, the dimensions of affection.
Rebels from the orthodox school have a new set of abstrac-
tions : psychophysical functions and modes of behavior which
constitute, in their view, the proper object of psychological
investigation. In the meantime, the self which perceives and
feels and thinks, which functions and behaves, though so
obviously cast by nature for the title role, plays no part at
all on the stage of most psychological systems. More liter-
ally stated: instead of studying concrete wholes — namely,
knowing, feeling, willing, functioning, and behaving selves,
persons or /'s — the psychologist is wont to concern himself
with one or another part or aspect of this concrete whole,
often with a schematized, artificial construct, invented for
purposes of easy description or in order to bring psychological
method into artificial correspondence with that of the
physical sciences.
The chief reason for this ominous neglect of the self in
psychology is unquestionably the fact that I am always
conscious of myself and that I am therefore inattentive to this
"ubiquitous" self. "Just as, if I were asked to report fully
my sensational experience at a given moment, I might well
forget to name the sensations of pressure from my clothing
simply because they are so constant"2 so in my introspection
1 A paper read November 24, 1916, before the Philosophical Union of the Uni-
versity of California. Part IV. has been enlarged and rewritten and the paper has
been revised.
2 Quoted from a paper by the writer, 'The Self in Scientific Psychology,' Amer. /.
of Psychol., 1915, 26, p. 521.
278
THE CASE OF SELF AGAINST SOUL 279
I simply forget to name myself. A second explanation of this
deplorable tendency to ignore the fundamental fact of
psychology is found in the traditional confusion of 'self
with 'soul.' The chief business of this paper is to compare
and to distinguish the two concepts.
It is necessary at the outset to agree, roughly at least, on
the meaning of our terms. By 'self I denote the object of
the observation expressed in the words 'I am conscious of
myself.' Obviously, this statement is no definition, and for
the best of reasons: the self, or I, is, in truth, indefinable,
since it is sui generis, in a class by itself. But though inde-
finable it is not, therefore, elemental and indescribable.1 On
the contrary, the self is a highly complex being which may be
described by an enumeration of its characters. Among these
characters of the self the following are surely fundamental:
First, the self of each of us to some extent persists: I am in a
true sense of the word the 'same' self who cut a philosophy
class in her senior days at college in order to take a drive
behind a pair of Vermont horses. In the second place,
however, the self, with all its persistence, truly changes,
develops: though an outgrowth from that frivolous self, and
identical with her, I am yet a changed self. Third, and very
significantly, I am a unique self: there is only one of me; I am
an individual; no one, however, closely she resembles me, is I.
The possibility of this enumeration shows, in the fourth place,
that I am a complex self, a unity of present with past — yes,
and with future — self and a totality, also, of many differ-
ent experiences; I am a perceiving and remembering and
thinking and feeling self. These different experiences or
aspects of me do not, however, exist apart from me; I ob-
viously am not what Hume called me, a bundle of perceptions,
but each of the perceptions or emotions or thoughts is the
expression of me who am inclusive of them. Finally, I am a
self related to the world in which I seem to myself to be placed;
my fundamental relation to the world I call my consciousness
1 For the unjustified implication that the indefinable must therefore be elemental
cf.E.B. Holt 'The Concept of Consciousness,3 pp. 73-74: "Either consciousness is a com-
plex entity not fundamental but definable in terms of simple entities that are not
consciousness ... or else consciousness is fundamental and simple."
280 MARY W. CALKINS
of it, and within my consciousness I distinguish different
forms of relation, as activity and passivity, and different
complexes of relation, as perception, emotion, and the rest.
All these characters, it must be added, are immediately ex-
perienced. The self, thus described, is observed and not
merely inferred; is, therefore, a psychological datum which is
taken over into philosophy when reflection discloses that it is
the unique fact which can neither be denied nor even doubted
without being at the same time asserted.
We have next to distinguish the concept of 'soul* from
that of the immediately perceived, the persisting, changing,
unique, complex and related ' self.' And here we find ourselves
involved in great difficulties of interpretation. For, whereas
the relatively late term 'self is primarily psychological, the
word 'soul' is common to many philosophies and to pre-
philosophic levels of civilization; so that the attempt to un-
ravel its meaning is fraught with subtlest difficulties.
Among these varying meanings are three of commanding
importance which for convenience we may designate as the
biological (or vitalistic), the metaphysical (or immaterialistic),
and the psychological conceptions of the soul. The first two
are alike in that both define the soul with reference to the
body, but they conceive the relation of soul to body in sharply
opposed fashions, (i) According to the earlier, the vitalistic,
view soul is equivalent to life. Every reader of early liter-
ature knows that the soul is often virtually identified, now
with the blood, now with one or another of the vital organs,
and again with the breath. This conception of the soul as
breath differs, it must be noted, from the identification of
soul with blood or heart or kidneys. For the breath though
it gives life to the body leaves the body at death. In truth,
the doctrine of the soul as animus or spiritus forms a sort of
bridge, or transition, between the 'vitalistic' view and the
immaterialistic conception of the soul. (2) This 'metaphysical,'
or immaterialistic, doctrine regards the soul no longer as in
positive relation to the body but rather as negatively related
to it, contrasted v/ith it, opposed to it. The early conception
of the soul as shade, or ghost, is an approach to this meta-
THE CASE OF SELF AGAINST SOUL 281
physical doctrine, the Orphic, Platonic doctrine which persists
to our own day, that the soul is an immaterial, simple, un-
changing being. These characters, it will be noted, are the
bare opposites of the materiality, the compositeness and the
constant flux of the body so that the soul, thus metaphysically
conceived, may well be defined as not-body. Along with this
theoretical distinction of soul from body there goes a change
in valuation. Whereas the Homeric shade is an insubstantial
and a pitiable being, the immaterial soul of the philosopher is
regarded as immeasurably more worthy than the body.
(3) The third conception of the soul is as conscious being, as a
something which perceives, feels and thinks. In truth, by
this psychological conception soul and self are virtually
identified. It is significant, therefore, to find that frdm ear-
liest days men seem to have held this view of the soul along
with one or both of the other conceptions. Thus, the Homeric
^uxi?, primarily a vital phenomenon, has sense experiences,
and the Homeric %t6s knows, feels and wills; [the Platonic
soul, though predominantly immaterial, unites sense experi-
ences and apprehends ideas and loves wisdom; and Descartes's
very * metaphysical' soul has the principal 'attribute' of
thought, that is, of consciousness. This fusion of concepts
occurs, as will later be shown in detail, even when the char-
acters of the soul, for example, changelessness and simplicity,
are exact contraries of the observed characters of the self.
Our immediate task is to compare and to distinguish these
three conceptions as they appear and reappear in the history
of thought. It will be necessary to make rigorous selection
from a great mass of material; and we shall profitably begin
by a study of the soul of Plato's 'Dialogues.'1
I
Plato's doctrine of the soul, like all the philosophical and
scientific conceptions of the 'Dialogues,' is presented to us not
in systematic form as an articulated body of doctrine but
almost incidentally. One can hardly re-read the dialogues in
1 This paper, in its first form one of a series of lectures on Plato, treats his doc-
trine at what may seem disproportionate length.
282 MARY W. CALKINS
any unbiased way without noticing the relative rarity and
usual fragmentariness of Plato's references to the soul. We
are wont to stress these passages because later discussions
have lent them such transcendent significance, but an unpre-
judiced study of the dialogues always leads us to reaffirm
Paul Shorey's assertion that Plato is a dramatic artist and an
impassioned moral and religious teacher, not a scientist or a
metaphysician. Yet in spite of the incidental character of
Plato's philosophical and psychological teaching, we may,
I think, gain a fairly accurate notion of his conception of the
soul; and we shall find that it contains all three of the factors,
already noted, of the traditional soul-doctrine. That is to
say: Plato somewhat confusedly combines the vitalistic, the
immaterialistic, and the psychological conceptions of the soul.
This summary account of his position has now to be justified.
I. The Homeric concept of soul as a bodily principle or
activity was, of course, familiar to Plato. He obviously takes
over this primitive view when he makes Socrates say, in the
'Cratylus': "I should imagine that those who first used the
name tyvxq meant to express that the soul when in the body
is the source of life, alnov rov £rjv, and gives the power of
breath and revival (ava\l/vxov) and that when this reviving
power fails then the body perishes and dies."1 Here we have
a conception of the soul as specifically the breath-giver, a
notion which is strongly redolent of primitive thought.
Closely connected with this is the repeated teaching that
the soul is mover of the body. This, again, is a theory
adopted not invented by Plato. Thales, as we learn from
Aristotle, endowed magnets with souls on the ground that the
magnets occasion the movement of attracted particles; and
doubtless the restless motion of fire fits it, in the view of
Heracleitus, to define the soul. Thus, Plato is following an
accepted tradition by his constant description of the soul as
moving principle and — he adds — self-moved. "The soul,"
Plato says, in the 'Phaedrus,' "is ever in motion. The
body," he continues, "which is moved from without is soul-
less but that which is moved from within has a soul."2 And
1 'Cratylus,' 399, D-E.
2<Phaedrus,' 245, E.
THE CASE OF SELF AGAINST SOUL 283
this, it may be noted, is the teaching of the 'Phaedrus,' a
dialogue which can not well fall later than in the early-middle
period of Plato's thought. But proof of the permanence
of the conception is not lacking. For it reappears in Plato's
latest dialogue, the Laws, in which the Athenian stranger
speaks thus to Cleinias:
"Let us assume," he says, "that there is a motion able to
move other things but not to move itself — that is one kind;
and another kind is that which can move itself as well as
other things. . . . And which . . . ought we to prefer as the
mightiest?"
"I must say," Cleinias replies, "the motion which is able
to move itself is ten thousand times superior to all the
others."
"If then," the Athenian resumes, "as most of these
philosophers have the audacity to affirm, all things were at
rest, which of the . . . principles of motion would first
spring up among them?"
"Clearly the self-moving," Cleinias says, "for there could
be no change in them arising out of any external cause: the
change must first take place in themselves."
"Let us put a question," the Athenian continues, "If we
were to see this power existing in any earthy, watery or fiery
substance . . . how should we describe it?"
"You mean to ask," Cleinias replies, "whether we should
call such a self-moving power life. . . . Indeed, we should."
"And when we see soul in any thing," the Athenian in-
sists, "must we not . . . admit that this is life? . . . The
soul," he concludes, "is the first origin and moving power of
all that is, or has become or will be."1
These last words of the Athenian stranger and their
omitted context, as well as that of the passage quoted from
the 'Phaedrus,' indicate the content of Plato's cosmological
theory that the stars and "the whole heavens and all creation"
are moved by soul. But our discussion is limited to the study
of the human soul and must utterly reject the lure of this
reference to "all that is or has become or will be." On the
1 'Laws/ X., 894, £-895, C, 896, A.
284 MARY W, CALKINS
other hand it must stress the emphatic and repeated identi-
fication of 'soul' with 'life' as proof that Plato — here as
often the unacknowledged precursor of Aristotle — holds a
genuinely vitalistic conception of the soul.
2. Clearly distinguished from the vitalistic concept of the
soul as life, and often in the main inconsistent with it, is the
doctrine cherished and stressed by Plato, which I am des-
ignating as the 'metaphysical' concept of the soul. This
is the theory that the soul is utterly unlike the body, in-
corporeal or immaterial. The soul, Plato always asserts in
opposition to the "vain opinion" of the "physical investig-
ators," is not formed of fire and water and earth and air as
its first elements; rather the soul is the "first," not last, "and
before all bodies."1
It follows, Plato always assumes, that the inherently in-
corporeal soul is, as it were, a prisoner in the body or entombed
within it.2 Evidently, this is a perpetuation of the Pytha-
gorean teaching that "for a punishment the soul is yoked
with the body and buried in it as a tomb."3 And Plato's
assertions that the soul is separable from the body are simi-
larly related to that Orphic belief which traces its source to
the ecstatic Dionysiac rites in which "the possessed devotee
was set free for the moment from the tangled net of daily life,
gained for a brief time new and superhuman powers"4 so that
the soul seemed to be loosed from the body. Now the Orphic
and Pythagorean doctrine of the soul's immaterial nature
formed an inextricable part of the belief in the preexistence
of the soul. Both of these doctrines are enforced by Plato's
teaching, but he supplements their early ethical and religious
content by more narrowly metaphysical considerations. In
the Orphic Fragments the doctrine of the preexistence and the
reincarnation of the soul enforces ethical exhortation by
promising a period of reward or punishment in which "they
who are righteous beneath the rays of the sun when they die
^Laws,' 891, £-892, A.
2<Phaedrus,' 250.
3 Quoted by Clement of Alexandria, Stromata, III., 17. (Diels, ' Fragmente der
Vorsokratiker,' I., p. 245, Philolaos, B, 14.)
4 C. H. Moore, 'Religious Thought of the Greeks,' p. 50.
THE CASE OF SELF AGAINST SOUL 285
have a gentler lot in a fair meadow by deep-flowing Acheron.
. . . But they who have worked wrong and insolence beneath
the rays of the sun are led down beneath Cocytus's watery
plain into chill Tartarus."1 The myths of Plato perpetuate
this teaching and stress the individuality of the soul. Stripped
of the body which is "interposed as a veil before" his own
soul, the judge of the Gorgias "with his naked soul shall
pierce the other naked souls" who have left their brave
attire strewn upon the earth."2 Those who have "already,"
in the words of the ' Phsedrus,' " begun the heavenly pilgrimage
may not go down again to the darkness."3 But all the souls
suffer for the wrong which they have severally done.
Plato's ethical soul-doctrine, it should be observed,
consists not merely in an exhortation to the soul to escape
the prison house of the body but in an emphasis of the domi-
nation of soul over body. "When the soul and body are
united," he says, "then nature orders the soul to rule and
govern and the body to serve." The soul is the pilot of the
body; is at variance with the affections of the body; "coerces
the bodily elements ... as if talking to a thing which is not
herself." But Plato, though he thus emphasizes the ethical
aspects of the earlier soul doctrine, strengthens the conviction
of the soul's freedom from the body's mortality by a philo-
sophical (in a way a psychological) consideration. We are
constantly, he points out, estimating people and things as
good or beautiful or equal; and in our estimate we presuppose
an absolute standard of goodness, beauty, equality and the
like. Yet the objects we meet with in our actual human
experience are never more than relatively good, beautiful or
equal. It follows, Plato insists, that the soul must have
existed without the body,4 that "every soul of man has in
the way of nature beheld true being: this was the condition
of her passing into the form of man."5
1 Fragment 154 A. (Quoted by C. H. Moore, op. cit., p. 56, from Abel's edition
of the Orphic fragments.)
2 'Gorgias,' 523, D-E.
8 'Phsedrus,' 256, D.
6 'Phsedrus,' 249, £-250, A.
286 MARY W. CALKINS
So, the Platonic teaching about immortality preserves the
ethical and religious coloring of the Orphic and the Pytha-
gorean beliefs, on a background of philosophical conception
and argument. Some of the arguments for immortality are,
to be sure, drawn from that vitalistic theory of the soul with
which, as will later appear, the immaterialistic concept is not
actually to be harmonized. Thus, that latest argument of
the 'Phsedo' which Socrates brought forward like "a general
rallying his defeated and broken army" is based on the ad-
mitted truth that "whatever the soul possesses, to that she
comes, bearing life."1 And in the 'Phaedrus' immortality is
argued from the self-motion of the soul: the self-moving soul,
Socrates asserts, as true beginner of motion cannot itself be
begotten.2
For the most part, however, the 'Phsedo' argues immor-
tality from the immateriality of the soul, its unlikeness to the
body. Thus, when Cebes begs Socrates to argue them out
of their fears, to persuade the "child within" them "not to
t>e afraid when he is alone in the dark," Socrates answers thus :3
"The composite may be supposed to be naturally capable
of being dissolved; but that which is uncompounded and that
•only must be ... indissoluble. And the uncompounded
may be assumed to be ... unchanging, whereas the com-
pounded is always changing. Now . . . equality, beauty-
are these essences always what they are ... not admitting
of variation at all?"
"They must be always the same," Cebes answers. And
Socrates continues:
"Let us suppose that there are two sorts of existences — one
seen, the other unseen. The seen is the changing and the
unseen is the unchanging. Further, one part of us [is] body
and the other soul. And to which class may we say that the
body is more alike and akin?"
"Clearly to the seen — no one can doubt that," Cebes
replies; and Socrates asks:
14Phsedo,' 105.
2 'Phsedrus,' 245.
3 Condensed from the 'Phaedo,' 78-80, with the omission of most of the brief
replies of Cebes.
THE CASE OF SELF AGAINST SOUL 287
"Is the soul seen or not seen?"
"Not by man, Socrates," Cebes answers.
"Then," Socrates says, "the soul when using the body as
an instrument of perception is dragged by the body into the
region of the changeable. . . . But when returning into her-
self she reflects, then she passes into the other world, the
abode of purity, and eternity, and immortality, and unchange-
ableness. . . . And consider the matter in another light.
When the soul and the body are united, then nature orders
the soul to rule and govern and the body to obey and serve.
Which of these two functions is akin to the divine? and which
to the mortal?"
"The soul,' Cebes agrees, "resembles the divine and the
body the mortal." And Socrates concludes:
"Then . . . the soul is in the very likeness of the divine,
and immortal, and intellectual, and uniform, and indissolu-
able, and unchangeable; and the body is in the very likeness
of the human, and mortal, and . . . multiform, and dis-
soluble, and changeable."
We are not at all concerned with the force of the argument
nor with the difficult problem of the relation of soul to the
essences or ideas. What we have to note are the characters
of uncompoundedness and unchangingness here attributed
by Plato to the soul — regarded, in metaphysical fashion, as
immaterial being — predicates which are sharply opposed, or at
least unrelated, to the characters of life and self-motion which
are basal to Plato's vitalistic concept of the soul.
3. We have thus summarily outlined two Platonic con-
ceptions of the soul, distinguishable if often uncontrasted :
first, his conception of the soul as vital principle of the body,
its breath, inciter to its movements, and second, what we
have named his metaphysical conception of the soul as a
changeless uncompounded individual being, radically dif-
ferent from body and variously thought of as dominating the
body or as marred and hampered by communion with it.
But Plato conceives, or at any rate treats, the soul in still a
third fashion as a conscious being, as subject of experience.
To begin with, the soul recollects. For Plato's well-known
288 MARY W. CALKINS
doctrine of reminiscence is not merely a metaphysical specu-
lation about reincarnation and an epistemological doctrine
about absolute knowledge, but a psychological study of rec-
ognition and memory. In other words, the soul not only
"calls to mind all she ever knew"1 as the 'Meno' has it, but,
in the words of the 'Philebus,' she has "the power ... of
recovering when by herself some feeling which she experienced
when in company with the body"2 — that is to say, the soul
remembers sense-experiences. The soul also conceives and
compares and reflects. " By a power of her own," Socrates is
made to say, in the These tetus,3 "the soul contemplates the
universals . . . comparing in herself things past and present
with the future."4 And Plato's conception of the conscious
soul is not purely intellectualistic. The soul, he teaches,
feels pleasure and pain, and " pleasure and desire" may
"exist in the mind only apart from the body."5 The soul
finally aspires* and — as figured in all the myths of judgment
— makes genuine choice. Moreover, the soul which perceives
the physical and understands universals has also, Plato im-
plies, a knowledge of itself. Thought is the conversation of
the soul with herself; the knowledge of universals is a recol-
lection of her former experience; her character is, as we have
seen, the outcome of her own choice. The soul, finally,
through her ' instruments/ the senses, perceives. In the words
of the 'Philebus' there is in perception "union of soul and
body in one feeling or motion";7 and there are affections
(irad-findTa) which vibrate through both soul and body and
impart a shock to both and each of them.8 It is significant
to find in this account of the perceiving soul a denial of the
sharp opposition of soul to body.
The fundamental protest of this paper, it will be remem-
bered, is against the confusion of the concepts of soul and
l'Meno,'8i, C.
2'Philebus,'34, B.
''Theaetetus,' 185, E.
4 Ibid., 1 86, B.
« 'Philebus,' 34, C.
e 'The<etetus,' 186, A.
7 'Philebus,' 34, A.
'Ibid., 33, D.
THE CASE OF SELF AGAINST SOUL 289
self. If, however, philosophical convention is to be trusted,
there is no room for the confusion in Plato's teaching since —
so we are told — the concept of the self dawns late on the
metaphysical horizon, and is descried at the earliest in the
speculations of Plotinus. But whatever may be true of the
term ' self/ however long the period before the self was formally
introduced in metaphysical circles, philosophers, as well as
other men, were conscious of themselves though — also like
other men — they often forgot to mention the fact precisely
because they were so accustomed to it. To return to Plato.
Unquestionably, he uses the reflexive pronoun eavros-eavrd
to refer to impersonal as well as to personal realities; and
certainly he does not explicitly refer to ego or self; but on the
other hand he over and over again refers to the man (avQpuiros)
or to ourselves (i7/zeTs) as being conscious. Thus, in 'Phi-
lebus,' 36, A, Socrates inquires whether one of us (TLS I?/KOJ>)
who is hungry may not at one time have a sure hope of being
fed and at another time feel in despair. "Each of us," he
says in 'Republic,' III., "has many wants." Indeed,
throughout the discussion of the virtues in the 'Republic,'
the state is contrasted not with the soul but with the man
(avrjp) or with the individual (ftccurros); whereas the soul is
referred to only incidentally and conventionally.1 In these
allusions to 'a man' and to 'us' it is at least possible to descry
the germ of a psychological concept of the experienced self,
no longer wholly overlaid by the conventional soul-concept.
And if we are right in this interpretation we have here an
early illustration of the greater empirical accuracy of state-
ments about the self. Influenced by his dualistic prepos-
session to stress the contrast between soul and body, Plato,
as we have already noted, scruples to say that 'the soul'
perceives, whereas he clearly realizes — as everyone who,
simply and without metaphysical bias, observes himself
must realize — that 'we' perceive as surely as we think.
The estimate of self and soul doctrines is, however, best
postponed to a point further on in our discussion.
1 Cf. 'Republic,' 435, C. r6v tva . . . TO. avra fldij tv T% avr
290 MARY W. CALKINS
II
1. The conception of the soul most emphasized by Aris-
totle is, as everybody knows, the biological or vitalistic. The
soul he says is the first entelechy, or realization, of an organic
body — in a word it is the body's capacity for life. It follows,
Aristotle teaches, that there are animal souls as well as human
souls. And by this he does not mean to attribute conscious-
ness to plants and animals but rather to attribute to them
life. Like Plato and the earlier Greek thinkers, Aristotle also
makes of soul "the source of local movement."1
2. Of the traditional immaterialistic conception of the
soul Aristotle, on the other hand, shows hardly a trace. "We
should no more ask," he observes,2 "whether soul and body
are one than ask whether the wax and the figure on it are
one." Yet the conception of soul as life is an integral part
of Aristotle's metaphysical system — the theory of the universe
as ascending from primitive, unformed matter (pure poten-
tiality) through successive stages (of which each is, viewed
from below, realization, or form, and viewed from above
itself potentiality) until at the end pure activity is attained:
completely realized being, mind in contemplation of itself,
mind which is principle of all motion in that it is goal of all
desire and ultimate object of all yearning. The discussion of the
Aristotelian philosophy, thus baldly suggested, would, of
course, lead us far afield. What is important to our purpose
is to notice that there is here no hint of the specifically im-
materialistic conception of soul — no description of it as simple
and uncompounded. And, so far from conceiving the soul
as unchanging, Aristotle explicitly defines it as the progressive
realization of growing capacities, or possibilities.
3. That Aristotle, beside conceiving the soul in the manner
of the vitalists, certainly regards it after the psychologist's
fashion, is shown by the slightest examination of the contents
of *De Anima.' In Book II., he distinguishes from the
vegetable soul (ri 6ptimK.ii) the animal soul which he here
forthwith describes as perceptive (mo-flrjriKi?). The remainder
1 'De Anima,' II., Chap. 4, 20, 415 b.
2 Ibid., Chap. I, 7, 412 b.
THE CASE OF SELF AGAINST SOUL 291
of the treatise is given over to the discussion of the senses,
sight, hearing, touch, and the others ; of imagination (<pavTaala) j1
of conception (uTroXry^ts) ; thought (vo£v)-, desire (TO bp^rinbv) .2
All these are over and over again referred to in terms common
to Aristotle's zoological and ethical as well as to his psycholog-
ical writings as either parts (ju6pux) or capacities (dwd/zeis) of the
soul. Thus, the discussion of reason (TO voelv) in Book III.,
Chapter IV., of the 'Psychology' is introduced by the words:
"We must next investigate that part (popiov) of the soul by
which it knows and reflects (vpoveiv) ." These capacities or parts
are differently enumerated in different passages; those most
often named are the perceptive, the thinking (Xoyio-ruoj' or
KOV or diavoriTLKov) and the impulsive or voluntary (TO
or povXtvTiKov or tiridvMTiKov). More than once, indeed,
Aristotle says that the soul is defined by thinking, judging,
and perceiving, and that both by thought and by sense " the
soul distinguishes and knows real things." 3 The simplicity
and precision of the words just quoted ("the soul distin-
guishes," "the soul knows") bring into clear relief the arti-
ficiality of Aristotle's more frequent descriptions of conscious
experience in terms, very often reproduced, of personified (yet
to the end more or less fictitious) 'faculties' — imagination,
reason, and the rest. More frequent than Aristotle's simple
statements that the soul knows or thinks are his references
to the man (avdpuiroi) who hears, smells or thinks; and still
more frequent are his assertions that 'we' see and touch
and perceive;4 that images present themselves to 'us,' that
'we' judge (KpLvonev) , affirm or deny (aKydevofjiev ij ^evdofieda) ,
feel fear.5 Here, as in Plato's parallel expressions, we may
perhaps fairly conclude that Aristotle is forgetting the sup-
posed necessity to record his experience in terms of his phi-
losophical doctrine and is conceiving a self, rather than a
soul, which is conscious.
It should be noted that Aristotle's profound doctrine of
1 'De Anima,' III., Chap. 3, 7, 428*.
2 Ibid., III., Chap. 9, 3, 4326.
•Ibid., III., Chap. 3, i.
4 Ibid., III., Chap. 3, 6, 4280.
6 Ibid., 5, 427^.
292 MARY W. CALKINS
substance, as elaborated in the * Metaphysics/ plainly con-
nects itself with the conception of the soul. To trace the
connection would be to involve ourselves in a long dis-
cussion, but we can not avoid the conclusion that Aristotle,
for whom the ultimately real was always neither matter nor
form but the individual, meant to endow the soul with this
character: uniqueness or individuality.
Ill
By Stoic, Judeo-Alexandrian, Neo-Platonic, Patristic, and
Scholastic thinkers the triple conception of the soul has been
handed down to modern psychologists and metaphysicians.
The soul-doctrine of the pre-Christian schools belonged, as
is well known, to a pantheistic philosophy. The soul was
regarded by the Stoics as part of the Universal Reason, by
the Neo-Platonists as emanation through the world-soul from
the One. But all — Hellenic, Jewish, Alexandrian and Christian
thinkers — united to emphasize the worthlessness of the body
and the supremacy of the 'soul burdened with a corpse.'1
It is not necessary to consider in detail most of these
theories but it would be impossible to pass over Augustine's
teaching. The conception of soul and self are, as everyone
knows, central in his philosophy; and he conceives the soul
in each one of the three traditional ways, (i) Although he
repudiates the essential materialism of the primitive vitalistic
conception which identifies the soul with the blood he yet
everywhere insists that the conscious soul is a living being.
And from the attribute of life he argues, in Plato's fashion,
the immortality of the soul. "Those," he says, "who have
held [the soul's] substance to be some kind of life . . . have
striven also to prove it immortal, since life can not be without
life."2
(2) More often, however, Augustine sets forth the im-
materialistic theory of the soul. His teaching is very close
to that of Plato. The soul, he asserts, in words perpetuated
for centuries, is a spiritual substance:3 it is incorporeal — and
1 Attributed to Marcus Aurelius (quoting Epictetus).
2'DeTrinitate,' X., 7.
3 'De Trinitate,' XL, i; 'De Quantitate Animse,' Chaps. II., III.
THE CASE OF SELF AGAINST SOUL 293
he argues this, as Plato has argued it, on the ground that the
soul discerns incorporeal things (cernit incorporea),1 such, for
example, as mathematical quantities. The incorporeal nature
of the soul, Augustine also teaches, implies its'unextendedness —
a negative character which he stresses very likely in opposition
to Tertullian's materialistic view. Over and over again, in
1 De Quantitate Animae '2 and in ' De Immortalitate,' Augustine
repeats the statement that the soul feels throughout the body
(per totum corpus) but not with the body (cum cor pore). The
reiterated statement that the soul is simple has the same
meaning, that the soul acts as a whole (tota operatur). And
from its unextendedness and its simplicity he deduces, after
the traditional fashion, its unchangingness, imperishability
and immortality. He argues in detail that though the soul
seems to grow, to develop with the body, it is none the less
unchanging. Non mutatur, he reiterates in passage after
passage.3 Finally, in his account of the relation between
soul and body Augustine follows Plato with especial fidelity.
The soul, he everywhere asserts and implies, is the ruler of the
body (rationis particeps regendo corpori).
(3) Up to this point we have nothing radically new. The
vitalistic doctrine of the soul is set forth less convincingly by
Augustine than by Aristotle; the immaterialistic doctrine in
no wise advances upon Plato's. But Augustine's treatment of
the soul as subject of consciousness or knowledge is like new
wine in new bottles — an outpouring of first-hand observations,
a formulation of acute and independent reasonings. The
pages in which he discusses loving and knowing, remembering
and attending, thinking and willing, seem to me, though
written so long before the rise of scientific psychology, com-
parable only with the introspective passages of William
James for their luminousness, their revealing quality, their
ability to cleave asunder the most tortuously intertwined
human experiences.
The most emphasized and most truly novel contribution
1 'De Trinitate,' X., 7; 'De Quantitate Animae,' Chaps. XIII., XV.
2 Cf. Chaps. XXIIL, XXX.
3<De Immortalitate,' Chap. I. Cf. 'De Quantitate Animae,' Chaps. II., XV.,
XVI.
294 MARY W. CALKINS
of Augustine to this psychological conception of the soul is his
insistence on the fact that the soul knows itself. We are not
here concerned with the implications of this fact for philosophy
— the significance of the discovery of this one reality, myself,
which is self-evident because in doubting or denying it I
assert it. What we are here remarking is the enlargement of
the concept of knowing soul by the explicit recognition of its
self-knowing character. " ' Know thyself,' " Augustine asserts,
"is not said to the mind as is 'Know the cherubim and the
seraphim'; for they are absent, and we believe concerning
them. . . . Nor yet as it is said 'Know the will of that man';
for this it is not within our reach to perceive at all, either by
sense or understanding, unless by corporeal signs actually set
forth; and this in such a way that we rather believe than
understand. Nor again as it is said to a man, 'Behold thy
own face'; which he can only do in a looking-glass. For even
our own face itself is out of the reach of our own seeing it.
. . . But when it is said to the mind, Know thyself: then it
knows itself by that very stroke by which it understands the
word 'thyself; and this for no other reason than that it is
present to itself."1
With equal subtlety Augustine discusses all forms of
knowledge. At times, but only apparently by accident, or
in theological exigency, he follows Aristotle in apotheosizing
the abstractions 'memory,' 'thought' and 'will.' Usually,
however, he says concretely that the mind remembers, thinks,
judges, wills; or else he says even more concretely, I remember,
/ will. For the truth is that Augustine uses the self-words,
'I,' 'he,' 'self,' 'man,' nearly if not quite as often as he uses
the terms 'soul' or 'mind,' and seemingly he employs the two
sets of terms interchangeably. An example of this inter-
change is selected at random from Chapter XI. of Book X.
of ' De Trinitate.' Augustine is arguing that though memory,
understanding and will are three yet each is contained by
each. He begins with the statement that "The mind is
certain of these three things concerning itself . . . memory,
understanding and will," but ends with the assertion: " What-
i 'De Trinitate,' X., 9.
THE CASE OF SELF AGAINST SOUL 295
ever I understand I know that I understand and I know that
I will whatever I will and whatever I know I remember."
Here the term 'I' replaces the term 'mind* within one and
the same closely argued proof.
But though Augustine thus identifies soul with self he does
not criticize the one conception by the other. He never
turns upon the obscure conventional attributes of the soul
the searchlight of introspective observation; he never recon-
ciles the divergence between the teaching of Augustine, the
observer, "When the human mind knows and loves itself it
does not know and love anything unchangeable"1 — and the
dogma which Augustine, the theologian, uncritically takes
over from Plato and the Fathers, "The soul is unchanging and
simple."
Mediaeval philosophers, so far as I know them, neither
make important contributions to the conception of soul nor
markedly diverge from it. Two instances will suffice for our
purposes. Alcuin, as quoted, compresses an assertion of all
three of the soul's traditional attributes into a brief sentence:
"The soul is a rational spirit ever in motion, ever living,
capable of good and evil." St. Thomas also expressly rec-
ognizes all three characters of the soul. He describes it as
principle of life and of bodily movement — a clearly vitalistic
description. He also asserts, in accord with Plato's doctrine,
that the soul is simple substance; and he argues elaborately2
(here, as in his doctrine of the individuality of souls, following
Aristotle) that in spite of being the mover of the body, the
soul is itself unmoved. Finally, St. Thomas teaches, in the
second Article of this same Question 75, that "the soul knows"
or rather (he corrects himself) "man knows through the soul."
Questions 77 through 80 develop this doctrine of the knowing
soul by enumerating and analyzing, somewhat in Aristotelian
fashion, the * powers of the soul,' that is to say, the five senses,
the intellect, memory, appetite and will. Like his predeces-
sors, Thomas obviously uses the three time-honored concepts
of the soul in unharmonized juxtaposition.
KDeTrinitate/IX., 6.
2 'Summa Theologise,' Pars Prima, Question 75, first Article.
296 MARY W. CALKINS
IV
The soul makes its way into modern thought encumbered
with its three more or less conflicting duties : to be life-giving,
to be immaterial, and to be conscious. It is not, however,
left undisturbed in the quiet possession of these traditional
characters. At the very outset Descartes threatens to deprive
it of one of them by destroying its pretension to be the life
and mover of the body. For Descartes formulates the me-
chanistic conception of life and effectively marshals the
evidence, from physiological observation and experiment, for
this mechanistic theory. Thus, he says that the circulation
"follows as necessarily from the very arrangement of the
parts of the body and from the heat which may be felt with
the fingers and from the nature of the blood ... as does the
motion of a clock from the power, the situation, and shape of
its counterweights and wheels." "The body," he repeats in
every possible connection, "is nothing more than a machine."1
Descartes, to be sure, is not daring enough to treat the human
body as a perfect mechanism He (i) still conceives the
human soul as vitally related to the body though he allows it
but a precarious seat in the pineal gland and tries to hold it
down to the function of mere direction, not initiation, of the
body's motions. In entirely uncritical fashion, (2) in the
second place, Descartes adopts the immaterialistic doctrine,
asserting after the manner of Plato, Antoninus, Philo, Augus-
tine, and Thomas the simplicity of the immortal soul and its
utter unlikeness to the body. But he (3) also stresses the
psychological conception of the soul as conscious being; and
here, once more in agreement with his predecessors and in
particular with Augustine, he uses the term 'soul' as virtual
synonym for 'self or 'I.'2 Even more often, however, he
describes the self, or I, purely as a conscious being, without
reference to the traditional characters of the soul. Thus he
says, in reply to the question: "What then am I?" "I am a
thinking thing . . . that doubts, understands, affirms, denies,
1 'Discourse on Method,' V.
2 'Meditation,' II., paragraphs 4, 7; et al. For Descartes's narrower use of the
term 'think' cf. paragraph 5 of 'Meditation? II; et al.
THE CASE OF SELF AGAINST SOUL 297
wills, refuses, that imagines also and perceives."1 There is
here no mention of 'motive power' or simplicity or imma-
teriality: I am conscious self.
This divorce of self from soul, as will presently appear,
becomes explicit in the teaching of Locke, (i) Locke follows
Henry More2 in stressing the conception of the soul as
"capable of motion," as "changing distance" and as "moving
or quieting corporeal motion."3 He argues simply that
"every day's experience clearly furnishes us with" instances
of "will or a power of putting body into motion by thought."4
This is, of course, a mere reassertion of the classic vitalistic
doctrine of soul. (2) The immaterialistic soul of the meta-
physicians is subjected by Locke to more radical treatment.
He resolutely strips it of all the characters of consciousness
and then he rightly insists that the "substance of spirits,"
thus abstractly conceived, "is unknown to us." "If any
one," he repeatedly asserts, "will examine himself concerning
his idea of * substance' he will find he has no other idea of it at
all but only a supposition of he knows not what support of
qualities."5 "Our idea of substance," he says in another
passage, referring definitely to spiritual substance, or soul,
"is but a supposed I-know-not-what to support ideas."6
(3) To the self, or I, Locke now gives over explicitly most or
all the concrete characters which soul and self (before his time
uncritically identified) once shared between them: conscious-
ness, moral responsibility, and even identity. "The Self,"
1 Descartes, ' Meditations,' II., paragraph 7.
2 More breaks sharply with tradition by insisting that the soul, though immaterial,
is not therefore unextended. There are two kinds of extension, More teaches : the first*
material and impenetrable, the second a penetrable, "subtle and immaterial extension
. . . whereof we have an innate ingrafted idea." The soul is, therefore, a thinking,
extended substance ('Enchiridion Metaphysicum, Chap. XXVIIL, 3 et al.) and is
"intrinsically endowed with Life and the Faculty of Motion."
8 'Essay Concerning Human Understanding,' Book II., Chapter XXIII., para-
graphs 20, 19, 15 et al.
4 More, on the contrary, with his conception of spirit as extended and penetrable,
finds nothing impossible in the view that a spirit may literally and "easily pass through"
and influence "anybody."
» 'Essay,' Book II., Chapter XXIII., 2.
*Ibid., Chapter XXIII., 15. Locke accords with More in this doctrine. Cf.
More's 'The Immortality of the Soul,' Lib. I., Axiom 8: "The subject or naked essence
or substance of a thing is utterly inconceivable to any of our faculties."
298 MARY W. CALKINS
he declares, is "that conscious, thinking thing (whatever
substance made up of, whether spiritual or material it matters
not) which is sensible, conscious, capable of happiness or
misery, concerned for itself."1 And by 'person' — a term
which he uses as synonym for 'self ' — Locke means f a think-
ing, intelligent being that . . . can consider itself as the
same thinking being in different times and places.'
To speculate on the curious tenacity with which Locke
clings to his empty shell of a soul (expressly reduced to an
'I-know-not-what' and replaced by a self to which he attri-
butes persistence, individuality, and consciousness) is irrel-
evant to our main purpose. Evidently, even Locke's inde-
pendent spirit could not withstand the force of the traditional
doctrine which received from Plato its classic form — the
conception of a being endowed with chiefly negative char-
acters, a not-body, uncompounded, unchanging, immortal;
and he seems also to have clung to the soul as a resource in the
self's frequent lapses of forgetfulness;2 but modern students,
unhampered by the classic convention, have not scrupled to
take the step from which Locke instinctively recoiled: they
have rejected altogether the merely inferred, 'immaterial,'
'life-giving' soul — but unhappily most of them have failed
to take the preliminary precaution of transferring from the
soul to the self the actually experienced characters of per-
sistence, individuality, and consciousness.
It is thus evident that our study of the concept of soul has
provided us with an answer to the question from which we set
out: why do not all psychologists (and philosophers) ac-
knowledge the existence of the self? One answer to this
question was suggested in the beginning: psychologists, like
other men, are naturally inattentive to the permanent
background of all experience; but a second, a historical, ex-
planation of this pertinacious prejudice emerges from our
study. From the very start the self has been confused with
the soul; that is to say, the soul has been conceived not only as
life and as immaterial substance but also as conscious being.
Now when modern biologists, following Descartes's lead, hold
1 ' Essay,' Chapter XXVIL, 17.
2 Ibid., Chapter XXVIL, 10.
THE CASE OF SELF AGAINST SOUL 299
that organic processes can be mechanically explained — these
mechanists in rejecting the concept of soul as life, reject with
it the wholly different conception of the conscious being. And
when psychologists and philosophers alike discard the partly
mythical, partly empty conception of the soul as simple, im-
material substance they too reject also the conscious being.
In their justified but undiscriminating attack on vitalism and
immaterialism many scientists have thus unwisely rejected
the basic fact of psychology: the directly experienced, changing
yet persistent, individual, and complex conscious self. And,
in similar fashion, the philosophic critics, from Hume down,1
have been wont to attack not the experienced self but the soul.
The remedy is obvious and has already been suggested. The
reinstatement, imperatively needed, of the self in psychology
requires, first, that the self take over from the soul the ex-
perienced characters which comport so oddly with the in-
ferred attributes of soul; and second, that the soul thus
despoiled of consciousness, be dismissed from psychology —
unless, indeed, the term be retained as mere synonym of
conscious self or I.2 The reasons for this expulsion of the
soul from psychology — and indeed from philosophy — need
hardly be re-stated. The metaphysical soul or incorporeal,
simple, unchanging being is clearly unjustified by experience,
an empty abstraction except for the characters — persistence
and individuality — which as truly belong to the self. With
equal vigor, the vitalises soul, or guiding entelechy, — opposed
by Descartes but warmly upheld in our own day — must be
rejected. For some of the facts on which vitalism rests its
case, the occurrence, for instance, of life-furthering instincts,
may be mechanically conceived; and the rest are not vital
phenomena at all but experiences of the conscious self.3
1 Yet Hume distinguishes soul and self. Cf. 'Treatise,' Book I., Part IV., Sections
V. and VI.
2 In this sense, for the most part, Leibniz and Berkeley use the term (cf. the phrase
"mind, spirit, soul or myself" in 'Principles of Human Knowledge/ II.) In the doc-
trines of Wolff, Baumgarten and other rationalists of the eighteenth century the
vitalistic and the immaterialistic conceptions of soul recur. Cf. Baumgarten, 'Meta-
physica,' III., II, I, § 750, Wolff, 'Psychologia rationalis,' §§ 48, 49.
3 It is interesting to find that some of the modern biologists who challenge the
extreme claims of mechanism, virtually identify vitalism with idealistic and personal-
300 MARY W. CALKINS
There is, in a word, no middle term between the mechanical
phenomenon and the self. This study, therefore, culminates
in one insistent conclusion: the soul must go. As a historic
concept of immense influence it will always retain its prom-
inent place in the history of ideas; as a term of modern
psychology it has outlived any use it may once have had and
has become a source of mischievous confusion.
istic philosophy. Thus, H.V. Neal writes: "Biologists must . . . accept the idealistic
assumption"; and again "the organic individual is in reality spiritual," ('The Basis
of Individuality in Organisms; a Defense of Vitalism.' Address before the American
Society of Zoologists, Science, N. S., XLIV., pp. 82 if.) Other biologists, of whom
J. A. Haldane may serve as type, reject conventional vitalism along with pure me-
chanism and count themselves with the personalists ('Mechanism and Personality').
RELATION BETWEEN STRUCTURAL AND
BEHAVIOR PSYCHOLOGY
BY A. P. WEISS
Ohio State University
The purpose of this paper is to show that the problems in
the stud/ of mind which are ordinarily the problems of the
structural psychologist, may be studied from the behavioristic
point of view, in accordance with the methods employed in
the natural sciences and with a greater degree of simplicity
than is possible from the structuralistic point of view. No
attempt will be made to determine whether behaviorism is or
is not psychology. The title of this paper might also have
been 'The Value of the Objective as Compared with the Sub-
jective Method in Psychology' were it not for the fact that
the terms 'objective' and 'subjective' because of their un-
certain connotation, are in evil repute both in psychology and
in all the social sciences.
PROBLEM OF STRUCTURAL PSYCHOLOGY
To state the structuralists' problem is not an easy one if
justice is to be done to all the psychologists who are repre-
sentative of this school. There are, however, certain funda-
mental conceptions which are characteristic of structuralism
that are generally accepted, and it is to these that the dis-
cussion will be limited.
By the term structural psychology is meant the type of
investigation which assumes that there is an existential datum
called mind or consciousness, within the totality of which
ultimate differences may be discriminated which are given the
names of (i) sensations, (2) images, (3) affections. These are
usually regarded as the elements into which consciousness may
be analyzed. The aim of the structural psychologist who
accepts this analysis is to describe consciousness, in all its
301
302 A. P. WEISS
complexity, in terms of these three fundamental classes of
consciousness. There is, of course, considerable difference of
opinion between structuralists as to just how these classes are
to be defined, and indeed, whether there are no more and no
less than three elementary categories. But for our purpose
these details may be disregarded. It is also irrelevant for
us whether it is held that consciousness is present in a situation
in which sensations, images and affections, as such, are absent.
It is sufficient if we recognize clearly that the structuralist
aims to describe the structure of the mind or consciousness,
under the manifold conditions of present-day life, in terms of
whatever elements he may have set up, and that he considers
the descriptive phase of his investigation as complete when
he has done this. The explanatory phase of the structuralist
problem is the determination of the neural correlates of the
conscious states which his analysis reveals.
The direct method by which the structuralist analyzes a
conscious complex into its elements is that of introspection.
This cannot be done satisfactorily under normal conditions.
It is therefore customary to create an experimental situation
similar to the one to be analyzed, in which the observer is
asked to report in language the mental states that were
present during the experiment. These reports of the mental
states are then treated statistically. The experimenter
assumes that the verbal reports of mental states are not the
mental states which the observer introspected, but that the
reports are merely expressions which describe the character
of these mental states.
There is considerable difference of opinion among struc-
turalists themselves as to whether introspection gives equally
valid results in all cases where it may be used. Some maintain
that even the most complex mental states may be analyzed
by the introspective method, while others maintain that in-
trospection modifies complex conscious experiences so that
the elements which it reveals may not be present in the actual
experience which is being investigated.
STRUCTURAL AND BEHAVIOR PSYCHOLOGY 3°3
ORIGIN AND PROBLEM OF BEHAVIOR PSYCHOLOGY
As the applications of structural psychology became more
numerous in the various applied fields of psychology such as
education and medicine, the question as to the relation
between conscious states and action arose. An increasing
number of psychologists believed that sensations, images,
memories, thought, emotion, affection, etc., were the causes or
invariable antecedents of certain forms of human behavior.
This naturally led to the question as to how consciousness was
related to changes in the form of human behavior. Now,
since no one denied that, in order to modify behavior, a cor-
responding change in the neural correlates of consciousness
must take place, this inquiry necessarily came to be stated as
follows: How can consciousness change the direction of a
nervous process? In its most general form this question is
an inquiry into the relation between mind and body. In
animal psychology the mind-body problem is not an acute
one for the zoologist-psychologist who studies animal be-
havior; and the success which these investigators have had in
explaining some of the most complex forms of animal be-
havior in purely neural terms (tropisms, reflexes, instincts)
has led some of the psychologists to ask whether human
behavior could not be placed in the same category with animal
behavior and explained without the introduction of a mental
or conscious factor.
The attempt to answer this question by the application of
the behaviorist point of view to such relatively simple forms of
human behavior as reflexes and automatic action was crowned
with considerable success, but for the more complex behavior
of learning the strictly neural explanation is not readily
accepted by psychologists. It will not be necessary at this
stage to state the behaviorist problem in detail except to
indicate that the behaviorist is concerned with determining
the properties and laws of the neuro-muscular system, of
which the introspective reaction is a part.
Briefly, we may state the problem of the structuralist as
an attempt to answer the question: What are our mental states
304 A. P. WEISS
and how do they come about? while the behaviorist problem is:
What are our actions and how do they come about?
The following paragraphs are devoted to a discussion,
from the behaviorist point of view, of those factors of struc-
tural psychology which give it its distinctive character:
namely, the relation between mind and body or the character
of consciousness; and the character of introspection.
BEHAVIORISM AND THE MIND-BODY PROBLEM
Discussion by psychologists of the mind-body problem has
led to the formulation of the three following positions :
1. Consciousness does not enter as a causal agent into
such actions as reflexes and instincts, but it does function in
what is usually known as intelligent or voluntary action.
2. Consciousness cannot be regarded as the invariable
antecedent to any kind of action whatsoever.
The first position is the one usually taken by popular
psychology and the group of psychologists known as "func-
tionalists," while the second position is the one maintained by
behaviorists. The position of the structuralists is usually
known as the double-aspect view, in which there is,
3. Parallelism between conscious processes and neural
processes without a causal relationship between them.
This position is best described by the following quotation :
"Our own position has been that mind and body . . . are simply two aspects of
the same world of experience. They can not influence each other because they are
not separate or independent things. For the same reason, however, whenever the two
aspects appear, any change which occurs in the one, will be accompanied by a corre-
sponding change in the other."1
The third view may be regarded as an intermediate posi-
tion between I and 2. That is to say a functional re-
lationship does exist between mind and body but only in the
mathematical sense of the term function, as when we speak
of the volume of a sphere as a function of its radius.
If this functional relationship between consciousness and
behavior is a simple one it would be useful for the behaviorist,
because consciousness could then be used as a measure of
1 Titchener, E. B., 'A Text-Book of Psychology,' 1912, p. 9.
STRUCTURAL AND BEHAVIOR PSYCHOLOGY 305
behavior, no matter what the ultimate connection between
mind and body might be.
Setting aside all speculation about this connection we find
that the facts in the case may be stated in the following
propositions:
1. Consciousness (the totality of our sensations, images
and affections) is a purely personal experience and has no
scientific value or validity unless it is expressed in some form
of behavior, such as speech or other form of representation.
2. Many forms of behavior (reflexes, automatic action) are
not accompanied by consciousness which can be unambigu-
ously analyzed.
3. The consciousness which does accompany a given form
of behavior varies from one observer to another, and at dif-
ferent times for the same observer.
4. Complex mental processes, such as reasoning and in-
vention, are more than the mental states into which they may
be analyzed. They have a social reference which no amount
of introspection will reveal. This social reference is measured
by the comparison of the individual's behavior, either with
his ordinary behavior or with that of other individuals.
From the above propositions the behaviorist feels justified
in making certain deductions:
From proposition I : If consciousness must be expressed in
some form of behavior before it becomes a scientific datum,
then consciousness is in the last analysis a classification and a
study of behavior.
From proposition 2: If there are many forms of behavior
(reflexes, automatic action) which are important from a social
or individual standpoint, which are not accompanied by
consciousness, then consciousness is not a function (mathe-
matical) of all forms of behavior.
From proposition 3: If the consciousness which does ac-
company a given form of behavior, varies from one person to
another, or for the same person at different times, then con-
sciousness is either an independent variable of the behavior,
or a dependent variable of highly complex formula. In
neither case is it any better measure of behavior than the
306 A. P. WEISS
behavior itself, and consciousness then becomes superfluous
as a means for predicting behavior.
From proposition 4: If complex mental processes such as
reasoning or invention are more than the mental states into
which they may be analyzed, then in this case too the analysis
of consciousness does not make it possible to predict the in-
dividual's behavior.
In view of the above implications, the very simplest as-
sumptions as to the quantitative relations between sensations,
images and affections which characterize an individual's con-
scious complexes, it is impossible to predict what his mental
states will be after an hour. An individual's behavior,
however, can be predicted with much greater certainty than
can the character of the conscious processes which go with it.
A daily photographic record, for example, will show greater
uniformities in the behavior of a given individual so far as
the socially significant factors are concerned, than will the
daily introspective descriptions of the consciousness which
accompanies his behavior.
It would seem then that the sensations, images and affec-
tions which one experiences are more variable than is be-
havior, since by the term behavior we usually refer to only
the grosser bodily movements and to the speech reactions.
The finer changes in muscle contractions, as in respiration,
digestion, vascular and glandular changes, are ordinarily not
considered as socially significant.
For the behaviorist, then, the double-aspect view of the
structuralist is of no value, since the function (consciousness)
of the behavior is more variable than the behavior itself.
CONSCIOUSNESS AN INFERENCE DERIVED FROM THE INTRO-
SPECTIVE REACTION
The structuralist is usually disposed to consider the in-
vestigations of the behaviorist as quite outside of his field,
because consciousness or mind is something entirely distinct
from the contraction of a system of muscles. In the last
analysis, however, the consciousness of his observer which
the structuralist regards as the unique phase of his investi-
STRUCTURAL AND BEHAVIOR PSYCHOLOGY 3°7
gallon, is only an inference from the verbo-motor behavior
which is called introspection. Structuralists' experiments are
controlled reactions to stimuli just as in any behavior experi-
ment. The sensations, images and feeling that are described,
are not experienced by the structuralist. They are inferred
from the behavior of his observers.
When for instance, an introspective reaction reads, "My
attention fell upon the central projection, and vocal-motor
imagery of 'Gosh, that's ugly' occurred. I was aware of con-
traction of my brows and unpleasantness." The structuralist
infers that these speech reactions are descriptions of the con-
sciousness that he has when he makes the same speech re-
actions. To every introspective report of his observer, the
structuralist adds, at least implicitly, the speech reaction,
namely: "The introspective report of my observer completely
describes certain conscious processes of which I am myself
conscious or aware, when I make the same speech reactions."
In other words, psychologists have agreed among them-
selves that an introspective report is more than a reaction; it
is a verbal reaction plus some kind of a conscious process,
which may be either sensorial, imaginal or affective. The
behaviorist may well ask of what scientific value is this habit
of supplementing, in a methodological sense, the observer's
reaction by another reaction of an entirely different type
(consciousness) which the observer did not report. By in-
ferring a conscious correlate to the introspective report — the
experimenter is not able to derive anything more from the
report than is expressed in it. The manual and introspective
reactions of the observer in a psychological experiment can be
classified just as minutely and scientifically without inferring
a conscious correlate, as they can with it.
BEHAVIORISM AND INTROSPECTION
From the behavioristic standpoint, introspection may be
regarded as an example of habit formation or learning. An
untrained observer cannot make these introspective speech
reactions in a manner which is constant and uniform enough
to permit of statistically treating the results. In order that
308 A. P. WEISS
these supplementary speech reactions may become uniform
enough to admit of statistical analysis, two methods of
simplifying the results are usually employed:
1. The observer passes through a preliminary period of
training in which the reaction time of relevant speech reac-
tions (report of imagery, kinesthesia, etc.) is reduced to the
experimental requirements, and those speech reactions which
are irrelevant are given an opportunity to disappear. The
relevancy of a given speech reaction is usually determined
beforehand according to the specific aim of the experiment.
2. The experimental situation is so modified by supple-
mentary situations in the form of instructions, questions or
other controls that the required uniformity of the introspec-
tive reactions is secured.
The first method is the one with which we will especially
concern ourselves in the following discussion.
As already indicated, the total reaction of an organism to
even the simplest situation is very complex. Respiration,
secretion, circulatory effects, incipient and minute bodily and
visceral movements, are always added to the particular
reaction (the discrimination of two points, for instance) which
might be called the major reaction of the experiment. The
minor reactions by themselves would escape notice, but the
speech mechanism is a rather sensitive index as to the char-
acter of these minor reactions. Permitting a free verbal
report after each trial or series of trials in an experiment,
reveals more as to the character of the total reaction than can
be learned from the major reaction alone.
Organic and kinesthetic reactions which would escape
observation entirely are thus easily revealed by the speech
reaction. If, for instance, an observer who has been trained
according to method (i) reports a * kinesthetic sensation of
movement of the neck' that report from the behaviorist
point of view would mean, 'an incipient or slight movement
of the head,' provided his methods of measuring such move-
ments were accurate enough. In this way the verbal reaction
* kinesthetic sensation of neck movement' really indicates
that the receptors in this case are located in the neck muscles.
STRUCTURAL AND BEHAVIOR PSYCHOLOGY 309
The speech reaction may therefore be regarded as a function
of particular kinesthetic and organic reactions which ordi-
narily would escape observation.
To report kinesthetic and organic conditions through
speech is really a highly specialized example of habit forma-
tion. Indeed introspection, whatever its kind, from this
point of view may be regarded as a continual process of
training in the formation of habits. For this reason many
years of training are necessary even for those who have
special aptitude in this direction, and many persons never
learn to introspect well.
The training of an individual in introspection is, therefore,
a process of:
1. Training in discrimination so that weak stimuli in ob-
scure receptors lead to verbal reactions.
2. Substituting for the usual anatomical and physiological
descriptions (head, neck, arms, muscles, viscera, etc.) the
special terminology of the structuralist (sensations, images
and affections).
The relation of these obscure receptors to the major re-
action of the experiment may be of a highly adventitious
character. A trained observer does not react exclusively to
the experimental situation — one is almost tempted to say
that he reacts to everything but the experimental situation.
An experiment with trained observers measures principally
the training of the observers, it does not necessarily reveal
what the normal individual could be expected to do under
similar situations or conditions.
From the behavioristic standpoint, then, introspection
may be regarded as a greatly augmented reaction to a given
situation. The observer not only reacts to the situation, for
example, by pressing a key, but he is also asked to react by
speech to the stimulation of many secondary and obscure
receptors, to which, under normal conditions, no such reac-
tion is made.
310 A. P. WEISS
HEURISTIC VALUE OF INTROSPECTION
When the structuralist experiment is stripped of its con-
scious reference and regarded from the standpoint of objective
science, its twofold character immediately comes to light as
may be seen from the following quotation which is part of the
conclusion of a long and painstaking series of experiments
"On the Analysis of a Phase of the Process of Classifying."
"The essence of the process of classifying, as this process occurred in our experi-
ments, consisted in the manner of our observer's perceiving the object which he had
been instructed to classify. This manner of perceiving consisted in the fact that the
region of essential group features — were stressed in consciousness, and these regions
behaved in consciousness in a fashion which depended upon their resemblance or lack
of resemblance to the corresponding features in the group members. In the former
event, the regions in question passed in and out of consciousness in rapid and ready
fashion, without retarding the course of attention. In the latter case, on the other
hand, the course of attention was arrested sharply; these regions often persisted in
consciousness, and they were frequently accompanied sooner or later by more or less
focal and intensive kinesthetic, organic and affective contents which functioned in their
conscious settings as definite rejectings of the figure."1
Will the above conclusions, we may ask, enable a botanist
or zoologist to classify his specimens more effectively, or will
the scientist know any more about classifying than he did
before he read the article? The question may even be asked,
whether this particular experiment is really an investigation
of the process of classifying. From the behavior point of
view at least, the investigation would be regarded as an ex-
periment in discrimination and habit formation, carried out
under the following conditions:
An experimental situation was created which approached
the conditions under which people make classifications, except
that the classifications are rather more than ordinarily dif-
ficult. A number of observers who had been carefully trained
to respond verbally to the weak stimulation of obscure re-
ceptors, were asked to perform a double task; to classify the
cards, and also to react to the obscure receptors by verbal
1 Fisher, Sara C., 'An Analysis of a Phase of the Process of Classifying,' Am. J. of
Psychol., 1917, 28, p. 115.
I have chosen this particular illustration because I regard it as one of the best
structuralist experiments that has appeared for some time and I need hardly say that
no criticism is intended either of it or the method by which the investigator carried it
to completion.
STRUCTURAL AND BEHAVIOR PSYCHOLOGY 3"
reactions which conformed to a more or less uniform termin-
ology. An analysis of the author's conclusion quoted above,
in behaviorist terms, may be stated in the following proposi-
tions :
1. The observers were able to classify the cards.
2. The observers reacted either to the similar or to the
dissimilar features of the cards. The reactions to the similar
features were relatively simple; the reactions to the dissimilar
features were supplemented by verbal reactions to obscure
kinesthetic and organic receptors.
The whole emphasis of the experiment is placed upon what
the behaviorist would call the reactions to obscure receptors.
The actual process of classifying the cards was merely a device
to obtain some degree of uniformity in the verbal reactions to
these obscure receptors. From these considerations we can
see why the experiment has so little significance for the sci-
entist who really wishes to increase his classifying efficiency.
The development of the ability to react to the weak stimu-
lation of obscure receptors may be of value in the above ex-
periment as a means of determining when the investigation
of the classifying process was being supplanted by the minor
reactions which have no direct bearing on the process of
classifying.
That is to say, the presence of sensations, images and
affections, indicate that there is something wrong with the
experiment.
In ordinary scientific observation, the aim is to eliminate
the unessential or obscure so that the effect of the major con-
dition can be observed in isolation. Under behavioristic
methods, however, if it is desired to investigate the minor
reactions independently, then every effort is made to isolate
them and they of course then become the major reactions.
But the introspective reactions in this particular experiment
would be regarded by the behaviorist as a disturbing factor.
In our effective adjustments to our environment, we do not
have sensations, images or affections. When we stop to
introspect as to the character of our consciousness in a given
situation, our reactions to that situation become to this
312 A.P.WEISS
extent irrelevant. The writer has frequent occasions to
design modifications in apparatus used by the students
working on original problems. Having also some episte-
mological interest, it often happens that he begins to intro-
spect on the method by which the designing process is going
on. When this happens, he might as well stop. Introspec-
tion seriously interferes with the designing activity. The
reactions, incipient though they be, which are called intro-
spections, change the character of the designing activity to
such an extent that it can no longer be called the designing
activity. This is gone; he is merely reacting to a new situ-
ation by making aimless sketches and many incipient verbo-
motor processes which if recorded might read, "Bronze
contact, visual image of tuning fork, kinesthetic sensations of
neck muscles, slight unpleasantness, image of Mr. M. before
chronoscope, visual image of Prof. X. rushing upstairs, etc."
None of these reactions would have occurred had he
actually completed the original design. This is, of course, not
a unique situation. Under the ordinary conditions, sensa-
tions, images and affections are absent. The average man
probably never has them; he must be trained to have them
in the same way that we must be trained to read and write.
Introspection is only one of the ways by which we may
react to a situation. How the reactions themselves originate,
or what the conditions are under which they become modified,
is not revealed by introspection.
TAUTOLOGICAL CHARACTER OF INTROSPECTION
By selecting a somewhat simpler activity than that of
"classifying" it can be shown that the introspective records
are merely special terms for particular classes of reactions.
Suppose we take the process of recognition from the struc-
turalist standpoint. We may assume that the recognitive
consciousness is made up of a characteristic pattern of sen-
sations, images and affections. Suppose the experiment is
one in which we are to select a card (A) which has been pre-
viously examined, from a series of similar cards which have
not been previously examined. We wish to know whether
STRUCTURAL AND BEHAVIOR PSYCHOLOGY 3*3
the card (A) has been recognized. To do this, the behaviorist
would inspect the introspective reactions of the observer.
Suppose the verbal reactions to the card (A) to be the follow-
ing. "I have seen it before; I have a visual image of it as it
appeared when I saw it for the first time; I have the feeling of
familiarity; I have organic and kinesthetic sensations
(images?), which mean that I am handling it in a more ener-
getic and discriminative fashion than the other cards; etc."
To the new cards the verbal reactions might be as follows:
"I have never seen these cards before; the kinesthetic and
organic sensations are different than for card (A); they do
not carry the meaning of familiarity; the visual and auditory
imagery is different than for card (A)', etc." It is not sup-
posed that this is the complete introspective record, and we
disregard for the present such objective measures as the
reaction time, or of the bodily movements which might have
been made.
If we regard the introspections from the purely objective
side as being muscle contractions of the speech mechanism
which produce the various sounds which we call words or
speech, we can immediately see that the observer has reacted
differently to the card (A) than to the other cards. It is not
necessary to interpolate a conscious process (of recognition)
in order to see this. The fact is, that if the observer has used
the identical words to describe both the (A) and the new cards,
we should be obliged to conclude that there was no difference
in the observer's consciousness of the (A) and the other
cards. That is to say, a difference in consciousness can only
be inferred from a difference in behavior. Moreover, the
interpolation of consciousness does not explain how it came
about that the card (A) was reacted to in a manner dif-
ferent from that of the other cards. The introspective record
merely reiterates the fact that the observer did react differ-
ently. This could have been seen directly by regarding the
report of the introspective reactions as what, in fact, they
are — objective reactions of the speech type. Because the
observer reacted differently to the (A) than to the other
cards, the experimenter says he has given a recognizing
314 A-P- WEISS
reaction. We cannot say he reacted differently because he
recognized, but that he recognized because he reacted dif-
ferently. In this sense the inference of the conscious process
of recognition is tautological.
STRUCTURALISM A PHASE OF BEHAVIORISM
The structuralist point of view can, of course, be con-
sistently maintained. There is a justification for inferring
the existence of a conscious correlate for at least some of our
actions, but the heuristic value of this assumption seems
doubtful when it is shown that behaviorism is not less dis-
criminative or descriptive than structural psychology, and in
addition it has the promise at least of a system whose parts
are causally related to each other in the sense that a temporal
description of a given form of behavior is more uniform than
the temporal description of the conscious states which are
said to go with it. If this is true then the law of Parsimony
will operate to eliminate the unnecessary assumption of a
conscious correlate for behavior (implied by introspection).
When we recall how we are being exhorted to recognize the
unconscious, subconscious, higher thought processes, unan-
schaulichen Bewusstseinsinhalt, one wonders whether this
process of elimination has not gone further than we suspect.
For the behaviorist the structuralist's classifications do not
lead to a solution of the problem, What are our actions and
how do they come about? or How do the manifold social
adjustments of the adult grow out of the relatively simple
reactions of the child ? Even if the objective side of the struc-
turalist problem be further developed, the neural correlate of
conscious processes will only be a very special problem in
behavior. Sensations, images, affections, emotions, will, the
self, recognition, etc., so far as science is concerned are only
special instances of receptor-effector activity. While this is
of course very generally recognized by psychologists, the
further assumption that these forms of receptor-effector
activity are highly significant for our understanding how man
is able to make his effective adjustment to his environment, is
not warranted, because the analysis and isolation of such a
STRUCTURAL AND BEHAVIOR PSYCHOLOGY 3*5
neural correlate does not indicate how the neural correlate
itself came to have the configuration that it has.
This, of course, can only be done by a genetic study of the
particular type of reaction. For example, when the structur-
alist has determined the neural correlate for the process called
"recognition," his explanation is complete. For the be-
haviorist this is merely a type of reaction whose effector-
receptor phase is described by the proposition: To objects
to which we have reacted once, we do not react in identically
the same way the next time they are presented, even though
the stimulus conditions approach identity for both occasions.
This is merely a statement of fact which becomes valuable
only when stated quantitatively.
The more important scientific question: How does this
action come about, can only be answered by increasing our
knowledge of neural function.
It is in the above sense that we may regard the intro-
spective reaction of the structuralist as only a part of the
total problem of determining how man makes his manifold
adjustments to his environment.
BASIS OF BEHAVIORISM
In order to state the position of the behaviorist more con-
cretely, the following assumptions as to the explanation of
human behavior are presented.
1. The resistance of neurons varies with function.
2. Every receptor is directly connected by a neural chain
with a restricted effector system, and through varying degrees
of indirectness to many other effector systems.
3. One nervous process will modify the character of other
nervous processes that occur together with it.
4. Certain configurations of neural connections between
receptors and effectors are inherited, some are acquired.
These assumptions are not essentially different than those
formulated by Max F. Meyer.1
From these four laws or assumptions it is evident that
there is a possibility of greater variations in the response of an
1 'The Fundamental Laws of Human Behavior,' 1911.
316 A. P. WEISS
organism than is usually included under what the structur-
alists call the study of mind.
Each of the four propositions represents a series of problems
which may be experimentally handled in the same manner
that an experiment in any other science would be conducted.
There need be no quarreling about consciousness or as to
whether there is or is not such a thing, since it will make no
difference in the effectiveness of the work of the behaviorist
or any scientist how the question may finally be settled. In
their present form the propositions represent little more than
a program for future work. Each proposition will be divided
and subdivided and coefficients will be supplied to the various
derived propositions. How fruitful the analysis will be
cannot of course be predicted, but if structural 'psychology,
with all the painstaking and careful work that has been done
on sensorial and imaginal processes, has not yet developed a
definition of sensation upon which two psychologists will
.agree, then behaviorism can expect a considerable degree of
tolerant good will.
SUMMARY
1. The close academic relationship between structural
psychology and behaviorism is due to the popular belief that
mind and body are related in such a way that mind produces
or modifies behavior.
2. A conscious state which is not expressed in some form
of behavior is, so far as science is concerned, non-existent.
The inference of consciousness is, therefore, unnecessary,
since in the last analysis behavior is the only thing that can be
classified.
3. For the behaviorist the introspective reaction is only
the habit of being able to react by speech, more or less adven-
titiously, to the weak stimulation of obscure receptors.
4. Even if it is admitted that the special introspective
habits are correlated with conscious processes, the latter are so
variable that they cannot be used as an index from which to
predict behavior which is socially or scientifically significant.
5. The method of introspection favors the reactions to
obscure stimuli and in a corresponding degree decreases the
STRUCTURAL AND BEHAVIOR PSYCHOLOGY 317
reliability of the major reaction for the investigation of which
the experiment is designed.
6. Behaviorism presents as manifold possibilities of an-
alysis and classification as does structuralism, and has the
added advantage that its phenomena can be represented as
a causal series in the same sense as we now speak of causal
relationship in the natural sciences.
7. When the behaviorist solves the problem: How do our
actions come about? all the problems of the structuralist will
also be solved.
8. The relation between structuralism and behaviorism is
such that they may exist side by side. This will occur only
when the fundamental conceptions underlying both methods
are not very closely scrutinized or where the individual is not
particularly interested in the 'theoretical implications of either
the one or the other method.
DISCUSSION
MEANING AND IMAGERY
In the discussion of imageless thought one must bear in mind
the fact that there is more than one way in which it might be held
that thought is imageless.
1. One might for instance say that thought is not only not
identical with imagery, but that it is always and under all circum-
stances independent thereof, that imagery is in all subjects and at
all times subsequent to thought and never helps in the solution of
problems that can be thought about. This would be an "out and
out imageless position," but one which, so far as I know, no one has
yet taken, though some approach it. It would also contradict
various pieces of experimental work.1
2. One might hold that thought is imageless in the sense that it
is a mental process sui generis, distinct from imagery, but never
imageless in the sense that it is ever present in the mind unaccom-
panied by imagery. One taking this stand might even maintain
that the thought process is dependent on previous mental imagery —
in the sense that the image is the source from which the thought
process is derived.
Such seems actually to have been the theory of Aristotle who,
while distinguishing between and even contrasting kirurriiijai and
ctla6f}<nsi maintained that the mind never thinks without (pavraffiJiaTce.
In fact whenever the mind sees, it is necessary for it to behold
imagery — OTOLV re Oeupfi avajKfj a/za (pavraff^a n Oeuptiv. The next
sentence shows that he is really speaking of imagery rather than
the forms of sense perception : ra yap (pavraffnara axnrep alffdrjuaTa
kffTL ir\fiv avev vArjs.
So also St. Thomas maintains that the human intellect cannot
actually understand anything without recourse to images. It is
interesting to note that this position was taken on the following
empirical grounds:
(a) When cerebral lesions are sufficiently serious to affect mem-
1 Cf., for example, George Herbert Betts, 'The Distribution and Functions of
Mental Imagery,' Teachers College, Columbia University, New York, 1909.
2 Trepi tvxw, III., viii, §3.
318
MEANING AND IMAGERY 3 1 9
ory and imagination they also interfere with the orderly course of
thought.
(&) An appeal to introspection — "quia hoc quilibet in seipso
experiri potest" — viz., "when one attempts to understand some-
thing he forms to himself certain images by way of example."
Likewise when we want to explain a matter to someone else we give
examples from which he can derive images that will lead to under-
standing.1
3. Without adopting either of these positions, one might say
that while the image is frequently very useful, nevertheless, a thought
process can be present to the mind without simultaneous imagery
and that it is not necessary for us to draw all our meanings and
thoughts from the contemplation of imagery. Thus thought would
be imageless in the sense that it is distinct from the (pavTaffnara.
It would often be imageless, too, in the sense that it would not be
dependent on the ipavracr^ara. or preceded or accompanied by
them.
That the extent to which one makes use of images, and the
rapidity with which they come, should be subject to individual
variations is certainly to be expected.
In an article on the 'Temporal Relations of Meaning and
Imagery,' 2 I pointed out that with eight out of nine subjects I found
that meaning preceded visual and kinsesthetic imagery in the per-
ception of printed words, but that in memory, on the contrary,
imagery precedes meaning. I was thus led to adopt the last of the
three positions I have above outlined.
I did not maintain that this relationship of meaning and imagery
in perception was universal. I was nevertheless inclined and still
am inclined to believe, mainly on empirical grounds, that it will be
very rare that individuals will be found who present genuine excep-
tions. The one subject who proved an exception in my experi-
ments, did not learn to refrain from reacting to nonsense words,
so that one could not be sure of just what he was reacting to, when
supposed to be responding to the task meaning or imagery.
I remember talking the matter over with Biihler, who thought
that exceptions would prove more frequent than I supposed.
Since then, I have often wondered what experiments on a larger
number of subjects would reveal.
In the March issue of the PSYCHOLOGICAL REVIEW, Edward C.
1 Cf. 'Summa Theologica,' I., Q. Ixxxiv, § 7.
2 PSYCHOL. REVIEW, May, 1915.
320 T. V. MOORE
Tolman has undertaken such an investigation on 49 subjects. He
attempts a preliminary sorting out by having his subjects react to
words representing black or white objects: (a) when they knew
whether it was black or white irrespective of the way in which this
was known; (#) when they could see it from their visual image. All
but fourteen of the subjects gave shorter times for knowing than
for visualization. I regret that this modification was made use of,
because it changes the whole situation. It introduces a new problem
and one in which visual imagery might be very useful and would,
therefore, be more likely to be used than in the former situation.
I am inclined to believe that I would be ranged among the visualizers
in attempting its solution.
Unfortunately this preliminary sorting cannot be compared to
the word perception test and we thus still lack evidence of the rel-
ative frequency of the cases in which imagery precedes meaning in
the process of perceiving printed words.
Having selected fourteen subjects by this preliminary experi-
ment as likely to prove an exception to the usual order he
subjected them to the same type of experiment that I used in
Munich. He finds, nevertheless, that three of these gave the
same order of sequence that I found in Munich. Three subjects
gave doubtful results. Eight gave results distinctly (?) pointing
to the opposite sequence, so that with them reaction time to
meaning was longer than to imagery. In the last group he finds
two subdivisions. The first comprises those who as a rule tend to
distinguish between meaning and imagery.
Let us consider this group a moment and see why their reaction
times to meaning are longer than to imagery. In every one of the
introspections given (pp. 133-134) the meaning is analyzed and more
or less roughly defined and does not correspond to the 'simple
meaning' spoken of in my paper. Sometimes the meaning is evi-
dently a 'concept of purpose' for which I also found much longer
reactions. I have not time here to discuss 'analyzed' and 'un-
analyzed' meanings, but hope to do so soon in a monograph on
'Perception and Memory' now almost ready for publication.
Meaning is a process which undergoes development. It is not
surprising that some subjects wait until it has developed and they
feel sure of a definite analyzed meaning before reacting. In these
subjects one is likely to get reaction times that are longer than those
for visual imagery. This expectation seems to be realized in the
reaction times of this set of Dr. Tolman's subjects and confirmed by
their introspections.
MEANING AND IMAGERY $21
The second subgroup of subjects did not distinguish between
meaning and imagery and their introspections show that when
reacting to what they called * meaning' they were as a matter of fact
reacting to images. In only one member of this group is the dif-
ference between the averages greater than the mean variations.
In none of the others is it even half as great. It might be safer,
therefore, to place all but one of these subjects in the doubtful
group. Whether the task is meaning or imagery they react to one
and the same thing and naturally give in either case about the same
results. The general slight tendency to a longer reaction time with
meaning and the single pronounced tendency in that way with one
subject might readily be accounted for by their being called upon
to pick out what they have not been accustomed to label.
Dr. Tolman's assumption that with them meaning is imagery
supposes too great a distinction of types. It is possible, but scarcely
probable, that one set of normal men understand and visualize,
whereas the other visualize but never understand. I do not doubt
that other experiments would show that these subjects were not
wholly devoid of understanding as something distinct from imagin-
ing.
Dr. Tolman's study does little more than raise the question: Are
there really any genuine exceptions to the hitherto observed relation
in the sequence of meaning and imagery? But what would be the
import of finding exceptions — and, let us say, even numerous ex-
ceptions to the rule? I do not see that the demonstration of even
a large number of exceptions would weaken the argument, drawn
from temporal sequence, to establish a distinction between meaning
and imagery.
Let us suppose that two events moved usually with the same
velocity, so that we could scarcely ever detect one without the
other. If, however, under certain circumstances one moved more
slowly or more rapidly than the other, so that we could observe
one without the other these cases alone would enable us to say that
the two things are not really identical.
If meaning is identical with imagery then the two events must
always come together. If it is not then you will find some subjects
in whom their rate of development is not identical and one will
come before or after the other.
If meaning is dependent on imagery, e. g., in the Aristotelian
sense, then imagery will always precede meaning.
We find on the contrary that meaning as a rule precedes imagery.
322 DISCUSSION
It is, therefore, not identical with it and as a rule does not depend
upon it in the process of the perception of printed words representing
objects easily visualized.
Dr. Tolman's paper confirms this very important fact. At the
same time it presents no absolutely conclusive proof of subjects
who constitute an exception to the rule. It is quite possible,
however, that some subjects will be found capable of distinguishing
between meaning and imagery, in whom the development of images
is so rapid that their mental pictures will usually precede their
simple unanalyzed meanings. Such a fact would be significant in
the study of types of individuals, but would have little bearing on
the more theoretical problem of the existence of imageless thought.
THOMAS VERNER MOORE
CATHOLIC UNIVERSITY OF AMERICA
SOME EXPERIMENTS IN MOTOR REPRODUCTION OF
VISUALLY PERCEIVED FORMS.
Some little time ago I had occasion to observe experimentally
certain cursive figures of an unusual type while engaged in an ex-
periment along lines quite different from that of the experiment
here reported. These figures were presented by means of the
Whipple tachistoscope. They were to be learned by means of
frequent repetitions, so that they could be reproduced with pencil
on paper. The experiment had to do with the relative advantages
of learning figures by means of copies in strong relief traced by the
fingers of blindfolded subjects, and by means of copies of similar
figures exposed to visual perception by the tachistoscope. I acted
as a subject for both forms of learning, and had opportunities to
observe closely other subjects at work.
Early in the course of this experiment I became interested in the
fact that the behavior of the subject at the time of attempting to
reproduce the figure was an important factor in determining the
accuracy of the reproduction. I found it to be true of myself, and
it seemed to be true of other subjects, that if one watched what he
was doing while attempting to make the reproduction it seldom
happened that the result was as satisfactory as if the reproducing
was done with eyes shut or averted. After casual inspection I
found that practically all the subjects concerned were far more
satisfied that their reproductions resembled the model closely when
these reproductions were made without the subjects seeing what
MOTOR REPRODUCTION OF VISUALLY PERCEIVED FORMS 3*3
they were doing. The point seemed to me to be of sufficient interest
to warrant some investigation.
One may have a pretty definite idea of just what he wishes to
draw, and may take his pencil in hand with absolute certainty that
he will draw it quickly and easily. But as he draws it his eyes
follow the outline being reproduced. Does the sight of what he is
drawing aid him in completing the figure accurately? Or do the
visual perceptions interfere with and destroy the guiding memory
images? It is to be understood that the figure which serves as
model is not in sight during the reproducing process.
The problem finally assumed the following form: "If one is
reproducing on paper an outline which he has just memorized what
effect does the visual perception given by the figure in process of
being reproduced exert on the visual or kinaesthetic images which
are directing the process?"
A series of figures was made which could be presented on the
tachistoscope. It was not particularly easy to find figures which
were sufficiently new to really require memorization. It would not
do to have the figures so made that one could easily describe them
to himself in terms of association with other forms or objects. We
had recourse to the Arabic alphabet, and, by adaptation and
combination, constructed therefrom several figures suitable for our
use. There were ten of these figures finally selected, and they were
arranged in two groups of five each. In each group these figures
were arranged in order of difficulty; or rather, we attempted so to
arrange them. We did not discover any principles which would
give an adequate measure of difficulty, and were forced to be content
with arranging them according to the general complexity of the
figures, their number of curves, loops and line crossings. We were
careful to vary the order of presentation of the figures so that if any
were markedly harder than others that fact would not influence the
results in one direction more than another.
There were two methods of reproducing the figures after they
were learned. In the first method, hereafter called the "sight"
method, the subject would watch what he was doing while drawing
the figure. In the second or "blind" method a screen was inter-
posed between the eyes of the subject and the paper so that he would
not see what he was doing while trying to draw the remembered
figure.
The following directions were given to subjects before reacting
according to the sight method:
324 G. R. WELLS
Fixate cross. Observe the figure which appears, and after the smallest possible
number of exposures draw it on the paper in front of you. Try to get both direction
of turns and general proportions of figure correct.
Subjects about to react according to the blind method were
instructed as follows:
Fixate cross. Observe the figure which appears, and after the smallest possible
number of exposures draw it on the paper in front of you, but keep eyes on the screen
between you and the paper while you are drawing it. Try to get both direction of
turns, and general proportions of figure correct.
One half of the subjects followed first the sight method on one set
of figures, and then the blind method on the other set of figures.
The other half of the subjects used the blind method first and then
the sight method. In half the cases the figures used for the sight
process were those of Set L, and in the other cases the figures were
those of Set II. So that it is true that the same figures were some-
times used for reproduction with sight and sometimes without sight.
Also, the same set was at times presented first, and sometimes second.
It does not seem that the results could have been influenced by any
possible differences in difficulty between individual figures, nor that
any possible practice effect was exerted on one reproducing process
more than another.
The exposure time averaged 65/roo1 of a second in length. The
subject was told that he could have as many exposures as he wished,
but was urged to request as few as possible. It was made very
plain that he must make every effort to keep the number of exposures
few in number.
There were twenty subjects. All but one were students in
Oberlin College, the exception being an instructor in economics. All
but three were taking courses in the psychological laboratory. The
exceptions were the instructor just mentioned, and two students who
had had a good deal of training in art. The fact that they had done
practical work in drawing suggested that their inclusion in the list
of subjects might furnish some interesting variations in results.
In point of fact their reactions were in no respect different from
those of the other subjects.
1 The working time of the Whipple tachistoscope varies quite a good deal, but this
was the average time. In the tachistoscope in the Oberlin laboratory we have replaced
the usual cardboard discs with light metal (ordinary sheet tin) discs, and we time the
length of exposure by means of two make-and-break contacts, so arranged that when
the disc revolves a circuit is broken at the point at which the exposure begins. This
circuit is remade when the disc reaches the place at which the exposure is completed.
A magnetic marker is arranged in series with these contacts. This method may leave
something to be desired from the physical standpoint, but, after all, it furnishes one
with some real information concerning the exposure time.
MOTOR REPRODUCTION OF VISUALLY PERCEIVED FORMS 325
Eleven subjects used the sight method of reproduction for the
figures of Set I., and the blind method for Set II. Nine subjects
used Set II. for the sight method, and Set I. for the blind method. It
was merely accidental that there were not ten subjects in each
group.
When the time came to sum up the result of the trials, it was
necessary to find some method of determining the relative accuracy
of the figures as reproduced. No completely satisfactory method of
measuring occurred to me and I finally arranged arbitrary principles
of scoring, and applied them to all the figures alike. The rules
which governed the scoring were as follows:
1. Score I for each exposure required for learning.
2. Score i for a curve in the wrong direction.
3. Score I for very great error in the proportion of parts.
4. Score I for the insertion of any marked alteration.
5. Score i for omission of an essential part of figure.
(Do not score 2 if an essential omission has been part of the result of a curve in the
wrong direction.)
It can be seen that on this scheme the lowest score is the best,
that there can be no score lower than one, and that there is no
maximum score. It should also be noted that the number of ex-
posures required plays an important part in the final score.
The total score made by the twenty subjects when seeing what
they were drawing was 381. This gives an average of 19.05 for
each person, or of 38.1 for each figure.
The total score made by the twenty subjects when drawing
without seeing what they were drawing was 326, an average of
16.3 for each person, and 32.6 for each figure.
It may be objected that the number of exposures should not have
counted in the summing up of the results. Some of the subjects may
have asked for more exposures than they needed, or might have
made equally good reproductions with fewer exposures than they
actually requested. I feel that the number of exposures must have
had some effect upon the accuracy of reproduction, though I am not
quite satisfied with the particular method used for exposing the
figures, that is, by means of the tachistoscope. But if we sum up the
results and neglect the number of repetitions entirely, the relations
of the figures for sight and blind methods is not thereby changed,
although the values of the figures alter somewhat.
The total score for the sight method, not taking the number of
exposures into account, is 130, and the total for the blind method,
under the same conditions, is 120. The average in the first case is
326 G. R. WELLS
6.5, and in the second case 6. The difference between the two
methods is not so great as when the number of exposures is included
in the score, but in both cases the score for reproducing the figures
without sight is lower, that is, is better than when sight is used.
That is to say, the scoring methods used show, beyond doubt, a some-
what marked difference between the accuracy of reproductions made with
vision and those made without vision, in favor of the latter.
Furthermore, the opinions of the subjects, recorded at the time
of the experiment, largely agree that, on the whole, it is easier to
record the remembered figures without vision than with vision.
This agreement was recorded in spite of the fact that many of the
subjects were sure that the records made without vision were far less
accurate than those made with vision, although the final results did
not bear out this conclusion at all.
Thirteen subjects agree that the method of reproducing without
sight is easier than with sight, and a small minority of them feel that
it is the more accurate method. Sometimes subjects would try to
explain the greater ease of one method, usually the blind method, by
saying that one set of figures was easier than another. But it is in-
teresting to note that in just one half of the cases where this state-
ment was made one set was found to be harder, and in the other half
of the cases it was the other set which was found to be more difficult.
Evidently it was not the figures which were difficult, but the method
of reproduction.
Seven subjects found that reproducing the figure with vision
was easier than without. Several of these subjects referred particu-
larly to a difficulty of getting the proportions of the figure correct
when they could not see what they were drawing. It is probably
true that it is easier to get proportions correct by the sight method
than by the blind method.
But of the seven subjects who felt that the easier method was the
sight method, two essentially modified their statements, if they did not
actually contradict them, as follows: One said, "The sight method
is easier than the blind method, because it is easier to visualize when
you can see where the hand is going." But during the course of the
experiment he had been heard to say in slight exasperation, while
working according to the blind method, "Pshaw, I looked under-
neath it (i. e., the screen) and lost it."
Another subject summarizes the situation by saying: "It is easier
to see what one is doing, one gets an idea of where one has gone.
But in the blind method I did it quicker, and drawing with sight the
MOTOR REPRODUCTION OF VISUALLY PERCEIVED FORMS 327
distraction caused the image to be lost." It is significant that "dis-
traction" should be the word used to describe the effect of seeing the
figure he is reproducing while the reproducing is in progress.
In its results the experiment seemed to indicate that in reproduc-
ing such cursive forms as I used, to see what one is doing is a hin-
drance, rather than a help. This conclusion is established in several
ways. Scoring the results in terms of their accuracy shows an
undoubted greater accuracy in the cases of the reproductions made
without sight. This would be true with any fair scoring method,
I think, but it becomes extremely evident if the number of exposures
of the figures used as models be taken into account as well as the
accuracy of reproduction. Further, the large majority of the
subjects, after working with both methods, definitely prefer the
method without sight.
The observations here reported are part of a more extensive study
being made in the Oberlin laboratory of the comparative ease and
accuracy of learning cursive forms by visual and by kinsesthetic
perception. This experiment, still in progress, seems to show that
ease and accuracy of learning stand in nearly direct ratio to the
amount of movement which is involved in the learning process.
This may mean that kinsesthetic imagery has peculiar advantages as
regards retention over visual or other imagery, at least for many
people.
If this should prove true, it is further true in all probability that
the distracting effect of observing the result of one's effort while
trying to reproduce a figure recently seen is due to the substitution
of visual for kinaesthetic guidance.
GEORGE R. WELLS
OBERLIN COLLEGE
VOL. XXIV. No. 5 September, 1917
THE PSYCHOLOGICAL REVIEW
AN ATTEMPTED FORMULATION OF THE SCOPE
OF BEHAVIOR PSYCHOLOGY1
BY JOHN B. WATSON
The Johns Hopkins University
COMMON SENSE PROCEDURE
Psychology a Science of Behavior. — Psychology is a division
of science which deals with the functions underlying human
activity and conduct. It attempts to formulate through
systematic observation and experimentation a series of
principles or laws which will enable it to tell with some degree
of accuracy how an individual or group of individuals will
adjust themselves to the daily situations of life as well as
to the uncommon and unusual situations which may con-
front them. It is equally a part of the function of psychology
to establish laws or principles for the control of human action
so that it can aid organized society in its endeavors to prevent
failures in such adjustments. It should be able to guide
society as to the ways in which the environment may be
modified to suit the group or individual's way of acting; or
when the environment cannot be modified, to show how the
individual may be moulded (forced to put on new habits)
to fit the environment. It must be understood at the out-
set, though, that psychology at present has little to do with
the setting of social standards of action and nothing to do
with moral standards. It does lie within her province to tell
1 The material presented in this article is essentially that which will be used in
the first chapter of my forthcoming book, 'Human Psychology.' It is published
now in the hope that helpful criticism will be furnished the writer both as regards
the level of presentation, that is, its textual fitness, and as regards the completeness
with which the scope, as here outlined, touches the main points of interest in modern
psychology.
329
330 JOHN B. WATSON
whether the individual can act in accordance with such stand-
ards and how we may control him or lead him to act in harmony
with them. These laws of control or training must be general
and comprehensive since social standards are constantly
changing.1
Psychology when looked at in this way is seen to be some-
thing which everyone has been using more or less all his life
without calling it psychology. We learn by failures and
successes how to run our business, how to get along with our
colleagues and associates. We teach our children to act in
certain ways. They must eat with a fork, learn to dress
themselves, to treat their playmates as they themselves like
to be treated, to master the three R's, and then later a trade
or profession. We skillfully or bunglingly steer them on
their course from infancy until they no longer need our
guidance.
The Ancient Origin of Psychology. — Indeed a glance at
the mythology, folk-lore, or history of any given race will
show that the practical psychology of control began as soon
as there were two individuals on the earth living near enough
together for the behavior of one to influence the behavior
of the other. The serpent controlled Eve's behavior by
offering her the delectable apple. Eve learned her lesson
quickly and tempted Adam in the same way. Atalanta, the
swiftest of runners, was beaten not through the superior
agility of Hippomenes but by the fact that she could not
resist the temptation to stop and pick up the golden apples
thrown by her suitor. The taboo system, the initiation
ceremonies and the machinations of the medicine men all
serve to illustrate progress towards the control of group
and individual behavior.
Common Sense Procedure in Securing the Control of Be-
havior.— Long before the dawn of modern scientific psy-
chology society found that by roundabout, hit-and-miss
methods she had secured a fairly serviceable body of data
as to what man can do — his complement of acts; the appro-
priate situation for calling out any given act; and crude
1 For a more scientifically worded discussion of the province of psychology see p. 336.
SCOPE OF BEHAVIOR PSYCHOLOGY 331
training methods whereby the repertoire of the acts them-
selves might be enlarged. A brief glance at the practical
procedure in securing control of individuals and groups
may serve to illustrate both how such data are obtained and
how they are used.
The situations or devices for drawing crowds for whatever
purpose show the greatest development of skill in practical
psychology. If one or two individuals happen by chance
to gather around a patent medicine vendor the vendor's
momentary success is assured. The small crowd is an
irresistible stimulus and soon gathers unto itself a larger
crowd. For this reason initial buyers, listeners, applauders,
and, yes, even sometimes suitors and mourners are provided
and paid for before the public is allowed to act. The adver-
tisement of a 'fire sale' is likewise an ancient and honorable
device for gathering a crowd. The announcement at James-
town of a shipload of virtuous women to be sold as wives
apparently, if historical report can be trusted, brought out
the whole of the unmarried able-bodied male population of
Virginia.
After the crowd has collected > devices for controlling the
individual are numerous. Chief among these we find the
various lottery schemes; many individuals will not give
twenty-five cents for a charitable purpose, but at any charity
gathering they will eagerly take one of a dozen twenty-five-
cent chances on almost any object the total value of which
need not be greater than the cost of a single chance. So
uniform is the response to lottery schemes that they have
oftentimes become national mediums for raising government
funds.
Organized society displays much ingenuity in devising
situations for the control of reaction: the clergymen, having
found empty pews under an orthodox type of routine, and
that the young people were not engaging in church activities,
began to try out in a similar way what could be done by
altering the old austere situation and by creating for the
church entirely new situations. The exteriors of the churches
were vastly modified, the interiors decorated, rest and play
332 JOHN S. WATSON
rooms added as well as gymnasia and playgrounds. A certain
type of action was expected and the church arranged a situ-
ation to bring it about. We see the same attempt at control
illustrated in governing bodies: state legislatures, depressed
by the prevalence of drunkenness and crime, establish a new
situation by prohibiting the sale of drugs and spirituous
liquors in the hope that legislation will prevent such conduct.
Finally, mention may be made of recent changes in prison
methods. The more advanced prisons, becoming dissatisfied
with the amount of insubordination, vice, sloth, and ignorance
of all forms of government among the inmates, are trying a
new situation with good results, viz., that of letting the
inmates establish a miniature republic. This republic makes
its own laws and metes out its own punishments. Society has
developed a system of standards of action but it does not know
just what situations will produce the needed responses. The
situations are set up in the social field by trial and error;
they are modified, changed, etc., until the desired reactions
take place, or until they are despaired of.
Watching the Act to Obtain Data on the Situation. — We
thus, as we see, get a part of our knowledge of the factors
underlying behavior by the trial and error method of manipu-
lating the situation and noting the reactions that take place.
This gives us a body of usable data on what to expect of men
when they are placed in certain situations. We deal here
with situations of our own contrivance. Equally serviceable
results are obtained by taking an individual who is performing
some act (action not regulated by the observer) and ex-
amining immediately into the situation which led to that
act — the act is known, the situation which led to it must be
investigated. This method is of course supplementary to
the first. It extends our knowledge of situations and what
to expect from them, and at the same time adds to our
knowledge of man's repertoire of acts. Illustrations of the
results of this type of practical procedure are numerous;
it is hard to convince a mother that she cannot predict1 what
1 By prediction we mean nothing but the common-sense, scientific, and logical
use of material gathered from observation and experimentation. The planet Neptune
was predicted, Mendelian ratios are predictable, coat color or eye color of animals
can, within certain limits, be predicted before the birth of the animal.
SCOPE OF BEHAVIOR PSYCHOLOGY 333
the situation is which leads her baby to give a certain cry.
Depending upon the variations in the cry she will say that
the 'baby is hungry, wet, or has colic' (her conclusions are
often wrong, be it said). Advancing somewhat in the age
scale we see, in passing through a forest, a youth trained to
hunt firing upward into a tree, and we note that the dog he
has with him has ' treed.' An observer responds to this
picture by telling his companion that the boy is hunting
squirrels. But if he sees the boy fire in another way, say
horizontally, and sees a dog in the act of pointing, he will
state that the lad is hunting quail. If the boy fires towards
the ground and has a hound with him, he is shooting rabbits.
Finally, if he is seen in the hunting fields on horseback, with
no gun but accompanied by a pack of hounds, in full cry,
our observer remarks that the boy is hunting a fox. Watching
his actions and taking note of all attendant circumstances
enables anyone to predict with some degree of probability
the immediate situation leading to the boy's actions. Our
ability to observe an act and predict the possible or probable
stimulus to that act depends upon the fact that often during
our past life, when we have seen individuals doing certain
things, we have immediately investigated the situations
which led to the acts.
Need of Acquaintanceship with an Individual' 's Past. —
When we come to deal practically or scientifically with in-
dividuals too much emphasis cannot be laid upon the extent
to which acquaintanceship with their past life will gradually
afford the basis for making serviceable predictions as to
their probable ways of acting and as to the situations which
will call out any given act. This can possibly be most easily
illustrated in the animal world. We soon come to the
prediction stage with our horses and dogs and can map out
with some certainty what they would do under the various
situations which might confront them. After watching two
monkeys for several years I found after repeated observation
that B would not touch food until / had finished and left the
dish, and that / would eat and stuff three bananas into his
cheek pouches and drag off another with his right forefoot;
334 JOHN B. WATSON
that / would attack a problem box rapidly and in a rough
and harum scarum way — pulling the box towards him,
turning it over and maltreating it generally; that B would
approach cautiously, moving the parts slowly and with no
violence; but as a rule would solve the problem before /.
But with an organism so highly developed as man's, the
prediction of his actions is not always so easy. An individual's
actions in everyday situations depend upon such complex
factors as his heredity, his past success in adjustments and
his failures of adjustment, the responses he has just had to
make, as well as upon the permanent and temporary con-
ditions of his organic mechanisms (digestion, circulation,
sleep, etc.). A badly cooked dinner, an insufficient amount
of food (as in the case of a person who is dieting), extreme
heat, etc., may so change the state of the organism that the
response reasonably expected is not forthcoming. A forth-
comin^jnarriage,, _ gradujlicin^__£epa ra tiojij^joffer ^situations
which , b^M^iejLr in^iil^^ t
completely disrupt forjthe timejbeing the everyday systems
oTTesponses which are customary with a givenindividual.
In such a complex setting a man may show a temporary
breakdown; he may make blunders or show a general in-
ability to go through with his ordinary routine. A bad
dream or a slight rebuff at the hands of a friend likewise
may upset a man's reactions for a whole day. Further on we
shall see that many of the habits formed in childhood and in
adolescence, now long since discarded, such, e. g., as attach-
ments to early playmates, to members of the family, early
love affairs, may have had a prepotent influence in shaping
the whole course of adult acquisitions.
If we are called upon to predict what a stranger of twenty-
five years of age will do when confronted in a dark alley by a
burglar, we are almost, but not quite, helpless so far as
prediction is concerned. Now give us an opportunity of
systematically studying the make-up of the man, of knowing
something of his reactions in past dangerous situations, the
stability of his emotional tendencies, and we shall be able
at least to make a crude but serviceable prediction, viz.,
SCOPE OF BEHAFIOR PSYCHOLOGY 335
that he will quietly throw up his hands and let the burglar
go through his pockets. He will neither become hysterical,
attempt to attack the burglar, nor will he suffer any severe
after-effects by reason of his experience. The chances are
good that he will report his mishap to the police, confess to
his wife or friends that he has been held up, and then will
cease to be further troubled by the experience. In another
type of individual, whose heredity is questionable, whose
reactions are unstable, who is generally excitable and liable
to over-reaction, we venture the prediction that even if he
does throw up his hands (which he does not always do)
and allow himself to be robbed, he will go to pieces after
the experience and may suffer some serious and lasting conse-
quences.
Common Sense a Crude but Genuine Psychology. — Most of
our illustrations have involved little or no technical psy-
chology and yet they do illustrate a genuine psychological
procedure. The business man, the artist, and the artisan
have built for themselves rather definite rules of psycho-
logical procedure without ever calling it psychology. The
church and the theater illustrate this equally well. It is
possibly even a debatable question whether common sense
has not kept closer to the fundamental truth underlying the
psychology of reaction than has the too detached psychology
of the laboratory. But even those who are the best practical
psychologists realize that common sense methods can never
produce universal or widespread progress in psychology.
Our great military leaders, our great religious leaders, the
demagogues and the politicians have accomplished their
results by their very wide acquaintanceship with the reaction
tendencies in man and by their happy accidents in creating
the situations which will call out such reactions. By reason
of the fact that occasional success has been obtained by
crude methods and happy accidents, we must not conclude
that psychology should not attempt to discover and analyze
and bring under scientific control the factors which have
occasionally made such successes possible. Because there
has been an occasional business leader who knew how to
336 JOHN B. WATSON
pick out and keep good men, we are offered no reason why
we should not seek to understand and control the processes
involved in picking and keeping good men. The same may
be said of the factors involved in keeping men out of crime,
keeping them honest and sane, and their ethical and social
life upon a high and well-regulated plane.
This brief summary of the everyday uses of psychology
should convince us of two things; first, that common sense,
while a reasonable method so far as it goes, does not go
far enough and never can; and secondly, that in order to make
progress, the phenomena of human behavior must be made
an object of scientific study. We shall attempt next, then,
to gain some impression of this systematic psychological
procedure.
SCIENTIFIC PROCEDURE
The Detailed Subject Matter of Scientific Psychology. — As a
science psychology puts before herself the task of unravelling
the complex factors involved in the development and regu-
lation of human behavior from infancy through old age.
At first sight it may seem that this program leaves out
many of the factors with which psychology ought to be
concerned. Historically considered this is true, but when we
are confronted both with the practical and scientific needs
of life we are ready to admit that after all what we seek to
have psychology busy herself with is just this matter of
environmental adjustment; what can man do apart from his
training; what can he be trained to do, and what are the
best methods for training; and finally, how, when the varied
systems of instincts and habits have sufficiently developed,
can we arrange the conditions for calling out appropriate
action upon demand? To answer such questions we must
necessarily study the simple and complex things which call
out action in man; how early in life he can react to the
various simple and complex sense stimuli; at what age he
can put on the various instincts and what are the situations
which call them out. Just what are the patterns of his
instinctive acts, that is, does the human being, apart from
training, do any complex acts instinctively as do the lower
SCOPE OF BEHAVIOR PSYCHOLOGY 337
animals? If so, what is man's full equipment of instincts?
When does emotional activity manifest itself and what are
the situations which call it out, and what special acts can be
observed in emotional behavior? How soon can we observe
the beginnings of habit in infants? What special methods
can we develop for rapidly and securely implanting and
retaining the body and speech habits which society demands?
Do we find special and individual equipments in infants and
do these develop and later form the basis for their entering
one kind of vocation or another, or developing into one or
another type of personality? Are there such factors as
habit and instinct conflicts, distortion of habits and emotions?
How do they manifest themselves, and is it possible to
develop methods for shaping the environment of the indi-
vidual so that such conflicts will not arise?
Stimulus and Response. — This general description of the
subject matter of psychology helps us very little as regards
the analysis of particular problems in conduct and behavior.
In order to plan an experimental attack upon any problem
in psychology we must first reduce it to its simplest terms.
If we look over the above list of subject matter and at our
practical examples we see that there are common factors
running through all forms of human acts. In each adjust-
ment there is always both a reaction or response and a stimulus
or situation which calls out that response. Without going
too far beyond our facts it seems possible to say that the
stimulus is always .provided by the environment, external
to the body, or by the movements of man's own muscles
and the secretions of his glands: finally, that the responses
always follow relatively immediately upon the presentation
or incidence of the stimulus. These are really assumptions,
but they seem to be basal ones for psychology. Before we
finally accept or reject them we shall have to examine both
the nature of the stimulus or situation, and of response.
If we provisionally accept them we may say that the goal
of psychological study is the ascertaining of such data and
laws that, given the stimulus, psychology can predict what the /
response will be; or, on the other hand, given the response, it
can predict the nature of the effective stimulus.
33 8 JOHN B. WATSON
Use of the Term Stimulus. — We use the term stimulus in
psychology as it is used in physiology. Only in psychology
we have to extend somewhat the usage of the term. In the
psychological laboratory when we are dealing with relatively
simple factors such as the effect of ether waves of different
lengths, the effect of sound-waves, etc., and are attempting
to isolate their effects upon the adjustments of men, we
speak of stimuli. On the other hand, when the factors
leading to reaction are more complex, as, for example, in the
social world, we speak of situations. A situation is, of
course, upon final analysis, resolvable into a complex group
of stimuli. As examples of stimuli we may name such things
as rays of light of different wave-lengths; sound-waves differ-
ing in amplitude, length, phase and combination; gaseous
particles given off in such small diameters that they affect
the membrane of the nose; solutions which contain particles
of matter of such size that the taste buds are thrown into
action; solid objects which affect the skin and mucous mem-
brane; radiant stimuli which call out temperature response;
noxious stimuli such as cutting, pricking, and those injuring
tissue generally. Finally, movements of the muscles and
activity in the glands themselves serve as stimuli by acting
upon the afferent nerve endings in the moving muscles
(P- 341)-
It must be emphasized here that only under the rarest
experimental conditions can we stimulate the organism with
a single stimulus. Life presents stimuli in confusing combina-
tions. As you write you are stimulated by a complex system
— perspiration pours from your brow, the pen has a tendency
to slip from your grasp. The rays of light reflected from
the paper focus the physical image of the words upon your
retinae. The chair offers stimulation — the noises from the
street, etc. But far more important, delicate instruments
would show that though you are not speaking aloud your
vocal mechanisms — tongue, laryngeal muscles, etc., are in
constant motion: moving in habitual trains, these laryngeal
movements serve largely as the stimuli for releasing the
writing movements of the hands. The fact that you are
SCOPE OF BEHAVIOR PSYCHOLOGY 339
here in the lecture room facing your instructor and surrounded
by your classmates is still another very important element.
The world of stimulation is thus seen to be exceedingly
complex. It is convenient to speak of a total mass of stimu-
lating factors, which lead man to react as a whole, as a situ-
ation. Situations can be of the simplest kind or of the
greatest complexity. It should be noted here finally that
there are many forms of physical energy which do not directly
affect our sense organs. As examples we may cite the facts
that ether waves longer than 760 juju or shorter than 440 juju
do not lead to visual reactions, and that many of the wave
motions in the air are of such length or amplitude that they
do not produce auditory stimulation. The inability of the
human organism to respond to many possible forms of stimu-
lation will be discussed later.
The General Nature of Response. — In a similar way we
employ in psychology the physiological term response, but
again we must slightly extend its use. The movements
which result from a tap on the patellar tendon, or from
stroking the soles of the feet are * simple' responses which
are studied both in physiology and in medicine. In psy-
chology our study too is sometimes concerned with simple
responses of these types, but more often with several com-
plex responses taking place simultaneously. In the latter
case we sometimes use the popular term 'act' or adjustment,
meaning by that that the whole group of responses is inte-
grated in such a way (instinct or habit) that the individual
does something which we have a name for, that is, * takes
food,' * builds a house,' 'swims,' 'writes a letter,' 'talks,'
etc.1 In working over the distinctions among the various
types of acts the speculative psychologists have introduced
many needless technicalities and metaphysical concepts,
such as purpose, end, etc. Psychology is not concerned
1 But it should be well understood that whatever the man does under stimulation
is a response or adjustment — blushing, increased heart beat, change in respiration,
etc., are definite adjustments. We have names for only a few thousands of the total
possible number of such adjustments. The term adjustment is used by most writers
to refer to the doing of one of these named acts. In this volume the terms adjustment,
response, reaction, etc., are used almost interchangeably.
34° JOHN B. WATSON
with these distinctions. Because a man fails by his separate
acts to get his food, to build his house, to work out his mathe-
matical problem, or to live in harmony with his wife, is no
reason for rejecting him as a psychological subject. We study
him for his reaction possibilities and without prejudice-:' the
discovery of the fact that he will make only abortive attempts
to meet and control certain aspects of his environment is an
important part of our task; just as important as being able
to state that he can make certain other types of adjustment.
'Successful' adjustments, 'good' acts, 'bad' acts, are terms
really which society uses. Every social age sets up certain
standards of action, but these standards change from cultural
epoch to cultural epoch. Hence they are not psychological
standards. Reaction possibilities, however, on the average
probably remain about the same from eon to eon. It lies
well within the bounds of probability that if we were able
to obtain a newborn baby belonging to the dynasty of the
Pharaohs and were to bring him up along with other lads in
Boston, he would develop into the same kind of college youth
that we find among the other Harvard students. His
chances for success in life would probably not be at all differ-
ent from those of his classmates. The results obtained from
the scientific analysis of reaction in the human being should
fit any cultural age. It is part of the function of the psycholo-
gist to tell whether a given individual has the reaction possi-
bilities within him to meet the standards of that cultural
age, and the most rapid way of bringing him to act in ac-
cordance with them. The fact that social values (group
mores) change puts ever new burdens upon the psychologist
because every change in the mores means a different situa-
tion to which man has to respond by a different combination
of acts, and any new set of acts must be incorporated into
and integrated with the rest of the action systems of the
individual. The problems put up to psychology are those
of deciding whether the individual can meet the new standards
and for determining and developing methods of instructing
him.
Motor and Glandular Indicators of Response. — What is it
SCOPE OF BEHAVIOR PSYCHOLOGY 34*
that the psychologist can observe? Behavior of course. But
behavior on analysis is the separate systems of reactions
that the individual makes to his environment. When we
come to study the mechanics of such adjustments we find
that they depend upon the integrations existing among the^
receptors and the muscles and glands.
The unicellular organisms have no separate muscular or
nervous systems. Yet a part of their one cell must be
specialized in a motor as well as in a sensory way, since these
organisms do move in response to stimuli — to light, gravity,
heat, cold, electricity, etc. As you pass higher in the scale
special sense organ tissues (receptors) develop and along „
with them both motor or effective organs, and neurones
connecting receptors and effectors. Action in such cases
becomes sharper, more localized, more immediate, and at the
same time more sustained. Furthermore, as we pass still
further up the scale, glands begin to develop. Glands like
muscles are responsive organs and special glandular action
takes place whenever motor action takes place. The activity
of the glands in turn reacts back upon the muscular system
and affects its functioning (p. 338). Furthermore, there are
two kinds of muscles, striped and unstriped. The striped '
muscles move the arms, legs, trunk, tongue, larynx, etc.
The unstriped muscles control largely the blood vessels,
intestines, lungs, etc. Usually when we speak of response
we mean that the organism goes forward to right or left,
or retracts as a whole, that it eats, drinks, fights, builds
houses, or engages in trade. But these patent and easily
observable changes do not exhaust the term response, as
we pointed out on p. 339. We should mean by response J
the total striped and unstriped muscular and glandular
changes which follow upon a given stimulation. Our problem
of the moment determines which movement shall be studied
in relative isolation; in man, though, interest has been
largely centered in the integration of separate responses;
in getting him to form some habit — that is, to do something
with arms or legs or vocal cords. It is important to get at
the outset a comprehensive notion of response. A child or
342 JOHN S. WATSON
animal may stand stock still under stimulation, but we
should not say that there was no response. Close observa-
tion shows that there are changes in the tension of the muscles,
in respiration, in circulation, and in secretion.
General Classification of Responses. — The various possi-
bilities of reaction are thus seen to be vast; so vast indeed
that it would seem at first sight as though any classification
would be impossible. We can at least find a convenient
grouping which will serve us well both for discussion and for
setting experimental problems. Most reactions may be
looked upon as falling into one of four main classes :
1. Explicit habit responses: as examples we cite unlocking
a door, tennis playing, violin playing, building houses,
talking easily to people, staying on good terms with the
members of your own and the opposite sex.
2. Implicit habit responses: ' thinking,' by which we mean
subvocal talking, general body language habits, bodily sets
or attitudes which are not easily observable without instru-
mentation or experimental aid; the system of conditioned
reflexes in the various glands and unstriped muscular mechan-
isms, as, for example, conditioned salivary reflexes.
3. Explicit instinctive responses: including man's observ-
able instinctive and emotional reactions as seen, for example,
in grasping, sneezing, crawling, walking, etc., and in fear,
rage, love.
4. Implicit instinctive responses: this includes of course
the whole system of endocrine secretions, changes in circu-
lation, etc., so largely studied by physiology. Here again
instrumentation or experimental aid is necessary before
observation can be made.
These various types of response will be studied in detail
in later chapters. The classification as a whole should be
clear with the possible exception of 2, implicit habit responses.
This group is so important and so generally neglected in
discussion that we shall single it out here for brief mention
in advance of the chapter in which it is entered into with
some care.
What Man is Doing when Not Overtly Acting. — With a
SCOPE OF BEHAVIOR PSYCHOLOGY 343
highly specialized organism like man even careful observa-
tion often fails to show any overt response. A man may
sit motionless at his desk with pen in hand and paper before
him. In popular parlance we may say he is idle or 'thinking,'
but our assumption is that his muscles are really as active
and possibly more active than if he were playing tennis.
But what muscles? Those muscles which have been trained
to act when he is in such a situation, his laryngeal, tongue,
and speech muscles generally.1 Those muscles are as active
and are carrying out as orderly a system of movements as
if he were executing a sonata on the piano — they are doing
it well or ill depending upon the training he has had along
the particular lines which engage him. While we cannot
at present watch the play of this implicit stream of words
there is no reason for hypothecating a mystery about them.
Could we bring ' thinking' out for observation as readily
as we can tennis playing or rowing, the need of 'explaining'
it would disappear. We shall see later that efforts have
been made to bring such responses under experimental control.
But entirely apart from our present unreadiness to make
observation on implicit habits, we find a certain way of ar-
riving indirectly at the same end : implicit language habits, v
by methods which we shall study, come to issue finally in
overt action. By watching the easily observable explicit
habits and instincts of an individual keenly enough, and for a
sufficient stretch of time, and under varying enough con-
ditions, we can obtain the necessary data for most psycho-
logical requirements.
Scientific Methods Contrasted with Practical Procedure. —
Having now examined at some length into the general nature
of both stimulus and response, we should be prepared to
understand the object of a psychological experiment and to
contrast the scientific procedure with the common sense or
practical procedure which we discussed at the beginning of
the chapter. We shall take up almost at random some
definite illustrative psychological problems and the methods
of solving them. Our first problem is to find out what the
1 Indeed the whole glandular and muscular systems are contributory.
344 JOHN B. WATSON
reactions of a six-months-old infant are to living furry animals.
We first arrange the situation (complex group of stimuli).
The infant is held by its mother in a well-lighted room. We
observe first that the infant is smiling and comfortably dis-
posed. Then one after another we present a white rat, a
dog, a cat, a white rabbit, beetles, and a snake. We next
record accurately and separately the responses to these ob-
jects. The infant, which has only learned to reach out for
objects a short time before, slowly puts out first one hand
and then the other. The smile leaves his face but no crying
or withdrawing of the hands or external secretions follow.
These are only the more easily observed responses. Other
changes take place undoubtedly, in the internal glands,
circulation, respiration, etc. It depends upon our immediate
problem as to where the emphasis in observation shall fall in
our record of reaction changes. In this case our problem
was to determine whether there were any overt instinctive
tendencies on the baby's part to react against or withdraw
the hands or whole body from live animals. Our problem
might very well have led us into observing the changes in
the eyes, respiration, blood pressure, salivation, or in the
endocrine glands, or in several of these at once. Again it
should be noted that our problem is not so simple as it seems
at first sight. Suppose we had found that the baby did with-
draw from the objects, began to cry, void urine, or attempt to
hide behind the mother's clothing — could we have concluded
that there was an instinctive reaction against live furry
animals? Not without delving into the baby's past. If we
had had the child under constant observation and found no
record of previous acquaintanceship with live animals, our
answer would be that the observable responses were probably
instinctive. But if on the other hand we found that the child
had been severely bitten by a cat only two days before our
test, our conclusions would have to wait upon more extended
observation. Nor can we, from the behavior of this one child,
draw any conclusions as to what other children of the same
age will do, or what this child might do at a slightly different
age or when tested under different conditions; before general-
SCOPE OF BEHAVIOR PSYCHOLOGY 345
izations can be made many children should be brought under
systematic observation.
As another example of a somewhat more restricted type,
let us take the case of a man whose everyday behavior has
led us to suspect the normality of his responses to mono-
chromatic (colored) light. Common sense has nothing to
say; it can give no adequate report upon him. His mistakes
may be due to one or many things. We take him into the
laboratory where monochromatic light is under control and
we put him in situations where he has to react to the lights in
pairs, and where each one of the lights can be widely varied
in energy. In the course of the investigation we find that
when there is a certain energy relation obtaining between
the red and the green lights he can no longer react to them
differentially (that is to say, they do not offer different
stimulating values). We note further that we can find a
white light of a certain intensity to which he reacts as he
does to either of the monochromatic lights. But at no
energy relation between any other two colors can we break
down his differential responses. We conclude after this
careful study that the man is red-green blind, that is, that
he reacts to red and green as he does to certain intensities
of white light.1 Let us take another example, and this time
from the field of vocational psychology. Suppose that the
telephone directory of a large city is getting entirely too
bulky and complex for men to handle easily. What is the
best method for obviating this? The telephone people and
the psychologists work together. The psychologist may
suggest printing in smaller type and four columns to the
page instead of three. These and many possible suggestions
may lead to a solution of the problem. But the matter
has to be put under severe trial both before individuals
trained to look up names in a directory and before individuals
having no more training than has the general public. Syste-
1 If we find by repeated tests that the anomaly is more than temporary, we are
right in advising this man that he will be handicapped if he enters certain occupations,
e. g., locomotive and marine engineering, geology, advertising, etc. In other words,
the results of psychological experimentation are as immediately practicable as are
results in any other scientific field.
346 JOHN B. WATSON
matic trial and error is the procedure here with statistical
treatment of the results. In the end it is found that a four-
column page with a certain amount of spacing between the
lines of print makes the directory not only 20 per cent,
less bulky but also one in which the subscribers can find
names 10 per cent, more rapidly.
THE DIVISIONS OF PSYCHOLOGY AND THE RELATION OF
PSYCHOLOGY TO THE OTHER SCIENCES
The Various Fields of Psychology. — It is just as difficult to
draw a hard and fast line between the different branches
of psychology as between the different branches of biology
and physics. Practical and theoretical interests determine
where a man will throw the emphasis of his observation.
All scientific psychology is experimental, or is at least carried
out under such conditions that rigid and controlled observa-
tion is possible. All psychology is 'genetic' in the sense
that we have to go back to the child and contrast it with
animals in order to determine what native systems of inte-
grations belong peculiarly to man. For purposes of special-
ization we speak of human psychology as being made up of
individual, vocational, child, folk, educational, legal, pathological,
and social psychology. For our purposes we need not enter
into a separate characterization of these special branches.
The remaining chapters in this book attempt to deal generally
with the simpler results, problems, and methods in common
use in psychology. We shall not emphasize, except here and
there, the particular branch to which such material belongs.
Relation of Psychology to Physics. — Both physiology and
psychology are dependent (as is every other science at bottom)
upon physics for the control of apparatus and of stimulus.
It is essential for a research student in psychology now to
know the general facts about wave motion; as, for example,
heat, sound, and light. It is important to know how to
install and use simple electrical instruments, galvanometers,
thermal couples, and photometers.
Relation to Neurology. — It might be supposed that psy-
chology would lean most heavily upon neurology. Indeed
SCOPE OF BEHAVIOR PSYCHOLOGY 347
this has been the general assumption in the past. Psycho-
logical texts have been overburdened with cuts and descrip-
tions of the nervous system and we have many works which
claim on their title pages to be physiological psychology.
Gradually we are gaining the point of view that the psycho-
logical laboratories cannot teach both psychology and
neurology. Where a neurological laboratory is at hand
training in neurology should certainly be included, but it is
doubtful if much can be gained by a psychological student
from merely looking over cuts and listening to lectures on
the subject. Some notion of the elements involved in reflex
arcs is certainly essential — the way sense organs are con-
nected with the central nervous system and the central
nervous system with the muscular and glandular systems.
In a later Chapter we touch upon some of the more elementary
features connected with the arrangement and functioning of
reflex pathways.
Relation of Psychology to Physiology. — It has been claimed
by some that psychology is really physiology. That this
is not the case appears from even a casual examination of the
respective scopes of the two provinces. Physiology teaches
us concerning the functions of the special organs. For
purposes of experimentation and exposition the heart, liver,
lungs, circulation, respiration, etc. are isolated, or are at
least discussed as isolated functions. All of the functions
of the bodily organs are gone over in this way. Muscle-nerve
preparations are taken out and their properties investigated.
It is not meant to assume that physiologists deal wholly
with organs in isolation. Certain combined processes are
studied, such as metabolism, digestion, effects of poisons,
etc., but nowhere in physiology do we get the organism, as it
were, put back together again and tested in relation to its
environment as a whole.
From our discussion of the scope of psychology we are
now prepared to see what when the physiologist has learned
all that he can about the functioning of the separate organs
of the body of man, he has encroached upon our field only
in a very slight degree. Our task begins only when the
348 JOHN B. WATSON
physiologist puts the separate organs together again and
turns the whole (man) over to us. The physiologist qua
physiologist knows nothing of the total situations in the
daily life of an individual that shape his action and conduct.
He may teach us all there is to know about the mechanism
of stepping, but it is not his task to determine whether man
walks before he crawls, the age at which walking begins,
whether walking begins earlier in boys than in girls, or
whether defective children walk at a later age than normal
children. Again, he may teach us a great deal about the
functions of the kidneys, the bladder, and of the sphincter
control of the latter; but of the special situations (outside of
disease entities) which may lead to incontinence in children,
his science teaches him nothing, nor of methods of controlling
this mal-adjustment. In studying psychological functions,
for example, the emotions, it does not help him very much
to try to picture what chemical and neural processes go on
in the brain. It has often been asserted, e. g., that the
thalamus is operative in emotional disturbances. We do
not get very far, though, by trying to picture such activi-
ties, or by speaking of what goes on in the individual
neurones. We get a very incomplete but a somewhat better
view if we consider what goes on in glandular action
during emotional states. But even glandular action is
not easily observed by methods which are known today.
We can, however, study the reaction states we popularly
call sadness, elation, moroseness, rage, fear, love, etc., from
the standpoint of what the organism can do in these states
and as to whether the smooth running of the general system
of organized habits is facilitated or disturbed by the presence
of emotional activity. We can, further, often determine
by a study of the life history of the individual how fre-
quently such disturbances come about and can trace out
the causes or factors leading to their onset. Physiology has
nothing to tell us of the character and personality of different
individuals nor of their emotional stability or lack of emo-
tional control, nor as to what extent their present place in
life is dependent upon their upbringing. Physiology tells
SCOPE OF BEHAVIOR PSYCHOLOGY 349
us nothing of man's capacity to form and retain habits nor
of the complexity of man's habit organization. Hence if
we wish to predict whether an individual is capable of rising
above the environment to which he is not adjusted, we
should have to go to psychology and not to physiology for
our answer. In thus emphasizing the entire theoretical
independence of the two fields let us not set up a false im-
pression of antagonism. Physiology is psychology's closest
friend among the biological sciences. We can hardly move a
step in psychology without using physiological data. But
in this we are not different from the other biological sciences,
or indeed from medicine itself.
Overlapping of the Two Fields. — Occasionally we find
physiologists who have dealt with functions which overlap
the field of human behavior. As examples, we cite the
work of Cannon on the bodily effect of violent emotional
disturbances, and of Carlson and others on the question of
the reactions which are present in the stomach in the absence
of food. Where the two fields overlap most, however, is
probably in the study of the nervous, muscular, and glandular
systems, and in the realm of sensory physiology. This latter
topic no longer seems seriously to interest the physiologists,
and where they have shown interest in it, in this country
at least, it has been mainly pedagogical. Most of the work
in sensory physiology has been done by psychologists. Until
the recent work of Pavlow and Bechterew and their students
physiologists have shown little interest in the study of habit
formation, which general topic is one of our central ones.
In general it may be said that there is some overlapping in
the two fields, but that this does not keep them from being
separate disciplines. In cases where there is an overlapping
the methods and points of view of the two sciences in no
wise differ.
Relation of Psychology to Medicine. — Up to the present
time psychology has been of only slight service to psychiatry
and medicine generally. It should form a background for
the whole field of medicine. But it has dealt hitherto so
largely with speculations and with philosophical considera-^
350 JOHN B. WATSON
tions that its usefulness for this purpose has been seriously
restricted.
The physician, whether medical specialist or general
practitioner, would like to know something about the method
of approaching and handling his patients. He must en-
counter— and he must be prepared to encounter — such things
as stubbornness and unyieldingness in his human subjects,
and he must learn to study his patients in relation to their
present environment, and to go back into their life history
for an understanding and explanation of such attitudes. He
must learn how to size up his patients and to get at the
details of their individuality and characteristics. He must
be able to tell whether the patient can do what he is told
to do, and whether he has sufficient assets to meet the en-
vironment in which he has to live, and whether he has suf-
ficient assets to rise out of the environment which is un-
satisfactory to him. These facts on character adaptation
cannot be expressed in any other terms than behavior terms.
These are, to be sure, factors which concern everyone who
has to deal with his fellow-man, but on account of the intimate
relationship existing between the patient and his physician
they are of especial importance to the latter. The psychi-
atrist has not neglected these factors; indeed, it has been due
to him that they have been emphasized at all, and it is largely
through his efforts that we have a well-developed and syste-
matic technique for isolating the factors of importance in
the life history of the patient. In so far as psychiatry is
concerned I think we can say that the psychology the psychi-
atrist uses is not different from the psychology we are trying
to study. The psychiatrist has to be both a physician with a
specially developed therapeutic technique, and a psychologist
with special interests in certain divisions of psychology.
Psychiatry has no special need for detailed studies on reac-
tions to sensory stimuli. Much of the detailed work on
habit formation and on the separate analysis of instincts
is not of special use to him. On the other hand, any of the
material which the psychologist may offer on the subjects
of attachment and detachment of the emotions, on the
SCOPE OF BEHAVIOR PSYCHOLOGY 35 1
genesis of instincts and habits and their interrelations, on the
effect of age, drugs, etc., on habit formation and retention,
upon false reactions and failures in reactions, on the effect
of lesions of the central nervous system in trained animals
and the resultant success that comes from retraining them,
can be utilized by the psychiatrist at once, both in a specific
way and by reason of its value in helping him to size up his
patients. Most psychiatrists will admit that when the
proper kind of psychology is developed they can utilize
directly a large part of both of our methods and of our
materials. This appears clearly when we examine the various
tests which have been devised by psychologists for evaluating
the general behavior levels of individuals. Such tests in one
or another form are in common use in every psychiatric clinic.
Topics such as * general behavior,' * stream of talk,' 'attitude,'
'orientation,' 'retention' of recent and past happenings,
'general information,' the emotional level at which acts can
be carried out, etc., are discussed in relation to every patient
admitted to a psychiatric clinic.
Preparation for Psychology. — In dealing with the native
equipment of man the student of human psychology will
find a background of study of animal behavior a helpful one.
As a further preparation for this part of his work he will
find that he needs some equipment in physiology and experi-
mental zoology. His work in habit formation leads him
again into physiology and pharmacology for such factors as
the effect of age, drugs, etc., upon the human organism. The
consideration of habit and instinct conflicts, abortive reac-
tions and failures of adjustments generally which we see so
well emphasized in tics, sympathetic chorea, hysteria, ob-
sessions, etc., leads the psychologist into the psychiatric
clinic if he wishes to prepare himself to the fullest extent.
Business and law are making ever and ever larger demands
upon him. Some familiarity with legal and business problems
is almost essential. Finally, in order to handle adequately
experimental data some training in the use of statistical
methods is needed. If a start is made early enough by the
student who is preparing for psychology he can obtain the
352 JOHN B. WATSON
above related branches before he begins his special study of
psychology. While today is a day of specialists it should
not be a day of narrow specialists. The tendency to have
information about one small corner of psychology should
not be encouraged. It leads to such anomalies as pure
"mental testers," psychotechnicians, and the like.
RELATION BETWEEN FUNCTIONAL AND
BEHAVIOR PSYCHOLOGY
BY A. P. WEISS
Ohio State University
The concept of evolution in biology supplemented the
purely descriptive and systematic accounts of plants and
animals by the introduction of a genetic and developmental
factor. So fruitful did this new outlook prove to be and
with such audacity and effectiveness did it probe into the
cherished beliefs of orthodoxy that even man's mind became
the object of prying research. Psychology was no longer
content to study the structure of mental states, but its
interests expanded so as to include the development and
genesis of mind. As experimental methods became more
prevalent, the practical needs of pedagogy stimulated interest
in the relationship between mind and action, and this in
turn resulted in a more critical analysis of behavior or con-
duct. Finally the axiomatic character of the proposition
'Mind controls action' was challenged and this introduces
one of the controversial points of modern psychology.
PROBLEM OF FUNCTIONAL PSYCHOLOGY
The causes which led to the point of view called func-
tional psychology were the need for a more dynamic principle
in the explanation of human behavior than was offered by
the descriptive and systematic accounts of the nature of
mind as given by the structural psychologists. That the
study of the mind was of great value in understanding human
behavior had never been questioned, but the lack of uni-
formity in terminology and the controversies on methodology
made it practically impossible to formulate the problem of
the relationship between mind and action in such a way
that it could be critically discussed by any considerable
number of psychologists. As the investigations of the sense
353
354 A- P- WEISS
organs and of the nervous system increased the body of
fact available for the development of hypotheses, the original
definition of psychology as the science of consciousness was
extended, in spirit at least, to include the neural correlates
of conscious processes.
William James1 was one of the earlier psychologists in the
country who clearly foresaw that the investigations into the
relationship between the psychological faculties and human
conduct had been neglected. The shift of the emphasis from
the systematic treatment of consciousness to the investiga-
tions of the conditions under which consciousness manifested
itself, may be regarded as the starting point of the functional
point of view.
Briefly, the problem of functional psychology may be
expressed in the question: How does consciousness function
in human behavior or conduct?
PROBLEM OF BEHAVIOR PSYCHOLOGY
The origin of behavior psychology was largely due to the
fact that functionalism failed to indicate the manner in which
a conscious process could be regarded as controlling behavior.
Neither the parallelism nor the interaction of conscious
processes and physiological processes did more than indicate
a possible relation. These theories did not show how behavior
was actually modified. The question, whether an idea or
conscious state could bring about action appropriate to the
idea was denied by many. Some psychologists even con-
cluded that no form of consciousness whatever could alter
the direction of a neural flux, and consequently mere con-
sciousness was not a factor in behavior or conduct. This
inevitably led to the question as to how human behavior did
actually come about and those psychologists who questioned
the causal effectiveness of consciousness, formulated their
problem in the following manner: How may the behavior
of man (or animals) be described as solely due to receptor-
effector processes in the neuro-muscular system.
1 'The Principles of Psychology,' 1890, Vol. 2, p. I.
"They (sensations and perceptions) are therefore names of different cognitive
functions not for different sorts of mental fact."
FUNCTIONAL AND BEHAVIOR PSYCHOLOGY 355
CONSCIOUS PROCESSES AS METAPHORS
The lack of precision and accuracy in the terminology of
the functionalists was one of the contributing causes in
differentiating behaviorism from' functionalism. The func-
tion of consciousness in behavior was accepted as a self-
evident fact that needed no proof. "That they (mental
phenomena) lead to acts is of course the most familiar of
truths, etc."1 This was, of course, merely a modified form
of faculty psychology in which the term mind or conscious-
ness was substituted for the term faculty. The following
quotation is representative of the attitude of many function-
alists so far as the practical applications of psychology are
concerned.
"The formation of the elements of the process of knowledge
and the inauguration of the control of our movements in
accordance with the mandates of experience — these are the
two great functions of perception."2
If this quotation is taken literally, then "perceptions"
must be regarded as entities which inaugurate and control
our movements. A faculty of * perception' is, however,
just as unscientific as were the entities of reasoning, poetry,
foresight, etc., of the faculty psychologists.
We are warned, however, that the quotation is not to be
taken literally but that the faculty implication is merely
assumed to avoid an involved and cumbersome terminology.
"Let it be understood once and for all that wherever we
speak, as occasionally we do, as though the mind might in a
wholly unique manner step in and bring about changes in
the action of the nervous system, we are employing a con-
venient abbreviation of expression which harmonizes with
ordinary everyday methods of thinking and speaking about
these relations. The real fact appears to be, that whenever
we have mental activity we also have neural activity in the
cerebral cortex. The basal distinction in the two kinds of
nervous action to which we are referring in this chapter
(mind, neural action and habit) is, therefore, not primarily
1 James, Wm., 'Principles of Psychology,' 1890, Vol. I, p. 5.
2 Angell, J. R., 'Psychology,' New York, 1908, p. 171.
356 A. P. WEISS
between a form in which the mind suddenly produces changes
in the nerves as against one in which it does not, but rather a
distinction between certain kinds of neural activity overtly
involving consciousness, e. g., cortical activity of the cerebrum,
and certain other kinds not overtly involving it, e. g., spinal
reflexes. To use on every occasion the long modifying phrases
necessary to precise accuracy on this matter would evidently
be unduly cumbrous, and so the commoner modes of ex-
pression are employed, but the fundamental facts which lie
behind these convenient metaphors must not be forgotten."1
One might very well condone a lapse from the "long modi-
fying phrases necessary to precise accuracy" into the "oc-
casional use of convenient metaphors," but when the precise
accuracy is restricted to a few paragraphs and the occasional
metaphors make up the body of the book, one is led to wonder
whether the principle of faculty psychology may be con-
sidered repudiated.
Passing aside the question as to whether a textbook
written in metaphors can be said to present its subject matter
scientifically, the more important question as to what are the
principles which underlie the functional point of view as
implied in the preceding quotation, may be formulated as two
propositions:
1. There are certain neural processes which overtly in-
volve consciousness; e. g., cortical activity in the cerebrum.
2. There are certain neural processes which do not overtly
involve consciousness; e. g., spinal cord reflexes.
It is difficult to see how these two propositions may be
used to convert the metaphors of the following quotation into
the precise accuracy that is necessary for scientific under-
standing.
"Perception enables its possessor to register in conscious-
ness the particular object momentarily presented to the
senses. But if consciousness never advanced beyond the
merely perceptual stage it is apparent that we could never
develop any highly systematized and intelligent movements
of response to environmental demands and opportunities.
1 Angell, J. R., 'Psychology,' New York, 1908, p. 59.
FUNCTIONAL AND BEHAVIOR PSYCHOLOGY 357
Intelligent deliberation would be impossible. We should
always live in the immediate present and our minds could
consciously look neither backward nor forward. Now it is
in the image with its ability to carry such prospective and
retrospective meanings that we find the psychical mechanism
for accomplishing both these highly important functions."1
The writer does not wish to imply that the functionalists
are the only group of psychologists who write essays rather
than scientific expositions. On the contrary the preceding
quotation on the function of perception is considerably clearer
than many of the explanations of the functions of mental
activity which are to be found in the literature of educational
and applied psychology. Nevertheless, just as long as we
excuse ourselves on the plea that others do it, and persist in
substituting rhetoric for science, we should not complain if
our work is regarded with suspicion by the biologists.
CONSCIOUSNESS AND BEHAVIOR
Returning again to the fundamental principle underlying
functional psychology, namely, 'that it is mental activity
rather than mental structure that is of immediate significance
for thought and conduct' we will make an attempt to illustrate
just what sort of a relationship must exist between conscious-
ness and behavior if the claims of the functionalists are to be
substantiated by science.
The form of representation in the following figure is
that usually used to illustrate the theory of psychophysical
parallelism in which action is the result of the stimulation of
a receptor (S) of some sort. The nervous excitation which
results is transmitted over neurons indicated by solid lines
in the direction of the arrows, to effectors (M) which may be
either muscles or glands. When the nervous excitation
reaches these effectors contraction occurs in the case of a
muscle, and secretion in the case of a gland. Consciousness
in the form of sensations, images, affections, volitions, emo-
tions, etc., are indicated by the dotted lines. This conscious-
ness is thought of as accompanying the neural processes.
10p. cit.t p. 215.
358 A. P. WEISS
The illustration thus represents the neural processes and the
conscious processes which are said to be correlated with them.
The diagram below represents the neural and conscious
conditions in the following situation: The subject is passing
along a familiar road and suddenly comes upon a loose wire
dangling from a telegraph pole. Let the solid lines between
Sa-Ma represent the neurons which connect the eyes with
the muscles that bring about the activity of walking. The
conscious processes which may occur with the walking are
represented by the dotted line parallel with Sa-Ma. If the
Wire Stopping
Sight of Road WalKing
FIG. i
activity of walking is completely habituated there may be
no conscious processes which can be readily analyzed. Let
the solid lines Sb-Mb represent the neural processes from the
eyes to the muscles which result in stopping, supposing this
is what happens as soon as the wire is seen. These muscles
may of course be partly the same as those used in walking,
but since walking and stopping represent different forms of
behavior, they have been differentiated in the diagram.
The dotted line parallel to Sb-Mb again represents the
conscious processes which are correlated with the neural
processes. For any given individual the elementary con-
stituents (sensations, images and affections) of the conscious-
ness will vary but by referring to it as the 'idea of danger'
(from an electric shock) its characteristics are sufficiently
FUNCTIONAL AND BEHAVIOR PSYCHOLOGY 359
described for our purpose. Let this 'idea' be represented
by (d) on the diagram.
If consciousness controls behavior the functionalist must
show how the idea (d) acts on the neural processes Sa-Ma
and Sb-Mb so that most of this neural flux will go to Mb
instead of Ma as was the case before the 'idea' appeared in
consciousness. To do this (d) must be regarded as acting
on the neural processes at some point, say at (c). The
problem is simply this: What is the character of the psychical
mechanism which will enable us to understand how a psychi-
cal process (the idea) can influence or change a neural process.
The functionalist cannot consistently accept the principle
of psychophysical parallelism in which the conscious process
merely accompanies a neural process without acting on it,
since this would repudiate the possibility of consciousness in-
fluencing action. Some of the functionalists accept the
principle of psychophysical interaction and believe that the
4 idea' does in some way * switch' the neural processes, though
they frankly confess that they do not know how it is done.
It might be well to call attention to the fact that the
diagram is general in character and that (d) may represent
any conscious process whatever. In the illustration (d)
is regarded as a perceptual process, but some functionalists
maintain that the affective processes (satisfaction and annoy-
ance) are the only ones which modify behavior; others would
ascribe this modifying capacity to the emotions (fear, rage,
sentiments, etc.); while still others substitute volitional
processes (will, desire, wishes). The problem, however, to
show just in what way these conscious processes act upon
the neural processes is the same for all and for us it is only
necessary to call attention to the solidarity with which they
one and all agree that they do not know how this interaction
takes place.
The most comprehensive attempt to show that conscious
processes should be regarded as having the power to modify
or control neural activity has been made by Wm. McDougall,
who finally reached the conclusion that "Of the limits of the
power of mental control over organic processes of the body
360 A. P. WEISS
we are altogether ignorant, and new evidence, much of it
ill reported and therefore valueless, but much of it above
suspicion, repeatedly warns us against setting up any arbi-
trary limit as to what may be effected in this way."1
Most psychologists would be content to determine the
limits of the power of mental control over organic processes
after it had been demonstrated that such control was an
actual fact. It seems strange that the functionalists have
never recognized clearly that they cannot expect any con-
siderable degree of scientific recognition until they have
developed a working hypothesis which will enable them to
analyze any experience into those mental and physical com-
ponents which are the essential condition of the experience.
This has been a rather difficult task for those who have
tried it and a few functionalists have gone so far as to re-
pudiate psychophysical interaction. For the latter only two
alternatives remain: First, the investigation of conscious
processes as such, without reference to behavior; or the point
of view of structuralism. Second, an investigation of be-
havior independently of consciousness; or the viewpoint of
behaviorism. Functionalism can only claim an independent
point of view when it has shown how a mental process may
act upon a neural process.
THE NEURAL CORRELATES OF CONSCIOUSNESS AND
BEHAVIOR
The relationship between consciousness and action toward
which functionalism seems to be moving regards the neural
correlate of consciousness, rather than consciousness as such,
as the factor that modifies action. In other words, when the
functionalist states that the function of perception is to con-
trol action, this means that it is the neural correlate of per-
ception that controls action.
Referring to the diagram of Fig. I, this would mean that
the individual does not stop because the 'idea of danger'
(d) acts on the neural processes so that most of the flux
reaches Mb (stopping) but that the 'idea of danger' had as
- 1McDougall, Wm., 'Body and Mind,' New York, 1911, p. 375.
FUNCTIONAL AND BEHAVIOR PSYCHOLOGY 3^1
its neural correlate the processes Sb-Mb and it is this neural
process (not the idea) which brought about the stopping.
This is more clearly illustrated in that phase of functionalism
in which the affective processes1 (satisfaction and annoyance)
rather than the perceptual functions are regarded as sig-
nificant for behavior. The neural correlate for satisfaction
is held to be the * readiness of a neuron to conduct' and the
degree of conductivity is measured by the " relief of inter-
ference with the life processes of the neurons concerned."2
This theory seems to imply rather clearly that it is the
neural correlate of satisfaction (readiness of the neuron to
conduct a nervous process) rather than the subjective satis-
faction that modifies the behavior.
It is rather difficult to represent this view on a diagram
such as Fig. I. The conscious process of satisfaction may be
represented by (d) and the * readiness to conduct' would then
refer to the reduced resistance over the path Sb-Mb. How-
ever, that there is a " relief of interference with the life proc-
esses of the neurons" can only be maintained when it has
been demonstrated that the reaction Mb does actually occur.
Since we have no direct way of predicting the metabolic
conditions of the neurons, to infer from the reactions that
there has been a " relief of interference with the life processes
of the neurons " does not help us to understand how the
increased * readiness to conduct' has been brought about.
This, of course, is the problem.
Whether we regard the neural correlate of either cognition
or affection as significant for behavior, the fact remains that
the conscious processes themselves cannot be regarded as
significant, because they cannot occur until the neural con-
ditions have been prepared. That is, at best the conscious
1Thorndike, E. L., 'Animal Intelligence,' New York, 1911, pp. 244.
This quotation has been selected because it represents a point of view which
ascribes causal effectiveness to the conscious processes of pleasantness and unpleasant-
ness rather than to perceptual processes. The actual working hypothesis of Thorn-
dike is however much better expressed by the following quotation, which is however
behavioristic rather than functional in principle:
"Every response or change in response of an animal is then the result of the
interaction of its original knowable nature and the environment," op. cit., p. 242.
2 Thorndike, E. L., 'Original Nature of Man,' 1913, p. 225.
362 A. P. WEISS
processes merely indicate that the neural conditions for action
have already been established. The mental processes them-
selves are no* assurance that the appropriate action will
actually take place nor do they indicate how the neural
conditions for the action which actually does take place
have been prepared.
In other words, the conscious processes follow the neural
processes, they do not lead them.
To learn how the neural correlates of conscious processes
come to have the configuration which they do, can only be
done by considering the properties of the neuro-muscular
system.
Some of the functionalists recognize this clearly enough
as a general principle, but in the actual development of their
subject matter they drop into the terminology of psycho-
physical interaction so naturally and with such abandon
that one cannot help but feel that the introductory emphasis
on neural function is a protective measure against criticism
rather than a working hypothesis.
THE NEURAL CORRELATE OF CONSCIOUSNESS AND THE
NEURAL CORRELATE OF BEHAVIOR
The element of greatest confusion in the relationship be-
tween consciousness and action is probably the fact that
consciousness has always been regarded as an existential
datum entirely distinct from action or behavior. Whatever
may be the metaphysical principles involved in the mind-
body problem, from the standpoint of science, consciousness
must be regarded as a reaction. When an observer reports
that he is conscious of * danger' this only means that the
muscles of his speech mechanism have contracted in such a
way that sounds which we call words and sentences are
produced. If the observer does not react in some way, we
cannot infer that he is conscious. That is, unless the mental
process is expressed by speech (or some equivalent action)
this mental process can never become available for science.
Consciousness, as a scientific concept, may be regarded
as merely a supplementary reaction of a specific type to a
FUNCTIONAL AND BEHAVIOR PSYCHOLOGY 363
given situation. If for instance I am asked "What is seven
times sixteen?" I may, after some hesitation, simply pro-
nounce the words "One hundred twelve." If then I am
asked to give my introspections, I may add: "I had a visual
image of the figures 7 X 10 followed by the visual image
of 70 written on a blackboard; this was followed by the
auditory imagery — seven times six ... forty two . . . ;
then I had auditory images with kinesthesis of the speech
mechanism — seventy . . . forty . . . hundred ten . . . two
. . . hundred twelve . . . visual image of 112 written on the
blackboard."
We must note that there are two sets of stimuli: (i) What
is seven times sixteen, (2) Give your introspections. We
should therefore expect two different reactions. Why should
we say that the second reaction (introspection) is of an
altogether different type than the first? Every one will
immediately admit that it is absurd to say that the response
to the second stimulus controlled the reaction to the first.
The facts in the case are more scientifically stated when we
merely regard the introspective reaction as only one of the
responses which might be expected in a situation of this
kind. It no more controls or determines the many other
Sight of Wire Speech "Danger" Stopping
^ *
Speech
Sight of Road \|v)x Description WdlKing
FIG. 2
responses that might have been made, than they control it.
The above diagram (Fig. 2) shows the road-wire situation
illustrated in Fig. I, but drawn as a number of reactions
without any hypothetical conscious processes.
364 A. P. WEISS
When the stimulus is Sb the observer not only stops walk-
ing (Mb) but he may also say: "That wire looks dangerous,
I have had one electric shock; last summer Mr. X was killed
in this way;" as indicated by the speech reaction Md. It is
not necessary to conclude that being able also to react by
speech (introspection) has anything to do with stopping
(Mb). The only essential condition for stopping is the
stimulus Sb. The speech reaction Md has nothing to do
with it.
Before showing the relationship between the neural corre-
late of consciousness and the neural correlate of behavior we
will consider the terms ' major' and 'minor' reaction. We
will assume that every stimulus or situation to which an
organism adjusts itself, may result in more than one reaction.
Major reaction: In the first place we have the reaction
which is regarded as the appropriate reaction or adjustment
to the particular situation. In our illustration this would
be 'stopping' (Mb) when the wire is seen.
Minor reaction: Secondly, there are also other reactions
such as vasomotor; respiratory; changes in the accommoda-
tion of the sense organs; speech reactions, both actual and
incipient; etc. In our illustration these minor reactions may
be a verbal exclamation or introspection (Md); decrease in
respiration rate; increase in pulse rate; or any of the many
bodily and visceral changes which might be described.
Simply stated the major reaction is the one which is sig-
nificant for behavior, while the minor reactions are those
which are usually disregarded. In this sense the introspec-
tive report would be called a minor reaction. In the diagram
Fig. 2, Ma and Mb are major reactions; while MX, Me, Md,
are minor reactions. The difference between Fig. I and
Fig. 2 lies in the fact that the dotted lines in Fig. I which
are supposed to represent hypothetical conscious processes
have been replaced by neural processes. In Fig. 2 the
conscious process of the 'idea of danger' (d, Fig. i) is repre-
sented as the speech reaction Sb-Md. That it must be
some kind of a reaction if it is to be regarded from the sci-
entific standpoint seems clear when we reflect that conscious-
FUNCTIONAL AND BEHAVIOR PSYCHOLOGY 365
ness which is not expressed in some form of reaction, can
never become available for science. It is absurd to say
that we may have conscious states to which we do not react.
The statement itself is a reaction. If we admit this, then
every conscious state or process is a reaction. The state-
ment that I have the 'idea of danger' or am * conscious of
danger' only means that in addition to reacting to the sight
of the wire by stopping I also react to it by saying, " I must
be careful; that wire may be charged; etc."
That this speech may not actually take place is due to
the fact that under ordinary conditions we do not react by
speech when we are alone. If we have some one with us, a
child perhaps, the reaction would take place as a warning
or protective movement of some sort. The minor speech
reaction Sb-Md (which is the only way in which the term
consciousness can have a scientific meaning) need have no
control or influence over the major reaction Sb-Mb. In fact
either Md or Mb might very well occur independently.
They are related only through the receptor Sb.
From the preceding we can conclude that every stimulus
which effects the organism results in more than one reaction.
Only one group, however, is usually regarded as socially
significant (behavior) and we have called it the major reac-
tion. Along with this major reaction there are many minor
reactions and those of the speech type are very numerous.
There is no need, however, to believe that these minor reac-
tions are necessary for the adequate functioning of the major
reaction.
CONSCIOUSNESS AND INTROSPECTION
From what has preceded, introspection is merely a name
for a group of speech reactions which conform to a particular
terminology. While we are introspecting we not only react
to a given situation in a manner appropriate to the situation
(pressing a key for instance) but we also react by speech in
the terminology of psychology. The functionalists have not
recognized that the introspective reaction is a minor reaction.
The social significance of the major reaction has obscured
the fact that introspection itself is only one of the reactions
366 A. P. WEISS
to the particular situation. They have assumed that in
some way it reveals what is taking place in the neural corre-
late of the major reaction. The development of this intro-
spective reaction is a process of habit formation just as any
other variation of response. It is not the expression of a
metaphysical entity (consciousness) that has been added to
the major reaction.
The only difference between psychological observation and
observation in the natural sciences lies in the fact that in
psychology the introspective reaction is regarded as the major
reaction, while in the natural sciences it is regarded as a minor
reaction or ignored entirely. However, neither the major
nor the minor reactions can be said to control or modify each
other and in this sense there is no mind-body problem.
FUNCTIONALISM OBSCURES THE NATURE OF THE PSYCHO-
LOGICAL PROBLEM
It is not supposed that the nature of the preceding dis-
cussion will be considered especially new by the function-
alists. Their excellent experimental work in pedagogy and
the emphasis they place upon habit formation irrespective
of any subjective implication, indicates that practically at
least they are following behavioristic methods.
The importance for general psychological theory as to
whether human behavior is regarded as the result of the
interaction between a hypothetical consciousness and neural
processes or as the result of neural mechanism only, lies
primarily in the fact that the experimental program will
reflect which of these points of view is adopted.
If, as the functionalists assume, consciousness can modify
behavior, then to bring about socially acceptable behavior
in the child, for instance, it is only necessary to bring about
those forms of consciousness which are effective. The prob-
lem then becomes one of teaching * ideals.' From the peda-
gogical standpoint an ideal must be regarded as a plan of
action or conduct. If the expressions of consciousness are
used as a test as to whether a given set of ideals has been
established there will be a tendency to place undue emphasis
FUNCTIONAL AND BEHAVIOR PSYCHOLOGY
upon verbal expression, since where a serious attempt is
made to determine the character of consciousness intro-
spection must play an important part. The problem of
teaching ideals then becomes one primarily of establishing
formal speech reactions to given situations. While for many
situations in life a speech reaction is the adequate reaction
there are, however, many other situations in which speech
alone is inadequate, as for instance in many ethical relations.
In such instances the verbal expressions of the ideals are no
assurance that the socially valuable reaction has been learned.
On the other hand the socially desirable reaction cannot be
learned by merely establishing the speech reaction which
describes the appropriate behavior. JThe behaviorist main-
tains that it is better to disregard the concept of conscious-
ness altogether and the pedagogical problem then becomes
one of determining exactly how socially acceptable behavior
is developed directly from the properties of the neuro-
muscular system. If a verbal reaction is part of the
appropriate adjustment, well and good; it must be learned.
However, after it has been learned there is no advantage
in assuming that this verbal reaction is the function of a
hypothetical psychical process.
BEHAVIORIST PROGRAM
Perhaps the distinguishing difference between the func-
tionalist and the behaviorist lies in the fact that the behavi-
orist disregards the entity which the functionalist calls con-
sciousness. This does not mean that the behaviorist ignores
those problems in behavior with which the concept of con-
sciousness is usually associated. On the contrary by regard-
ing man as an organism he believes that even the most
complex problems can be described and explained without
assuming the existence of any causes which are not already
accepted by science in general. The behaviorist's principal
interest will be the movements of man whether these move-
ments are of the skeletal muscles which carry his body from
place to place or the movements which result in the com-
position of a symphony. He regards the neuro-muscular
368 A. P. WEISS
system as the means by which the organism adjusts itself
to its environment, just as the heart, the lungs, digestive
tract, are means to keep the organism alive. To withdraw
the hand from the scorching candle is a movement of only
less complexity than the movements of the pen that signs a
treaty between nations. The real object of worth for the
behaviorist is not the thrilly, fascinating, esoteric, pseudo-
problems of the mystic, but the permanent, measured and
describable adjustments of the race.
SUMMARY
1. The axiomatic character of the statement 'Mind Con-
trols Action' is questioned by the behaviorists.
2. The metaphorical explanations of the functionalists
obscure the 'faculty' character of their 'mental activity.'
3. The functionalists have never shown how mental
activity may control action.
4. The conscious processes of the functionalists actually
follow the conditions which bring about a modification in
behavior and hence cannot be said to control behavior.
5. Consciousness and introspection to have any scientific
value must be regarded as speech reactions which are re-
stricted to the terms used in psychology. The presence or
absence of these classes of verbal reactions in no way influence
the socially significant reactions.
6. The difference between the natural sciences and psy-
chology lies in the fact that psychology restricts its observa-
tions to those special speech reactions (introspections) that
conform to the terminology of psychology. In the natural
sciences this class of verbal reactions is ignored.
7. The concept of a hypothetical conscious process ob-
scures the real problems of psychology.
8. Behaviorism regards the introspective reaction as only
one of the ways in which an individual may react to a situ-
ation.
THE RELATION BETWEEN EMOTION AND ITS
EXPRESSION
BY HARVEY CARR
University of Chicago
This paper proposes a conception of the nature of an
emotion in relation to its expression which constitutes some-
what of a compromise between the theory of James and the
older view which it displaced.
The popular view assumed that some inner or central
emotional experience follows the act of perception, and that
this emotion is succeeded by a complex series of organic
disturbances. The inner activity is the emotion and the
resulting organic change is the expression of that emotion.
James denied the existence of any centrally conditioned
process intervening between the perception and the organic
activities which may properly be termed an emotion. James
asserted that the terms emotion and expression must refer to
distinctions within the series of organic activities aroused by
the perception, and that these activities are sensory and peri-
pheral in character. As is well known, James dichotomized
these psychophysical activities into their conscious and their
material aspects. The experiential or non-material aspect is
termed the emotion, while the material or physiological aspect
is the expression of that emotion. As a consequence of this
usage of terms, and of the acceptance of the prevalent doc-
trine as to the relation of consciousness to afferent and effer-
ent nervous impulses, James was forced to the paradoxical
conclusion that the emotion is not the cause but the result of
its expression.
Our view agrees with that' of James in maintaining that
the terms emotion and expression must refer to distinctions
within the total series of organic activities, and that these
processes are peripheral and sensory in character. We
369
370 HARVEY CARR
shall, however, adopt a radically different mode of division
from that employed by James.
These organic activities may first be divided into three sets
of psychophysical processes — the act, the emotion, and in-
cidental by-products of the emotion. In anger, the term act
is applied to those activities immediately concerned in combat
— the fighting activities. Flight or running away is the act
in fear. The term emotion refers to all those prior and ac-
companying organic processes whose function it is to render
the act more efficient. The emotion and the act are to some
extent independent variables; theoretically one can fight
without being mad, and one can become angry without
fighting. Likewise one can run away without being afraid,
and fear without indulging in flight. The nature of the
emotion and its functional relation to the act have been well
depicted by Cannon. In general the emotion consists of
those processes by means of which the total energy of the
organism is mobilized and concentrated for the service of the
act. The function of anger is to increase the efficiency of
the fighting activities. A cause and effect relation obtains
between the emotion and the act. The increased efficiency
and sometimes the initiation of the act are thus a result of the
emotion. The total series of processes involved in the organic
disturbance may also contain other components which are
to be regarded as the incidental but necessary by-products
of the emotion or the act, but which contribute in no way to
the efficiency of either. The trembling of anger or certain
digestive and nutritive disturbances incident to the vasomotor
shift may be adduced as hypothetical examples.
These three groups of activities — the act, the emotion,
and their by-products — constitute the whole of the organic
processes. There is no fourth class which can be termed the
act of expression. The term expression implies a dichotomy
of the same organic activities from another standpoint, viz.,
their relation to some observer. An emotion can express
itself only by producing some effect upon an observer. With-
out an observer the term expression is without meaning.
The emotion can express or manifest itself to an observer
RELATION BETWEEN EMOTION AND ITS EXPRESSION 37*
in three ways: (a) It expresses itself indirectly through its
observable effects upon the act. Such characteristics as the
determination, vigor, and persistence of the fighting act are
observed and become the sign and symbol of the emotion
of anger which produced them. The increased efficiency of
the act thus constitutes both a result and an expression of
the emotion, (b) Any of the observable by-products of the
emotional situation also constitute a mode of expression.
To an observer they may symbolize the existence of the inner
emotional disturbance of which they are a result, (c) The
emotion also manifests itself to an observer in a more direct
fashion. Certain essential components of the emotional
processes, such as the flushed face, the frown, and the deeper
breathing in anger, are directly observed and constitute
another mode of expression. The emotional process as a
whole is not observed; only certain surface aspects of the
total process are perceived. These perceived aspects are
interpreted by an observer in terms of his experience and
knowledge and hence become the visible symbols or mani-
festations of the emotion as a whole. A part thus becomes
the symbol of the whole, and a symbol is a mode of expression.
Good usage, I think, will justify this meaning of the term.
We may thus legitimately assert that the emotional activity
manifests or expresses its nature to an observer by means of
these surface or observable features.
Our conception may now be compared with that of James.
Both are actuated by the same purpose. Both attempt an
expository definition of the popular meaning of two terms.
Both attempt a definition of emotion and expression in de-
scriptive and empirical terms. Both agree that emotion and
expression must refer to certain aspects of the organic activi-
ties involved in the emotional situation. The two views
ascribe radically different contents to these terms. James's
analysis was dominated by the subjective conception of the
province of psychology which prevailed at that time. An
emotion as a psychological phenomenon must be defined in
purely conscious terms; the physiological aspect of the process
must be discarded. This subjective emotional experience
37 2 HARVEY CARR
can naturally express itself to an observer only through be-
havior or physical means; the behavior, material, or physio-
logical aspect of the process must then constitute the only
avenue of expression. With this conception, emotion and
expression must refer to the psychic and the physical aspects
respectively of the organic reaction to the emotional stimulus,
and in virtue of these definitions the emotion must be a result
and not a causal antecedent of its expression. Our hypothesis
rejects the purely subjective point of view in psychology, and
consequently discards this psychophysical dichotomy of
James. Both emotion and expression are regarded as psycho-
physical processes, or rather they are regarded as real func-
tional activities of a human organism irrespective of the fact
whether they do or do not contain a conscious component.
According to our analysis, these terms have been so defined in
relation to each other that one can say that the expression is
a result of the emotion.
James's theory of emotion contains two more or less
distinct doctrines whose validity must be separately esti-
mated. One of these, to my mind, is correct, and the other
fallacious. The first doctrine asserts that the term emotion
refers to certain aspects of the organic activities and that these
activities are essentially sensory in character. We have sub-
scribed to this feature of the theory, and we believe that it
constituted a genuine and important contribution to the psy-
chological thought of the time. This aspect of the theory,
it is well to note, is open to empirical verification; James's
various factual proofs and the recent experimental attempts
at a disproof are relevant to this aspect of the Jamesian
doctrine. Needless to say, we believe that the factual
evidence at the present confirms James's contention.
The second aspect of James's doctrine consists of the
following features: (i) the assumption that emotion and ex-
pression refer to the psychic and the physiological components
respectively, (2) the acceptance of the prevalent assumption
as to the relation of the psychic to its physiological correlates,
and (3) the final conclusion that the emotion is the result of
its expression.
RELATION BETWEEN EMOTION AND ITS EXPRESSION 373
It must be at once admitted that this conclusion is logically
valid and unassailable from the standpoint of the prior assump-
tions. Neither is the proposition susceptible to experimental
proof or disproof. James's formidable list of factual proofs
is not relevant to this phase of the argument. His conclusion
is logically implicit in his assumptions; it represents merely
the result of a deductive analysis of what was contained in
his premises. James assumed the truth of this conclusion
when he made his preliminary definitions. A rejection of
one of these assumptions constitutes the only avenue of
escape for those who dislike the final conclusion.
This aspect of James's theory contains, to my mind, two
essential defects, (i) The conclusion contradicts common
sense, and this contradiction is due to the fact that James
ascribes to the term * expression' a meaning which is directly
antagonistic to the significance usually attached to it. As
previously noted, the term expression popularly signifies
some effect of the emotion upon an observer, and James has
arbitrarily so defined the term as to reverse this causal
relation. (2) James's analysis and conclusion are also lacking
in pragmatic value. His mode of treatment gives us no
analytical comprehension of the functional interrelations
between the various constituents of the organic activities,
nor of their nature and significance in relation to mental life
and conduct. The conclusion resulting from James's mode
of analysis is logically true and valid, but the knowledge it
represents lacks genuine significance and worth.
In spite of varied criticism, James's theory has enjoyed an
enviable reputation for many years. Several factors have
probably contributed to this result, (i) Unless the two
phases of the argument are differentiated, one is compelled
either to adopt or reject the theory in toto, and undoubtedly
to many minds the advantages of the theory outweigh its
deficiencies. (2) As we have noted, James was driven to his
psychophysical distinction by the adoption of the subjective
conception of the province of psychology. Likewise, James's
theory will of necessity make a strong appeal to those whose
thought is dominated by this attitude of mind, and the
374 HARVEY CARR
conventional definitions of the subject matter of psychology
have, until recent years at least, been couched in subjective
terms. If emotion is a psychological phenomenon, it must be
defined in conscious terms. Expression, on the other hand,
must be conceived in behavioristic or physiological terms,
because expression must refer to some effect upon an observer.
(3) The paradoxical character of the conclusion is a third
factor. The popular mind is somewhat prone to judge the
value and worth of a science upon the basis of the novelty
and startling character of its discoveries. The wonders of a
science are paraded in proof of its amazing progress in attain-
ing its ends. These discoveries are frequently wonderful and
startling simply because of their novelty and unexpectedness,
—because they contradict or modify prevailing conceptions
and opinions. Psychology as a young and growing science
must also produce its miracles in order to secure popular
acclaim to its worth and greatness, and what can be more
wonderful and miraculous than the discovery and labored
empirical demonstration of a truth which directly contradicts
the common sense opinion of mankind? As psychologists,
I fear, we have been somewhat susceptible to this influence,
and this suggestible attitude of mind has probably been streng-
thened to some extent by our experience in the classroom. I
know of no doctrine in psychology which is comparable with
that of James's theory of emotion from the standpoint of
inculcating in the mind of the average undergraduate a
wholesome awe and respect for the achievements of our
science. This chapter of James almost invariably makes a
profound impression upon the student mind, and the reason
is not far to seek, for it consists of a very clever and brilliant
exposition, and a persuasive logical and empirical demon-
stration of the truth of a proposition which many students
accept with some degree of mental reservation.
In place of James's psychophysical analysis, this paper
suggests a threefold division of the organic activities on the
basis of their causal interrelations. The act refers to those
processes of adaptation to the objective situation. The
emotion refers to those activities which increase the effective-
RELATION BETWEEN EMOTION AND ITS EXPRESSION 375
ness of the act. The remaining processes consist of incidental
by-products of the emotion or the act. There is no coordinate
fourth group of processes which can be termed the expressive
activities. The term expression implies a dichotomy of
these same activities from a different standpoint — their
relation to an observer. The larger and more important
portion of the emotional group of processes can not be directly
observed; their nature and existence must be inferred from
those aspects of the organic activities which are susceptible
to immediate observation. Certain aspects of all three of the
previously enumerated classes constitute a sign or symbol of
the existence of these hidden operations, and consequently
become the means by which these latter manifest or express
themselves to an observer.
THE THEORY OF THE SOCIAL FORCES
BY H. G. KENAGY
University of Minnesota
It has been said that "the corner stone of sociology must
be a sound doctrine of the social forces."1 With certain
limitations this statement may be allowed to stand. In the
first place, pure science, as such, does not admit of the use
of the term * force' in the sense of the sufficient cause of
any phenomena, physical or social.2 Taken to mean, how-
ever, only an active factor in a given situation, 'force' is in
common usage in the physical sciences and has been borrowed
by the social sciences for want of a better term. At best it
is only a helpful analogy. In the second place, a 'sound
doctrine' of the social forces must not be taken to mean
or imply an 'exhaustive, logical or psychological classifica-
tion' of the factors active in social life (as has generally
been done by the older sociologists), for the progress of
scientific knowledge of human society is not dependent upon
success in any such classification. Indeed, as Dr. Bernard
has pointed out, "the most accurate possible classifications
[of social forces] mark only the most elementary stage in the
analysis of social phenomena."3 Professor Ross's statement,
therefore, interpreted or limited in these respects, comes to
mean simply that an understanding of what are the active
factors in social life is necessarily basic in any scientific
study or discussion of social situations and institutions-
What are the active factors in social life? Some sociolo-
gists would limit them to forces which are social in their
origin; others would include all factors, whatever their origin,
which are socializing in their effects. Professor Baldwin
1 Ross, E. A., 'Foundations of Sociology,' p. 181.
2 See article by Prof. E. C. Hayes, 'The Social Forces Error,' in Amer. J. of Soc.,
16, 613-625.
3 Bernard, L. L., 'The Transition to an Objective Standard of Social Control,'
p. 74.
376
"HE THEORY OF THE SOCIAL FORCED
limits his definition to 'only those psychical products, called
desires, which influence individuals in their social relations';1
Professor Ellwood, much more sanely, uses the term to in-
clude * every factor which has some degree of active influence
in shaping and molding the forms of association and the inter-
action of individuals.'2 Accepting this latter definition, as
the more adequate of the two, we are forced to include among
the social forces such physical factors as climate, soil, and
other geographical conditions. To primitive men these
environmental conditions were much more important than
they are today, for civilized man has brought his environment
largely under his control.3 Heredity and variation must also
be included as active factors. True, they are not direct
factors, and, assuming a strict psychological interpretation
of society, they become factors in the organization of society
only by setting up conditions or limits within which, and
only within which, the more strictly social forces may act.4
At any one moment, perhaps, the forms of social life seem
to depend much more upon mental elements than upon
physical factors, but when one surveys human groups over
long periods of time the influence of physical factors is more
apparent. The environment has acted indirectly, selectively,
upon man's hereditary equipment, but also directly, through
such agencies as temperature for example, to modify instinc-
tive and habitual responses of masses of individuals and the
interaction between individuals.
Coming now specifically to a discussion of the psycho-
logical factors or forces in association, we may agree at once
with Dr. Ellwood that these consist of man's innate impulses
(instincts), his feeling states, and his cognitive or intellectual
processes.5 This assumes, apparently, the existence of a
1 Baldwin, J. M., 'Social and Ethical Interpretations,' p. 484.
2 Ellwood, C. A., 'Sociology in its Psychological Aspects,' p. 278.
3 Thomas, W. I., 'Social Origins,' pp. 130 ff.
4 See Baldwin, op, cit., Appendix H v, p. 573.
8 Op. cit., p. 282. The feelings and intellectual processes may enter, however,
only in their physical aspects as attitudes called out by stimuli and thereby influencing
other responses in the same individual and in others. They cannot function in this
connection purely as feelings and as awareness. We can know of psychic states in
others only by the behavior of those individuals. These facts will be borne out later.
378 H. G. KENAGY
Asocial mind' and therefore 'social feeling' and 'social
thought' which have the same functional relationship to
social activity which individual feeling and thought have
to individual activity — since social activity is due simply
to the interaction and coordination of individual activities.
Necessarily, therefore, we must decide first what is the func-
tional relation of feeling and thought to individual activity.
First of all, however, we must assign to instinct its proper
role in determining activity.1 Here there is more or less
agreement. The instincts are of first importance because
they are primary in man and serve in the initiation of action.
Human conduct can never exceed the limits of these native
proclivities which assign for man the ends of action and
alone make any action worth while.2 Man's native dis-
positions to activity, however, do not remain long unmodified,
but, through interaction with the environment and by
training, become overlaid with a mass of habits which come
to function, under proper stimulation, as readily and as
actively as did the original propensities to activity.
The moot point of the whole theory of social forces enters,
however, at this point of the discussion. The issue centers
around the part which feeling plays in determining the direc-
tion and extent of the modification of man's instinctive
impulses to activity. The older view, held by the hedonists
from the time ol Hobbes to the present, assigns to feeling the
function of a primary force, as lying behind these instinctive
activities. All action, individual and social, was explained
on a basis of pleasure and pain, as the springs of activity
were to be found in calculations of agreeable or disagreeable
sensations.3 The late Professor Ward, the father of Ameri-
1 By instinct is meant the inherited disposition to respond in certain ways when
appropriately stimulated. This definition is used so that instinct may be taken as
synonymous with the whole "original nature of man," and, for the purpose in hand,
is sufficiently exact.
2 See Veblen, T. B., 'The Instinct of Workmanship,' Introduction.
3 See Hobbes, 'Leviathan,' Works, III., p. 42; Locke, 'Essay Concerning Human
Understanding,' Book II., Ch. XXL, Sec. 33, 41; Bentham, 'Principles of Morals
and Legislation,' Chap. I., Sec. I.; Spencer, 'Principles of Psychology,' II., p. 541;
Baldwin, 'Handbook of Psychology,' pp. 301-303; Angell, 'Psychology,' p. 273;
Patten, 'Theory of the Social Forces,' Chap. I., Sec. I.; Ward, 'Psychic Factors in
Civilization,' pp. 52, S4> ^6, and 'Pure Sociology,' p. 132.
THE THEORY OF THE SOCIAL FORCES 379
can sociology, was particularly at fault in this regard. To
quote short passages from his works: "The dynamic agent
consists wholly of feeling;"1 "Feeling is a true cosmic force
. . . and constitutes the propelling agent in man and ani-
mals."2 "In the associated state of men, it (feeling) is the
true social force."3 "The thinking faculty is not a force;
but feeling is a true force and its various manifestations con-
stitute the social forces."4 True, Ward speaks of the desires
also as a true social force, but he uses ' desire' in the ' feeling'
sense,5 saying that desire is a form of pain.
There are two other types of writers, only partially or
not at all hedonistic, but individualistic for the most part,
who treat of social forces as causes of activity in one way
or another. One class regards feeling as one, but only one,
of the determining factors in activity. Those of the other
type hold that feeling can never be such a cause.6
What then is the actual part which feeling plays in
initiating activity? The answer must first of all settle the
question of what part consciousness plays, if any, in the
process. The better psychology at present holds that mental
processes are conditioned by changes within the organism,
notably within the nervous system.7 The more immediate
physical qualities of mind lie within the brain and are deter-
mined (i) by stimulus and (2) by disposition or tendency,
which latter indicates that the "neural functions are deter-
mined by the residues of earlier function (impressional,
associative, determining and habitual tendencies, and general
cortical set)."8 The facts of perception are mainly to be
1 ' Pure Sociology,' p. 256.
2 Ibid., p. 39.
3 Ibid., p. 99-
4 Ibid., p. 101.
5 'Psychic Factors in Civilization,' pp. 53-54.
6 Among writers of the first class are Titchener (see his 'Outlines of Psychology/
p. 250), Thorndike (see his 'Elements of Psychology,' p. 284). Among writers of the
second type are James (see his 'Psychology,' II., pp. 559, 580), Dewey (see Dewey
and Tufts, 'Ethics,' p. 270), McDougall (see his 'Introduction to Social Psychology,'
P- 43)-
7 Bentley, Madison, 'A Preface to Social Psychology,' in one of a series of articles
entitled 'Studies in Social Psychology,' in PSYCHOL. MONOG. No. 92, June, 1916,
pp. 10 ff.
s Ibid., p. 10.
380 H. G. KENAGY
explained by stimulus and associative tendency; passive
memory and imagination by associative and impressional
tendencies; emotion and action by stimulus and determining
tendency; skillful performance by habitual tendency; and
thought by dispositions of the determining sort. In view of
these facts we cannot admit that imagery is necessary, in
any causal way, even to voluntary activity. Neither can
we regard it as a superfluous or parallel process. Rather,
imagery is the evidence of associations and neural activity
between stimulus and response. It is these processes self-
aware in a fashion. The complete determinant of voluntary
activity is nothing more or less than the total set of the
nervous system of the moment, plus the stimuli. The total
cause of any act is certainly more than the conscious part
of it. A percept or image, coming about when the neural
pathway or the act is forced by interference to run probably
through the more complex channels in the cortex, is not the
cause of the act but only the sign of the whole act of which
it is a part.
The same is true of feeling — it is not the cause of the
act but only the sign of the whole act of which it is a part.
Says Dr. Bernard: "Feeling modes are resultants of internal
neural adjustments, . . . which correlation probably is made
in the cortex only when feeling is experienced. It is absurd
to speak of these feeling modes as the cause of such neural
correlations."1 Again: "Feeling ... is the result of the
correlation, that is, the supplementation or interference, of
nervous processes in such a way as to increase the neural
activity along a ... given pathway. Where a nervous
process is augmented, pleasantness is experienced, and where
a nervous process is weakened or diminished, there is un-
pleasantness."2
Professor Ellwood professes to be in sympathy with these
views as being simply more exact statements than the crude
evolutionary view of feeling which he advances, and recon-
1 Op. dt., p. 37.
2 Ibid., p. 1 8. This theory is only a more accurate statement of the theory of
feeling given by Max Meyer in his articles on 'The Nervous Correlate of Pleasantness
and Unpleasantness,' in PSYCHOL. REV., 1908, 15.
THE THEORY OF THE SOCIAL FORCES 3Sl
cilable with it.1 Feeling to him is however 'an organic
valuation of our activities.'2 He says: "Society is made up
of biological and psychological individuals, and these indi-
viduals are thinking, feeling men whose actions are mediated,
guided and controlled by feelings and ideas."3 Again: "Feel-
ings and ideas are not coextensive with activity," but 'rela-
tively new and independent elements' which 'appear within
physiological activities at certain points to evaluate them,
mediate and control them.'4 Finally: "... social phe-
nomena are in the nature of responses to stimuli, and these
responses are modified, in the mature individual at least,
by complex series of feelings and ideas."5
To such statements must be raised the fundamental
objection that they do not in any tangible way explain how
feeling operates to accomplish its evaluating function. Why,
for example, does the pleasurable act survive over the other
acts? Can a successful result act backwards and strengthen
the impulses leading up to it and stamp out the unsuccessful
impulses? Hardly. Rather, it happens that "by the actual
overlapping of many tendencies to respond in diverse ways
the erroneous tendencies are directed into the successful ones
and the latter are strengthened by reinforcement. Without
such overlapping of various impulses in the same general
response, the inhibiting effects of the successful upon the
unsuccessful or irrelevant tendencies are incomprehensible."6
The pleasure accompanying the successful act as a com-
plete response is not itself a cause or natural antecedent of
the surviving act but only the inner or 'felt' aspect of it and
therefore valueless in explanation. The selectiveness of the
organism, which Ellwood would call 'feeling control,' is
simply its more easy adaptation to certain direct and indirect
stimuli than to others, due to inherited and acquired nervous
correlations. The pleasurable tone accompanying certain
1 Op. cit., footnote on pp. 112-13.
2 Op. cit., Chapter X., p. 247.
3 Op. cit., p. 250.
4 Op. cit., p. 250.
5 Ibid., p. 251.
6 Peterson, J., 'Completeness of Response as an Explanation Principle in Learn-
ing,' PSYCHOL. REV., 1916, 23, 153-162.
382 H. G. KENAGY
activities is only a subjective indication that the response,
up to a certain limit, follows the line of least resistance.
Certain acts are 'chosen,' because they are on the whole the
most natural to the organism under the circumstances, not
because they are pleasant. That they are pleasant, in the
main, indicates subjectively that the response is relatively
'complete' and in harmony with one's inherited and acquired
organization.
As has been shown experimentally, some acts may be
both pleasureable and painful, painful alone, or without
conscious results. We may, again, experience pleasantness
from higher sensory or ideational processes at the same
time with pains from lower neural processes. Finally, any
act may be made pleasant or unpleasant through habit.
Consequently feeling modes cannot be effective guides to
individual or social adjustment and control. Ideas, images
or other subjectivistic criteria are not always valid, says
Bernard, and are only dependable when checked up by ob-
jective reference. "Feeling," he states, "as the conscious
part of mere correlation, i. e., supplementation and inter-
ference of neural processes, is the least able to be so checked
up and is consequently the least reliable of all subjective
criteria or evaluations of action in an objective and social
world."1 Feeling is a purely personal and individualistic
phenomenon.
A similar problem to that of determining the function of
feeling in connection with the forms of association and indi-
vidual interactions is presented in establishing the role which
the intellect or the cognitive elements of mind play in these
processes. Professor Ellwood holds that "the distinctive
character of our social life is due to the modifying influence
of intellectual elements,"2 and he maintains that the intellect
plays a decisive role not only in adapting the individual
organism, in man at least, to his environment, but also in
bringing about those 'higher adaptations which characterize
civilized societies.'3 Again, he says that the intellect, the
1 Op. cit., p. zS, Practically the same view is set forth by Dr. A. F. Bentley.
See his 'The Process of Government,' Chapters I. and II.
2 Op. cit., p. 261. 3 Op. cit., p. 261.
THE THEORY OF THE SOCIAL FORCES 383
cognitive, objective side of mind, 'evaluates activities with
reference to the environment and functions to mediate and
control them with reference to environmental factors.'1
And finally: "While the intellect seems to have been de-
veloped chiefly as an aid in carrying out the instincts and in
satisfying the demands of feeling, in its higher reaches it
can and does act more or less independently of them."2 By
this last statement he means that the intellect modifies
instincts, substituting habits for them which become as
strong as the original instincts.3
Such passages seem, almost, to set up the human mind
as a separate, supernatural entity which acts as a sort of
dictator to instinct and feeling and to motor responses.
This idea is brought out more forcibly when he says that
" more and more the process of living together needs the
interference of reason."4 What Professor Ellwood means,
however, seems to be only that ideas have come largely to be
the stimuli to activity among civilized peoples. Reflective
thought, which played an insignificant role in primitive
society, is now, he thinks, the decisive element because
"upon it depends the control, not only of the forces of physi-
cal nature but also, ... of the feelings and impulses of
human nature."5 For this reason, and in this sense, he
argues that ideas are entitled to be called forces, "since they
at any rate become active factors in the later stages of social
evolution and absolutely decisive, ... in the making of
the more complex adjustments."6 These ideas, in civilized
man at least, come in time to constitute for the individual
and society a sort of 'subjective environment ' and to this
environment "the mass of individuals respond quite as they
do to stimuli in the objective environment."7 Professor
1 Op. cit., p. 263.
2 Op. cit., p. 263.
3 Practically the same view of the function of the intellect was held by Ward.
See 'Pure Sociology,' Chapter XVI.
4 Op. cit., p. 264.
6 Op. cii., p. 264.
« Ibid., p. 277.
7 Ibid., p. 265. The same position is argued at length in his recent article: 'Ob-
jectivism in Sociology,' in the Amer. J. of Soc., December, 1916.
382 H. G. KENAGY
activities is only a subjective indication that the response,
up to a certain limit, follows the line of least resistance.
Certain acts are * chosen,' because they are on the whole the
most natural to the organism under the circumstances, not
because they are pleasant. That they are pleasant, in the
main, indicates subjectively that the response is relatively
'complete' and in harmony with one's inherited and acquired
organization.
As has been shown experimentally, some acts may be
both pleasureable and painful, painful alone, or without
conscious results. We may, again, experience pleasantness
from higher sensory or ideational processes at the same
time with pains from lower neural processes. Finally, any
act may be made pleasant or unpleasant through habit.
Consequently feeling modes cannot be effective guides to
individual or social adjustment and control. Ideas, images
or other subjectivistic criteria are not always valid, says
Bernard, and are only dependable when checked up by ob-
jective reference. "Feeling," he states, "as the conscious
part of mere correlation, i. e., supplementation and inter-
ference of neural processes, is the least able to be so checked
up and is consequently the least reliable of all subjective
criteria or evaluations of action in an objective and social
world."1 Feeling is a purely personal and individualistic
phenomenon.
A similar problem to that of determining the function of
feeling in connection with the forms of association and indi-
vidual interactions is presented in establishing the role which
the intellect or the cognitive elements of mind play in these
processes. Professor Ellwood holds that "the distinctive
character of our social life is due to the modifying influence
of intellectual elements,"2 and he maintains that the intellect
plays a decisive role not only in adapting the individual
organism, in man at least, to his environment, but also in
bringing about those 'higher adaptations which characterize
civilized societies.'3 Again, he says that the intellect, the
1 Op. cit., p. 28- Practically the same view is set forth by Dr. A. F. Bentley.
See his 'The Process of Government,' Chapters I. and II.
2 Op. cit., p. 261. 8 Op. cit., p. 261.
THE THEORY OF THE SOCIAL FORCES 383
cognitive, objective side of mind, 'evaluates activities with
reference to the environment and functions to mediate and
control them with reference to environmental factors.'1
And finally: " While the intellect seems to have been de-
veloped chiefly as an aid in carrying out the instincts and in
satisfying the demands of feeling, in its higher reaches it
can and does act more or less independently of them."2 By
this last statement he means that the intellect modifies
instincts, substituting habits for them which become as
strong as the original instincts.3
Such passages seem, almost, to set up the human mind
as a separate, supernatural entity which acts as a sort of
dictator to instinct and feeling and to motor responses.
This idea is brought out more forcibly when he says that
" more and more the process of living together needs the
interference of reason."4 What Professor Ellwood means,
however, seems to be only that ideas have come largely to be
the stimuli to activity among civilized peoples. Reflective
thought, which played an insignificant role in primitive
society, is now, he thinks, the decisive element because
"upon it depends the control, not only of the forces of physi-
cal nature but also, ... of the feelings and impulses of
human nature."5 For this reason, and in this sense, he
argues that ideas are entitled to be called forces, "since they
at any rate become active factors in the later stages of social
evolution and absolutely decisive, ... in the making of
the more complex adjustments."6 These ideas, in civilized
man at least, come in time to constitute for the individual
and society a sort of 'subjective environment' and to this
environment "the mass of individuals respond quite as they
do to stimuli in the objective environment."7 Professor
1 Op. cit., p. 263.
2 Op. cit., p. 263.
3 Practically the same view of the function of the intellect was held by Ward.
See 'Pure Sociology,' Chapter XVI.
4 Op. cit., p. 264.
6 Op. cit., p. 264.
« Ibid., p. 277.
7 Ibid., p. 265. The same position is argued at length in his recent article: 'Ob-
jectivism in Sociology,' in the Amer. ]. of Soc., December, 1916.
3 86 H. G. KENAGY
If the mind is conditioned, as was shown above (p. 5),
by stimulus and disposition or tendency, the use of such
expressions in connection with social activity as: "mind
'influenced' by another mind," "Man is * suggestible' or
* imitative,'" "one mind 'rules' or 'dominates' and another
'acquiesces,'" etc., is wholly outside the plane of scientific
explanation. A stimulus, as said before, is a physical agent
and sets up a series of concrete organic processes. 'Sug-
gestion,' 'imitation' cannot be accounted such agents when
used to explain the destructive activities of a mob, nor can
'domination' or 'sense of power' be given as the cause of
the acts of a railway trainmen's union. Used in such a
fashion such terms are wild abstractions used as forces,1
and are closely akin to the 'faculties' of two centuries back.
'Faculties,' of course, is unacceptable for the purposes of
scientific explanation, but we also lack a concrete meaning
for 'mental dependence' or 'mental interaction' in such a
statement as "individuals tend to believe and to think and
to feel, etc., in mutual dependence." "It is obvious," says
Bentley, "that the mind of my neighbor is not to be added,
as a condition of my mental processes, to the sober and
authenticated facts of stimulus and disposition. If my
neighbor speaks with the voice of authority and decision
and so convinces me that I should attend the meeting of
the Municipal League, my mental processes are set up,
after all, just as they would be if I found a blight upon my
fruit trees and decided to destroy the orchard. Auditory or
visual stimuli and associative tendencies account for the
perceptual part of either experience, and determining and
habitual tendencies for the performance.
"The only thing that is unique about the conditioning
factors in social or mental dependence is the fact that the
presence of other persons ... or the assumption of them
. . . touches off certain dispositions or neural tendencies,
giving to our 'social experiences' a certain kind of sig-
nificance. The sight of the blighted fruit trees and the sight
and sound of my persuasive neighbor are psychological
1 Illustrations of this are to be found in Ellwood, op. cit., pp. 283 ff., 288; Ward,
'Pure Sociology,' pp. 256 ff., 457 ff.; Ross, 'Social Psychology,' p. 13.
THE THEORY OF THE SOCIAL FORCES 387
events of the same order. There is not, in the one instance,
the mere apprehension of an object; in the other, the opera-
tion of a subtle and mysterious force through the agency of
which my mind is wrought upon by my neighbor's. Because
of my constitution and my history the two things are differ-
ently apprehended, have different significance, and lead to
unlike performances."1
Classifications of the social forces have been attempted
by sociologists from Hobbes, Fourier and Spencer to many
present-day writers, and the majority unite in placing
emphasis upon the psychic factors or forces, though the
later writers reject the old hedonistic criteria and adopt the
functional view. However, with only one important ex-
ception,2 all classifications have been subjective in that the
content is lodged in the individual consciousness as the source
of activities. What is needed is a classification of social
forces which does not stop with consciousness, real or imagin-
ary, and, in so doing, cover up the real and objective sources
of stimulation to activity.
The most complete, accurate, and objective classifications
of the social forces must constantly aim to point out, according
to Bernard, (i) how the individual acts or behaves, the
organs he uses and how he uses them when stimulated in
known or unknown ways, and (2) how a group acts or be-
haves, the types of control which are exercised over individual
activities or behaviors, in known or unknown ways. In the
individual these may be instinctive or acquired (habitual)
tendencies; in the group they may have grown up uncon-
sciously through custom, or they may have been consciously
legislated into existence, or taken on through the pressure
of public opinion, or as a result of scientific investigation.3
The * causes' of activity under (i) have been called 'social
1 Bentley, op. cit., pp. 11-12.
2 Recently a very valuable and suggestive inventory or classification of the re-
sponses of the original tendencies in man to various classes or types of stimuli has
been made by Professor E. L. Thorndike in 'The Original Nature of Man,' 1914. He
regards this, however, as only a beginning — as indicating the direction which further
inquiry must take and the subject-matter with which truly objective classifications
will have to deal.
3 See his discussion, op. cit., p. 73 ff.
388 H. G. KENAGY
forces' and the subjectivists have traced them back to the
individual consciousness and lodged them there, * because the
individual is usually conscious of his socially most con-
spicuous acts, and when he is not thus conscious, conscious-
ness is ... assumed.'1 Thus the early sociologists stopped
at the sacred threshold of consciousness and ended their
search for social forces in the forms of consciousness. Only
recently has psychology pointed out that consciousness is
not ultimate, but is caused, and is only one factor in adjust-
ment. The social behaviors under (2) have also, by analogy,
and with the same subjective emphasis, been called 'social
forces,' being abstracted for this purpose from the unified
social situation and made the product of individual activity.
The distinction was not made, between 'interests' and 'de-
sires,' nor was it recognized that the latter type of behaviors
is to be termed 'social forces' with more reason than the
former, since they necessarily go behind the individual
consciousness to some extent.
The passage from an introspective to an experimental and
biological psychology, with its analysis of the conditions of
consciousness and its functional activities in making adjust-
ments to the environment, has caused us — forced us — to
look back of the mere forms of consciousness in studying
social causation. Our search for 'social forces' must under-
take to account objectively for the activities of the individual
and of the social group. In actual practice we have adopted
this method. Thus, for example, we have 'ceased treating
disease on a demonistic basis, or attempting to cure national
ills by public prayer.' We still, however, practice retaliatory
methods in our criminological procedures, and limit morality
in general to the scope of consciousness or intention.2
The old classificationists present the subjective social
forces as only forms of consciousness by which the conscious
person is made aware of his own activities, while their formu-
lations of objective social forces are only abstractions for
presenting to ourselves the social processes. Says Bernard,
1 Ibid., p. 74.
2 For a critical discussion of this point see Bernard, op. cit., Introduction.
THE THEORY OF THE SOCIAL FORCES 389
in speaking of these abstractions: "They are not forces; at
the most they are partial indices of social * forces' or processes.
Nor have they constant equivalents; for conscious processes
and our statements of social processes have at different times
different activity equivalents. . . . They are qualitative
rather than quantitative indices. They merely invite to
always further analysis and re-analysis of the objective
social situation; and it is on the bases of these analyses that
all our problems are to be comprehended and effectively
solved. When a situation is once adequately analyzed,
when the forces lying back of the forms of consciousness or
the abstracted and generalized types of social and individual
activity are understood, the method of the solution of the
problem is simply that of the application of common sense.
The only mystery that there is about the treatment of social
problems is that which we make by being content to stop
with the forms of consciousness in our analysis. We talk
about the riddle of personality as an impregnable barrier to
an adequate understanding of social conditions, because we
are attempting to work out a logic of forces and activities
from the kaleidoscopic presentations of our conscious proc-
esses."1
The problem ahead of the sociologists, then, according to
Bernard, " is to push farther back the analysis of objective
phenomena." Until sociology abandons its subjective criteria
it cannot attain to true scientific efficiency. "As psychology
retreats from its introspective analysis of the solipsistic self,
and as ethics gives up mere intention as the criterion of
morality, so sociology must turn from a subjective classi-
fication of 'social forces' and study the functioning of objec-
tive social processes as they operate in individuals and
groups."2
SUMMARY
Sociology, ethics, and the other social sciences are in
need of more tangible explanations of individual and social
action than are commonly given, and they have a right to
look to psychology for a true statement of the facts so that
1 Op. cit., pp. 75-6.
2 Ibid., p. 76.
39° H. G, KENAGY
all can work on the same plane. The terminology now in
common use implies or postulates the possibility of stimula-
tion by such subjective factors (frequently referred to as
'forces') as 'feelings' and 'ideas,' though the way in which
they become stimuli has never been satisfactorily explained.
This conception seems to have failed in the explanation of
behavior as much as the 'faculties' now supposed to be
discarded from psychology.
All tangible stimuli must be of such a nature as to set
up nerve impulses. They may come from within or from
without the individual. He may be stimulated by external
things — by other persons, by spoken or written words of his
language, by books, etc., — or by organic processes and
muscular changes within his own body. Organized stimulus-
response systems as developed by habits are important in
determining the direction and the extent of his responses.
It is inconceivable, therefore, how conscious states can serve
as stimuli; for all conscious states are but imperfect, sub-
jective representations of stimuli in the objective world.
The customary usage thus involves one in a vicious circle.
In group relations, therefore, explanations must hark
back to stimulus, disposition, and response instead of halting
upon the plane of 'mental interaction,' 'suggestion,' 'social
consciousness' and the like. The uniform action of indi-
viduals in society is due to common sources of stimulation,
to common associative tendencies, and to common habits,
overlapping in the generations. Divergent reactions are
due to differences in organic set or total nervous organiza-
tion, which are the products of more or less unlike heredity
and experiences. Finally, the 'intellect' can not be set up
as a separate entity operating as a force or cause in inducing
action or in mediating activity. As an aspect of the sum
total of nervous correlations in the higher brain centers, the
intellect can not be considered as a causative agent, nor can
an idea, which is always the outcome, though remotely so,
of sensory stimulation, be used as the cause of individual or
group activity.
THE MENTAL WORK CURVE
BY DANIEL STARCH AND I. E. ASH
University of Wisconsin Ohio State University
Practically all studies in the field of mental work and
fatigue have employed, in one form or another, one of two
general methods of investigation. According to one method
the subject is required to perform certain tests at the begin-
ning and again at the close of a period of work, or at intervals
during its progress. The differences between the successive
applications of the tests are regarded as measures of the effect
of the intervening work. According to the second method
instead of making comparisons of the results of short tests at
the beginning and at the end of the period of work, a study is
made of the progress of the work itself, and the factors which
affect its progress.
Instead of regarding these two methods as different ways
of dealing with the same general problem, they should rather
be regarded as two distinct problems, since they differ not
only in procedure, but also in the result sought. The former
tries to secure results that shall be free from the influence of
such factors as practice, "warming up," and variation in
attention and interest and to obtain an index of fatigue only
as it is developed in the course of the work. The latter,
instead of trying to eliminate the influence of these factors,
tries to take particular account of them and to assign to each
its specific value in the work curve.
It is very essential in studies of the latter type that the
task done be of uniform difficulty throughout the entire
period of work. Otherwise we shall not know whether the
variations in the curve are due to conditions in the work or
in the worker. It is likewise very desirable to employ a form
of activity which involves a minimum of physical effort and
sensory strain, so that the mental factors may be isolated as
fully as possible. The kind of work which has proved most
391
392 DANIEL STARCH AND I. E. ASH
satisfactory, and which has been most extensively used, is
some form of mental computation.
The Experiments. — In a recent study of mental fatigue1
results were secured from forty-five hours of mental calcula-
tion done under uniform experimental conditions and care-
fully checked both as to the number of errors made and the
amount of work done in each successive half-minute. Twenty-
three persons took part in the experiment. The work was
done in forty-one periods varying in length from thirty
minutes to two hours and a half. The type of work done was
that devised by one of the writers and described elsewhere.2
This method was essentially as follows:
The subject was seated comfortably in a quiet room. The
experimenter began by giving him a number consisting of
two digits. The subject added mentally, six to this number,
then seven to the sum thus secured, then eight to this last
sum, and then nine to this result, and then again six, seven,
eight, and nine in rotation, adding in every case to the sum
resulting from the previous addition. The subject spoke his
answers aloud. At the end of thirty seconds the experimenter
announced a new number and the subject immediately
dropped the series on which he was working and began with
this new number by adding six, seven, eight, and nine in
rotation as before, for another thirty seconds when another
starting number was given, and so on without interruption for
the entire period of work. The advantages of this type of
addition are that it reduces all physical and sensory elements
to a minimum, that it taxes mental effort to its full extent, as
was witnessed by every person who took part in the experi-
ment, that it is almost perfectly continuous, giving no op-
portunity for relaxation, and that it is composed of small and
fairly uniform units of work in terms of which the results can
be measured accurately. Even adding columns on paper
involves more muscular and sensory activity than the present
type of addition involves. The subjects were usually allowed
1Ash, I. E., 'Fatigue and Its Effects upon Control/ Archives of Psychology.
Vol. V., No. 31. The results here discussed were not presented in that monograph.
2 Starch, D., 'Experiments in Educational Psychology,' pp. 172-181.
THE MENTAL WORK CURVE 393
to make two or three practice series of the additions before
the regular experiment was begun.
When the sum of successive additions had reached one
hundred or more the hundreds digit was dropped and the
additions continued with the units and tens digits. Thus, if
the number first given was 72, then the successive sums should
be 78, 85, 93, 102, 108, 115, etc., but instead of giving these
last sums as 102, 108, 115, they were given simply as 2, 8,
15, etc.
To enable the experimenter to keep the results accurately
mimeographed sheets of paper were prepared in advance
which contained all the numbers used as starters1 and, in
columns directly under them, the correct sums of twenty
additions. Whenever the subject gave an incorrect sum, the
experimenter wrote it down beside the correct sum. A line
was drawn under the sum resulting from the last addition
in each thirty-second period, so that an exact record was
kept of the amount and accuracy of the work done. During
the experiments, the subjects either closed their eyes or sat
in such a position as to be unable to see the experimenter in
order to avoid distractions and to concentrate to the fullest
extent upon the task of the experiment. Neither subject
nor experimenter spoke a word during the experiment except
the "starting numbers" which were given by the latter, and
the results of the successive additions by the former. No
time was allowed to elapse between series.
Results. — The data afforded by these experiments may be
treated from several different angles, and will throw light on
a number of important phases of the problem of mental
work and fatigue. The first questions to suggest themselves
very naturally related to the general form of the curve for
work periods of varying lengths. How soon after beginning
work does one reach his maximum efficiency, and how long
will he be able to maintain it? How closely will the curve
of work for one day represent that of another? How widely
1 The numbers used as starters comprised all the numbers between 10 and 100 except
those whose right-hand digit is o or which will produce sums whose right-hand digit is o.
All these were omitted since the additions in such cases are decidedly less difficult.
394 DANIEL STARCH AND I. E. ASH
do different persons vary with respect to the form of their
work curves? How greatly and in what general respects do
the work curves of those who work slowly differ from the
curves of others who work rapidly?
The results are shown graphically in Figs. I to 4. The
points in the curves represent the average number of additions
made in each five successive thirty-second series. The curves
for the errors are shown in the lower part of each figure.
Fig. I shows the composite curves for three two-hour records
obtained from two subjects. Two records were obtained
from one subject. Fig. 2 gives the results for nine one-and-a-
half hour records obtained from eight subjects. Fig. 3 re-
presents twenty-one one-hour records obtained from fourteen
subjects. Fig. 4 represents eight half-hour records obtained
from four subjects.
These figures show that the number of additions per unit
of time gradually increases during the first twenty-five to
thirty-five minutes, when the maximum speed is attained.
This maximum speed is maintained for another thirty or
thirty-five minutes, at which time the number of additions
begins gradually to decrease and continues to do so until
the work ceases. The curve of errors takes, on the whole, the
opposite direction to that of the work curve. That is, the
number of errors is the largest when the number of additions
is the smallest. There is greater "zigzagging" or irregularity
in the work curves when the rate of addition is highest. The
initial period of increase in the rate of additions is not found
in Fig. I. That is, the rate at the beginning of the experiment
is as high as at any time during its progress. This is probably
due to the fact that the subjects were quite familiar with the
work of the experiment since they had made shorter records
before.
While the curves in Figs. I to 4 show clearly that there
are fewer additions made in those periods in which the
number of errors is greatest, they do not show the full differ-
ence in the number of additions made in those series in which
errors occurred and those in which no errors occurred. In
order to determine just how great this difference was, the
THE MENTAL WORK CURVE
395
396 DANIEL STARCH AND I. E. ASH
number of additions in those series which contained errors,
and in those which contained no errors was counted.
Before doing this, however, the data of all the experiments
were divided into two groups. Group I. included those
experiments in which the subjects averaged more than twelve
additions in each half-minute period or series. Group II.
included all those records in which the average was less than
twelve. The experiments of Group I. covered 25 hours of
work and therefore included 3,000 half-minute series of addi-
tions. In 2,333 °f these series, no errors were made; while in
the other 667 series 920 errors were made. The average
number of additions per half-minute period in the series in
which no errors were made was 17.2. The average for the
667 series in which errors were made was n.8. In other
words, an average of 46.2 per cent, more additions were
made in those series in which no errors occurred than in those
in which errors did occur.
The experiments of Group II. covered 20 hours of work
and included 2,400 half-minute series of additions. In 1,441
of these series, no errors occurred. In the remaining 959
series, 1,784 errors occurred. In this group, the average
number of additions per thirty-second period for the series
which contained no errors was 9.71 and for those in which
errors did occur the average was 7.8. In this group there
were, on an average, only 23.8 per cent, more additions made
in those series in which no errors occurred than in those in
which errors occurred, as against 46.2 per cent, in Group I.
These results indicate that those who worked rapidly worked
more accurately than those who worked slowly. The more
significant fact, however, brought out in these results is that
it requires a longer time for any one, whether he work rapidly
or slowly, to make an incorrect addition than a correct one.
This fact, overlooked by many investigators, obscures the
signs of fatigue, which have been sought by the method of
continuous work. Fatigue is unquestionably developed in all
such experiments but it is not shown in the actual output of
work.
These figures, however, do not express the real, or full
THE MENTAL WORK CURVE 397
differences between the number of correct additions made in
any unit of time and the number of incorrect additions for
the same time; or the effects which an error, or those factors
which cause one to make an error, have upon one's speed in
mental work. In order to show the full significance of the
errors, or of the factors which caused them, in slowing up the
work, a computation was made of the number of additions in
those half-minute series which contained errors, before any
errors were made, and also of the number including and fol-
lowing the first error in each of the different series. In all
of the 667 series of Group I. which contained errors, the total
number of additions before the errors occurred was 5,714.
The total number of those including and following the errors
for all the series was 2,177. We may reasonably assume that
the rate for those additions which were made before any
errors occurred would be the same as for those series which
contained no errors. If this assumption be correct, then the
amount of time required to make these 5,714 additions which
were made before the errors took place would be about equal
to 331 thirty-second periods, leaving 336 periods to make the
2,177 additions which included and came subsequent to the
errors. This would make an average for these additions of 6,5
per thirty-second period, or slightly more than one third the
rate when no errors occurred. In the 959 series of Group II.
which contained errors, 3,846 additions were made before
any errors occurred, while the number including and following
the errors was 3,665. Treating these numbers as we did those
in the preceding group we find the average rate for the in-
correct additions and those subsequent to them in the series
in which they occur to be 6.5 per thirty-second period, or
about two thirds the rate for those series in which no errors
occurred.
That the assumption is substantially correct is indicated
by a special record made on one subject in which the exact
time was noted when the first error in each series occurred.
In this record the time per addition after the error occurred
was 3.78 seconds, or 31 per cent, longer.
Still another factor in the distribution of errors and the
DANIEL STARCH AND I. E. ASH
effect of this distribution on the rate of additions is the
relative number of series containing errors at different stages
in the experiment and the number of errors in each of these
different series. This distribution is shown in the following
tables. (Table I. giving the results of Group L; Table II.
those of Group II.) The top horizontal row gives the suc-
TABLE I
GROUP I
I.. .
i
2
3
4
5
6
7
8
9
10
II
12
13
14
IS
2. .
20
2O
20
20
20
20
7
7
7
2
2
2
i
I
i
3- •
4- •
400
i«;6
400
129
400
116
400
IOO
400
84
£
I40
,40
140
5
40
O
40
I
40
I
20
o
20
I
20
O
I::
159
1.0+
129
1.0
127
i.i —
129
1.3
124
i.")-
/.&
40
5-°
&
39
7-8
o
o.o
6
6.0
6
6.0
o
o.o
7
7.0
O
O.O
7- •
39
32
29
25
21
IS
5-7
4-3
3.6
0.0
2-5
2-5
o.o
5
O.O
TABLE II
GROUP II
I
i
2
•:
4
c
6
7
8
9
10
n
12
2
21
21
21
M
i-z
M
5
5
5
i
i
I
•5 . .
42O
42O
42O
7*0
760
265
IOO
IOO
IOO
20
20
2O
A. .
2O2
T*"
176
I7O
106
ICK
no
74
77
17
i
i
2
I::-:::::
7
342
1.7-
48
324
1.8+
42
347
2.0+
40+
165
4i-
1 68
1.6-
40+
194
1.8-
42-
61
1.8-
34
2+
33
60
3-5+
17
n
3-7-
iS
8
8
5
II
5-5
10
cessive ten-minute work periods. The second row gives the
number of records concerned in the experiment; the third row
gives the total number of half-minute addition series made in
the successive ten-minute periods; the fourth row gives the
number of half-minute series in which error occurred;
the fifth row gives the total number of errors; the sixth row
gives the average number of errors in those series in which
errors occurred; and the seventh row gives the percentage of
series containing errors out of the total number of series made.
The reason why the numbers in the second row decrease
from left to right is that not all records were of equal length.
Twenty records in Group I. extended through the first six
ten-minute periods, seven through the next three, etc.
THE MENTAL WORK CURVE 399
It will be seen from the foregoing tables that not only
does the number of errors decrease as the work proceeds,
row 5, but more particularly does the number of the series
which contain errors decrease, rows 4 and 7, while the number
of errors in those series increases, row 6.
These facts very naturally raise the question: What causes
these errors in the additions, and why should they become
fewer as the work proceeds? It is clearly evident that these
errors did not occur because the subjects did not know the
sums of certain numbers to be added. The errors undoubt-
edly resulted from a kind of interruption. Some other
thought, coming into the focus of consciousness, momentarily
crowded out the number to be added, or the previous sum to
which this number was to be added.
Now if we take mental fatigue to mean (as we have shown
in the article previously referred to)1 a loss of control over the
direction which any particular nervous excitation within
the brain shall take, and a growing inability to inhibit or
repress irrelevant or obtruding ideas and suggestions, and
analyze the work from that point of view, we may see how
fatigue can be developing all the time during the work and
yet the gross results show little or no effects of it. At the
beginning of the work all one's faculties are alert and ready
to respond to the slightest suggestions, or the intrusion of
any idea or impression. But as the work proceeds the facul-
ties become, as it were, insulated to extraneous suggestions
and intruding impressions. While from two fifths to one
half the series of additions at the beginning of the work
contained errors, at the end of an hour and a half or two
hours only about one twentieth to one sixth contained
errors. At the beginning of the work we are required not
only to make the additions but also to combat the host of
intruding ideas which are striving for a place in the focus of
consciousness. Every one who engages extensively in mental
work, especially if it be varied in character, knows that a
certain amount of time is required to "get settled" to any
particular kind of work. We say we can not concentrate at
1 'Archives of Psychology,' No. 31.
400 DANIEL STARCH AND I. E. ASH
the beginning of work, which is only another way of saying
that we cannot successfully combat irrelevant ideas which
are seeking to intrude themselves into consciousness.
To attempt to measure mental fatigue by such experi-
ments as are usually employed in those studies which employ
the methods of continuous work is like having the subject do
a number of things at the same time and then measure his
efforts by what he accomplishes in one of them. Such a
method would not be very far wrong if the relative difficulty
of the different activities remained constant throughout the
work period. But if the unmeasured activities can be shown
to grow constantly less difficult as the work progresses, then
it is evident that more could be accomplished of that which is
being measured without, on the whole, the expenditure of
greater effort or more energy, or the same could be accom-
plished with the same expenditure of effort.
It has been shown in muscular work and fatigue that, as
the muscles become fatigued, there is developed a resistance
to motor impulses in the nerve tracts leading to the fatigued
muscles. This resistance protects the muscles from complete
exhaustion which would occur if every motor impulse reached
the muscles without having any of its force or strength neu-
tralized by this nervous resistance which arises as fatigue of
the muscles develops. Just so in mental fatigue. As the
mind becomes fatigued by mental work its "receptive"
faculties become less responsive. Fewer impressions and
suggestions enter the mind, and as a result fewer demands are
made upon it as fatigue develops, and more of our mental
energies can be devoted to dealing with those ideas and
impressions which we are consciously and purposely in-
troducing.
The question may have arisen: Why is it that there are
fewer errors at the close of a period of work than at its be-
ginning, if fatigue means the loss of control over the processes
of mental associations and mental elaborations? Adding
two numbers is simply making an association between two
numbers to be added and a third which is their sum. The
answer is that this loss of control affects first the most recent
THE MENTAL WORK CURVE 4O1
and hence the least familiar forms of mental associations.
In the matter of adding simple numbers the associations are
old and well formed. The association is so completely formed
that the stimulus of the former will naturally bring forth the
response of the latter or the sum, unless that stimulus be
confused by the presence of another idea or impression.
That these irrelevant thoughts and impressions are less
easily controlled when the mind is fatigued, if they succeed
in entering the focus of consciousness, is shown by the fact
that near the end of our experiments, when an error did occur
in an addition series, it was usually followed by four or five
more. In other words, when the intruding impression got
into the focus of consciousness and hence crowded out the
numbers to be added, it required a longer time and more
effort to suppress the former and recall the latter.
SUMMARY
1. The number of additions per unit of time gradually
increases during the first twenty-five to thirty-five minutes
of work. This maximum is maintained for approximately
thirty to thirty-five minutes. Then the number of additions
gradually decreases until work ceases.
2. The curve representing the errors has in general the
opposite course. The number of errors decreases as the speed
of addition increases, and then increases again as the speed
decreases. An incorrect addition occupies considerably more
time than a correct addition does.
3. The half-minute periods in which errors occur becomes
considerably less numerous during the progress of work even
toward the end of long work periods, when the speed of adding
again decreases.
4. But when an error does occur it is followed immediately
by other errors more and more frequently as the period of
work continues.
5. The explanation offered for this fact is that as work
continues the mind and the neural processes involved become
more and more insulated against distracting stimuli accom-
panied by a decrease in controlling the direction of mental
402 DANIEL STARCH AND I. E. ASH
energy. The decrease in the occasion of errors would seem
to indicate the former and the increase in the number of
errors in immediate succession would seem to indicate the
latter.
6. Slow workers make relatively more errors than rapid
workers make. The difference is considerable.
7. In some respects the most striking fact brought out
by these experiments is that mental work, even of a difficult
nature, and when continued without interruption for as long
as two and a half hours, seems to produce a much smaller
lowering in speed or accuracy than is commonly supposed.
INDIVIDUAL DIFFERENCES IN A NORMAL
SCHOOL CLASS
BY ROBERT A. CUMMINS
State Normal College, Bowling Green, Ohio
The data which form the basis of this article were obtained
in connection with an advanced course in educational psy-
chology given at the Bowling Green, Ohio, State Normal
College during the summer session, 1916.
TABLE I
PHYSICAL DATA
No.
Pupil
Age
Sex
Weight
Vital
Capacity
Vital
Index
Height
Standing
Height
Sitting
Relation
Ave..
60.8
3 600
CQ.2
165.6
87.7
52 3
I
28
M
59-0
j >wwv"/
4,700
j y ~
80
*V3 *v
178
v/ •/
93
52
2
30
F
57-7
3,600
62
163
89
54
3
39
M
75-5
4,800
64
169
93
55
4
32
M
70-5
4,800
65
172
88
Si
5
25
F
52.5
3,000
153
83
54
6
28
F
45-2
3,000
66
159
82
52
7
31
F
52.7
2,600
50
161
83
52
8
28
F
59-5
3,400
58
161
88
54
9
19
F
69.6
3,800
55
171
89
52
10
29
F
56.7
3,200
56
171
93
54
ii
35
F
62.2
3,100
So
156
84
47
12
25
M
63.0
3,900
63
170
90
52
13
31
F
66.0
2,900
44
169
87
Si
No.
Pupil
Head
Girth
Head
Length
Head
Width
Cephalic
Index
Grip R. H.
Grip L. H.
Relation
Strength
Pull
Ave.. . .
55.6
I8.S
14.8
80.2
89
81.8
81.6
60.8
i
54-o
18.5
14.0
76
IOO
IOO
IOO
70
2
57-o
19.0
IS.O
79
70
66
94
55
3
56.0
18.5
IS.O
82
148
133
90
no
4
56.5
18.5
IS-0
81
120
IOO
83
62
5
55-o
18.0
IS-0
83
67
58
87
58
6
54-5
18.5
14-5
78
70
65
93
40
7
56.5
19.0
15-0
80
80
65
75
S3
8
55-5
19.0
14.0
73
IOO
80
80
58
9
58-5
19-5
15-5
79
85
88
103
60
10
55-o
18.0
IS.O
83
70
65
92
So
n
55-5
18.5
I4.S
78
72
75
103
58
12
54-5
17-5
16.0
IOO
IOO
IOO
70
13
54-5
17-5
14.0
80
75
68
9i
48
403
4°4
ROBERT A. CUMMINS
The great variation among the members of this class with
respect to age, experience, physical inheritance, etc., offered
an exceptional opportunity to illustrate some of the more
essential facts of individual differences by the measurement
of the class itself.
Then, too, it was desired to demonstrate to the class the
value of the experimental method of teaching a subject like
psychology.
TABLE II
PSYCHOLOGICAL DATA
No. Pupil
Log.
Mem.
Rote
Mem.
Digit-
Symbol
Symbol-
Digit
Free
Asso.
Oppo-
sites
Add. O.
I
Add. O.
II
Add. O.
Ill
Genus-
Species
Ave
38.3
93-o
29.2
29.9
24.2
26.O
27.7
27.7
H
I5.8
i
25
89
28
27
23
24
26
26
12
16
2
40
9i
31
42
24
28
28
2O
16
H
3
44
85
21
23
23
26
2O
28
22
22
4
25
82
24
32
24
22
2O
26
14
14
5
28
91
27
24
24
26
28
36
4
IO
6
43
93
38
33
27
30
30
32
18
22
7
44
112
31
30
30
34
34
32
20
22
8
30
83
32
35
17
30
34
30
8
4
9
43
IOI
27
28
2O
28
30
28
18
20
10
49
I O2
31
32
28
18
24
28
10
18
ii
39
1 08
30
32
30
26
30
26
10
10
12
30
62
24
24
2O
18
22
2O
12
14
13
58
109
36
27
25
28
34
28
18
20
Av. men
38 3
866
2Q.-2
a«
4.2.2
22.1
1C. I
Av. worn. .
juo
40.1
87.4
"*r*J
32.2
j j
31-3
T*^ ~
38.3
22.4
3
IS-5
(These averages taken from data of Professor Pyle.)
No. Pupil
Add.
G.-S. I
Part-
Whole
Add.
P.-W. I
Ink Blot
Cancel-
lation
Cane.
Accur.
Word-
Build.
Add.
W.-B. I
Puzzle
Box
Ave. . . .
21.8
I9-S
26.3
II-7
21.4
93
17.2
15-2
5-07
i
24
16
22
10
26
99
14
IS
5-00
2
3
4
3
26
20
%
28
32
24
II
12
16
32
2O
13
92
IOO
94
19
19
13
10
21
13
3-19
2-00
5-06
5
20
26
26
8
IS
92
18
16
2-45
6
28
24
32
ii
29
98
17
19
H-54
7
26
30
30
17
29
98
22
18
6-35
8
14
H
22
4
19
87
17
IO
14-00
9
26
24
28
14
23
98
IO
12
3-06
IO
22
18
26
19
23
86
2O
IS
1-07
ii
16
14
28
10
23
92
18
16
1-25
12
14
18
5
7
74
IS
IS
2-10
13
28
24
26
IS
19
98
21
18
5-00
Av. men.
18 c
10 6
22.2
18.6
22 7
Av.wom.
io'5
19.7
9.8
23.0
21. 1
22.O
(These averages taken from data of Professor Pyle.)
INDIVIDUAL DIFFERENCES
405
Table I. exhibits the physical data for the 13 members
of the class through the traits named, Table II. gives the
psychological data, while Table III. gives the data for an
experiment in the learning process and the transfer of training
with the same class.
TABLE III
SHOWING DATA FOR INITIAL AND FINAL TESTS IN ADDITION AND DIVISION, WITH
THE AVERAGES FOR THE CLASS. THE AMOUNT OF GAIN is ALSO SHOWN
Speed
Accuracy
No. Pupil
Addition
Division
Addition
Division
Init.
Final
Init.
Final
Init.
Final
Init.
Final
Test
Test
Test
Test
Test
Test
Test
Test
I
53
71
108
137
81
91
96
98
2
30
47
91
89
89
IOO
99
95
3
32
41
61
91
84
87
96
97
4
37
48
97
126
82
90
98
99
5
25
24
44
78
83
72
IOO
IOO
6
42
61
82
88
92
95
97
98
7
45
62
122
127
98
IOO
IOO
IOO
8
20
27
32
49
93
99
IOO
9
39
43
114
115
88
93
95
95
10
44
77
92
90
88
95
99
ii
31
5i
66
94
95
96
97
12
21
26
45
56
43
75
93
13
70
88
90
IOO
96
IOO
IOO
Av..
37-2
48.0
10 8
78.3
94-5
84.6
88.8
A 2
95-8
98.0
2 2
Gross gain. . .
Per cent
29.0
22.O
Inasmuch as all the measurements and tests are described
in detail in various publications, it is not deemed necessary
to offer here more than a brief explanatory sentence con-
cerning each.
Age — Recorded to the nearest birthday.
Sex — 'F' represents female and "M" represents male.
Weight — Taken in the metric system and recorded to the nearest tenth of
a kilogram.
Vital Capacity — Taken with a wet spirometer and recorded to the nearest
hundred cubic centimeters.
Vital Index — Computed as the ratio between the vital capacity and the
weight.
Height Standing — Taken in the metric system and recorded to the nearest
centimeter.
Height Sitting — Taken and recorded the same as height standing.
Relation — The ratio of the sitting height to the standing height, recorded
in terms of per cent.
406 ROBERT A. CUMMINS
Head Girth -.
Head Length ?• — All taken in the metric system and recorded to the nearest
Head Width one half centimeter.
Cephalic Index — The ratio of the width of the head to the length, expressed
in terms of per cent.
Grip, Right Hand 1 — Taken in the English system and recorded to the
Grip, Left Hand j nearest pound.
Relation — The ratio of the grip of the left hand to that of the right, expressed
in terms of per cent.
Strength, Pull — The number of pounds the subject can pull with both hands,
the dynamometer being held near the chest, but not touching
the body.
Logical Memory — The material used was "The Marble Statue."
Rote Memory — The material used was that given in Professor Pyle's manual.
The score is the average for the concrete and the abstract lists
of words taken together.
Digit-Symbol, Symbol-Digit, Free Association, Opposites, with the addi-
tional tests I, II and III, Genus-Species, with additional test I,
Part-Whole, with additional test I, are all described in Pyle's
manual.
These tests were all given according to the instructions given
in the manual, the score recorded being the average for one
minute.
Ink Blot — The material used was the set of 20 ink blots by Whipple. The
score recorded being the number of suggestions written down in
two minutes.
Cancellation — The material used was the standard test beginning with the
letters h, p, 1, g. The score recorded being the number of a's
cancelled in one minute, with the per cent, of accuracy recorded
in the next column.
Word-Building — The material used was composed of the letters a, e, o, b,
m, t, with the letters e, a, i, r, 1, p, for the additional test. The
score recorded is the number of words made in 2| minutes.
Puzzle Box — The Healy puzzle box was used in this test. None of the mem-
bers of the class had ever seen the box. Each one was allowed
two minutes to examine the box and was then given the button
hook and directed to open the box as quickly as possible. The
time recorded is the number of minutes and seconds required to
perform the task.
In Table I. the averages at the top are for the entire class,
men and women taken together.
In Table II., in addition to our own averages, the averages
found by Professor Pyle for adults are also included at the
bottom, insofar as these are comparable.
The slight discrepancy between the adult averages given
by Pyle and those obtained from our class in the case of free
association was noted by the class at the time and this test
INDIVIDUAL DIFFERENCES 4°7
was re-checked, but with no better results. It will be noted
that the class showed no better ability upon the whole in
writing words from free association than was shown in writing
easy opposites. This fact may be explained upon the ground
that in either case the test was not a test of ability to asso-
ciate ideas so much as it was a test of ability to write down
words, i. e.j a test in motor speed.
In the experiment in the learning process the materials
used were the Thorndike addition sheets (single column
addition examples of ten digits each, no zeros or ones in-
cluded) and the division sheets of Thorndike, which were
used by Kirby in his experiment with pupils in the Children's
Aid Schools of New York.
These examples consist of columns of division problems
arranged as follows: "48 equals ... 55 and . . . remain-
der," the task being to fill in the blanks with the proper
numbers, in this case the numbers being 9 and 3.
In the initial test 10 minutes were allowed for addition
and 5 minutes for division, the score being the number of
examples done correctly in the time given, which is called
the speed. The score for accuracy represents the per cent,
of the total number of examples done that was done correctly.
The final tests were given in the same way as the initial
tests, but after an intervening daily practice of 5 minutes at
addition for eight successive days, Sundays being omitted.
There was no practice in the case of division.
RESULTS OF THE EXPERIMENT
The class, composed of 4 men and 9 women, showing an
average age of 29.2 years to the nearest birthday, showed
an average initial ability of 37.2 columns added correctly in
10 minutes, with an average accuracy of 84.6 per cent.
As a result of 40 minutes of practice distributed over 8
days the class showed a gross gain in speed of 10.8 columns
added correctly in 10 minutes, or a percentile gain of 29.0
per cent., with a gain of 4.2 per cent, in accuracy.
The average ability of the class in division in the initial
test was 78.3 examples done correctly in 5 minutes, with an
average accuracy of 95.8 per cent.
ROBERT A. CUMMINS
The final test, with no intervening practice in division,
but with 40 minutes of practice in addition, showed a gross
gain in speed of 16.2 examples done correctly in 5 minutes,
or a percentile gain of 22.0 per cent., with a gain of 2.2 per
cent, in accuracy.
SUMMARY AND CONCLUSIONS
1. Individual differences due to heredity are best illus-
trated in a study like this by such traits as height, weight
and other traits not here recorded, as for example, color of
hair, color of eyes, shape of features, complexion of skin, etc.
2. Individual differences due to sex are best shown by
such traits as vital capacity, strength of grip, strength of
pull, together with certain tests of motor capacity such as
tapping, not here recorded.
3. Individual differences between persons of the same
sex are often greater than the difference between groups of
opposite sex. This appears in the case of imagination as
tested with the ink blots, hard opposites, and the genus-
species tests.
4. Individual differences due to school training are well
illustrated by the addition and division tests, as it was evi-
dent that certain members of the class had mastered these
combinations, while others evidently had not done so. In a
general way it may be said that equal amounts of school
training will produce somewhat the same results in a number
of individuals, as for example, three years of school training
will enable the majority of a class to read with the same
general ability from, say the third reader, yet if the same
class were tested with psychological tests it might be found
that individual differences had been augmented by the three
years of school training and experience.
5. Individual differences due to other causes are not so
well shown in this particular class, since they were all of one
race, one occupation, one nationality, and none was possessed
with any defect of body or mind due to accident.
6. One of the most striking and perhaps the most sig-
nificant examples of individual differences, from the stand-
point of the teacher, was brought out in the puzzle box test.
INDIVIDUAL DIFFERENCES 4°9
The puzzle box designed and used by Dr. Healy in con-
nection with his work at the Psychopathic Institute in
Chicago, was the one used in this experiment. The origin,
improvement, construction and use of this puzzle box are
set forth in detail by the above author in an article appearing
in Vol. XIII., No. 2 of the PSYCHOLOGICAL MONOGRAPHS.
The puzzle box is about six inches square by five inches
deep. The lid is glazed and is hinged and fastened with a
hasp, which in turn is held in place with a bolt hook and this
is made fast with a ring attached to a string, and so on with
a series of strings and rings fastened to metal pins within the
box. By proper manipulation the fastenings may all be
removed by means of a button hook and thus the box may
be opened without the necessity of forcing at any point.
The following paragraph from the monograph of Dr.
Healy will give a fair understanding of the operation of
opening the box: "The color of the strings is, of course,
arbitrary and is made different in order to facilitate the
tracing the sequence of events necessary in opening the box.
One removes first the ring over the post K and pulls out the
staple from its holes in the back of the box, releasing the
attached ring. Next the ring over the post G is lifted off,
which loosens the short orange colored string so that the
ring on the arm of post D can be readily removed. This
then so loosens the blue string that the final ring can be
pushed over the curved arm of the bolt hook and the latter
may be withdrawn, the hasp lifted and the box opened."
With regard to the purpose and the significance of this
test, Dr. Healy says: "The purpose of the test is obvious.
It may bring out abilities or defects in the manipulative
powers, in the ability to analyze a slightly complicated
situation, in powers of attention and continuity of effort.
... It is obvious that the general results obtained from this
test must vary greatly, but there seem to be three main
types of approach to the problem: first, random trials;
second, intelligent profiting by the experiences of trials
and successes or failures; third, conscious analysis of the
puzzle as a whole with recognition of the relation of the
410 ROBERT A. CUMMINS
parts. Of course, on account of the differences in strength
and manipulative power there would, other things being
equal, be considerable difference in the times taken by the
subjects. Indeed, altogether it has seemed to us that the
method employed by the subject is of more significance than
the time. Most of our twelve-year-old subjects have opened
the box in from one and a half to nine minutes, but a certain
number have finally failed."
While the above test was designed especially for use in
the classification of juvenile delinquents, yet it seems evident
that it tests the same abilities in any one who might under-
take to open it. The individual differences brought out in
our class of thirteen adults were vastly more significant than
might be inferred from the examination of the data pre-
sented in the table accompanying this article, since we have
here only recorded the time that was required for each person
to open the box.
Observation of the efforts made by the different members
of the class in trying to open the box showed that no one
of the thirteen worked out the solution of the problem by
the use of reason before attempting to open it. Most of
them began by random trials. Two or three of them made
definite moves which, as they thought, were correct, although
most of these efforts were useless.
One significant fact was that most of the class failed to
profit by mistakes made until the same mistake had been
made over a number of times, in one case as high as seven
times. The one step in the solution of the puzzle that caused
most trouble was the removing of the staple at the back of
the box after the first ring had been removed. In their own
words, they "did not think that the staple would come
out."
The experiment was carried a little further and each person
was asked to retrace all the steps and thus lock up the box,
after having succeeded in opening it. In almost every
instance an effort was made to replace some of the rings or
other fastenings without first closing ike lid! Now no one but
a child would fall into the error of locking a trunk, for ex-
INDIVIDUAL DIFFERENCES 411
ample, before closing the lid, and when this fact was called
to their attention, then each one recoiled with chagrin that
he had not been able to think of that himself.
One member even replaced the bolt hook and all the rings
and fastenings complete and failed to notice that the hasp
had not been placed over the staple! Subsequent experi-
ments upon a number of adults at random have brought
similar results in almost every instance.
7. Insofar as comparisons will admit between our data
and that reported by Professor Pyle, there appears to be a
very close correspondence, with the exception of the free
association test.
An irregularity, however, in our data which could hardly
be due to the small number of subjects, is that our subjects
make a better showing in writing the additional tests in
restricted association and invention than in the first tests
given. The additional tests are supposed to be more difficult,
hence one would naturally expect the record to be lower.
Our explanation of this reversal of expectation is that the
influence of practice in the first tests more than offsets the
difference in the difficulty of the tests.
8. The outcome of this experiment tends to confirm the
writer's contention that the experimental method of teaching
psychology is the best method, even for short term courses.
Of course it is understood that the regular textbook and
reference reading, together with systematic notebook work
were all carried on by the class.
9. In the experiment in the learning process and the
transfer of training the improvement shown by the class
was about the same as that reported by other investigators
and corresponds very closely to unpublished results obtained
by the writer from a group of 157 adults.
10. The apparent transfer of ability from the function
of adding to that of dividing is partly explained by the fact
that the division examples were arranged in a novel way
which of course means a lower initial score and a higher
score in the final test, due largely no doubt to the influence
of the practice in the tests themselves.
413 ROBERT A. CUMMINS
On the other hand, a part of the gain in division may be
accounted for by reference to improved habits of work
brought about by the practice in addition, since none of the
members of the class had been practicing with such material
just iirthis way.
VOL. XXIV. No. 6 November, 1917
.
THE PSYCHOLOGICAL REVIEW
ADVANCE ADAPTATION IN BEHAVIOR
BY S. BENT RUSSELL
St. Louis, Mo.
In the study of behavior we find that the animals that
have the most highly organized nervous systems are the best
adapted to changes in their environment. When we observe
behavior that is adapted in advance to changes in the en-
vironment, we call it purposive behavior. It is thought that
from the behaviorist point of view it would-be more con-
sistent to call it ' advance adaptation,' but purpose and
purposive are very convenient expressions. We shall there-
fore use these terms in this discussion. It is to be under-
stood, however, that this does not mean that we recognize
consciousness as a factor in the operations.
A study of the broader aspects of purpose in behavior has
recently been presented in a series of articles by Prof. Howard
C. Warren.1 The following extracts from them will serve
in some measure as a basis for our arguments: "For some
biologists purposive activity remains an insuperable obstacle
to the complete acceptance of the mechanistic standpoint."
"The vitalists . . . assert that a nonmechanistic factor . . .
is needed ... in purposive phenomena." "It remains to
be seen whether the mechanistic interpretation is adequate
to account for the purposive character of behavior."2
Let us consider now what part of purposive behavior can
be traced to nervous mechanisms. Recent writers on be-
havior have shown how we may account for such behavior
1 Howard C. Warren, 'A Study of Purpose,'/, of Phil. , Psychol.t&c.t 1916, 13,
Nos. i, 2 and 3.
2 Loc. cit., pp. 31 and 32.
413
4H S. BENT RUSSELL
as performing a definite series of movements. They have
shown that each movement of the series excites kinesthetic
impulses which follow certain paths in the nervous system
because those paths have been opened up, as it were, by
frequent or recent use. In this way these impulses excite
the next movement of the series.1
In order to provide a good foundation for a discussion of
the subject before us, let us try to get a further understanding
of such behavior as performing a definite series of movements
and of the nerve mechanisms that account for such behavior.
As already stated, movements make impulses that excite
other movements. Such impulses play a most important
part. In many cases impulses from the environment assist
in exciting the particular movement. When a man takes
his accustomed morning bath, he makes movements in a
definite order,' each being prompted by the joint action of
the environment and the preceding movements through the
medium of the nerve paths.
When the movements are repeatedly made in a certain
order, the lot may be termed a memorized series. The nerve
paths or fibers that determine the movements may be dis-
tinguished as association fibers. They are developed by
practice. Movement A may send kinesthetic impulses
through association fibers to the muscles that cause move-
ment B, or movement A may make a change in the environ-
ment and the change may affect some receptor which sends
impulses that provoke movement E. In either case we
may say that a counter signal from movement A excites
movement B.
Each movement of a series is excited by a counter-signal
and each produces counter-signals for movements to follow.
That is not, however, the whole story. Let us suppose that a
soldier has been trained to make a series of movements
which we will call A B C D E F G, each letter representing a
movement. Counter-signals from movement A will directly
cause movement B. Moreover, counter-signals from A will
tend to excite C and more faintly D and still more faintly E,
1 Watson, John B., 'Behavior,' New York, 1914, p. 275.
ADVANCE ADAPTATION IN BEHAVIOR 415
and so on. In the same way each movement will partially
excite every movement that follows in its train. Now when,
for example, a counter-signal from A reaches the nerve path
leading to the motor ending for Z), it in some way facilitates
the later impulse from C to D. It opens the common path
or lowers the resistance so that movement D is thereby
insured.
To explain further, we will assume that there is a dual
common path that is open to impulses from A and C only
and which leads to the motor ending for movement D.
This path has been developed further each time an impulse
from C has followed in the wake of an impulse from A.
As a result of this path-making, an impulse from A leaves the
path so wide open for a time that an impulse from C can
effectively reach the motor ending for movement D. This
explains why movement D follows after A and C. In like
manner each other movement makes counter-signals that
facilitate each movement that follows in its train.1
To express it another way, each movement makes a certain
impression on or change in the nervous system which takes
part in determining what movements shall follow. This
impression must last a considerable time, but gradually
diminishes. Let us now suppose that these impressions
on the nerve fibers remain after the whole performance is
completed and last until the performance is repeated. We
see that new impressions will be imposed on old ones.
In passing on these theories of nervous operations it is
well to remember that the simplest movement is caused by
the contraction of one or more muscles and that each muscle
is a group or community of muscle fibers. Leading to these
muscle fibers is a nerve made up of nerve fibers. Many of
these fibers are motor fibers. A single motor fiber may
branch to a number of muscle fibers. We know, of course,
that a muscle contracts with varying strength when stimu-
lated by impulses coming through the nerve. On the other
hand, physiologists have made experiments in recent years
1 A discussion of this theory is given by the writer in an article on ' Compound
Substitution in Behavior,' PSYCHOL. REV., 1917, 24, 62-73.
4J6 5. BENT RUSSELL
which indicate that in an individual nerve fiber the size of
the propagated disturbance does not vary with the strength
of the stimulus.1 In this connection, the great number of
nerve fibers that may be concerned in each operation has
an important bearing.
Modern behaviorist theories of nervous mechanisms rest
on the principle that recency and frequency of previous
excitations determine the extent that the resistance is
lowered in a given nerve path, to the passage of an impulse.
It may be noted here that a memorized series is learned by
one lesson after another. In the first lesson the situation
is more or less new or unfamiliar. With each additional
lesson, of course, the situation becomes more familiar. In
childhood man learns a vast number of movement series,
i. e., he acquires numerous habits. In maturity his behavior
is largely governed by habits. Language habits and other
social habits are important factors in man's daily life. In
new situations behavior may be largely instinctive, in familiar
situations habit governs. Let us put it down as an established
fact that such behavior as performing a memorized series
of movements can be satisfactorily explained in terms of
nervous mechanisms made up of association nerve fibers.
The object of this article is to show that purposive be-
havior can just as well be explained in the same terms. At
least, we can show that some forms of behavior that would
be considered purposive can be so explained. We shall begin
our demonstration by pointing out what behavior may be
termed purposive. We shall then call attention briefly to
the correspondence in the growth of purposive behavior with
the growth of habits and language movements, and move-
ments associated with measurement of time. The part
played by teaching will be pointed out. The effect of reward
upon actions will be discussed. Recognition of purpose will
be considered. We shall then take up the development of
forethought, so to speak. It will be pointed out that mem-
orized movement series and series of incipient movements
may be combined so as to provide for and give rise to pur-
posive behavior.
1 Adrian. /. of Physiol., 1914, 47, 460-474.
ADVANCE ADAPTATION IN BEHAVIOR 4'7
In looking back over a series of behavior phenomena, we
see that some of the changes in the environment were pro-
vided for in advance just as "Coming events cast their
shadows before." For example, we see that Mr. Gray,
who went out for a walk at three o'clock, at starting borrowed
an umbrella. At four he returned in the rain, sheltering
himself with the umbrella. We see that the act of borrowing
the umbrella at three was adapted to the need for shelter
at four. It is therefore advance adaptation or in common
terms purposive behavior. Let us take another case. We
see that Dick said to Ralph, "I want to get a canoe and go
paddling. Will you come too?" Ralph replied, "Yes, I
will go with you," and later on we find them on the water in
their canoe. The question and reply, we find, were adapted
to later conditions when they arrived at the dock. So this,
too, is purposive behavior. Let us take a third case. A boy
saw a strange dog and stooped and picked up a small stone.
The dog saw the movement and dashed around the corner
of the house so that he escaped the flying missile. We see
that the dog's action was adapted in advance to the boy's
act of throwing. This, too, is then purposive behavior.
To compare with this, let us think of a boy that slipped and
fell on a briar and hurriedly got up. This is not purposive
behavior, as there is no anticipation in it.
We may here observe that the acts of an educated man
are usually compounded of purposive and nonpurposive
elements. It is in the educated man that we find advance
adaptation the most highly developed. The best measure
of the degree of advance adaptation, it is thought, will be
found in the precision and definiteness of the correspondence
of the behavior with the ensuing changes in the environment.
Let us call this, for convenience, the degree of purpose. As
an illustration, a man who composes the score of a grand
opera is giving a good example of behavior having a high
degree of purpose. No argument is needed to show that in
the human species, the degree of purpose develops side by
side with the accumulation of knowledge, i. e., with the
formation of memory associations. The same rule seems to
4i8 5. BENT RUSSELL
hold with the brutes. The fox that has been hunted before
is harder to capture because, as we say, he is more wary.
On consideration, we find good reason to believe that in race
development and in individual development, the growth
of purposive behavior has been in proportion to the growth
of habits.
Let us now enquire briefly into the effect of language upon
purposive behavior. In considering adaptation in animals
we must take account of their social environment as well as
of their physical environment. The individual must adjust
his behavior to meet the actions of others of his species and
also those of his natural enemies and it may be those of the
creatures he preys upon. If the animal has even the most
primitive nervous system, it will be affected by the move-
ments of other animals. Natural selection brings about the
evolution of language habits in many species. The sense of
hearing must have been developed mainly by the demands
of social environment. By means of sound waves there is
communication between animals so that the movements of
one individual will excite a response in another individual.
When a quail is flushed, the sound of his wings provokes the
flight of the rest of the covey. The senses of touch, smell
and sight are also used for communication. It is probable
that sign language is much used in many species. Dogs
and cats show by the movement of their tails what their
feelings are and no doubt their companions are observant
of the movements and are guided accordingly. When a big
dog growls at little dogs, he says in dog language, "Keep
away, or I will bite you," and when the little dogs retreat,
they say with their tails, "Come away or he will hurt us."
On consideration, we find that the evolution of language
habits must have proceeded side by side with the evolution
and differentiation of species and with their social evolution.
In the individual we find that language habits develop from
birth up. In this connection let us note that nerve paths
are lines of communication from sense organs to muscles
and from muscles and joints to other muscles. Let us note
furthermore that by means of distance receptors the lines
ADVANCE ADAPTATION IN BEHAVIOR 4*9
of communication are extended into the environment.
Language movements serve for communicating with other
individuals. Hence the excitation of a sense organ of one
individual is conveyed to another one and the latter responds
accordingly. This control of one creature by another must
have been, from the beginning of its evolution, concerned
with advance adaptation.
In the human species we find that the degree of purpose
keeps step with the increasing use of language. The naval
architect making plans and specifications for a submarine
boat, is employing language movements and, at the same
time, is behaving with high-degree purpose. In fact high-
degree purpose seems almost dependent on language. Let us
remember that language movements in man are the result
of training, i. e., of the development of habit mechanisms in
the nervous system.
For high-degree purpose, a knowledge of the meaning of
time is necessary. Until a child has learned the meaning
of time, his actions show little definite purpose. He must
be taught to note the passage of time. He must learn that
it takes a minute to walk to the barn and it takes an hour
to walk to the railroad station. He is taught to count the
days of the week and the weeks before vacation. Let it
be remembered that such knowledge of time is largely due
to the development of association or habit mechanisms in
the nervous system. By these mechanisms, certain move-
ments come to be associated with the measurement of time.
When the individual has acquired language habits and
time measuring habits so to speak, he begins to take part
in the joint activities of the family and other social groups.
More purposive behavior in the individual is favorable to
greater adaptation of the social group to its environment.
It follows that natural selection, acting through the group,
brings about the evolution of purposive behavior. We can
say then that man owes his purposive powers to his social
position.
In highly cultured communities, the children are trained
up to purposive behavior. This is done by means of language
420 S. BENT RUSSELL
mechanisms and habit mechanisms. A child is made to
tell what he is going to do and what good will come of it.
He is trained to plant the seed so as to get the flowers or
fruits in due season. As a result the individual not only
comes to be highly purposeful but he also comes to think of
human behavior as being largely purposeful. Hence he is
apt to think that he himself or another man is guided by
purpose when he is really guided by habit. He is apt to
think also that the purpose that guides his movements is
created spontaneously within him, quite independently of
his organs of sense or movement.
Let us now go back to more primitive behavior. The
animal trainer rewards his dog with a tempting bit of food.
Every time the dog performs well, he gets the reward. So
the dog associates the reward with the movements that
make up the trick. Hence we may say that this behavior
is purposive in a sense. He responds to the trainer's signal
so as to secure the reward. Now it can be shown that
association mechanisms will account for such behavior.
Experiments by PawlowV method prove that after proper
experience, a dog's mouth can be made to water at the sight
of a green light or other stimulus. Such a response is known
as a conditioned reflex. It is plainly a case of association
mechanism. The sight of food by a hungry dog has a
decidedly animating effect. The effect on his nervous system
is evidently widespread. By means of association mechan-
isms, other kinds of stimulus may be substituted. An animal
will run to the feed box when it hears the keeper coming.
Any stimulus that frequently occurs a short time before the
getting of food will soon establish a sort of conditioned reflex,
so that the stimulus will have a marked animating effect
on the animal. So when the trainer gives the signal for a
trick, it has much the same effect on the trained animal as
if food were shown it. The animal is alert and responds
readily to the accustomed signals according to the estab-
lished habit. It then receives a bit of food as its accustomed
reward.
1 J. B. Watson, 'Behavior,' Holt, 1914, p. 65.
ADVANCE ADAPTATION IN BEHAVIOR 421
Now it may be while the brute is performing that he has
something in mind like an image of the food that is to come
after the trick, but, whether he has or not, we see that the
association nerve paths provide for the proper responses.
We see then that although the animal's behavior is in some
degree purposive, there is no active power that comes from
within, as it were. There is nothing beyond the impulses
aroused by the environment and by the trainer's signal.
The association nerve paths govern the behavior. The
trained animal is then only a piece of machinery, although
we must admit that it is in advance of the machinery built
by man, for as yet no machine has been constructed by art,
even in appearance, that will be influenced by a reward to
follow its operations. It is probable that the most intelligent
behavior of animals is governed by past experience and
present environment and not by what is going to be. So
far as animals go then, purpose is, in one sense, only a name,
as it were. And yet the dog that goes after a stray sheep
and drives it back to the flock seems to have as much purpose
as there is in most actions of a human child.
Let us now observe that purposive behavior in animals
includes what we may call the faculty of recognition of
purpose in others. When you pick up a stone to throw at a
dog, he appears to know what you are about to do and
slinks off. People used to say that such behavior was due
to association of ideas. There is no way to tell when the
response is due to association and when it is due to conscious
recognition of purpose. When a hunter takes down his
gun from the rack, his bird dog shows by her demonstration
of joy that she knows her master is preparing to take her
out for a hunt in the field. A critical view, however, indi-
cates that it is another case of association mechanisms. Just
as the sight of food can cause increase of animation, so can
the sight of her master's preparations, by virtue of con-
ditioned reflexes, do the same. The dog wags her tail,
leaps and perhaps barks, thus saying in dog language that
she is delighted at the prospect.
Let us now review briefly what progress we have made.
422 S. BENT RUSSELL
We have considered the effect of language habits and time-
measuring habits upon purposive behavior. We have found
that men are taught to have purpose. We have considered
the way reward acts in advance adaptation. In all these
items, we have found that associative memory is a most
important factor. In view of these relations, we may allow
that purposive behavior is underpinned on all sides by
association nerve mechanisms. Let us now enquire how
the central foundation of such behavior is constituted.
When an animal or child has learned to go through a
certain series of movements in a given order, the performance
is, as before stated, a memorized series. We have already
established that such a series is the operation of nerve mechan-
isms and that kinesthetic impulses link the movements
together. Each one of the movements in turn is provoked
by a volley of effective nervous impulses coming by way of
converging association fibers, located, we will say, in the
cerebral cortex. These fibers have been recently prepared
by scout impulses, as it were, that came from the receptors
in the muscles that made the movements which went before
in the series. Each association fiber received a stimulus or
scout impulse from a certain movement which lowered its
resistance for a time. It was thus prepared for the effective
impulse that came soon after. In this way, we see how each
movement in the memorized series is determined by the
movements that go before, by means of association fibers
which have been developed by previous training. Let us
now think of the series so modified that each movement is
incipient only, but that it causes afferent impulses that excite
other incipient movements in order, so that the incipient
movements correspond to the actual movements in the pri-
mary series. Let us for convenience call this modified form a
secondary series. We see that a secondary series is some-
thing like a train of thought.
As above stated, we must presuppose that an incipient
movement can in some way provoke an appropriate afferent
nerve fiber, so that we shall have an impulse that corresponds
to a kinesthetic impulse provoked by an actual movement.
ADVANCE ADAPTATION IN BEHAVIOR 423
To explain more definitely what takes place, we will recognize
two alternative assumptions. The first one is that an im-
pulse is conducted by a motor nerve, although it is too weak
to cause contraction of the muscle and is then conducted
slowly through the muscle to a sensory terminal of an afferent
nerve fiber. Hence a faint discharge from the brain to a
muscle is followed by a counter-signal from the muscle to
the brain. As an alternative the assumption is that at some
nerve junction or synapse there is a short circuit from the
motor nerve to an afferent nerve fiber and that the impulse
is delayed appreciably at this junction. Hence a faint motor
discharge is followed by a counter-signal that reaches the
cortical centers and serves in place of a kinesthetic impulse
from an actual movement. With either assumption, we
see how one incipient movement will provoke another one
and there will be a series of impulses running back and forth,
from and towards the cerebral cortex. As in the case of the
primary series, the counter-signals will follow the common
nerve paths where the resistance has been lowered by recent
or frequent previous impulses.
We now have a fair understanding of both the memorized
series and the secondary series and the relation between
them in behavior. Let us therefore attack the main question
before us and determine whether such behavior mechanism
operations will account for responses that are adapted in
advance to changes in the environment.
Let us take a definite memorized series made up of move-
ments which we will call ABCDEFGHI, and suppose
that it is sometimes replaced by a secondary series, a b c d e
f gh i. When a child has learned these series, if he is induced
to make the movements A B C, the other movements D E F
G H I will be made from habit and if he is induced to make
the incipient movements a b c, the other incipient movements
d e f gh i will tend to follow. We may believe that the same
association mechanisms are concerned in both series. When
the primary series is followed by the secondary series, we
may term it a recall. On consideration we see that it may
sometimes happen that the secondary series will come in
424 BENT RUSSELL
between the movements of the primary series like this :
A B C D E abc def ghiFGH I.
We see too that here we have something resembling fore-
thought, for the incipient movements / g h i show a corre-
spondence with the actual movements F G H /, made after-
wards. In this way we find that an elementary case of
forethought can be accounted for by simple association
mechanisms. It is the first step that counts, so it is not an
unfair presumption to claim that we could advance by
degrees and account for all purposive behavior by association
nerve mechanisms. It will be seen that where there is a
series of incipient movements that corresponds with a series
of actual movements to be made later, a foundation is pro-
vided for advance adaptation.
The intelligent animal knows the future by the past.
What was the future yesterday morning is now the past.
Looking back one sees what he might have looked forward
to yesterday morning. From this he may conclude what
he may look forward to this morning. Hence the experience
of yesterday becomes the forethought of today. The en-
trance to the path that led to food yesterday will suggest
the best path to take today to the hungry animal. If he
takes the path and it leads to food again we have a case of
advance adaptation. In considering the effect of reward,
we saw that such behavior as that is explained by association
mechanisms.
Let us now think of a child that makes the incipient
movements / g h i, as a result of previous operations. Sup-
pose that the afferent impulses from these movements pro-
voke associated language movements as in uttering the words,
"I am going upstairs." Then the actual movements F G H I
of the memorized series are made, such as proceeding up-
stairs. We see that the uttered words exhibit advance
adaptation. Let us note that this demonstrates how a
secondary series together with association accounts for
purposive behavior. In a similar manner it is thought that
behavior equally purposive but not including language
ADVANCE ADAPTATION IN BEHAPIOR 425
movements can be accounted for. For example, taking a
key out of your pocket to unlock a trunk. Of course the
necessary common nerve paths must be there.
We may conclude from this demonstration that the
secondary series provides a foundation for advance adapta-
tion. It is obvious then that purposive behavior can be
explained as due to association nerve mechanisms. On
further examination we shall find in very many cases, espe-
cially with man, that the secondary series is made up of
incipient language movements. A man who is given to
talking will hold a silent conversation with himself, before
acting in a situation that is not quite familiar. From our
point of view, it is all due to nervous impulses that follow
the paths most open to them as determined by previous
nerve muscle operations. We shall also find in human be-
havior many cases where movements or incipient movements
that are associated with measurements of time, constitute
the secondary series that anticipates the future situation, so
to speak. One can as easily plan for next Christmas as for
his next meal. The same kind of nerve mechanisms are
employed in either case, and in both the responses are deter-
mined by one's past experience. It should be remembered,
however, that thanks to the advantages of spoken and
written language and to education, a man can make the
experience of others serve as his own. That is where the
social environment comes in.
The more highly developed the nervous system and the
greater the number of association nerve fibers, the more
precise and definite will be the anticipation of future needs.
An old fox is warier than a young one. An educated man is
more purposeful than a savage.
To sum up our conclusions briefly, if habit forming can
be explained as due to association nerve mechanisms, advance
adaptation can be accounted for in the same way. If our
view is the right one, mechanistic interpretations are adequate
to account for the purposive character of behavior.
RELEVANT AND IRRELEVANT SPEECH
INSTINCTS AND HABITS
BY P. F. SWINDLE
Research Fellow in Psychology, Harvard University
INTRODUCTION
The audible vocal responses of certain birds and of the
human being will be the matter of chief concern in this paper;
but in order to discuss these responses intelligently, and in
order to include all the movements which may be used as
means of social communication, it will be necessary to in-
vestigate a number of other bodily movements. Excepting
the vocal responses, the most important ones for speech
are the mouth movements which modify the vocal responses
directly. Movements of minor significance are those of
the hands, feet, head, eyes, wings, tail, and the like, which
may influence in any way the vocal responses, or which may
themselves serve as gestures.
All of these responses are instinctive, and all of them
may be at the same time habitual; a habit is an instinct which
occurs more frequently than originally. The frequency of
occurrence of an instinct is necessarily increased if it is
associated with, and is regularly conditioned by, at least one
other response of the same individual.
Relevant speech instincts and habits are those responses
of the individual which serve as stimuli to call forth pre-
dictable responses in other individuals. Some speech re-
sponses are at the outset relevant, while others are made so
through a process of training of the individuals who are to
produce, perceive, and respond to them. In other words,
many irrelevant speech responses must. become convention-
alized, if they are not to remain superfluous expressions in
ordinary speech.
For the sake of simplicity in dealing with vocal responses,
426
SPEECH INSTINCTS AND HABITS 427
I shall make use of the vowel names, a, e, i, o, and u, as
abbreviated means of speaking of the very large number of
vowel sounds which an organism in question may utter.
The qualitative nature of any given vocal sound could be
indicated by marking diacritically one of these five letters,
but it would be unnecessary and perhaps confusing if the
attempt should be made in the following pages. It will be
necessary, however, to indicate in a limited number of cases
the specific qualitative nature, e. g., of an a\ but in general
the reader will be privileged to image or utter any a sound
which occurs to him at the time.
In discussing the consonants, I shall also speak primarily
of some of the simpler consonant names of the alphabet as
the most familiar representatives of the very large number of
consonant sounds which the organisms utter.
EXPERIMENT AND DISCUSSION
To understand conventional speech, it is advisable to
study carefully some simple forms of speaking organisms.
I shall for this purpose begin with the barbet, or, as it is
usually called, the * bearded bird.' It frequently speaks, or
more strictly sings the letter a somewhat as in art. When
the bird is not molested it utters a succession of #'s approxi-
mately in the tempo of 0.2 sec. These sounds form a great
number of <z-groups, the largest one of which contains not
less than five hundred <z's as its elements. When the group
is very long an observer can become aware of a progressive,
qualitative change of the a's. The change is usually so
gradual that it is not apparent in the shorter groups. Often
the mouth closes at the end of a group in such a way that
the final a is converted into a sound resembling the English r.
When an <2-group is thus terminated, I shall speak of an
ar- group-compound, or simply of an flr-compound. The
a- group-complex, or the <z-complex, which is the summation
of the quantitatively identical and different ^-groups, is the
most essential one which the vocal apparatus of this bird
manifests.
The mouth movement which occurs regularly at the end
428 P. F. SWINDLE
of a given fl-group and converts the final a into an r sound,
is a specific activity which at some previous time interrupted
a series of a*s and became permanently associated with the
final a of the a-group which it isolated from the longer series.
This mouth movement became the permanent conclusion
or final accent of the isolated tf-group. The following
experiment makes this view plausible.
Observing the bird carefully, I struck, whenever possible,
the glass of its cage at the fourteenth a. This stimulus called
forth the mouth movement which interrupted the series of
<z's and served to convert the fourteenth a into an r. It
became eventually so well associated with the final element
of the i4-<2-group that I no longer had to apply any stimulus
to the bird at this point of the series. The ^/--compound
which thus came to be uttered as regularly as the i4~<2-group
occurred, consisted in all of fifteen perceptible elements:
first, thirteen stutters, that is, the a's which were unnecessary
for the perfect pronunciation of the r\ secondly, the significant
a, the only one necessary for the r\ and, thirdly, the mouth
movement which made the audible difference between the
fourteenth a and the preceding ones. The bird was also
taught by the same method to utter an <2r-compound con-
sisting of thirty-three stutters and the r which was a modi-
fied a. No other habits were acquired by this animal under
experimental conditions.
The particular training to which the bird was subjected
habitualized the mouth movement; that is, it caused this
movement to occur more frequently than originally.1 After
the training, the mouth closed not only when the glass of
the bird's cage was struck, but also whenever the final a of
either the 14- or the 34-<2-group occurred. No portion of an
<2-series was caused to occur more frequently than originally,
and consequently not even the 14- and 34-a-groups can be
called habits. This training, at least, did not make them
habits.
1 For a more detailed discussion of instinct and habit, see my article, 'Ueber
einfache Bewegungsinstinkte und deren kiinstliche Beeinflussung,' Z. /. Sinnesphysiol.,
Bd. 49, 1915. See especially pp. 247-248 for my definitions of instinct and habit.
SPEECH INSTINCTS AND HABITS 429
The training did, however, cause certain a's to occur less
frequently than originally. For example, in the case of the
series from which the i4-<z-group became isolated, the four-
teenth a conditioned the mouth movement instead of the
fifteenth a; and with the mouth closed the bird could not
continue the series. This does not mean that the latter
part of the original series vanished completely, but only that
the fourteenth and fifteenth #'s became dissociated. The
most significant result of this complete dissociation was the
addition of one more group to the <z-group-complex.
The cockatoo is a bird whose speech instincts outnumber
those of the barbet. Its vocal apparatus functions in various
tempi, as is indicated by the differences in pitch of the many
tones; and the mouth moves in various tempi, directions,
and amplitudes. A large number of group-complexes is
manifested by the vocal apparatus and mouth parts. The
cockatoo, like the barbet, stutters; but the succession of the
elements of an audible series is usually so rapid that only a
continuous tone is perceived. It is sometimes a whistling
tone. The mode of whistling is, however, quite different
from that of the human being; for, while the human being
whistles mainly with his lips, the cockatoo, which is almost
void of soft lips, produces these tones primarily with its vocal
cords.
When any one of the vocal responses of the cockatoo is
interrupted by a mouth movement a compound almost
invariably results which might well be used in conventional
speech. Many mouth movements are of such slight ampli-
tudes that a vocal series which is in progress at the time may
not be interrupted completely, but only slightly modified
by them at the points of the series where they occur. Further,
these movements are often of considerable duration. I
learned that I could call forth a large number of these move-
ments in a desired order, and took advantage of the situa-
tion to teach a yellow crested cockatoo to whistle recognizably
a certain simple portion of 'The Wearing of the Green.'
The cockatoo had never heard this melody until it itself
produced it. The stimulus I used to modify the monotones
43° P. F- SWINDLE
which the bird for some unknown reason frequently whistled,
was a burning candle which I moved in various directions
and with various speeds before the animal. In whatever
direction the candle was moved, the cockatoo threw its head
or even its entire body toward the flame. For example,
when the candle was moved to the left, the bird often raised
only the left foot and thrust its body in that direction.
Many of these thrust-like movements were accompanied by
alterations in the size of the mouth cavity. Thus the tone
could be modified successively in various ways.1 The method
used to cause the cockatoo's tones to become melodious, also
caused it to dance to the movements of the flame. At other
times, i. e., in the absence of the candle, it danced as if to the
tune it produced. At these odd times the thrusts of the
head or entire body were not quite so pronounced and sudden
as when I called them forth with the flame, and the dancing
was accordingly more graceful and the whistling more
melodious.
I also used various other stimuli which called forth the
desired responses when properly moved before the bird,
but the candle was the most useful one I found. Some of
the others were a piece of ermine fur, a bundle of feathers, a
frog, and a dead mouse. The burning candle was especially
adapted for this investigation, not merely because the flame
was a stimulus which the bird's eyes generally followed, but
because its shape and size could be varied with the velocity
of movement. When moved rapidly, the flame became very
small and then flared when the movement ceased. I could
make the flare of the flame so disturbing to the bird that the
tone would be interrupted completely, which event I avoided
in this particular experiment.
There are some very brief mouth movements manifested
by the cockatoo which do not interrupt a vocal series com-
1 This particular bird was the most suitable subject I found for this experiment.
Other cockatoos responded to the flame in a similar way, but their movements toward
it were not so pronounced, and the tones were not altered in the same degree. I also
worked in the same way with a small owl, whose vocal utterances were more similar
to those of the barbet, and determined that its vocal responses became slightly modified
at many turns of the head.
SPEECH INSTINCTS AND HABITS 43 l
pletely, but only alter the tempo slightly and become inserted
between two of the elements. Such a compound would be
aaapaaaa, or eeetee.1 Any such mouth movement which can
be caused to occur once in a vocal series, may be caused to
occur twice, and if the bird in a given environment is stimu-
lated to insert a p in an ^-series, it will very probably be
stimulated in the same way soon again. This is a form of
natural training which is evidently responsible for many of
the compounds spoken by wild cockatoos. Some of the
more frequent of these are: aaapaapaa, aaaamaaamaa,
aakaaakaaa, uutuuutu, eeemeemeee? These activities are
not necessarily rhythmical, because the mouth movements
in any case do not necessarily occur at absolutely regular
intervals.3
It is not necessarily the case that a vocal response becomes
thus slightly modified two or more times by the same mouth
movement. The mouth is capable of executing a great
variety of movements, and it is scarcely conceivable that the
stimulus for only one of these should be present in any given
environment. Further, if the bird should take a step or
perhaps turn the head after the mouth has moved the first
time, and while the vocal response is still in progress, such
an act might create a sufficiently new environment for the
animal that it would now be exposed to a stimulus for a
decidedly different mouth movement. Some of the resulting
compounds might be: aaaa-paaataaaaa, aaaaaapaaateeeeeee,
1 Any series of like letters not separated by commas are meant to indicate ele-
ments of a vocal response which is only apparently continuous.
2 Perhaps Binet and Simon had such compounds in mind while investigating
the child's language. Alfred Binet et Th. Simon, 'Langage et Pensee,' Uannee
Psychologique, 1908, p. 310: "Comment pourrait-il prononcer des mots qu'il n'a
jamais entendu, et si on ne les lui a pas appris par une autre voie? II existe done,
au moment de 1'elaboration des fonctions, des relations nombreuses entre les fonctions
naissants."
3 Wundt calls special attention to the fact of the repetitions in the child's utter-
ances, but he seems to think the phenomenon is due to its appreciation of rhythm.
Wilhelm Wundt, 'Die Sprache und das Denken,' Essays, Leipzig, 1906, 2. Aufl., S.
281: "Eine primitivste Ausserung des Gefallens am Rhythmus ist wohl auch die
Neigung zur Wiederholung der Laute, durch die beinahe alle Worter der Kinder-
sprache zu Verdoppelungsformen geworden sind. Urspriinglich sind aber die Wieder-
holungen fast immer mehrfache, und erst allmahlich sind diese zu den gelaufigen
Verdoppelungen, Papa, Mama, Wau-wau, u. dergl., verkiirzt worden."
43 2 P> F. SWINDLE
eeeteeeepaaa, aaaaakaaakaaaatuuuu, and the like. In the
last instance, e. g., the w's follow the t because the initial u
of the w-group became at some previous time associated in
this order with the mouth movement responsible for the t;
and if the bird does utter the w's, it cannot utter the a's at
the same time. The ^-compound, or whatever is responsible
for it, which in this case inhibited a number of a's, served the
same purpose as the mouth movement of the barbet: the
barbet was unable to pronounce the a's while the mouth was
closed, and the cockatoo was unable to pronounce the a's
while uttering the u's.
One reason why so many of these words more often start
with a consonant than with a vowel, is that the very move-
ment which opens a closed mouth to utter a vowel group is
quite frequently one which is indispensable for many of the
consonants. If the bird should be stimulated to utter the
series uuuu at a time when the mouth is not already in the
proper position for this act, it might be opened suddenly
and thus permit the cockatoo to say tuuuu, puuuu, or some
other consonant. We should look upon the last three of the
u's in either of these cases as stutters, because they are not
necessary for the perfect pronunciation of the t or p. The
letter t, as it appears on paper, is only a symbol for some one
of the many forms of a vowel which is modified by a mouth
movement. We can write the letter t without the vowel
of which it is a modification, but while this exists as a visual
stimulus, it is an impossibility in audible speech. The
written t of the alphabet signifies the ^-compound. Another
reason for the fact that vocal responses so frequently start
with a consonant is that the brief mouth movements essential
for the pronunciation of many of the consonants occur
much more frequently than the vocal responses; no animal
speaks every time it opens its mouth. This means that the
proper mouth movement for a given consonant often occurs
and modifies a vowel sound, provided that upon the opening
of the mouth there occurs a vocal response to be thus modified.
Many speech compounds, however, must start with a vocal
response. For instance, any one of the consonants of our
SPEECH INSTINCTS AND HABITS 433
alphabet, which is not one or another of the vowel forms
initiated by and modified by at least one mouth movement,
is initiated by some vocal response, the last element of which
is modified by the mouth movement. This means that some
of our consonants would be impossible in verbal speech if
certain vocal responses could not be in progress when the
significant mouth movements occur. These consonants of
the alphabet are the ef-, ah-, el-, em-, en-, ar-, es-, and ex-
compounds. The existence of such compounds as these
and many other similar ones which do not appear in the
alphabet, illustrate clearly the impossibility of starting a
rather large number of our conventional speech compounds
with mouth movements, for, in these cases, the mouth
merely moves and modifies vocal responses already in prog-
ress.
The vocal and mouth movements of the cockatoo can be
associated under controlled conditions to produce not only
the various consonant names, but also great numbers of our
monosyllables, and these can in turn be associated with one
another to form larger compounds such as our polysyllabic
words. A monosyllable is for the speaking organism just
as simple as, or at least not much more complex than, most
of our consonant names. A monosyllable in speech is merely
a more or less compound utterance that maintains its identity
when associated with other speech compounds, more often
than do our consonants, which are also compounds. In
other words, a monosyllable is ordinarily hot as variable
as a consonant name, and is usually just as easily spoken.
The consonant names consist of a-, e-, i-, o-, or ^-groups
which are either preceded or succeeded by a mouth move-
ment that modifies at least one element of the group. This
mouth movement is of such magnitude and brevity that the
resulting audible effect is usually an explosive one. The
monosyllables consist of like vowel groups, which may, how-
ever, be preceded and also succeeded by the distinct mouth
movements.
The vowels, too, are groups which are often modified by
mouth movements; the mouth frequently moves before as
434 P> F- SWINDLE
well as after each vowel group, but these movements do not
produce such explosive effects as in the cases of the con-
sonants and monosyllables. Various tendencies which mani-
fest themselves when we attempt to pronounce the vowels,
deserve emphasis. In uttering a or i, the a-group or z-group
is frequently terminated in each case by an ^-group, while
the ^-group, on the other hand, instead of being terminated
by an ^-group, is frequently preceded by it. In the face of
these facts we are not justified in writing for the vowels only
the a-, e-y i-, o-, and ^-groups, but we must add the ae-^ ie-,
and ^-compounds, as well as others. These compounds
represent only three of the so-called diphthongs of speech;
and even they, depending in general on whether they occur
at the beginning, in the middle, or at the end of our larger
speech compounds called words, are under conditions differ-
ently pronounced.
While the majority of the mouth movements are such
as can modify vocal responses, it is to be borne in mind that
there are a great number of other bodily activities which
neither completely interrupt nor modify in any observable
way the audible vocal responses. Some of these are the
movements of the tail, wings, hands, feet, head, and eyes.
True enough, these may carry the body through space and
cause the speaking organism to become affected by a number
of different stimuli which call forth various mouth move-
ments. An eye movement may sometimes appear to alter a
vocal response which is in progress when the eye moves, but a
safer supposition is that it conditions a mouth movement
which interrupts the vocal response or modifies it in one
of the other ways previously discussed. Any movement of
the body can become associated either directly or indirectly
with a vocal response and condition it, be conditioned by it,
or merely accompany it; but the mouth movements alone
can interrupt it.1 Many of the larger movements of the
body which are associated with vocal responses, and which
appeal primarily to the visual sense, are called gestures; but
1 For reasons that ohould be obvious, I am considering only those vocal responses
which occur during the period of exhalation.
SPEECH INSTINCTS AND HABITS 435
there is really no fundamental distinction to be made between
these and the movements of the vocal cords, or of the very
slight movements of the various parts of the mouth. Any
given perceptible response of an animal organism becomes a
means of social communication when it stimulates other
beings to behave in a predictable way. Some of these
movements may be perceived with the ear, some with the
eye, and some with both eye and ear. It is a matter of
insignificance what sense organ they affect. Those per-
ceptible movements which do not call forth predictable
responses in other individuals, since they have not become
conventionalized, are nonsense expressions. The process of
conventionalizing responses for means of communication is
nothing more nor less than a process of training the indi-
viduals who are to utilize them to stimulate one another for
calling forth predictable responses.1
The human being often stutters in a manner similar to
the barbet, but, as does the cockatoo, in such a very rapid
tempo that for the hearing organism an apparently con-
tinuous tone results. He stutters thus either before the
interrupting mouth movement occurs or after it. The
tendency of the human being to stutter in this way is indeed
very great, as is illustrated by the manner in which the
infant * coos' and by the way it later utters such words as
m
maaaaaaaaaa, or paaaapaaaaaaaaa. In the earliest stages of
infancy the child does not often begin these words with m or p,
but, as the cockatoo often does, with a. They should accord-
m m
ingly be written somewhat as follows: aaaaaaaaaaaaaaaa^
aaaaaaapaaaaapaaaaaaaaa. The mouth movement of the
child which is responsible for the m sound within an ^-series
1 While this point seems obvious, it has not been given due attention by the
investigators of speech. Wundt, for example, in his 'Physiologische Psychologic,'
III., S. 285, says: "Alle Ausdrucksbewegungen geschehen selbst beim Menschen im
Anfang des Lebens unwillkiirlich; sie sind teils Triebhandlungen, teils reflectorische
Bewegungen. Allmahlich erst werden einzelne willkurlich gehemmt, andere hervor-
gebracht, und es entstehen auf diese Weise willkurliche Ausdrucksformen;" and also,
S. 452: "Auch die Sprache ist in gewissem Sinn eine form der Geberde. Sie entwickelt
sich wahrscheinlich teils als affectartige, teils als nachahmende Bewegung."
43 6 P. F. SWINDLE
merely occurs simultaneously with one or more #'s and does
not even alter the tempo of the series as does that movement
responsible for the p. When the m is uttered the ^-series
continues as a nasal sound, which does not occur with the
cockatoo.
Other prevalent compounds spoken by the child are:
m
aaaapaaakeeeee, aaaaaaaameeeee, aaapaapeeee, aadaaadeeee,
aaaabaabeee, aaaaapaaapuuus, and the like. And in reading,
the child says "Whaat iiis thiis? liit iiis aaaaa hooorse
aaaand aaaaa cooolt. Caaaan youuu riiide theeee cooolt?
Oooh, no! liit iiis toooo Kittle." Stimulate the child by
telling it to read more rapidly, and it will very often pro-
nounce the words so indistinctly that they can be under-
stood only with difficulty. One child which was thus stimu-
lated seemed to read the words as follows: "Wat is sis?
It is a hog and a shoat. Can you wide de shoat? Oh, no!
It is too witte."
There is moreover a second variety of stuttering which
is quite closely related to stammering. An adult once asked:
"I, i, i, i, is ti, ti, ti, this the ti, ti, train for A, A, A, A, A, A, A,
Aurora?"1 If one tries to stutter in this manner in pro-
nouncing Aurora, one easily observes that these are not the
same, but qualitatively and quantitatively different A's
which succeed one another; each of these is in reality an
y^-group, seven of which are nonsense or superfluous ex-
pressions. The eight different A's may be indicated as
follows: A A, AAAAA, A, A A, AAAA, A AAAAA, A,
A A Aurora, or, as follows: 2-A-, $-A-, 1-A-, 1-A-, \-A-,
6-A-, i-A-, 3-^-group. Although each group has its accent
which is either a mouth movement or a movement of the
respiratory muscles, the final accent of the 3-^-group is
most easily observed. The speech compound represented by
the letters ro was not associated with the 2-, 5-, I-, 4-, or 6-,
but with the final element of the 3-^-group, and other
^-groups merely preceded this one. The expectation arising
from this view that the individual would occasionally not
1 As the word Aurora was spoken the u was silent, and the initial A was pro-
nounced in the same way as the a of the alphabet.
SPEECH INSTINCTS AND HABITS 437
stutter in the indicated fashion in speaking Aurora, is realized:
he does not stutter thus when the 3-^-group is the first one
of the possible ^-groups to occur.
The same fundamental statements can be made con-
cerning the ti's that were uttered before the word this was
pronounced. In this case the mouth movements which
resulted in modifying the initial element of each of the
^-groups was more pronounced than any of those which
accompanied the ^-groups. When the compounds are as
distinct and as explosive as these ta's, I shall call them stam-
mers. There is not the ^-compound in the word this, and
the stammerer who has an image of the word simply has to
wait until he is properly stimulated to utter the compound
which begins with the very peculiar mouth movement re-
sponsible for the th sound. He would have no trouble in
speaking the word if he would in the attempt either sneeze
or imitate a sneeze. All of the various compounds which
appear as stammers, involve at least one vocal response and
at least one distinct mouth movement. The failure to recog-
nize the fact that even the simplest stammers involve in each
case at least one form of a vowel group which is modified by a
mouth movement, has led investigators of speech to very
serious errors.1
Many stammers are conventionalized compounds which
either merely occur at inappropriate times, or which are
unnecessarily repeated. They are accordingly not only
1The following quotation taken from C. S. Bluemel's 'Stammering and Cognate
Speech Defects,' L, p. 187, is one of many cases which I could choose to illustrate this
point. " The stammerer's difficulty is transient auditory amnesia: he is unable to recall
the sound image of the vowel that he wishes to enunciate. This then is the thesis of
the present monograph. . . . His futile struggles with the initial consonant are
directed solely by his kinsesthetic imagery, but he cannot pass to the vowel because
he cannot recall its sound, its peculiar or characteristic quality — in short, the vowel-
color. When he attempts to speak the word ten, he produces the t entirely by feeling;
but he cannot mentally hear the sound e, and is hence unable to proceed." My criti-
cism of the thesis of the monograph is that the stammerer utters the vowel whether
or not he recalls its sound; for, in his attempt to pronounce the word ten, his stutters
are generally a number of ^-compounds. The t which the author has written, is
only a visual stimulus symbolizing the vowel which is modified by a mouth move-
ment. The compound te can be spoken, as the stutterer does, but we cannot utter
a t apart from the e, e, or some other vocal sound.
43 8 P- F> SWINDLE
often superfluous, but even harmful in speech, in so far as
they distort the compounds which the stammerer is at-
tempting to use to call forth predictable responses in other
individuals. Take, for example, the compound 'you know.'
When one says, "Do you know the name of that mountain?"
or, "You know what I mean," there is no violation of a
conventionality of speech; but, when one says, "Did you
ever, you know, smoke a cigar in the wind?" the superfluous
'you know' is a non-conventionalized or stammering expres-
sion. Other common superfluous utterances are 'Well,'
'Oh,' 'Now look,' 'Now listen,' 'Now,' 'Don't you think?'
'Don't you know?' 'Isn't it?' 'Yes sir!' and so forth. We
often hear such meaningless expressions before or after a
sentence, or indeed often after mere phrases of it.
The simple fact that a speech compound is at one time a
nonsense and at another time a sense expression, means that
our training methods are not adequate to enable us to use
all the speech manifestations of the organism in all the
combinations they so frequently help form. It is important
that any nonsense expression can become conventionalized
by training the individuals of the social group to respond in a
definite way when stimulated by the expression. Our con-
ventional speech is accordingly reduced to a form of stuttering
and stammering, the stammers being the more pronounced
elements. Some stutters and stammers are common to a
greater number of individuals than are others and are conse-
quently more highly conventionalized. Generally, only the
more idiosyncratic utterances are superfluous in ordinary
speech.
A bald-eyed cockatoo was taught to stutter. It was taught
to repeat after me the word 'Kakadu,' the German for
cockatoo; but it did not say 'Kakadu' immediately: first it
said 'What,' 'Jako,' 'Kak,' 'Kak,' and then 'Kakadu.'
While the bird frequently uttered various compounds before
saying 'Kakadu,' the training to which I subjected it estab-
lished this particular order. It was also trained to stammer
in certain very definite, but silent ways after the word
'Kakadu' was spoken. My primary interest was in the
SPEECH INSTINCTS AND HABITS 439
speaking of the word 'Kakadu,' and I accordingly called all
those compounds which were uttered or produced in any
other way before and after this word, stammers. For a
second observer the word 'Kakadu' might have been only
one of the stammers which regularly occurred after the word
'Jako.' A wild cockatoo may utter a series of compounds
which become established through self-training; but any one
of these may be of the same interest to us as any other one.
If we should become able, by training ourselves for the
purpose, to respond in a definite way, or, in other words,
to attribute a definite meaning to the whole of such a series of
compounds, we should be inclined to consider no element of
it a stammer. Also, if the cockatoo should be taught a con-
ventional sentence with a subject, a predicate, and so forth,
tradition would scarcely permit us to speak of stammering
activities, in spite of the fact that the cockatoo may not
have been taught to use the sentence in the conventional way.
It is a noteworthy fact that well-established, convention-
alized sentences of the human being become frequently dis-
torted with stutters and stammers. It is peculiar, but
interesting, that students of ancient languages, who use
their knowledge of these to speak a more correct English, do
not as a rule speak fluently; they often pause in the middle
of a sentence and insert a few stammering expressions before
going further. These stammers are activities which inter-
rupt the established series of conventionalized compounds
and become associated with some of their elements just as
the mouth activity of the barbet interrupted a long series of
innately associated a's and became associated with the
fourteenth or final a of the 14-^-group. Those Greek and
Latin scholars who, according to their introspections, do not
attempt to use their knowledge of these languages while
speaking their mother tongue, may speak as fluently as other
people. The fact that our Greek and Latin scholars may and
usually do speak, provided we neglect their non-convention-
alized stammers, a more grammatical English, and also that
they may possess an extraordinarily large and varied vocabu-
lary, has nothing to do with our present problem. A student
44° P. F. SWINDLE
of logic may make fewer fallacies in his reasoning; that is,
his- responses may conform more with convention, by having
studied logic; but a logician who attempts to apply his
principles of logic and to speak fluently at the same time will
acquire habits of stammering. Stimuli which may cause
a person to stammer thus, may have many sources.
A non-excited person may say Aurora, but when in a
state of excitement he may say A, A, A, A, A, A, A, Aurora.
These A's represent eight qualitatively different ^-groups
which became isolated from different parts of an ^-series
which can be represented as follows with the potentially
adequate stimulus for each A underneath it: A A A A A A
a1 a2 a3 a4 a5 a6
AAAAAAAAAAAAAAAAAA. When
a7 a8 a9 a10 a11 a12 a13 a14 a15 a16 a17 a18 a19 a20 a21 a22 a23 a24
the stimulus a1 is presented the first A is conditioned by it,
the second A is conditioned by the first, the third by the
second, and so on. If a1 should be removed as soon as the
first A is conditioned, the after-response would consist of
the remaining twenty-three A's. If a5 were the stimulus
presented, the fifth A would be conditioned by it, and the
after-response would consist of the last nineteen A's. There
are experimental and also theoretical reasons for asserting
that a5 can be presented repeatedly without bringing about a
dissociation of the fourth and fifth A's, but in this discussion
we do not need to consider why it should be the case. If the
24-^-series appears more frequently than any of its parts,
this means either that a1 is a stimulus which affects the
organism more frequently than does any one of the other
stimuli, or that the first A is conditioned not only by a1, but
also by one or more frequently occurring responses which
have become associated with and precede the first A. It
may well be the case that of the potentially adequate stimuli
here represented for the different A's, a1 is the only one
present in a given environment. If then, while the series
is in progress, a mouth movement should occur and isolate
the first three A's from the remainder of the series, the
3-y^-group only will occur as long as the person remains in
the given environment.
SPEECH INSTINCTS AND HABITS 44 l
If the individual should shift to a new environment and
become affected by a4, the 2i-^-series will occur and become
habitualized, because the initial A of this series will become
associated with and later be conditioned by the response
which happened to occur immediately before it. This acci-
dental response will now serve as a substitute for the ^-A-
group which became previously isolated from the original
series. We could proceed in the same way until we get
the series broken up into a number of fragments or unitary
groups each of which maintains its identity in many different
environments by occurring more frequently as a whole than
in parts. These unitary fragments of the original series
may be the 3-^-, 5-^-, i-A-, 2-A-, 4-^-, 6-A-, 2-A-, and
l-^-groups. If it should be such a compound activity as
rora (I omit the u because it is silent in the spoken word
Aurora), which isolated the 3-^-group from the original
series and became associated with the final element of this
group, the individual would be able to pronounce the word
Aurora only when the 3-^-group occurs. If it should happen
that while he has an image of the word he should be stimu-
lated to produce the 5-^-group, which is not associated with
the 3-y^-group, at least one more effort would have to be
made before the word could be pronounced.
Let us suppose that the individual happens to be in an
environment in which the stimulus for only the 3-^-group
is present. He would succeed at the first effort to pronounce
the word. If he should pass to a new environment which
contains the stimuli for all the groups, he might by chance
have to make many efforts. Now it does not matter whether
he actually passes into a new environment or whether these
stimuli are introduced into the old one, leaving it otherwise
unchanged. It would be the same if they should be secreted
by his own body. Perhaps they are under certain conditions
secreted by some of the ductless glands of his body. Perhaps
some of the hormones secreted under conditions of excitement
are adequate stimuli to condition a number of the super-
fluous ^-groups while he has an image of the word Aurora.
It is thinkable that the hormone secretion could be of such a
443 P- f. SWINDLE
nature that if it should not call forth all the ^-groups simul-
taneously, it might cause them to occur in very rapid suc-
cession. In the first case a 'boisterous' A sound of relatively
short duration would result, and, if the final accents of the
superfluous ^-groups should be relatively insignificant for
the hearing organism, it might seem in the second case as
if one long A were being uttered before the word Aurora is
pronounced. A hormone which is, under certain conditions
of stimulation, thrown into the blood, may be analogous to
a response which a training in logic, Greek, or Latin has
caused to occur quite frequently and necessarily interrupt at
times either an instinctive vocal series or an habitualized
series of speech compounds.
Out of the previous discussions arises the practical problem
as to the surest and most economical means of breaking an
individual of stuttering or stammering. In order to solve
this problem we should become thoroughly acquainted with
the process of teaching a subject to stammer. I noticed
that when I stepped noiselessly to the bald-eyed cockatoo's
room and, before opening the door, said 'Kakadu,' the bird
said 'What,' and when I said 'What,' the bird said 'Kakadu.'
When I rattled the door knob or opened the door without
showing myself, the cockatoo said 'Jako,' when I jerked its
cage with a string it said 'Kak,' and when I suddenly ap-
peared in the room before the bird, it said 'Kakadu,' after
which it began beating a horizontal bar of the cage with its
beak, becoming then active in a number of other ways.
In the course of time I found many stimuli which called forth
the mentioned responses; but I made use only of the par-
ticular ones here given. Before the bird was aware of my
presence outside its door I called out 'Kakadu!,' opened the
door without showing myself, jerked the cage twice in suc-
cession, and then entered the room suddenly. After each
stimulus the bird made a verbal response, the entire series
of which was 'What, Jako, Kak, Kak, Kakadu,' after which
it beat the horizontal bar with its beak. It was necessary
for me to carry out this program several times each day for
about three weeks before the bird pronounced the series
SPEECH INSTINCTS AND HABITS 443
regularly without error when I spoke the word 'Kakadu.'
I discovered that after the series was well established I no
longer had to go through with the ordeal of stimulating the
bird by calling out the word 'Kakadu' while still outside the
door; I could remain in the bird's room for some time and
then utter the word 'Kakadu,' and the series would be spoken
just as perfectly as when I called through the door.
At the end of two months I employed the following means
to break up completely this series of associated speech com-
pounds. I said 'Kakadu' repeatedly and as rapidly as I
could for several minutes during each period, and after about
three days this stimulus no longer served to instigate the
series. However, when I rattled the door knob, the cockatoo
omitted 'What' and began the series with 'Jako.' I then
presented this stimulus repeatedly without long pauses until
it likewise became ineffective. As soon as this task was
accomplished, I repeatedly jerked the cage with the string
until this stimulus was no longer effective. Of the three
tasks, the last one was the most difficult to accomplish.
When the series was thus completely broken up and I spoke
the word 'Kakadu,' the bird generally said something, but
not always 'What,' as it did before the training, and it said
'Kakadu' no more frequently than it did 'Jako,' 'Kak,'
'Ohoh,' 'Adieu,' or almost any of the many other sense or
nonsense words which it could utter. Now, when it said
anything other than 'Kakadu,' I repeated the stimulus for
some time, and as soon as the bird said 'Kakadu,' I made a
long pause before restimulating it in the same way. This
procedure was necessary in order to make the cockatoo act
as if it were imitating me. Before I could bring about this
apparent imitation it was necessary for me to cause the
rather complex stimulus used to become ineffective for a
considerable number of stammers. I have applied this
method of dissociation with equal success in various other
fields of behavior.
If we now desire to apply the method to the human
stutterer or stammerer, we should proceed somewhat as
follows. Paradoxical as it may sound, we should present in
444 P- F- SWINDLE
rapid succession the adequate stimuli for the undesired
activities. If we cannot find or cannot control the most
adequate ones for these responses, we should use those less
adequate ones which we do have at our command. For
example, when a person, attempting to speak a word, stutters
or stammers, we should simply stimulate him by telling him
to do that same thing again and again, perhaps a hundred
or a thousand times in rapid succession. But care must be
taken that he does the same thing and not something which
may be only very similar to the particular undesired act.
After he has repeated the act a great number of times, he
should then be stimulated to pronounce the word which was
previously preceded by the undesired activity. It is essential
that he should make a pause of at least a minute each time
before he repeats the word; otherwise, this word, too, will
appear less frequently and will not always be recalled at the
significant places for it in ordinary speech. This process of
dissociation will not effect an absolute forgetting of the
undesired responses, but will cause them to occur less fre-
quently at the inappropriate places.
It is to be emphasized that this dissociation does not mean
an absolute forgetting of any activity. In the case of the
cockatoo, for instance, each of the relatively forgotten or
dissociated elements of the series of speech compounds, was
later spoken at times just as perfectly as previously. Ex-
cepting the word 'Kakadu,' each of the compounds occurred
less frequently than before the dissociations were accom-
plished. This relative forgetting brought about by con-
tinued stimulation is sometimes spoken of as 'forgetting
with experience.'1 We have no reason for supposing that
any of the organic structures involved in the so-called for-
gotten responses became annihilated. Owing to the fact
that the recuperation pauses were made as short as possible,
certain structures involved became at least temporarily
simplified to such an extent that they later functioned only
upon the presentation of very particular, and, naturally, a
smaller number of stimuli; the structures became 'immuned'
1 See for example, H. Pieron, 'L'oubli chez la Limnee,' Archives de Psychologie,
No. 33, Tome IX., 1909.
SPEECH INSTINCTS AND HABITS 445
to certain stimuli. The simple fact that a person can stutter
and stammer when properly stimulated to do so is significant
in this connection. The fact that an expert rope walker
returns to an amateur level of performance the very moment
he, as we ordinarily say, 'gives serious thought to what he
is doing/ is a phenomenon of the same category. The
simplification of certain of the organic structures involved
in a given, desired response to bring about its dissociation
from other undesired responses, is, I believe, the fundamental
principle underlying the facts of dissociation or relative
forgetting, the automaticity of certain responses, and a few
other related phenomena; but it would not be appropriate to
attempt the explanation in this paper. I have merely
inserted this view as a suggestion that my method of breaking
up such undesired activities as those of stuttering or stammer-
ing can be explained without assuming the annihilation or
permanent exhaustion of any of the structures involved.
CONCLUSION
Speech instincts and habits are the perceptible responses
manifested by an individual which serve primarily as auditory
and visual stimuli to call forth responses in other individuals.
These stimuli are vocal utterances, mouth movements which
modify audibly the vocal responses, and the larger bodily
movements which affect primarily the visual sense.
A vocal instinct is a series of innately associated elements
which presents a gradual, qualitative change from the initial
to the final element. Any vocal instinct may be called a
vowel series. There exist besides the a-, e-, i-, o-, and u-
series, a number of vocal instincts which may be indicated
by appropriate diacritical markings of the vowels of the
alphabet. But a marking which designates the sound of
the first few elements of a series is often quite inappropriate
for the last few elements of the same series. For this reason,
diacritical markings should also be employed to symbolize,
e. g., the different a sounds which can frequently be detected
between the initial and final elements of a given ^-series.
Theoretically, if the anatomy were such that an a-series
446 P- F- SWINDLE
would not be prematurely interrupted by exhalations, there
should occur within this series all the a sounds which the
organism can utter. The gradual, qualitative changes of the
fl's within the longest <2-series of the barbet support this
supposition.
A vocal response is a habit, as well as an instinct, when
it is caused to occur more frequently than originally. The
process of habitualizing a vocal response is a process of
training which causes the instinct to become the final accent
or conclusion of one or more responses which originally did
not regularly precede and did not condition the initial ele-
ment of the vocal instinct. The whole or any fragment of
an instinct is not necessarily a habit, but a habit is invariably
an instinct.
Vocal responses are essential for audible speech and are
accordingly speech instincts and habits.
Mouth movements serve to interrupt vocal instincts and
habits and isolate fragments or unitary groups of elements
from them. A given mouth movement which isolates a
unitary group from a longer series of a's becomes at the same
time associated with the final a of the isolated group and is
later conditioned by it. This instinctive mouth movement
which becomes the final accent or conclusion of the isolated
fl-group, is thus caused to occur more frequently than origin-
ally and is accordingly a habit as well as an instinct.
There are many mouth movements of such slight ampli-
tude and of such brief duration that they may not completely
interrupt a vocal response, but merely modify it slightly.
Such mouth movements may only modify slightly the tempo
of certain elements of a vocal series and become inserted
between two of them, as in aaaaaaapaaaaa, or they may
occur simultaneously with one or more of the elements, as in
m
aaaaaaaaaaaaaaa.
Mouth movements are as essential for audible speech
compounds as are the vocal responses and should likewise be
reckoned among the speech instincts and habits.
Larger movements of the hands, feet, head, and body in
general, serve to carry the organism through space and cause
SPEECH INSTINCTS AND HABITS 447
it to become affected by a variety of stimuli which may not
occur in any given environment. Many of these stimuli
call forth mouth movements and vocal responses which be-
come the final accents or conclusions of the movements of
transportation and are later conditioned by them. Further,
the less perceptible movements of the eyes serve the same
purpose.
All of the bodily movements which cause mouth move-
ments and vocal responses to occur more frequently, should
also be called speech instincts and habits.
Any instinctive or habitual response of an organism is a
means of social communication if it is used to stimulate
individuals of known training to behave in a way which can
be predicted from the nature of the stimulus. In discussing
speech instincts and habits, it was necessary to discuss some
problems of stuttering and stammering, because our con-
ventional speech is only a form of behavior of which stutters^,
stammers, and larger bodily movements are the elements^
When these elements of speech are conventionalized, tLey
are ordinarily called letters, monosyllables, words, sentences,
and gestures. A series of innately associated stutters which
is modified by one or more mouth movements to produce
such a speech compound as a consonant or a monosyllable,
is a stammer. A number of like or unlike stammers may be
combined through association to form larger speech com-
pounds, such as polysyllabic words and series of words; but
these are merely stammers of a higher order. The process of
conventionalizing a stutter, a stammer, a number of associ-
ated stammers, or any other simple or compound response,
is a process of training the individuals who are to utilize
these to call forth predictable responses in one another.
Any artificially combined series of speech compounds can
be interrupted in the same way as can an innately associated
series of simple elements; and a response which either com-
pletely interrupts a conventional series or merely becomes
inserted between two of its compound elements, is analogous
to a mouth movement which either completely interrupts or
merely changes slightly the tempo of movement of a vocal
448 P- F- SWINDLE
instinct. Those responses which occur most frequently are
most likely to interrupt the conventional series. A training
in ancient languages or in logic may cause them to occur
frequently. A hormone secretion may affect the speaking
organism in a similar way.
It seemed appropriate to present a method to diminish
the frequency of superfluous utterances, which are undesired
stutters and stammers. This method consists in presenting
appropriate stimuli to cause the patient to produce a given
nonsense or superfluous expression repeatedly without the
intervention of long pauses until at least a large number of
the stimuli for the superfluous act cease to call it forth.
This results not in an absolute, but only in a relative for-
getting of the undesired response. It was finally suggested
that the fact of relative forgetting of the superfluous response
is due to a dissociation of certain organic structures involved;
that this dissociation is in turn a result of a simplification
or purification of the structures; and that these phenomena
can be consequently explained without assuming the anni-
hilation or even the permanent exhaustion of any of the
structures which function to produce the nonsense expression.
A PRELIMINARY REPORT ON 'WORK WITH
KNOWLEDGE VERSUS WORK WITHOUT
KNOWLEDGE OF RESULTS'
BY GEORGE F. ARPS
Ohio State University
The primary purpose of this study is to secure a quanti-
tative statement of the influence of awareness as a factor in
work.1 To this end three sets of experiments were conducted
by means of the Bergstrom ergograph. In each of the sets
the observer at one time remains in comparative ignorance
(complete so far as it is possible to make it so) of the amount
and character of the work he is doing; at another time he is
given every opportunity to observe his work as it proceeds
and to study the results ad libitum. Each set of experiments
is therefore divided into two series, designated 'unknown'
and 'known' respectively. A series is made up of work
periods.
The mental complexes, antecedent to the work response
in the various series of experiments, are sufficiently diverse
to prompt an inquiry concerning the meaning of this diversity
in terms of work units.
Three observers, M, / and W, functioned in the experi-
ments; in the case of M, experimentation continued over
the greater part of three years. The method common to
ergographic work prevailed. In our case the hand and arm
of the observer were strapped into the machine in such a way
as to secure maximum freedom for the middle finger (the
work-finger of our experiments) of the right hand.
The instructions given each observer were simple. In the
first work period he was instructed to pull the ergographic
load on every beat of the metronome, sixty to the minute.
This load was 4 kilograms for the first two sets of experiments
1 The study will in no wise concern itself with the theoretical aspects of the con-
troversial question of the relation of the psychical to the physical.
449
45° GEORGE F. ARPS
and 6 kilograms for the third set. The observer was re-
peatedly instructed to make each lift represent his maximal
pull and to continue the lifting until he was no longer able
to budge the load. The work period closed when the ob-
server failed to move the load in two successive attempts.
In the second work period, forty-eight hours later, the ob-
server was instructed to lift the load for ten successive beats
of the metronome and then to rest for one second (one beat)
then to lift again for ten beats followed by another rest of
one second. This process was repeated until the observer,
as in the first work period, was unable to budge the load.
The third work period of this series differed from that of the
second in that the rest period was increased by one second
(one beat). In each successive work period of this series
the rest increased by one second. There are eleven work
periods in each series. This procedure prevailed as well for
the work periods in which the observer was relatively un-
aware of his accomplishments as for those periods during
which he was fully aware. The procedure for any ascending
series may be algebraically represented by the following
formula:
/F+£o' W+RS W + R^ W + Rlo'
in which P represents ten ergographic lifts; W + R^ a work
period without rest; W + jRi, one second rest, etc., until the
number of lifts and the number of rests are equal.
A 'set' of experiments is made up of (a) an ascending
series of work periods in which the periods differ increasingly
by one second of rest up to ten as a maximum, and (b) a
descending series in which the periods differ decreasingly by
one second until there are no rests. Graphically they may
be represented as follows:
-> Ascending Series, Known.
Rests = o, i, 2, 3, 4, 5, 6, 7, 8, 9, 10.
-> Ascending Series, Unknown.
— > Descending Series, Unknown.
Rests = 10, 9, 8, 7, 6, 5, 4, 3, 2, i, o.
-> Descending Series, Known.
WORK WITH AND WITHOUT KNOWLEDGE OF RESULTS 45 l
Care was taken to distribute as equitably as possible the
effect of practice. The disadvantage, for example, accruing
to the known in the ascending series of the first set was
practically neutralized by beginning the second set of experi-
ments with an unknown ascending series. Other advantages
and disadvantages are similarly neutralized, as, for example,
the disadvantage of the so-called phenomenon of ' breaks'
which occurs between the conclusion of certain of the series
and the beginning of certain other series. It will be seen
from the above graphically represented series that the con-
cluding work period of the ascending known series contains
an equal number of ergographic pulls and rests. The be-
ginning work period of the immediately succeeding series
contains no rests. This transition from a work period con-
taining ten rests in each of its subdivisions to a period con-
taining no rests is termed a * break.' Obviously it is im-
portant to distribute such breaks equally to the known and
unknown series.
In the three sets of experiments there are six unknown
and six known series, which together constitute 112 work
periods. In the third set of experiments the number of work
periods in each series is six. Each of these periods has an
absolute and a unit efficiency value. Both values are ex-
plained under the tables given below.
Tables I. and II. set forth the two values of the twenty-
two work periods of two of the series and other data which
may serve in presenting a survey of this study.
Some of the important results are as follows :
1. The final average of the absolute efficiency values of
all the known series of the first and second sets of experi-
ments exceeds that of the unknown by .2 per cent.
2. The final average of the unit efficiency values of all
the known series, excepting the third set of experiments,
excels that of the unknown by 7 per cent. In the known
work periods the observers work at a higher rate of speed.
3. The average absolute efficiency value of all the known
series exceeds that of the unknown by 10 per cent.
4. The average absolute efficiency value of the known
45 2
GEORGE F. ARPS
series of the first and second set of experiments exceeds that
of the corresponding unknown by 10 per cent.
5. The average absolute efficiency value of the known
series of the third set of experiments exceeds that of the
corresponding unknown by 10 per cent. (The ergograph
load in this set is 6 kilograms.)
TABLES I. AND II.
FIRST ASCENDING KNOWN. LOAD 4 KILOGRAMS (OBSERVER M)
c
III
C+III
B
0
i
0
<n
7
a*
S
8
</>
a
§
J
z>
w
w
~c
w
MWJ
WP
C+III
I
i
57
o
0
0
I
57
1.665
6.66
.0569
.0569
•0595
•0595
12
i
53
i
0
ii
2
4
2.073
8.29
.0733
.0669
.0805
.0737
II
i
46
2
o
20
2
6
2.330
9-32
.0879
.0740
.0878
•0735
12
i
56
3
o
33
2
29
2.276
9.10
.0785
.0610
.0854
•0338
21
3
37
4
I
24
5
I
4.877
19-51
.0899
.0649
.0828
.0605
24
3
59
5
I
55
5
54
5-710
22.84
•0954
.0646
.0883
.0614
23
3
49
6
2
12
6
i
5-447
21.79
.0949
.0603
.0859
•0549
32
5
16
7
3
37
8
53
6.144
24-57
.0776
.0461
.0789
.0481
32
c
27
8
4
8
9
3<5
8.482
33-93
.1008
.0589
.0908
.0627
54
8
59
9
7
57
16
s6
13.921
55-68
.1003
.0548
.0917
•0497
55
9
17
10
9
o
18
17
13-693
54-77
.0985
.0500
.0917
.0465
FIRST ASCENDING UNKNOWN. LOAD 4 KILOGRAMS
I
I
20
0
0
o
I
20
2.162
8.65
.1008
.1008
.0705
.0725
II
I
54
I
0
IO
2
4
2.642
10-57
.0926
.0851
.0893
.0788
13
2
18
2
o
24
2
42
2.905
11.62
.0843
.0717
.0876
•0793
15
2
25
3
o
42
3
7
3.101
12.40
.0856
.0664
.0747
•0595
21
3
35
4
i
20
4
55
4.382
17-53
.0816
•0593
.0707
•0579
33
5
26
5
2
40
8
6
5-734
22.94
.0703
.0471
.0714
.0479
21
3
35
6
2
0
5
35
4-339
17.36
.0806
.0518
.0708
•0454
40
6
37
7
4
33
ii
10
6.746
26.98
.0679
.0402
.0677
.0407
64
10
40
8
8
24
19
4
9.781
39-12
.0612
.0342
.0624
•0350
101
16
52
9
IS
o
3i
52
19.528
78.11
.0771
.0408
.0658
•0350
90
15
o
10
15
o
30
o
17-584
70.34
.0781
.0407
.0692
•0354
B
C
/
Ill
C + III
D
W
W
, TTT
Cx -p 111
MWJ
WP
Number of subdivisions of work periods.
Total time at work.
Length of rest period in seconds and number of the period.
Total rest time.
Total time of work and rest.
Distance load travels in meters.
Work done in kilogram meters.
Unit value per second of work done each period.
Rate of work including the rest time.
Average unit value (WIC} of all observers.
Average rate of work per second for each period including the rest time.
WORK WITH AND WITHOUT KNOWLEDGE OF RESULTS 453
6. The average unit value of the known series of all sets
of experiments shows a higher rate of work by 18 per cent,
over the corresponding value of the unknown series.
7. The average unit value of the known series of the
first and second sets of experiments shows a higher rate of
work (speed) by 5 per cent, over its corresponding value of
the unknown series.
8. The average unit value of the known series of the
third set of experiments shows a higher rate of work by 35
per cent, over the corresponding value of the unknown series.
\ /
\T\
C'mt nts-
Rest Periods .
INVOLUNTARY SUBSTITUTION OF IMAGERY FOR PERCEPTUAL
CONTENTS
The curious appearance of imagery during the work
periods of certain of the unknown series is an interesting
phenomenon in that it clearly indicates the observer's ten-
dency to avoid working blindly. The point of definite appear-
ance of imagery may be seen by reference to the curve given
above.
The relation of the unit values of the known and unknown
second ascending series of the second set of experiments is
454 GEORGE F. ARPS
especially interesting in that the fact of overlapping of the
curves is referred to the apparent functioning of imagination.
From the introspections it appears that the condition of
'ignorance of results' prevailed more completely in the first
set of experiments than in the second set.
The unknown series of work periods of the second set of
experiments approach that of the known through the em-
ployment, without set purpose on the part of the observer,
of visual and kinesthetic imagery reproductive of the per-
ceptual and kinesthetic experiences of the work periods of
the preceding known series. In the third work period of the
second set of experiments (see curve), the following intro-
spections are recorded: "The task of this period proceeded
with a fair degree of definiteness, with a comfortable degree
of orientation hardly comparable to any of the preceding
unknown periods. Certain individual lifts I pictured vividly;
in certain other cases I compared successive lifts. The com-
parisons were especially pronounced when the first evidence
of fatigue appeared." In the sixth period of the same series
the observer remarked that he "seemed to sense the efficiency
of the entire period in perspective more or less tangibly."
Imagery is also involved in the more or less abrupt closing
of the unknown period as contrasted with the gradual taper-
ing off closing of the known periods. On this point one of
the observers remarks as follows: "In closing an unknown
period I seem to let down suddenly in spite of all efforts to
avoid it. Short lifts have little meaning in that I fail to
image the pen marks which support me in the long lifts."
In all these cases evidently imagery of the sort here de-
scribed acts as an incentive to work. The presence of
imagery was first detected in the first work period of the
second unknown ascending series of the second set of experi-
ments. It is probable that the overlapping of the curves
of this set of experiments finds an explanation in the definite
presence of imagery. For, as the imagery content of con-
sciousness peculiar to the work periods of the unknown series
approximates that of the perceptual content characteristic
of the work periods of the known series, the differences in
WORK WITH AND WITHOUT KNOWLEDGE OF RESULTS 455
the awareness of results diminish. It is probable that the
more essential features of the perceptual experiences acquired
in the first set of experiments function in imagination in the
work periods of the unknown series of the second set of
experiments. If the mental complex operative in the second
set of experiments forms a close resemblance to that complex
functioning in the first set, and, if the efficiency differences of
the known and unknown work periods are primarily due
to the degree of * knowledge of results,' then we have an
explanation for the crossing back and forth of those parts
of the curve representing the unit values of the work periods
of the second set of experiments. This means that we have
here phases of the process of habituation in which the estab-
lished habits of response under perceptual conditions tend to
subordinate responses operating under imaginal conditions.
That the organism tends to gravitate towards its more ac-
customed channels of behavior is clearly indicated in succeed-
ing work periods. Each succeeding set of experiments shows
an increasing amount of crossing of the ' unknown ' and
'known' curves (see graph).
From the general character of the curves and the intro-
spective statements it appears that work carried on under
conditions of partial awareness of results loses in efficiency
and that such conditions are extremely difficult, if not im-
possible, to maintain when such work is followed or preceded
by, work of identical character under conditions of complete
awareness. Without set purpose the observers employ cer-
tain mental complexes during work under the former con-
ditions, which parallel in function certain essential features
of the mental complexes operative during work under the
latter conditions.
THE BEHAVIOR OF THE HUMAN INFANT
DURING THE FIRST THIRTY
DAYS OF LIFE1
BY MARGARET GRAY BLANTON
Madison, Wisconsin
In view of the fact that a large amount of experimental
work is being done by the psychological laboratory upon the
reflex and instinctive equipment of the babies in the maternity
ward of the Johns Hopkins Hospital, it seemed worth while
to make a preliminary observational study of them under
the conditions actually obtaining in the nursery. Most of
the studies now in the literature are of this character and
the present one has also the object of checking up others of
the same character. The present observations have been
made upon a very large number of infants and a thoroughly
objective viewpoint has been maintained throughout.
BUCCOPHARYNGEAL
I. Sneezing. — The earliest reflex noted. It was present
on one occasion while the infant was being lifted from the
mother. Subject L. sneezed before the birth cry appeared,
and Subject G. at 5 hours, from which time on it is common.
It was frequently noted on taking the babies into an over-
heated room where observations were made. It was not
noticeable on bringing them again into a cooler atmosphere.
Preyer2 says: "It demonstrated the existence of a very
firm connection, long hereditary, of the nasal branches of
the trigeminus with the motor expiratory nerves." He also
1This work was undertaken under the direction of Professor John B. Watson,
director of the psychological laboratory of Johns Hopkins University. Through the
courtesy of Dr. J. Whitridge Williams and Dr. Daniel Davis, the material in the
nursery of the maternity ward was placed at our disposal. I am also much indebted
to Miss Liphart and to Drs. McKee, Harris, and Sperry for many privileges on the
ward.
2 Preyer, 'The Mind of the Child,' p. 214.
456
BEHAVIOR OF THE HUMAN INFANT 457
quotes Darwin as saying that many healthy children at their
coming into the world do not cry, but sneeze.
2. Hiccoughing. — Subject Sp. at 6 hours hiccoughed for
three minutes. Subjects 7, 10, n, and 14 days old hic-
coughed after feeding. This list might be continued, for
over 50 cases were noted. It seems so well developed at 6
hours that there is little reason to believe that the possibility
of it does not exist at birth did conditions call it out. The
condition most commonly calling it out is a full stomach
(producing pressure on the diaphragm).
3. Yawning. — I have noted yawning twice within 5
minutes after birth and relatively few times during the first
month. Subject S. yawned at 5 hours, eyes tightly shut.
Subject F., ii days, yawned 6 times in I hour; her eyes were
squeezed tight and her chin trembled as she shut her mouth.
Subject R., 5 days, yawned so like a sigh that the movement
barely fell within the definition of a yawn. It was the only
occasion on which the remotest resemblance to a sigh was
observed.
4. The Beginning of Speech and Emotional Reactions. —
Crying, less than any other manifestation of a baby's exist-
ence, can be considered separately. During the birth proc-
esses the connection between the mother and infant are
made less efficient by the gradual displacement of the pla-
centa. The concentration of the blood increases, which
stimulates the respiratory centers and causes the intake of
air to the lungs. The birth cry takes place at this point.
Occasionally the nerve endings of the mucous membrane,
affected by the unaccustomed stimulation of the air, may
cause a light breath to be taken in and expelled in what we
know as a sneeze. In four observed cases in which hot and
cold plunges were necessary to stimulate breathing there was
also a sharp movement of the arms. The cry observed came
on the plunge into icy water and was sharp and short and
on an inspiration as in an adult under similar conditions.
The birth cry is rarely spontaneous, as the technique of the
delivery is to stimulate it at once by vigorous rubbing and
slapping on the back and buttocks. This cry, as I observed
458 MARGARET GRAY B LAN TON
it, differed in no way in timbre, pitch, etc., from other cries
of the first few days. The birth cries of different infants
were not alike, ranging from simple a (as in at) to u (as in
cut). Most of them were compound u (as in cut) followed
by wok (as in at); uh (cut); nga (at); and variations just as
in the later cries of infancy. This was true also of the three
spontaneous cries which I heard.1
Major2 says: "The beginnings of language are usually
traced to the reflex crying of the newly born babe." Both
Major and Preyer assert that "they are produced as well by a
child without a cerebrum as a child with one." I was not
present at the birth of Subject B., mentioned elsewhere, but
I was told by the surgeon that this infant did not cry out.
Only 4 cries were noted during its entire 10 days of life, all
of them taking place during a deep examination to determine
the extent of displacement of the bones of the skull. The
birth cry which is followed by an interval of quiet is rarely
followed by an interval of stillness.
Crying has been observed under the following conditions:
(i) hunger; (2) in response to noxious stimuli (including
rough handling, circumcision, lancing and care of boils, sores,
etc.), and (3) possibly from fatigue or lack of exercise. In
the subjects with which I worked I did not find the cries of
hunger, to noxious stimuli, to fatigue, and so forth, uniform.
There were differences of vowels and consonants, of timbre
and degree, but no one was used as response to one set of
circumstances that was not at some time used to others.
For example the cry in response to noxious stimuli was in
many cases an exaggeration of the cry usual when hunger
was present.
The cry of colic was the one exception. Starting abruptly
1 The dramatic element of the first cry has led to a most amusing amount of
poetic license. Kant's famous saying that "the outcry that is heard from a child
scarcely born has not the tone of lamentation but aroused wrath," has been subjected
to many and various comments. Preyer in quoting him says, "Kant wrote without
having himself observed children and animals just born." Dr. William A. White
quotes Adler as saying: "It is an expression of its overwhelming sense of inferiority
on thus suddenly being confronted by reality, without ever having had to deal with
its problems."
2 Major, 'First Steps in Mental Growth,' p. 282.
BEHAVIOR OF THE HUMAN INFANT 459
about 3 to 5 octaves above the adult female voice it slid
through a modified chromatic scale to within the range of
the middle octave. It was made with an accompanying
rigidity of the abdominal walls and thus of necessity varied
from the cry which included the activity of these muscles.
Subject L., 15 days, often started with a few short rapid
breaths, a few of which sounded like an * affected' cough and
ended with a two syllable cry, a (at), and yow (owl), long and
on a falling inflection. The whole is much like the word
meow with the m replaced by a. The cries did not vary so
much with the ages of the infants, during the first five or
six days, as with the weight. A baby of 2,300 grams at one
day, and 2,300 grams at 5 days will show little difference.
However, a baby of 4,000 grams first day and 4,000 grams
fifth day would show more difference, probably because
4,000 grams weight would indicate a better physical condi-
tion. In such a large group of babies any vowel or consonant
can at some time or other be heard on the first day of life
that will be heard on any other day during the first month.
A (what) will be heard in a first syllable, but I have not
heard it in a second syllable excepting in a series of short
grunts. This I suppose is because a maximum of activity
is reached on the second syllable, and the maximum activity
in the infant is accompanied with an open mouth and a con-
traction of those muscles which draw the inferior maxillary
inward and downward. This, it is seen, precludes the making
of the ah sound.
The crying of one baby can be distinguished with some
practice from the cries of another even in a nursery of 25,
the overtones varying just as in older people. Subject M.,
first day, u (cut), nah (at) accent on last syllable, u (cut),
wah (at), wuh (cut), ha (at). The * hunger cry' has generally
a well marked rhythm, the first syllable of preliminary sound
coming on the first part of the first beat, the second or ac-
cented syllable on the second part of the first beat and a
quick intake of breath as the third beat. This measure is
most often repeated in groups of 5 or 6, each slightly more
forceful than the preceding ones until the fourth or fifth,
460 MARGARET GRAY B LAN TON
the last one being softer. Thus also will the groups be
repeated. Each measure is also a trifle higher in pitch than
the one preceding.
Sounds Heard during the First 50 Days. — Consonant
sounds commonly heard are m in conjunction with a as ma
(at), n as nga (nat), g as in gah, h as in ha (at), w in wok (at),
r as in rah (at), r as in burr, very slight sound, and y as in
yah (at).
Vowel sounds are o as in owl, e as in feel, oo as in pool,
a as in an, and a as in father (relatively rare).
Of interest was the variety of animal cries simulated in
the nursery. The l pot-rack' of the quail, the cry of the
goat, the whine of the young pig, and the wail of the wild
cat, each had a close imitation.
PELVIC
5. Erection of Penis. — Subjects Le S. and R. had an
approximate erection of 50° at birth, Subject F. 80°, and
Subject Sch. at birth, a complete erection against the stomach
wall. These babies were, of course, lying on their backs.
Subject M. at 4 days, also lying on back, had complete
erection.
6. Voiding of Urine. — This occurs at birth in about 30
per cent, of the male infants observed.
7. Defecation. — This of course occurs often before birth,
as the condition of the amniotic fluid shows. It is also,
possibly, an occasional cause of suffocation, which would
seem to indicate that it occurs some time before birth^ In
Subjects S. and R. it was persistent and profuse immediately
after birth. The passage is of course the usual meconium.
If it. does not occur then the usual time is from 18 to 24
hours. No norms were kept, however. When the tempera-
ture thermometer is pressed into the anus there is very
often a resultant passage of feces. On five sets of observa-
tions the results were: (a) 7 positive, 8 negative, (b) 4 posi-
tive, 15 negative, (c) 13 positive, II negative, (d) 5 positive,
4 negative, (e) 8 positive, 16 negative, making out of 91
observations a total of 37 positive and 54 negative.
BEHAVIOR OF THE HUMAN INFANT 46*
OCULAR
8. Eye-movements. — While an inequality of eye-move-
ments is not uncommon it is not the rule. Subject J., 5,600
gr. birth weight, often had a hint of 'cross-eyes.' At 25 days
he occasionally had an unusual condition — his eyes being
entirely open he would drop the balls in such a way as to
show much more of the sclerotic coat above one pupil than
the other. In other subjects, 6, 7, 10, n and 15 days of
age, the sclerotic was noted as visible above the iris. Sub-
ject H., whose entire system of reflexes was markedly different
(the one"baby examined whose hands would not clasp on a
small round rod), had peculiarly symmetrical use of his eyes,
but his eyes were never seen to fixate the light nor to follow a
hand moved in front of him. Subject W.'s eyes worked inde-
pendently: starting to the left in unison the right eye would
move more rapidly than the left, and starting to the right
the left eye would move more rapidly than the right. A
large percentage of the babies fixated on the light at birth.
Subject S. continued to do so when removed to the side
of the room. When we turned her head away she turned
it back at once, bringing the light into her line of vision.
She would also fixate on the figure of a person who inter-
rupted the light. At 8 days of age she would gaze steadily
at the electric light in the nursery, rolling her eyes up, when a
change of position necessitated it, keeping the light in her
line of vision. An interesting 'by-product' was the per-
sistent paddling of her hands at such times and a recurrent
spasmodic smile. Subject M. at birth gazed without blink-
ing until her face was covered. Subject L., 34 days old,
held by mistake in bright sunlight, first shut her eyes, then
opened them, fixating directly on the sun. Her pupils con-
tracted to the size of pin heads. At 30 days of age she would,
when carried down the corridor, gaze at a high light on the
painted wall, arching her back and finally throwing back her
head as she was carried away. She would also, while holding
her head still, follow the figure of the nurse around the room.
At 32 days of age her eyes caught a red velvet bag held in
bright sunlight, and followed it, with her head held still,
462 MARGARET GRAY B LAN TON
around the nearly complete half circle of her vision. The
eyes of many of the infants followed a slowly moving hand a
few moments after birth. Subject S. would follow a hand
moved slowly 8 or 9 inches from her face without any accom-
panying movement of the head.
Subjects S., A., M., F., and J. gazed at the light above
the birth bed and also followed a moving hand. Subjects F.
and K., neither of whom gazed at light or followed hand at
birth, were seen to do both on the 8th day. Subject K. at
8 days, Subject R. at 10, and Subject L. at 26 days focused
first on one and then another face. Subject S., 13 days,
fixated a large paper bag which was being handled. A dim
light moved slowly at half a meter was followed by subjects
8 hours, 18 hours, 36 hours, and 3, 4, 5, 6, 14, 15, 21, and 30
days of age. Subjects which did not follow were aged 9 hours,
3, 5, and 14 days; 17 in all were tested.
Preyer observed in two cases, n and 13 days, a turning
of the eyes from one bright light to another. Miss Shinn's
niece1 fixated her eyes on a face first at 25 days. Preyer,
Rochlman, and Wilkowski observed 'real fixation' on the
tenth day. Dearborn2 notes seeing the eyes of a child 13
days old follow a slow-moving desk light at 18 inches, and
says had he the courage of his convictions he should say he
saw the baby's eyes follow a hand slowly on the first day.
Preyer's baby followed a desk light at one meter at 23 days,
much to his surprise, as he says "other children do not
follow a moving light until after many months." According
to him, however, Lacy saw this the second week and Darwin
the 45th day — but a majority of investigators say the 5th,
6th and 7th days.
9. Eyes during Nursing. — These babies, as a rule, had
their eyes open at the beginning of nursing. As they began
to get enough food the eyes would close. Some babies, how-
ever, were most active after nursing.
The following observations were made:
1 Shinn, Millicent, 'Notes on the Development of a Child,' Book No. I, p. 30.
2 Dearborn, G. V. N. ' Moto-Sensory Development.'
BEHAVIOR OF THE HUMAN INFANT 463
Sub ect O. K., 5 days old right eye open, left shut,
H. 3 " " eyes partially closed,
R. 7 " " near end of feeding, shut,
W. 7 " " near end of feeding, shut,
R. 14 " " both eyes open,
and subjects 4, 5, 5, 5, 9, 15, 20, 21, and 25 days old with eyes
open. Miss Shinn1 says: "I have watched and enquired
about several other babies and found none that nursed with
eyes open."
FACIAL
10. Tears. — On the testimony of Miss Liphart, Subject P.
cried tears at birth. Subject C., whom I saw, had tears in
her eyes at ten minutes after birth. As silver nitrate is
administered as a routine measure some time during the
first two hours, a reliable record of the activity of these
glands was not obtainable. Subject B., birth weight 5,600
gr., at 4 days cried profusely with right eye. He was an
overdue baby, breach delivery and with, consequently, no
head moulding. Subject B., an exceedingly cross-eyed
baby, cried tears on his 6th day. They were not observed in
him again, however, in the remaining 4 days he was in the
institution. Subject S. showed dampness in corner of eyes,
uniformly, on the I3th day, Subject L. on the I5th, and
crying with tears seemed well developed with her on the
34th day. The first average normal show of dampness
was 13 to 16 days. The first regular flow of tears 28 to 34.
11. Smiles. — Spontaneous smiles at an early age are rare.
Subject S., 4 days, smiled spontaneously while at the breast
after feeding; Subject O'K., 7 days, at end of bottle feeding;
Subject S., 13 days, while looking at bright light; Subject K.,
8 days, smiled and immediately regurgitated; Subject C.,
28 days, smiled repeatedly after feeding at the breast, but
the first smile had followed a light touch of the mother's
nipple on his cheek. Subjects I, 2, 4, and 6 days old smiled
slightly to a tickling touch under the chin when awake and
comfortable.
12. Facial Expression. — According to Preyer, the corners
of the mouth are not drawn down until the eighteenth week.
1 Shinn, op. cit., Book I, p. 12.
464 MARGARET GRAY B LAN TON
He quotes Darwin as noting it in the 6th week to the 2d and
3d month. Subject M., 10 minutes, pulled the corners of
her mouth down. Subject S., 30 minutes, rolled the lower
lip so as to show the inside surface, pulled the corners of
his mouth down, and after remaining in this typical pouting
attitude for a minute started crying. Subject J. did this
constantly as a preliminary to crying, sometimes accom-
panying it with a whimper-like grunt which made him seem
remarkably mature. He had also the square box-shape
mouth in crying, as had Subject L. on the 7th day and again
on the I4th, lyth, and 23d days.
Subject G., 6 days, had a horizontal line directly across
his nose from corner to corner of his eyes. In crying he
would pull down the inner corners of his brows making
wrinkles that radiated from the horizontal cross line. Sub-
ject C. at birth had the horizontal line between his eyes.
Horizontal lines across the forehead are not exceptional
when an infant looks up. It was seen in Subject M., 6 days,
Subject P., 15, 19, 21, and Subject L., yth day, and often
thereafter.
HEAD
13. Turning Head (when Lying on Face). — The ability to
turn the face (when the nose is firmly planted) in such a
manner as to get air is a disputed point. Preyer1 says:
"Newborn children cannot so much as free the face by turning
the head when one lays them on a pillow with face down-
ward." In making the experiment I used a pillow only once,
but as I had no way of ascertaining how hard the pillow was
which Preyer had used I made my tests on the lap of a nurse
and tallied each test on a hard table. In the one case of
the pillow a female child of 3 days freed her face at once and
held her head upright for several seconds. The tests made
on lap and table did not differ in results nor did those on
the birth bed. Subject C., 30 minutes old, rotated her head
in such a way as to leave her mouth and nose free. Subject
G., if hours, Subject B., ij hours, and Subject M., 9 hours,
could do equally well. On one occasion 15 healthy babies
1 'The Mind of the Child,' p. 266.
BEHAVIOR OF THE HUMAN INFANT 465
were tested and all were positive but one — Subject G., 8 days
old, who lay until turned. Subject L., 8 days old, not only
turned her head but raised it in the air also, entirely un-
supported, and held it so for 30 seconds by stop-watch,
when she was interrupted. At 15 days the same baby held
her head so for 5 minutes. This test was repeated time and
time again. Even the sickest babies could do it when the
air supply was cut off.
14. Holding Up the Head (when Held in Upright Posi-
tion).— Preyer1 says: "During the first ten weeks no trace
could be discovered, in the case of my boy, of an attempt to
hold the head in equilibrium." "In this important step is
expressed an unquestionable, vigorous act of will." There
has been some confusion among the different writers, due I
think to the fact .that the exact test conditions were not
given. In my tests the subjects were held on an open lap,
not against the body, were supported back and front for
two inches above the umbilicus. Subjects 2, 6, 7, 10, 10,
n, 13, 13, and 15 days old could support their heads for
times varying from I to 6 seconds. Subject L., 22 days,
held her head quite erect for 3 minutes, turning it from right
to left. Subject F. made numerous attempts with only the
shortest periods of success. Subject Y., 22 days old, badly
underfed and weak, made no apparent effort. Subject W.,
22 days old, child of retarded parents who is also weak and
undernourished, failed likewise. Any weakness in the neck
muscles must be in those which draw the head forward, as
numerous and successful attempts were made on the ability
to lift the head when laid on the face (see turning head, when
lying on face).
ARM AND HAND
15. Hand Movements at Birth. — Subject E. at birth spread
his fingers and closed his hand, repeating this 4 consecutive
times. Subject S. stretched his fingers with first joint of
first, second and third fingers bent. This was with both
hands at the same time.
1 6. Grasping Reflex. — Reflex closing of the hand to a
1 'The Mind of the Child,' p. 264.
466 MARGARET GRAY BLANTON
touch on the palm has been noted a very few moments after
birth as in the case of Subject C. who grasped the artery
clamps which held his umbilical cord. It was seen in Sub-
ject K.'s gripping of the doctor's finger to the degree of
slowing up the entire operation of caring for his umbilical
cord. Experimentally of course the first step was putting
a small rod into the hands of the infant to excite the reflex
and then to lift the rod so that the full weight would be
supported. My apparatus was a small skirt hanger with a
rod about a quarter of an inch in diameter bound with a very
thin wrapping of gauze. The prevailing custom of oiling a
baby at birth was counteracted by giving the hands of the
baby a good cleansing with alcohol. This reflex has been
described as one which is most pronounced at birth and
which diminishes rapidly afterwards. I have not found
this to be so. For whereas a baby inside of the first few hours
of life will cling I or 2 seconds, Subject Sm. on her 23d day,
weighing 4,260 grams, held on for 20 seconds and was then
laid down still gripping the rod. This reaction is very hard
to elicit from a quiet baby, very easy from a crying one,
.-and the crying when set up has the sound of what is called
in the nursery an * angry cry.'1 Subject A., 20 hours old,
clung 2 seconds and 10 seconds, Subject L., at 52 days,
supported herself with one hand on the rod for 42 seconds.
Abnormalities either mental or physical affect this reac-
tion less than any other. Subject B. clung tenaciously at
10 days even when life was practically extinct. He would
grip a rod put into his hand so strongly that it was necessary
to undo his fingers. Subject Q., 5 weeks old, dying of mal-
nutrition, lifted his entire weight and only relaxed on being
laid again on a solid support. An interesting by-product was
the climbing movement of the legs occasionally noted.
But these alternation movements of the legs are also the
movements of kicking.
17. Grasping during Nursing. — Major found in his child
that the finger clasp was firmest when the child was nursing.
1 Robinson says in "Darwinism and the Nursery" that at the beginning he found
that babies had grip and muscle power to sustain their own weight on a horizontal
bar, and he adds the comment that they even delight in the process.
BEHAVIOR OF THE HUMAN INFANT 467
My examination of 15 babies before, during the beginning of,
and after nursing did not seem conclusive. Subject W. gave
definitely harder pulls at the breast, according to the
mother, as the pull of my finger in his hand made him grip
firmer. One subject's grip was better before nursing; 2 were
the same all 3 times; 2 were the same before and during
nursing, and poorer after; I was negative before and after
and only very poor during nursing; I pulled only slightly
during and after but had about the same grip on both occa-
sions; i was definitely better during nursing and was absent
after nursing; 4 were best during nursing; and 3 were best
after nursing. No one baby was poorer during nursing
than before or after, but as will be noted in 3 cases it was no
better during nursing than at one of the other trials; and in
4 cases it was definitely poorer than on one of the other trials.
LEG AND FOOT
1 8. Creeping. — In one case (aged 7 days) there was a
creeping movement in which a backward movement of six
inches was accomplished in twenty minutes' time. By
marking the position of the head at the beginning and end
of the period the amount of movement was ascertained.
On each occasion the contracting and relaxing seemed mainly
of the abdominal and back muscles. The great hindrance
to successful movement in each trial seemed to be the insuf-
ficiency of the muscles of the arm used to draw the arm
from a position on a line parallel with the body to one at
right angles to it. The legs often at any time after birth
assume the crawling position. The head (see turning head)
can be held up out of the way, but the relaxed position and
relative weakness of the shoulder muscles make forward
progress impossible. This condition lasts beyond the 3Oth
day and certainly well into the 3d month, probably longer.
An interesting accompaniment of this creeping movement is
the opening of the mouth and the bumping of the head, set
up, it seems to me, by the well-organized face and mouth
reflexes. The accidental touching of cheek and chin to the
unyielding surface of the table and his inability to get any
468 MARGARET GRAY B LAN TON
part of it into his mouth, seemed elements in the arousal of
his crying.
19. Foot Movements and Reflexes. — Subject S., at birth,
held the foot rigid, raised all 5 toes uniformly and pulled
them down uniformly 4 times. The Babinski reflex was
found present in 10 cases tested, and apparently absent in
one. Subject S., when tickled on the center of her foot,
drew the two outside edges towards each other. The state-
ment is frequently made that babies can cling with their
feet. I have never seen this. Placing a fine wire under the
toes, a reflex movement downward caught the wire between
the pads of fat on the toes and the ball of the palm, but the
slightest pull would remove it.
20. Kicking and to a less extent moving of the arms is
almost continuous for from 15 to 30 minutes after delivery,
which period is commonly followed by a sleep of from 6 to
8 hours. The kicking was greatest on the side to which
there had been a rotation of the head, this rotation appar-
ently being decided by the position of the child in utero for
the last period before birth. For instance Subject M.,
whose head rotation was to the right contrary to the usual
rule, was a remarkably active child, and kicked excessively
with her right leg and waved her right arm. Her reflex to
the tying of the cord and to alcohol was however with the
left side of the body. After delivery it was noted that she
had a preferential side, which in this case was the left side,
and this preference for the left continued for the term of
observation. Her right-side activity immediately after birth
might be explained by the cramp of the right side in utero
and the relatively free position of the left.
GENERAL RESPONSES
21. Turning Over. — Subject M. at 10 minutes after birth
given a slight, unintentional advantage by the slanting of
the mattress on which she lay, turned from her face to her
back. Subject T., at 7 days, turned repeatedly from face to
back when not impeded by clothing. Placed face downward
on an unyielding surface her arms outstretched in line with
BEHAVIOR OF THE HUMAN INFANT 469
her body, she would immediately start crying. In crying,
relaxing and contracting of the legs, arms, abdomen, and
back muscles are natural accompaniments. Pulling her
knees under her and contracting her muscles generally, she
would, when relaxed, have her arms nearer by a fraction
of an inch. In 10 minutes, after 9 repetitions of this general
maneuver, her arms would be sufficiently near her side so
that with a final contraction she would roll over.
22. Reflex to Stimulation of Umbilical Cord. — There was
apparently a reflex to the clamping, cutting and tying of
the umbilical cord. There is no direct nervous connection
between this and the body. There are of course nerves in
the opening surrounding the cord at the end where it is
attached to the skin of the abdomen. The findings were
not uniform. I should say that the deliveries which I wit-
nessed were divided between three surgeons. Subject L.-i
straightened leg at clamping of cord with artery clamp.
Subject L.-2 (twin to L.-i) same; Subject S. straightened
leg to tying, not to clamping. Subject C. kicked at tying
of cord, no reflex to clamping; Subject B. kicked lustily at
tying at i| hours (I did not see clamping); Subject R., no
reaction to clamp, positive to tying; Subject R., positive to
tying; Subject S., positive to clamp only. Three babies
kicked vigorously when clamping and tying were done;
Subject F. and Subject M. were negative to both. The
only explanation was that in clamping and tying there was
some pulling and the reaction was from nerves lying next to
the umbilical opening. One of the surgeons kindly undertook
to hold the cord much laxer in all the operation than even
the technique called for. There was no reaction to clamping
but a distinct doubling up of the legs to the tight tying of
the tape.
23. To Dropping. — The best example of this reaction was
mentioned in the section on grasping. Watson and Morgan1
have described dropping as a stimulus to fear as follows:
"To suddenly remove from them all means of support, as
14 Emotional Reaction and Psychological Experimentation,' by John B. Watson
and J. J. B. Morgan, Amer. J. PsychoL, 1917.
47° MARGARET GRAY B LAN TON
when one drops them from the hand and allows them to be
caught by the assistant, the child being held over a bed on
which has been placed a feather pillow." Additional evi-
dence of reaction to dropping appeared in the course of this
study. On 3 occasions when the box in which newborn
babies lay was suddenly lowered there was a marked move-
ment of the arms and in one case brief holding of the breath.
This same thing occurs repeatedly when the infants are
being weighed. After the removal of a big weight the scale
pan in which the infant lies may drop suddenly; the result in
many cases was the same upward movement of the arms.
In this case the jar of the pan hitting the support is to be
taken into consideration.
24. Stretching. — This movement varies from the mere full
raising of the arms and a complete stretching of the legs
and toes, to arching of the back and abdomen and pushing
the arms until they trembled, accompanied by the bending
of the end flanges of the fingers. There was in one case a
movement of the neck and pulling forward of the shoulders.
The inferior maxillary was drawn inward and downward,
giving a most adult cast to the countenance. Infants stretch
with the greatest freedom at the removal of the clothing
and especially at the removal of the diapers now in use.
Subject L., 25 days, crying for a late feeding and very wet,
was hushed at once by being held to the fire with feet un-
covered. Here she stretched her legs into muscular combina-
tion after combination and finally juxtaposed the palms of
her feet and went to sleep. She accompanied every marked
change of position and tension of the muscles with an opening
of the mouth, and each relaxing with a closing of it. Sub-
ject S. stretched on being put with a bright light in her line
of vision. This was repeated 6 times, each time giving the
same result.
25. Response to Sound Stimuli. — The reaction to sounds im-
mediately after birth is unusual. Subject M., lying in a high-
walled tin box, on a metal table, resting on tiles set in cement,
reacted vigorously when one of the doctors in passing hit a
metal stool, a foot or so away, with the heavy door. The
BEHAVIOR OF THE HUMAN INFANT 47 l
baby was covered with a blanket. The stool did not hit the
table, and the only vibrations reaching him were through the
tiled floor or through the usual air vibrations. He threw
his arms forward forcibly enough to dislodge the covering,
at the same time moving his entire body. It seems reason-
able to conclude that his reaction was to sound. At two
days this baby jumped similarly when a metal tray was hit
with a fountain pen at a distance of one meter. The tray
was held stationary. Three other subjects the same number
of days old gave no response. Subject H., 6 days old, batted
his eyes when metal disk was struck with a hammer behind
his head. She responded thus the first two times, and the
third time she did not respond. Subject H., 6 days old, and
Subject T., 5 days old, were tested with tuning forks. The
forks were struck and the boxes held opposite the ear of each
baby when they were crying. There seemed to be a slight
diminution in the crying each time. Subject L., also, at 7
days, seemed to be quieted by the big forks. Subject L.,
20 days, and Subject H., 16 days, did not react to the Galton
whistle at any pitch. Subject P., 29 days old, turned her
face 5 times in the direction in which I was whistling softly —
about half a meter from her head. There were several ele-
ments to be considered, however, as I was sitting on the
side of her next an open fire, and, also, my breath may have
touched her face. Her general movements, her expression,
and the immediateness with which she turned each time
as soon as I began whistling, incline me to believe that there
was localization as well as auditory response. She was
also seen to turn three times in the direction of the basin
into which water had suddenly and forcibly been turned.
Subject S., 13 days, as mentioned under eye-movements,
turned at once in the direction of a heavy paper bag which
was being folded. Her face was away from the bag, turned
towards a mild fire, and her movement in turning was de-
cisive, her eyes catching the bag unhesitatingly and holding
it steadily. It was a very interesting example of the co-
ordination of audition and vision. Repeated experiments
showed that babies showing no reaction to other sounds would
472 MARGARET GRAY BLANTON
react positively to these rattling paper bags. In 13 trials
with subjects 4 to 14 days old, 12 were positive. Four were
hushed in crying, 4 turned in the direction of the bag, and 2
in the opposite direction, but since the movement of body
and eyes was markedly different to what it was when the
experiment started, those turning away were considered as
giving a positive reaction. Subject B., on the second day,
would jump at a whistle or the sudden sound of a voice on
the far side of the room. The sudden dropping of the scales,
6 feet away, would cause marked convulsive movements.
This condition persisted until about the 4th day, when his
response gradually became less marked, until on the 7th
he would not respond at all.
RESPONSE TO Noxious STIMULI
26. To Deep Pricking of Big Toe. — For laboratory pur-
poses it was necessary to draw blood from 3 babies at 2 times.
Two of the babies were pricked 2 times on the left big toe,
one once on the left and once on the right. At each time
the other foot went up at once with a pushing motion against
the other ankle. As this motion is also one of the motions
of kicking no conclusion could be drawn. There was no
pronounced cry but one of them was crying when taken up,
and all were given, between the two operations, a lifting test
which always elicits crying.
27. To Lancing of Infected Finger. — Subject Le S., crying
with hunger, had to undergo the lancing of a badly swollen
and infected finger. It was necessary also to hold the arms
firmly, which very uniformly has the effect of making the
subjects cry. His cry during the lancing and cleaning of
the finger continued to be of the same character, but much
exaggerated.
28. Circumcision. — Unfortunately the time selected for
circumcision was immediately before a feeding. The ele-
ment of hunger had therefore to be taken into consideration
as well as that of pain. A small gauze sponge dipped in
whiskey and sugar or wine and sugar was given the baby by
the Hebrew physician who performed the operation. The
BEHAVIOR OF THE HUMAN INFANT 473
crying seemed loudest while preparation was being made,
stopped slightly at handling of the penis, began again at
manipulation of the foreskin, and was slightly accelerated
after the cutting. The babies were in every case quiet during
the ceremony which followed, but this was insured by the
amount of alcohol administered.
RESPONSE TO DERMAL STIMULI
29. To Prick on Wrist. — On one occasion I tested 21 sleep-
ing infants by lightly pricking the wrist. I attempted to
use the same pressure on the pin each time, making it just
hard enough so that when tried on my own wrist it aroused
a pain response. Thirteen gave a movement of the hand
or forearm in response to the pin touch and 8 did not respond.
Subjects giving positive results were 2, 2, 2, 3, 6, 7, 8, 8, 8,
16, 17, 17, 19 days old. Those giving negative were 2, 2, 3,
6, 6, 8, 9, and 26. As an interesting note on the experiment
it is worth mentioning that of the 8 giving no response 6 were
the children of either one or two defective parents.
30. To Being Rubbed. — The cleaning of the newborn with
oil affords a most interesting opportunity for the observa-
tion of dermal and deep pressure reactions. The rubbing
of the head, and especially the rubbing of the back, brings
out the most active reactions which in all probability the
baby will be called on to make in the first month of life.
The screaming which invariably ensued was of such a volume
as is hardly attained in the first 6 weeks under ordinary
conditions. During the scrubbing of the head 15 of the
babies which I saw delivered impeded the progress of the
operator to a greater or less degree by a constant moving
of the hands. Subject K. caught the hands of the operator
through the whole of the cleaning process. Since the vernix
caseosa is most heavily deposited on the back, rubbing the
back brings out a response of the lustiest crying, and in 3
cases, when the infant was supported by the left hand of
the surgeon in a crawling position with hands and knees
touching the bed, crawling movements were elicited. In
neither of these cases of course is the stimulus entirely dermal,
474 MARGARET GRAY BLANTON
since a good deal of pressure is exerted on the muscles and
deeper structure; and certainly the squeezing of the body
as well as the pressure on the unattached skull bones un-
doubtedly introduced kinaesthetic factors. The cry elicited
is identical with the ' angry cry' of the nursery and is in
fact so called by a great many operators and nurses.
Subject F., 13 days, raised her right arm at each of seven
strokes of her nose, but did not approach the hand to the
nose. Subject S., at 30 minutes, moved both hands to her
nose when it was held, but on the second trial moved them
in the opposite direction.
31. To Dampness. — Immediately after the voiding of
urine the napkin has a higher temperature than usual.
After a certain length of time the wet diaper becomes colder
than usual. When the element of temperature is eliminated
there is little evidence to show that babies up to 6 weeks
show any particular response to wetness or dryness of their
diapers. In a nursery in which there were 16 wet babies,
10 were quiet and 6 crying. A few minutes after changing,
10 were crying and 6 were quiet. Substantially the same
results were obtained in another group of 19 infants. In
the course of these observations enough cases have been
investigated to show that it is extremely difficult to predict
whether a baby is wet or dry because of its cry. There are
nursery saws to the effect that babies 'cry because they are
wet' and are still 'because they are dry.' Rounds of the
nursery will often show that neither conclusion is justified,
as babies under 30 days void frequently, and out of a nursery
of 25, 15 to 20 wet babies can be found at any time within
30 minutes of changing. It is not uncommon to have them
void immediately on changing, stimulated perhaps by the
coolness of the fresh diaper. Subjects 4, 6, 10, and 14 days
old were on one occasion crying and wet, but the diapers
were quite cold to the touch as well, and subjects 4, 5, 9,
n, and 21 were wet and quiet and the diapers seemed about
body temperature.
32. To Warmth and Cool. — Preyer1 says: "Sensibility to
1 'The Mind of the Child,' p. 183.
BEHAVIOR OF THE HUMAN INFANT 475
contact is, in the first hour of life, much inferior to what it
is later; the sense of temperature does not exist." I have
noted 6 occasions on which there was marked shivering
within 15 minutes of birth, on two occasions it persisted
until the infant had been brought near to a hot water bottle
for some time. The skin of this baby was cold to the touch.
The shivering was so pronounced as to make a marked
movement of the toothless lower jaw which under other
conditions we call 'chattering.' The reaction to warmth
during the second day of life was marked. Uncovering the
lower part of the infant's body to the heat of a mild fire
caused curling of the toes both upward and downward.
Subject L., 7 days, when she had been some time with the
outside of her right leg next a rather hot fire, crossed it over
her left and thus away from the fire. Subject S., at 9 days,
when placed in a warm tub, moved her arms outward and
downward, palms backward, elbows straightened, and on
the forward movement, bent — much as in swimming. She
opened her mouth wide, with her tongue over the surface
of her lower lip; her breathing became noisy and accelerated,
the air being forced out of her nose in what might be called
a snort.
33. Alcohol. — An excellent test of the reaction to cold of
the skin of the very young was offered by the application of
alcohol dressings to the stump of the umbilical cord. The
surgeons delivering obligingly dropped small amounts where
I indicated. When dropped on the lower half of the ab-
domen kicking resulted uniformly; dropped on the lower left
part of the abdomen the left leg would respond; on the
lower right, right leg; lower center, both legs or either right
or left. Above the umbilicus the reaction was not so certain
but when obtained was a movement of the legs.
RESPONSE TO KINJESTHETIC STIMULI
34. Kincesthetic Sense. — Dearborn1 says: "It is one of
the anomalies of psychology and of physiology, that kinses-
thesia, most basal and important of the senses, has been so
1 'Moto-sensory Development,' p. i.
476 MARGARET GRAY B LAN TON
relatively neglected. Without it no infant could become
more humanly efficient than a plant." This sense is prob-
ably the earliest of all developed, coming as may reasonably
be supposed before kicking does in the fourth or fifth months
of life in utero. Miss Shinn1 refers, under Rhythm, to the
* superior quieting influence of a monotonous jarring as com-
pared with a smooth motion.' Experiments on placing
infants to the breast to quiet them were made as early as
the third day of life, and as the amount of milk gotten up
to that time is negligible it is fair to assume that changes in
the position or pressure was the cause of the resulting quiet.
Walking with the babies quieted them as early as the first
day. Ten subjects under 4 days responded by ceasing to
cry. One of these was crying with every symptom of ab-
dominal pain. It was noted that babies crying in the corridor
were exceedingly rare and the approximate number of trips
daily through the corridor with infants was 120. Of the
many dozens of trips made to and from the laboratory only
one child was known to cry the entire distance; one cried
the first third of the way and was quieted. Subjects at
6 hours, 18 hours, i, 3, 3, 4, 6, 6, 7, 7, 9, n, 14, and 18 days
who were crying in their cribs were quieted by lifting and
gentle pressure. Subjects at 10 hours, 4, 6, and 9 days were
quieted by pressing the infant's body between the hands.
Subjects I, 3, and 6 days were quieted by pressure on chest.
Subjects at 6, 8, and 10 days were quieted by firm pressure
around head.
REFLEXES CONNECTED WITH FEEDING
35. Cheek and Chin. — This test was a light touch on either
cheek in a straight line parallel with the mouth and about
an inch removed from the corners, and a light touch on the
chin exactly above the center. The subjects giving a positive
reaction moved their mouth in the direction of the touch.
Subject M., 5 hours, positive; Subjects Mon., 5! hours, K.,
at 7 hours, H., at 8f, and D., at 9 hours, were all positive.
Subject F. at 9 hours was positive to right cheek, negative
1 Note book II., Miss Milicent Shinn.
BEHAVIOR OF THE HUMAN INFANT 477
(did not react to touch) on left cheek and chin. Subject B.
(microcephalic in dying condition), 10 days, positive to right
cheek and negative to left and chin. Two hours later Sub-
ject H. (positive in above list) gave a complete negative
when in a very deep sleep. After feeding this reaction is
hard to elicit. During hunger it is easy and the infant
often moves with surprising quickness, catching the testing
finger in its mouth. One infant giving a positive reaction
at 5 hours showed a complete negative at 7. Inquiry showed
that through a mistake in routine the baby had been put
to the breast in the interim.
36. Lip Reflex (Thompson's).1 — This is gotten by tapping
lightly with the tip of the finger below or above the corner
of the mouth of a sleeping baby. The result is a closing and
pouting of the lips into a nursing position. Tests were made
on 14 sleeping babies an hour before feeding. The most
marked result was Subject R., 7 days, who pursed his lips,
protruded the tip of his tongue between them and sucked
it vigorously. Subjects 16 hours, 4, 5, 7, 9, 10, n, 19, and
21 days old pursed the lips and protruded the tongue but did
not suckle. Subject M., 9 days, pulled his head back quite
markedly and rolled out his under lip. There was no sug-
gestion of pursed lips or suckling. Two subjects at 10 days
jerked the head slightly but gave no other response. Sub-
ject Y. at 19 days gave no reaction. This baby was markedly
malformed and the child of retarded parents.
37. Tongue Reflexes, at birth and later. Movements
similar to the movements and position of the tongue and
lips in the lip reflex (Thompson) may be seen immediately
after birth. These movements which appear to be merely
unrelated and random are related to suckling. The tongue
protrudes from the mouth and the edges are curled upward
and over in such a way as to make the partial vacuum
essential to sucking, much easier than if the back part of
the tongue, only, was depended on. The use of the muscles
in the back of the mouth apparently requires a higher type
1 Thompson, John, 'On the Lip-reflex of Newborn Children,' Rev. of Neurology
and Psychiatry, Vol. I., 1903, p. 145.
478 MARGARET GRAY BLANTON
of coordination (see suckling and swallowing below). The
forming of the tongue into this protruding tube can be seen
in infants up to 30 days, at the time of feeding. If the food
is removed before satiety is reached there is often an active
sucking of the protruded tongue.
38. Sucking Fingers. — This seems to vary mainly with
the manner in which the fingers are gotten into the mouth.
Subject M. at 20 minutes put his thumb directly into his
mouth and began sucking it. Subject S. during the first 2
hours of life put his right first finger into his mouth 6 times.
His action was not fumbling. He touched his face either
below or above his mouth, which set up the sucking reflex
noted under lip reflexes (36) and the finger was pulled directly
into the mouth. Once the sucking had been started there
was rarely any fumbling. Three exceptions to this rule
also showed weak swallowing reflexes and each of them had
either one or two retarded parents. Subject S., child of
imbecile mother and retarded father, attempted for 15
minutes to get her finger in her mouth, but did not succeed.
These babies' mouth reflexes are given more fully under
swallowing and suckling below. Subject G., a 'blue baby'
two hours old, put his fingers directly into his mouth. Sub-
ject D., 2 days old, Caesarian delivery, very feeble, was seen
sucking two fingers so vigorously that it required a decided
effort to remove them. She put them back in at once with-
out trouble. She was also seen with all 4 fingers in her
mouth. Subject B., a mal-formed baby mentioned else-
where, 10 days old, in dying condition, put his finger in his
mouth after 4 trials. Both his sucking and swallowing
reflexes were moderately good. The former were better
than the latter.
39. Suckling and Swallowing. — A demonstration of the
presence of the ability to suckle immediately after birth was
attended by many practical difficulties on account of the
sterile operating field surrounding the subject. A sterile
nipple was supplied at each delivery and the surgeons very
kindly held them in the babies' mouths. As the process of
obtaining sterile nipples became difficult on account of some
I BEHAVIOR OF THE HUMAN INFANT 479
local conditions the expedient was devised by one of the
surgeons of cleansing his gloved hand in sterile water and
letting the baby suck his finger. Subject S., immediately
after spontaneous cry, was given the nipple. The reflex
faintly present was perceived as a slight movement of the
lips as in sucking, and as a swallowing movement in the
throat. Subject O'K., aged 23 minutes, suckled slightly.
Subject Sc., very blue, and with umbilical cord around neck
at birth, at 10 minutes sucked well, at 25 minutes sucked so
violently that it was necessary to hold the unattached nipple
with an artery clamp. Subject Le. S. sucked definitely but
not vigorously at 15 minutes. Subject M., 6 minutes, sucked
vigorously and at 30 minutes sucked very hard. Subject R.
sucked vigorously at 10 minutes. Subject S. sucked definitely
at 13 minutes and vigorously at 25. This list continues with
little variation through the whole series. Occasionally it
was not practicable to get the trial immediately after birth,
but in no case did the baby refuse definitely to suckle during
the first hour of life. Nothing but the most marked retarda-
tion or injury seems to affect this reflex. Swallowing, with
sucking, tongue, lip and cheek reflexes go to make up the
feeding reflex. The evidence is strongly in favor of the
fact that weakness in the ability to swallow is at least sug-
gestive of mental retardation whereas sucking is present at
birth in most infants. Interesting in connection with this
is the fact that certain forms of retardation which are accom-
panied by speech defects show a lack of coordination in the
same muscles which seem defective in swallowing — namely,
the control of the soft palate and back part of the tongue
and throat. Going back over this list, Subject S. was the
child of healthy and mentally well developed parents. Both
her swallowing and sucking reflexes were good. She was
dismissed at the end of 10 days and was considered excellent
in every way.1 Subject O'K., child of intelligent parents,
dismissed in 14 days, rated excellent. Subject Sc. sucked
violently at nipple but did not swallow well and was not
1 A discharge was not given these patients until the best possible feeding habits
had been established.
48° MARGARET GRAY B LAN TON
dismissed until 21 days old with a rating of moderately good.
Her mother was a high-grade moron. Subject L., whose
mother was graded as 12 years by the Binet-Simon intelligence
test, was dismissed at 18 days. This infant's swallowing
reflexes were very poor. Subject M., child of mother rated as
10 years mental age and who also showed a marked tendency
to certain psychopathic conditions, fed moderately well, but
at 2 months is not thriving as well as at 2 weeks. Subject R.
(child of deaf and dumb primipara of 32, who, on account
of her deafness, was not given a mental test, but whose
practical ability for learning was poor and who showed a
marked tendency to cruelty to her baby) left at the end of
3 weeks, her swallowing still faulty, with a grading of poor.
Subject S. whose sucking at birth was vigorous enough left
at the end of 3 weeks without having learned to swallow
at all. The only food that this baby obtained was that
which went down when it was carefully moved in such a
way as to assist gravity. This baby sucked actively but
ineffectually with his lips. At 2 months he was little more
than a skeleton and could not swallow. The mother of this
child was a low-grade imbecile who could not even put on
her own clothing and who had to be fed with a dull-edged
spoon. Subject P. (not on the list above) who possessed
the sucking reflex, but in whom the habit of regurgitation
was most marked, had an intelligent mother but the father
was cruel, a drunkard and an habitual deserter, who had
never been able to finish the lower grades at school. Twin
babies K., whose mother's mental age was 10, had to be fed
with 'Boston' feeders which push the milk into the mouth.
It was exceedingly hard to get them to swallow. Subject K.
remained in the hospital 2 months. He was the child of
incestuous relations, the mother a low-grade imbecile and
the father the same. In this baby even the sucking reflexes
were absent and tube feeding was necessary. At 4 months
he was just alive. Subjects W., Y., and K., whose swallowing
reflexes were exceedingly weak, had each been in the hospital
more than the usual time. Each regurgitated constantly.
The mothers of these three did not exceed 10 years mental
age.
BEHAVIOR OF THE HUMAN INFANT 4Sl
Of value, in this connection, will be a study of the eating
and speech habits of older children of retarded mentality.
SOCIAL BEHAVIOR
40. As the subjects were often found crying in groups,
observation was made to determine social influence as an
element. Note was made of each baby that cried whether
in a group or singly through a number of hunger periods.
There were 28 in groups of two or more, and 35 who cried
singly. In observation L, two babies were kept together
in a quiet room, and both slept. In observation II., Subject
W., 10 days, slept. Subject Sm., 7 days, cried 20 out of
the 60 minutes. Finding nursery conditions not under
control, graphophone records were made of Subject L. when
crying with hunger. These records were then played for 6
infants from I to 14 days of age. The results were negative.
In observation IX., Subject L., 8 days old, and Subject W.,
9 days old, were placed on a couch so that the conditions
would resemble the nursery with regard to vibrations. Sub-
ject W. cried but Subject L. remained quiet, and later
Subject L. cried and Subject W. remained quiet (and awake).
Observation X., Subject M., 7 days old, quiet and awake
(during graphophone record) made some suckling movements
with mouth and tongue and his breathing became a trifle
irregular — but afterwards, in an interval of quiet he made
the same movements. Result negative. Observation XL,
a final trial was made: smoked drum records of breathing
were taken on several infants while the phonograph crying
record was being run through. The resulting breathing
curves differed in no way from those made of breathing in a
quiet dark room. Conclusion: there seems to be no positive
evidence of social influence of this character on babies under
15 days of age.
SUMMARY OF RESULTS
In summing up the observations it will be seen that the
reflex and instinctive equipment of the child at birth is more
complex and advanced than has hitherto been thought. This
482 MARGARET GRAY B LAN TON
discrepancy is due perhaps to the rarity of the opportunity
for unrelated persons other than nurses or physicians to ob-
serve during this period.
During the first twenty minutes of life may be observed
sneezing, yawning, tears, sucking at nipple, fixating on light,
putting thumb in mouth, jumping to loud sounds, grasping,
crying with box-shaped mouth, crying with the corners of
the mouth pulled down, following a moving hand with the
eyes, turning of the head in such a way as to get air when
placed on the face, turning over when given a very slight
advantage, complete erection of penis, and most indicative,
perhaps, the cry of so-called anger immediately after birth,
justifying perhaps Kant's oft denied statement that the
cry of a child just born has not the tone of lamentation but
of aroused wrath.
The responses of the child, under twenty-four hours, to
pressure and the completeness and effectiveness of the food
reflexes as well as the value of the response to kinaesthetic
stimuli to this complex of food reflexes are most interesting,
and the possibility of a direct relation between the intelligence
of the parent and the swallowing ability of the child hinted
at in topic 39 suggests a field for further research which
would be of direct value to those interested in the develop-
ment of speech.
INDEX OF REFLEXES (Topic NUMBERS)
Alcohol (33); Arm and Hand (15, 16, and 17).
Buccopharyngeal (i, 2, 3, and 4).
Cheek and Chin (35); Circumcision (28); Creeping (18).
Dampness, to (31); Deep pricking of big toe (26); Defecation (7); Dermal Stimuli,
response to (29, 30, 31, 32, 33); Dropping, to (23).
Eye movements (8); Eyes during nursing (9).
Facial (10, u, and 12); Facial expression (12); Feeding, reflexes connected with
(35, 36, 37, 38, and 39); Foot movements and reflexes (19).
General responses (21, 22, 23, and 24); Grasping during nursing (17); Grasping
reflex (16).
Hand movements at birth (15); Head (13 and 14); Hiccoughing (2); Holding up
head when held in upright position (14).
Kicking (20); Kinaesthetic Stimuli, response to (34).
Lancing of infected finger, to (27); Leg and foot (18, 19, and 20); Lip reflex
(Thompson's) (36).
Noxious Stimuli, response to (26, 27, and 28).
BEHAVIOR OF THE HUMAN INFANT 483
Ocular (8 and 9).
Pelvic (5, 6, and 7); Penis, the erection of (5); Prick on wrist, to (29).
Rubbed, to being (30).
Smiles (u); Sneezing (i); Social behavior (40); Sound stimuli, response to (25);
Speech and emotional reactions, the beginning of (4); Stretching (24); Sucking
fingers (38); Suckling and swallowing (39).
Tears (10); Tongue reflexes (37); Turning head, when lying on face (13); Turning
over (21).
Umbilical cord, reflex to stimulation of (22).
Voiding of urine (6).
Warmth and cool (32).
Yawning (3).
DISCUSSION
A CRITIQUE OF THE YERKES-BRIDGES-HARDWICK
COMPARISON OF THE BINET-SIMON AND
POINT SCALES1
The readers of this journal are doubtless well enough acquainted
with the Yerkes-Bridges Point Scale to make unnecessary an
extended descriptive account. It consists of twenty tests, nine-
teen of which are selected or adapted from the Binet-Simon series,
but arranged, like the 1905 Binet series, in general order of diffi-
culty, without being grouped by ages. The score is given by
points, so many points being allowed for success in each test or
part of test. Much emphasis is laid on the fact that degrees of
success in meeting the several tests are distinguished, and some
credit given for lesser degrees of success, as distinguished from the
so-called 'all or none' procedure of the Binet-Simon Scale. The
highest score which can be obtained is 100. The score of any
individual is to be interpreted by comparison with a norm which
has been determined by the examination of a group of individuals
of the same sex, race and social level. Further details will be
commented on in connection with the comparison between this
scale and the Binet-Simon Scale.
Since the justification of the Point Scale rests confessedly on a
criticism of the Binet-Simon Scale, and since this criticism is severe,
and if it is accepted, must cause the abandonment of the Binet-
Simon Scale, it is worth while to consider this criticism in detail
and follow out the comparison of the two scales.
The criticisms of the Binet-Simon Scale are by implication
statements of advantages of the Point Scale, and the advantages
of the Point Scale are by implication faults in the Binet-Simon
Scale; hence we may bring under the same head the disadvantages
of the one and the advantages of the other. They make the fol-
lowing long list:
I. The Binet-Simon Scale falsely assumes that the mental develop-
ment of all individuals proceeds by similar stages.
1 ' A Point Scale for Measuring Mental Ability,' by Robert M. Yerkes, James W.
Bridges and Rose S. Hardwick. Baltimore, Warwick and York. 1915.
484
COMPARISON OF BINET-SIMON AND POINT SCALES 485
There has been a good deal of criticism of the Binet-Simon
Scale on the score of the assumptions which the critics find to
underlie it. The authors of the Binet-Simon Scale do not tell us
that these assumptions are involved, and even if some such assump-
tions as are laid down do seem to be involved it is most natural to
regard them as working assumptions and the scale as a working
instrument rather than a highly precise and accurate engine to
apply a precisely known law. It is perhaps no exaggeration to
say that the Binet-Simon Scale and the results of its application
has given us more information regarding the actual course of
intellectual development than all the other testing of children put
together.
But take the assumptions themselves. Must not any scale
which uses age norms make some such assumption as is indicated
in the first criticism? The very purpose of the scale is to measure
the deviation of an individual from a norm, and if there is no norm
the deviation has no significance. As a matter of fact, the Point
Scale makes this assumption in exactly the same degree as does
the Binet-Simon Scale.
2. The Binet-Simon Scale falsely assumes that the correlation
between different functions is the same for all individuals at a given
stage.
If no flexibility in scoring were allowed in the Binet-Simon
Scale the second objection would have some force, for there would
then be no opportunity to make up for a deficiency in one mental
function by unusual performance in another; but the possibility
of gaining advanced credits at least in part overcomes this ob-
jection. Besides, the probability is that the correlation in rapidity
of development of the different functions is fairly close.
3. The Binet-Simon Scale falsely assumes that each stage of mental
development corresponds in turn to a certain physical age, and that
there is a 'correlation between the different functions at different
stages of development'
Since any norm must be based on physical age, unless the age
principle is entirely excluded by finding functions which do not
develop with age, it does not seem possible to avoid the third
assumption — that is, the first part of it. The second part of the
statement is not amplified and the writer is not able to tell what
it means.
4. The Binet-Simon Scale falsely assumes that the mental develop-
ment of the Paris school children follows the above-mentioned course
of development.
486 F. N. FREEMAN
The fourth assumption is not fundamental to the scale, but
only to its first form. It can be obviated, and has been obviated
by revision.
5. As a consequence of these false assumptions the Binet-Simon
Scale employs the erroneous principle of age grouping of tests.
It follows from the above statements that the age grouping of
tests is not a fundamental point of difference between the Binet-
Simon Scale and the Point Scale, since both scales involve the age
principle, but that it is only one of the possible devices by which
the age principle may be applied. The Binet-Simon Scale uses
the age principle in the classification of the tests, and the Point
Scale uses it in the construction of the norm. The issue is purely
one of convenience. • If many norms are needed it is easier to have
them independent of the classification of the tests. Otherwise not.
This question we shall discuss in a moment.
6. As a further consequence of the age grouping the Binet-Simon
Scale employs the crude all-or-none method of scoring.
Considerable emphasis is laid on the distinction between the
all-or-none scoring of the Binet-Simon Scale and the partial scores
of the Point Scale. Here again the issue is not fundamental, and
the Binet-Simon not only can, but does use the device of partial
scoring by including several grades of difficulty of the same test
and putting them at different ages. This is done, for example,
with memory, interpretation of pictures, and weight discrimination.
The same type of organization could easily be extended without in
the slightest degree altering the general structure of the scale.
In passing it may be remarked that the Point Scale in its present
form assumes that the weighting of the various tests which it adopts
is correct without giving it an experimental basis.
7. Because of the age grouping of tests the Binet-Simon Scale
cannot make due allowance for different rates of development due
to sex, race and social or educational advantages.
The bearing of the seventh objection depends on the interpre-
tation which is made of the degree or basis of the sex, racial or
social differences. If these differences require different norms the
point scale method furnishes an easier device than the age grouping
method. But if sex, race and social level all are factors which
necessitate separate norms, the task, as we shall see in a moment,
is a very considerable one. Let us first ask concerning the necessity.
It seems to the writer that, instead of a case being made out for
the necessity of sex norms, the results of the application of the
COMPARISON OF BI NET-SIMON AND POINT SCALES 487
Point Scale indicate that no distinction of any clear sort can be
made out. The lines which represent the scores cross and recross,
and the crossing points in the English-speaking group are different
from those of the non-English-speaking group. The race com-
parison which is reported brought entirely negative results, so far
as total scores are concerned. Even if marked race differences
had been found, the question might at least be raised whether it
does not suit the practical purpose of an intelligence scale better
to measure different race groups, at least within the same com-
munity, by a single norm than by separate norms. For classification
in the school it would seem certainly better; and if feeble-minded-
ness is conceived in functional terms it would seem better for the
selection of feeble-minded individuals also. With reference to
social groups the chief question is whether the differences are
inherent or the product of environment. If inherent, no separate
norm is needed; otherwise it is. The probability is that the differ-
ences are at least in part inherent. In so far as they are not, the
application of different norms would be exceedingly difficult. How
many groups would be distinguished, what would be the basis of
the distinction, how would a particular child be placed in the
proper group, etc. ?
Let us assume that we do have norms for sex, race and social
level, as the authors propose, and that we have scales for the four
types of mental function. A short calculation will show that the
number of norms that would be necessary is sufficient to daunt the
hardiest investigator. If we provide for three races only, and for
three social levels, no less than seventy-two norms would be neces-
sary. We start with four scales. The sex distinction necessitates
eight, the racial distinction twenty-four and the social distinction
seventy-two. The derivation of these norms would require the
examination of at least seventy-two thousand children.
8. Since in the Binet-Simon Scale * later and more difficult tests
have no more weight in making up the score than do earlier and easier
ones, the same mental age may correspond to records far from equiva-
lent* (p. 33).
The feature which is criticized in the eighth article is given
as one of the chief advantages of the Point Scale. It certainly is
true of that scale, and is the direct opposite of the characteristic
which is criticized in the second article.
9. In its present form the Binet-Simon Scale does not give uniform
opportunity to the various mental functions at different levels.
488 F. N. FREEM4N
The ninth article is similar to the second, and does hold true
of the original Binet-Simon Scale, but is not essential to it.
10. The Point Scale, as contrasted with the Binet-Simon Scale,
is capable of giving results of ever-increasing reliability and precision
as data accumulate and norms are established.
Theoretically, the tenth point is well taken. Practically, the
establishment of a norm is not a very extensive process, after the
choice of tests has been made, and for practical purposes it is
desirable not to defer the establishment of a norm too long. It is
probable, for example, that the present Point Scale norm will be
used for some time, since ideal programs are frequently long deferred
in realization.
11. There is less influence of the personal equation of the examiner
in using the Point Scale.
If the examiner has a constant bias in the direction either of
undue conservatism or liberality in scoring the child's responses it
is clear that the error is greater the larger the units which are
represented in each judgment, and the larger the number of judg-
ments which are made. A judgment which determined a large
share in the child's rating would, if in error, cause a greater dis-
placement of his rank than one which had a small share in his
rating. Thus the breaking up of the Binet tests into subsidiary
tests in the Point Scale, by making each unit smaller, reduces the
error which is due to constant bias in either direction on the part of
the examiner. As was pointed out in discussing the all-or-none
method of scoring, this difficulty can be overcome in the Binet
scheme by breaking up the tests and placing the constituent parts
at different appropriate ages. The size of each error being deter-
mined by the value of the unit upon which it is based, the size of
the total error will be determined by the number of the individual
errors. An error can occur only in the case of those tests in which
the judgment of the examiner varies from the true judgment.
In the Binet Scale the range of such tests is confined to those of a
few ages immediately adjacent to the child's mental age. In the
Point Scale, because of the partial credit system and the scoring
by points, the range of tests from which the child's score is made
up is much greater. Hence the number of judgments which are
factors in his rating is relatively larger. For this reason there are
more opportunities for error which will affect the score. It would
appear from these considerations that theoretically — and thus
far we have no other basis for judgment — there is one respect in
COMPARISON OF BI NET-SIMON AND POINT SCALES 489
which the Point Scale offers less opportunity for error on account
of a constant bias on the part of the examiner, and one in which
the Binet Scale, even as at present constituted, offers less oppor-
tunity for error. If this (admittedly hazardous) reasoning is
correct the situation is about a * stand-off.'
12. The Point Scale 'works with a smaller amount of testing
material, and thus makes possible a better choice of the same.9
The twelfth point is not self-evident, particularly since the
program of the Point Scale calls for the multiplication of the
present number of tests four fold.
13. The Binet-Simon Scale cannot — and by implication the Point
Scale can — diagnose moral imbecility, dementia and intellectual
degeneration, and occasional phenomena of degeneracy such as im-
pulsions, obsessions and. delirium.
The present Point Scale cannot diagnose moral imbecility,
etc., any better than can the Binet-Simon Scale, and no proposals
are made which promise to make such diagnosis possible.
14. The statistical results of the application of the Binet-Simon
Scale are unsatisfactory, and a comparison of the results of the ex-
amination of the same 100 cases by the two scales are favorable to the
Point Scale.
The comparison of the statistical results of the application of
the two scales to the same 100 children do not seem to the writer
to be conclusive. In the first place, these 100 children were part
of the group the examination of whom formed the basis of the
Point Scale norms, and their Point Scale ratings were to some
degree tested by a comparison with themselves. Furthermore, the
mode of selection of the cases is not stated, and when the two
ratings do not agree we have no means of knowing which is more
nearly right. They are apparently not random selections, since
only one individual is rated as normal by each scale.
15. The Point Scale can be developed so as to give differential
rating in the different kinds of mental processes, namely, receptivity,
imagination, affectivity and thought.
The development of differential scales is highly desirable, but
the essential problem concerning it is the selection of tests, and
when this is done they can be arranged either on the age principle or
on the point scale principle, after the manner of the present scales.
This review does not attempt to pass judgment on the com-
parative merits of the two scales, but only to criticize the com-
parison which is made by the authors of the Point Scale. If this
49° F. N. FREEM4N
criticism is drastic it is not more so than the criticism of the Binet-
Simon Scale on which it comments. These comments may be sum-
marized in the statement that the criticisms of the Binet-Simon Scale,
which are made by the authors of the Point Scale are for the most
part not essential and may be obviated by revision, or are equally
applicable to the Point Scale itself.
FRANK N. FREEMAN.
UNIVERSITY QF CHICAGO.
ucari. HUU
PLEASE DO NOT REMOVE
SUPS FROM THIS POCKET