Skip to main content

Full text of "Language, thought, and reality; selected writings"

See other formats

i? ^. 






i — irf 
■:^ ^-^ ^, .:^ 


<rii33Nvsoi^^ ^a^AiNfiawv^ "^^c^Aavaaii-i^N'*^ '■^(■ 




-n I— ' ,-^ „ > io ■'■' 






Jd'O ' 


S ■< 














^ ^;^lllBRARY<:?r 

>?^ % 

^^^\E■^INIVER% ^.yl.OSANCEl^* -^tllBRARYQ^ ^IIIBJ 


"y^^ <;^ - 

%jnV3J0^ ■^J^lSDNVSQl^'^ '^/Sa3AINIl-3\<^'^ 





■■: O 



^ 5 

<f5l33NVSO)^ %ti3AINft-3iVV 



■:3a ^ 

33 -T- 

:^ ...-..■..■■ 










.^WEUNIVERS/^ ^^. 

..jrs!* s? 

O (i_ 


% i % ^^ 


'■^mhmW^ ^ ^'-^o-mm^-^ ""^nsmm^ 

iiRD,u'-,:,', '..-I I ci\/r;-,/ 

\g ^ 

uj/ »;iiii ji 1-- 

oF.rAiiFM: r.P-rMicv:. ^^EUNIVEk,.^^ inc.urnr 

3 . f3I» <:5 0? 

■S 5 



yLi X^- 

.'^' ''<>JiaJAIwi!Jk^^'' ^-^talTVD-ao^' '-'tonVj-jO-'^' %133KV-S01=^^'' 
^^WtUNIVER% ^lOSANGElfx^ AOfCAllF0% ^^;0FCAIIF0%. ,.5i\E l]NIVER% 

"^ O 

^i^3NYS0^-^- --^ > ^ - ^ "^UQNVSOl^ 






The Scanlon Plan: A Frontier in Labor-Management Cooperatiox 

Edited by Frederick G. Lesieur 
The Inflationary Spiral: The Experience in China, 1939-1950 

By Chang Kia-Ngau 
The Tao of Science: An Essay on Western Knowledge and Eastern Wisdom 

By R. G. H. Siu 
Soviet Education for Science and Technology 

By Alexander G. Korol 
The Economics of Communist Eastern Europe 

By Nicolas Spulber 
On Human Communication: A Review, a survey, and a Criticism 

By Colin Cherry 
Location and Space-Economy 

By Walter Isard 
Science and Economic Development: New Patterns of Living 

By Richard L. Meier 
Moscow AND the Communist Party of India 

By John H. Kautsky 
Language, Thought, and Reality 

By Benjamin Lee Whorf 

Edited by John B. Carroll 
The Terms of Trade: A European Case Study 

By Charles P. Kindleberger 
Machine Translation of Languages 

Edited by W. N. Locke and A. D. Booth 
An American Policy in Asia 

By W. W. Rostow and R. W. Hatch 
Nine Soviet Portraits 

By Raymond A. Bauer 
The Prospects for Communist China 

By W. W. Rostow and others 
Labor Mobility' and Economic Opportunity 

By Members of the Social Science Research Council 
Nationalism and Social Communication 

By Karl W. Deutsch 
Industrial Relations in Sweden 

By Charles A. Myers 
Pressures on Wage Decisions 

By George P. Shultz 
The Dollar Shortage 

By Charles P. Kindleberger 
Mid-Century: The Social Implications of S<.ientific Progress 

Edited by John E. Burchard 
Cybernetics: Or Control and Communication in the Animal and the Machine 

By Norbert Wiener 
The Movement of Factory Workers 

By Charles A. Myers and W. Rupert Maclaurin 

/?. 7f. »^ 




Edited and with an introduction by 

Foreword by STUART CHASE 

Published jointly by 

The Technology Press of 

Massachusetts Institute of Technology 


John Wiley & Sons, Inc. 

New York London 


Copyright © 1956, by The Massachusetts Institute of Technology 

All rights reserved. .This book or any part 
thereof must not be reproduced in any form 
without the written permission of the publisher. 

Library of Congress Catalog Card Number: 56-5367 
Printed in the United States of America 



^7 . 

Once in a blue moon a man comes along who grasps the relationship 
between events which have hitherto seemed quite separate, and gives 
mankind a new dimension of knowledge. Einstein, demonstrating the 
relativity of space and time, was such a man. In another field and on 
a less cosmic level, Benjamin Lee Whorf was one, to rank some day 
perhaps with such great social scientists as Franz Boas and William 

He grasped the relationship between human language and human 
thinking, how language indeed can shape our innermost thoughts. 

We are thus introduced to a new principle of relativity, which holds 
that all observers are not led by the same physical e\ idence to the same 
picture of the universe, unless their linguistic backgrounds are similar, 
or can in some way be calibrated. 

Indo-European languages can be roughly calibrated— English, French; 
German, Russian, Latin, Greek, and the rest; but when it comes to 
Ghinese, Maya, and Hopi, calibration, says Whorf, is structurally diffi- 
cult if not impossible. Speakers of Chinese dissect nature and the 
universe differently from Western speakers. A still different dissection 
is made by various groups of American Indians, Africans, and the 
speakers of many other tongues. 

Whorf was a profound scholar in the comparatively new science of 
linguistics. One reason why he casts so long a shadow, I believe, is that 
he did not train for it. He trained for chemical engineering at M.I.T., 
and thus acquired a laboratory approach and frame of reference. The 
work in linguistics was literally wrung out of him. Some driving inner 
compulsion forced him to the study of words and language— not, if you 
please, the mastery of foreign languages, but the why and how of lan- 
guage, any language, and its competence as a vehicle for meaning. 



As a writer, I have long been interested in semantics, sometimes de- 
fined as "the systematic study of meaning." It does a writer no harm, 
I hold, to know what he is talking about. Whorf, using linguistics as 
a tool for the analysis of meaning, has made an important contribution 
to semantics. No careful student of communication and meaning can 
afford to neglect him. One might add that no philosophical scientist 
or scientific philosopher can afford to neglect him. Linguistics, he 
boldly proclaims, "is fundamental to the theory of thinking, and in the 
last analysis to all human sciences." He is probably right. Every con- 
siderable ad\ance in science, such as quantum theory, involves a crisis 
in communication. The disco\erers have to explain first to themselves, 
and then to the scientific world, what has been found. 
iWhorf as I read him makes two cardinal hypotheses: 

First, that all higher levels of thinking are dependent on language. 

Second, that the structure of the language one habitually uses influ- 
ences the manner in which one understands his environment. The 
picture of the universe shifts from tongue to tongue. 



'I'hcre is a good deal of competent scientific support for the first hy- 
pothesis. I'he biologist, Julian Huxley, for instance, declares that "the 
evolution of verbal concepts opened the door to all further achieve- 
ments of man's thought." Language, observes Whorf, is the best show 
man puts on. Other creatures have developed rough communication 
systems, but no true language. Language is cardinal in rearing human 
young, in organizing human communities, in handing down the culture 
from generation to generation. Huxley goes so far as to venture that 
adaptation through the culture, depending, of course, on language, may 
be displacing the biological processes of evolution. When the next Ice 
Age moves down, for instance, instead of growing more fur, homo 
sapiens may step up the production of air-conditioning units. 

The power to reason constitutes the "uniqueness of man," to philos- 
ophers as well as biologists. Unprotected by claws, teeth, thick hide, 
flcetncss of foot, or sheer strength, homo sapiens has to think his way 
out of tight places. It has been his chief weapon for survival. 

Probably everyone experiences brainstorms too fast to be \crbal. In 
writing, I frequently have them. But before I can handle such bolts 


from the blue, I must verbalize them, put them into words for sober 
reflection, or discussion. Unverbalized brainstorms do not get any- 
where on paper. 

Perhaps driving a car furnishes a good analogy for Whorf's initial 
hypothesis. Light waves and sound waves are enough to guide the 
driver's hand on the wheel along straight roads. But threading his way 
through a cloverleaf intersection, or reading a road map, will require a 
good deal more than reflex action. The first, a \ery clever chimpanzee 
might learn to do; the second is forever beyond it. 


The Greeks, so active mentally, and so reluctant to exert themselves 
in observation post and laboratory,^ were the first to inquire into logic 
and reason. The Sophists were apparently the Madison Avenue boys 
of the Aegean, teaching young men how to capsize an opponent in de- 
bate or legal case, and to choose the most effective slogans in political 
campaigns. Aristotle inxented the syllogism, and fashioned his Three 
Laws of Thought, beginning with the Law of Identity, A is A, now and 
forever— against which we semanticists sometimes protest. 

The Greeks took it for granted that back of language was a universal, 
uncontaminated essence of reason, shared by all men, at least by all 
thinkers. Words, they believed, were but the medium in which this 
deeper effulgence found expression. It followed that a line of thought 
expressed in any language could be translated without loss of meaning 
into any other language. 

This view has persisted for 2500 years, especially in academic groves. 
Whorf flatly challenges it in his second major hypothesis. "A change 
in language," he says, "can transform our appreciation of the Cosmos." 

The day-by-day experience of skilled translators at the United Na- 
tions goes a long way to support him. Edmund S. Glenn of the State 
Department, for instance, aided by a grant from the Rockefeller Foun- 
dation, has waded through masses of U.N. transcriptions, looking for 
differences in concepts due to language.- An English speaker in one of 
Mr. Glenn's cases says "I assume"; the French interpreter renders it "I 

1 James Harvey Robinson, the historian, lays it to the large number of slaves. 
- Peter T. White, "The Interpreter: Linguist Plus Diplomat." New York Times 
Magazine, November 6, 1955. 


deduce"; and the Russian interpreter "I consider"— By that time the 
assumption idea is gone with the wind! 

After isolating twenty similar instances, Mr. Glenn concludes that, 
while the translation technique was smooth enough on the surface, 
"the degree of communication between the Russian and English-speak- 
ing delegates appears to be nil" in these cases. 

If there is thus some difficulty among Western peoples, all speaking 
varieties of Indo-European, it is not surprising that a much wider chasm 
yawns between languages from wholly different stocks— between the 
language of Hopi Indians, say, and English. This is the field which 
Whorf cultivated intensively, and on which he largely bases his concept 
of linguistic relativity. 

In English we say "Look at that wave." But a wave in nature never 
occurs as a single phenomenon. A Hopi says "Look at that slosh." 
The Hopi word, whose nearest equivalent in English is "slosh," gives 
a closer fit to the physics of wave motion, connoting movement in a 

"The light flashed," we say in English. Something has to be there 
to make the flash; "light" is the subject, "flash" the predicate. The 
trend of modern physics, however, with its emphasis on the field, is 
away from subject-predicate propositions. Thus a Hopi Indian is the 
better physicist when he says Re/i-pi— "flash"— one word for the whole 
performance, no subject, no predicate, no time element. We frequently 
read into nature ghostly entities which flash and perform other miracles. 
Do we supply them because some of our verbs require substantives in 
^front of them? 
-^^ The thoughts of a Hopi about events always include both space and 
^ time, for neither is found alone in his world view. Thus his language 
gets along adequately without tenses for its verbs, and permits him to 
think habitually in terms of space-time. Properly to understand Ein- 
stein's relativity a Westerner must abandon his spoken tongue and take 
to the language of calculus. But a Hopi, Whorf implies, has a sort of 
calculus built into him. 

"The formal systematization of ideas in English, German, French, 
or Italian seems poor and jejune"— in dealing with certain classes of 
phenomena, when contrasted with the flexibility and directness of 
Amerindian languages. Whorf demonstrates the trouble wc Westerners 


have with mascuhne and feminine genders, and with our built-in, two- 
valued, either-or logic. 

Does the Hopi language show here a higher plane of thinking, a more 
rational analysis of situations than our vaunted English? Of course it v 
does. In this field and in various others, English compared to Hopi ^ 
is like a bludgeon compared to a rapier. 

For other classes of phenomena English might be the rapier and 
Hopi the bludgeon. Both languages have been developed over the 
ages, largely unconsciously, to meet the experiences and problems of 
their speakers, and we cannot call one higher or more mature than the 
other. For, while human societies vary widely in their supply and con- 
sumption of artifacts, the human mind, reflected in language, shows no 
examples of primitive functioning. . . . "American Indian and African 
languages abound in finely wrought, beautifully logical discriminations 
about causation, action, result, dynamic or energic quality, directness 
of experience, all matters of the function of thinking, indeed the quint- 
essence of the rational." 

As you will see in Mr. Carroll's excellent biography, Whorf early in 
his Indian language studies noted similarities between certain Mayan 
inscriptions and that on an Aztec temple in Tepoztlan. I climbed to 
that rocky shrine in the same year, 1950, though not to study the hiero- 
glyphics. With Aztec he combined studies in Maya and then in Hopi. 
He found the last the most subtle and expressive of the three, and com- 
piled a Hopi dictionary, as yet unpublished. If he seems sometimes 
more affectionate than coldly scientific about his Indian tongues, it is 
easy to forgive him. 

Most of the above quotations I have taken from a monograph, also 
hitherto unpublished, which Whorf wrote in 1956. You will find it at 
page 65, and it deals with the thought processes of primitive peoples. 
He had planned to send it to H. G. Wells and H. L. Mencken, as well 
as to various distinguished linguists like Sapir. I wish that he might 
have done so, for it brings together all his remarkable qualities: his 
learning, his creative imagination, his idea of linguistic relativity, and 
his hopes for the future. What the essay says to me, a layman, is in 
essence this: 


There is no one metaphysical pool of uni\ersal human thought. 

Speakers of different languages see the Cosmos differently, evaluate 
it differently, sometimes not by mueh, sometimes widely. Thinking is 
relati\c to the language learned. There are no primitive languages. 

Research is needed to discover the world view of many unexplored 
languages, some now in danger of extinction. 

Somewhere along the line it may be possible to develop a real inter- 
national language. Some day all peoples will use language at capacity, 
and so think much straighter than we now do. 

Theoretically this might mean the end of linguistic relativity, but it 
would not mean the mountain had been climbed. The next great task 
would be to devise new forms of speech to bring us ever closer to reality, 
to move capacity on and up. "So far as we can envision the future, 
we must envision it in terms of mental growth." 

It is tragic for us all that the mental growth of Benjamin Lee Whorf 
was so prematurely interrupted. 

Stuart Chase 

GeorgetoMm, Connectictit 

November 23, 1955 







i \A07 



V 246 


Foreword, by Stuart Chase 

Introduction, by John B. Carroll 

On the connection of ideas (1927) ^ 

On psychology (date unknown) 

A central Mexican inscription combining Mexican and Maya 
day, signs (1931) 

The punctual and segmentative aspects of verbs in Hopi 

An American Indian model of the uni\ersc {circa 1936) 

A linguistic consideration of thinking in primitive communi- 
ties (circa 1936) 

Grammatical categories (1937) 

Discussion of Hopi linguistics (1937) 

Some verbal categories of Hopi (1938) 

Language: plan and conception of arrangement (1938) 

The relation of habitual thought and bcha\ior to language 

Gcstalt technique of stem composition in Shawnee (1939) 

Decipherment of the linguistic portion of the Maya hiero- 
glyphs (1940) 

Linguistic factors in the tcrminologv of Hopi architecture 

Science and linguistics (1940) 

Linguistics as an exact science (1940) 

Languages and logic (1941) 

Language, mind, and reality (1941) 


A. Published writings of Benjamin Lee Whorf 

B. Unpublished manuscripts (selected) 

C. Books and articles relating to \\niorf's writings (selected) 

1 Dates in parentheses are dates of composition. 


The career of Benjamin Lee Whorf might, on the one hand, be de- 
scribed as that of a businessman of speciahzed talents— one of those 
individuals who by the application of out-of-the-ordinary training and 
knowledge together with devotion and insight can be so useful to any 
kind of business organization. On the other hand, his career could be 
described as that of an unusually competent and diligent research worker 
in several otherwise almost completely neglected fields of inquiry — the 
study of the lost writing system of the Mayas and the study of the lan- 
guages of the Aztecs of Mexico and the Hopis of Arizona. Neither 
description, of itself, would mark him as a particularly engaging subject 
for biographical treatment. When it is realized, however, that he com- 
bined both these careers, achieving recognition in his business activities 
at the very same time that he advanced to high eminence in scholarly 
work— without even having undergone the usual preliminaries of formal 
academic study signalized by an advanced degree— and in addition in- 
jected into contemporary discussions on the study of man and his culture 
a challenging set of hypotheses concerning the relation of language to 
thinking and cognition, his biography becomes a matter of more than 
passing interest. 

He was born in Winthrop, Massachusetts, on April 24, 1897, the son 
of Harry Church and Sarah Edna (Lee) Whorf. He was a scion of an 
old American stock, his ancestors having come from England to settle in 



Provincetown and other parts of the Bay Colony soon after the landing 
of the Pilgrims. In England, the Whorf name had been found most 
frequently in West Riding, in Yorkshire, and there may be some obscure 
connection with the name of the Wharfe River in that locality. 

Benjamin was very much the child of his father, as were also his two 
younger brothers, each in his own way. Benjamin was the "intellec- 
tual," the more bookish and idea-centered. John, born in 1903, became 
a well-known artist, particularly noted for his watercolors. Richard, 
born in 1906, has distinguished himself as an actor and director on the 
legitimate stage and in motion pictures. 

Intellectual, artist, dramatist— the father was all three. After a brief 
career as a rather indifferent student at Massachusetts Institute of Tech- 
nology (it is said that he did not care to apply himself to his engineering 
studies), Harry Church Whorf drifted into commercial art, or what he 
liked to call "designing," an occupation which drew upon his talents in 
draftsmanship as well as his fertile imagination. In this work he was 
very successful. Among his productions which survive even today is 
the chain of little Dutch girls which encircles each tin of a well-known 
brand of cleansing powder. He made himself a master of the then 
rapidlv developing art of photolithography. But he was not content to 
remain within the confines of his occupation. He lent his artistic talents 
to numerous enterprises, stage designing being the foremost of these. 
He also wrote and directed plays for church groups and charitable or- 
ganizations, and he wrote the libretto for Bobby Shaftoe, a musical 
comedy which was once given a performance in Boston. He enjoyed 
giving illustrated lectures on various subjects, and apparently had a 
knack of pleasing an audience. At the time of his death in 1934, he 
was at work on a manuscript concerning the Massachusetts littoral— its 
geology, history, fauna and flora, and so forth. 

Even before the birth of their first child, Harry Whorf and his wife 
had taken up residence in a modest house in Winthrop, a residential 
suburb situated on a peninsula which flanks Boston harbor on the north. 
With the collections of drawings, books, manuscripts, chemicals, photo- 
graphic equipment, and odds and ends which the father had accumu- 
lated, the house pro\'ided a stimulating environment for three abnor- 
mally curious and inquisitive boys, all endowed with talents with which 
to take advantage of>it. Like his brothers after him, Benjamin early 
acquired some considerable skill in drawing, but the chemicals, dyes, 


and photographic apparatus intrigued him most. He loved to perform 
such experiments as the one in which hquids of various colors are made 
to form different layers in a single vessel. It may have been his early 
experiences with chemicals which led Benjamin later to choose to study 
chemical engineering at the Massachusetts Institute of Technology. 

He went to the public schools of Winthrop, through the high school, 
where we are told he did well. We are further told that he developed his 
powers of concentration at this time— even to the extent of apparent ab- 
sentmindedness. Once, on being sent to the cellar to fill the coal hod, he 
brought it, filled, all the way back to his room rather than to its place be- 
side the kitchen stove. (Later in his life, friends occasionally complained 
that he passed them on the street without even a sign of recognition.) 
While not especially strong, he had sufficient confidence in his physical 
prowess to protect his younger brothers from neighborhood bullies. Par- 
ticularly with John, who was six years his junior, he liked to play intel- 
lectual games. A favorite was the game of secret codes; Benjamin could 
nearly always solve even the most complex ciphers his brother could 
devise. In the meantime, while alone, Benjamin read voraciously and 
amused himself with composing humorous verse. 

After graduation from Winthrop High School in 1914, he entered the 
Massachusetts Institute of Technology, majoring in chemical engineer- 
ing (Course X). His academic performance there appears to have been 
of only average quality; his record shows no marks in the highest cate- 
gory (H, corresponding to what is now ordinarily called A), even in Eng- 
lish composition or in French. This is, of course, a commentary on the 
precarious relationship which exists between performance rated in col- 
lege and performance in later life. In the fall term of his senior year, 
a mysterious illness acquired in an ROTC summer camp forced Whorf 
to be absent from classes; the necessity of making up deficiencies the 
following summer delayed his obtaining the degree of Bachelor of 
Science in Chemical Engineering until October 1918. 

We do not know what sort of professional career Whorf planned for 
himself while a student at M.I.T. Most probably, he hoped to find 
employment as an engineer in some type of chemical production plant 
or factory. His professional career was to prove most unusual, for he 
emerged as a specialist in a line of work which, as he once complained 
in a letter to his M.I.T. alumni organization, was at the time hardly 
recognized as a distinct field of engineering even by his alma mater. In 


1919, not long after his graduation from M.I.T., he was selected as a 
trainee in fire prevention engineering by the company which employed 
him for twenty-two years, up to the time of his death. According to an 
account prepared by C. S. Kremer, Chairman of the Board, Hartford 
Fire Insurance Company, "he was selected by an officer of the company, 
Mr. F. C. Moore, who was himself an M.I.T. graduate and had charge 
of the underwriting and handling of the insurance on buildings equipped 
with automatic sprinklers." After graduating from the company school 
which Mr. Moore conducted for fire prevention engineers, Whorf was 
assigned to the home office of the company, in Hartford, to assist in fire 
pre\ention inspection of properties insured by the company in the north- 
eastern part of the countr}'. The company was starting to develop what 
was then a new idea in the business, namely, fire-prevention engineering 
inspections as a service to the property owner and policy holder. In this 
work, which necessitated constant travel, he became extremely skillful. 
"In no time at all," writes Kremer, "he became in my opinion as thor- 
ough and fast a fire prevention inspector as there ever has been. . . . 
He was intensely practical and taught what he knew as facts to engineers 
and skillful men in various manufacturing businesses." He specialized 
more and more in the inspection of plants which utilized chemical proc- 
esses in manufacturing. 

On one occasion while inspecting a chemical plant he was refused 
admission to a certain building on the ground that it housed a secret 
process. Even the head of the plant, to whom he was referred, insisted 
that no outsider could inspect this building. Whorf said, "You are 
making such-and-such a product?" The answer was "Yes," whereupon 
Whorf picked up a pad, quickly wrote down a chemical formula, and 
handed it to the head of the plant, saying, "I think this is what you are 
doing." The surprised manufacturer replied, "How in the world did 
you know, Mr. Whorf?", to which Whorf answered calmly, "You 
couldn't do it in any other way." Needless to say, he was admitted to 
the building which contained the secret process. 

He was so much respected among chemical manufacturers that his 
advice was eagerly sought. In an inspection of a very complicated chem- 
ical plant in Connecticut, he suggested to the management that a cer- 
tain process be abandoned until it could be made safer, and indicated 
how this could be done. Some time later, after the suggested improve- 


ments had been made, the management delayed starting the process for 
several days until W^horf could return to the plant and approve going 

He was admired not only for his technical skill but also, strange as it 
may seem to anyone who may know only Whorf's linguistic work, for 
his ability to attract business for his company. He was once asked to 
make a fire prevention inspection of some public schools on which the 
company had only a trifling amount of business. The recommendations 
which he submitted so impressed the school board that they decided to 
appoint Whorf's company as the manager of their insurance account, 
quite to the surprise of the local agent, who had found this particular 
school board difficult to approach. 

The value put on Whorf's services by his employer was signalized by 
his appointment in 1928 as Special Agent, and by his election in 1940 as 
Assistant Secretary of his company. It may have been that the company 
was proud of his accomplishments in linguistics and anthropology, and 
we know that it was liberal in granting him occasional lea\^es to carry 
on these activities,'^ but he was valued primarily for his actual services to 
his employer, which must have been of a high order, far beyond the 
ordinary. It is truly remarkable that he was able to achie\e distinction 
in two entirely separate kinds of work. During certain periods of his 
life, his scholarly output was enough to equal that of many a full-time 
research professor; yet he must have been at the same time spending 
some eight hours every working day in his business pursuits. His friends 
often speculated on why he chose to remain in his occupation. Although 
several offers of academic or scholarly research positions were made to 
him during the latter years of his life, he consistently refused them, say- 
ing that his business situation afforded him a more comfortable living 
and a freer opportunity to develop his intellectual interests in his own 

As if his insurance work, his linguistic studies, and his extensive read- 
ing were not enough to occupy him, he found time for certain com- 
munity activities, such as serving on a fire prevention committee of the 
Hartford Chamber of Commerce. From about 1928 on, he became 

1 Nevertheless, Whorf often combined business with science on these trips. In 
the course of the field trip to Mexico in 19^0, he inspected the Mexico City 
agency of the company and wrote a comprehensive report of his findings and 


increasingly popular as a lecturer before men's clubs, historical societies, 
and the like.^ 

In 1920 he married Celia Inez Peckham, by whom he had three chil- 
dren, Raymond Ben, Robert Peckham, and Celia Lee. In somewhat the 
same way that his father had done for him, he was able to arouse in his 
children, as if by magnetic induction, something like his delightful curi- 
osity and his unhesitating imagination. 

By his own account, Whorf did not become interested in linguistics 
until 1924, but one can trace a distinct succession of intellectual en- 
thusiasms which led him to this. Even as a boy, along with his pre- 
occupation with chemistry experiments, he was an avid reader. He 
became interested in Middle American prehistory through reading 
(several times, we are told) Prescott's Conquest of Mexico. On one 
occasion his father was engaged in doing the stage designs for a play 
which he had written about a Maya princess, and in this connection 
assembled all manner of books about Maya archaeology. Young Ben 
was intrigued with the resulting display of stage designs, which doubtless 
portrayed ornate fagades of Maya temples, and he may have begun to 
wonder about the meaning of Maya hieroglyphs. The interest in secret 
ciphers, mentioned earlier, may have reinforced this curiosity, but, if so, 
it lay dormant until a somewhat later period. Instead, he began to 
spend a good deal of time on a variety of scientific topics. He became 
interested in botany, and learned the English and Latin names for thou- 
sands of plants and trees. (This was a lasting interest; on his trip to 
Mexico in 1930 he took copious notes on Mexican flora, and as late as 
1936 we find him filling several pages of one of his linguistic notebooks 
with a "quiz" on botanical terminology and curiosities.) As if to con- 
trast with this, he was for a time intensely interested in astrology and 
amused himself with casting horoscopes for his friends. At some time 
in his adolescence he began to manifest what might seem an almost 
pathological graphomania, for at the age of 17 he began to keep a diary, 
a practice which he continued throughout his life. He contrived some 
sort of secret writing which he occasionally used to conceal some of the 

- My own acquaiiitniKC witli W'lioif dcxclopctl as a consequence of my attend- 
ing a lecture he gave at the Children's Museum of Hartford, December 1, 1929. 
The title of the lecture, announced as a "chalk talk," was "The Aztec and Maya 
Indians of Mexico." 


contents of his diaries, and which he also used to record his dreams in 
a series of "dreambooks." 

Shortly after settling in Hartford, Whorf became increasingly con- 
cerned about the supposed conflict between science and religion. It 
seems that he had been deeply impressed by the fundamentalist shad- 
ings of his Methodist Episcopal religious background, which at times 
seemed to controvert the current doctrines in science. He became so 
deeply preoccupied with this issue that he wrote a 130,000-word manu- 
script on the subject, described as a book of religious philosophy in the 
form of a no\'el. This manuscript, completed in 1925, was submitted 
to several publishers and as promptly rejected by them, even over his 
protests. Another, briefer manuscript prepared about this time was 
entitled "Why I have discarded evolution." An eminent geneticist to 
whom it was submitted for comment made a very courteous reply, start- 
ing with the admission that, although the manuscript at first appeared 
to be the work of a crank, its skill and perceptiveness soon marked it as 
otherwise, but continuing with a point-by-point rebuttal of Whorf's 

In the meantime, Whorf's reading led him to believe that the key to 
the apparent discrepancy between the Biblical and the scientific ac- 
counts of cosmogony and evolution might lie in a penetrating linguistic 
exegesis of the Old Testament. For this reason, in 1924 he turned his 
mind to the study of Hebrew. 

It may come as a surprise to some that Whorf's interest in linguistics 
stemmed from one in religion. The reader may incidentally be re- 
minded of the considerable connection which has long existed between 
linguistic and religious enterprises— the philological work represented in 
the Septuagint, in Ulfilas's creation of the written Gothic into which 
he could translate the Bible, in the study of hundreds of non-European 
languages by missionaries in the seventeenth and eighteenth centuries, 
and in the thoroughly scientific investigations being carried out by con- 
temporary linguistic missionaries. Whorf, however, was not interested 
in any translation of the Bible, at least not in any ordinary sense; he 
seriously believed that fundamental human and philosophical problems 
could be solved by taking a new sounding of the semantics of the Bible. 
Whether this conviction was independently reached by him we do not 
know. We do know that sometime during 1924 there came to his at- 
tention a book which could have buttressed his beliefs, and which at 


any rate drew him closer to linguistics. He himself gives testimony of 
this in a hitherto unpublished paper which appears in the present col- 
lection. This book, hardly known to contemporary scholars, was by a 
French dramatist, philologist, and mystic of the early nineteenth cen- 
tury, Antoine Fabre d'Olivet (1768-1825). It was entitled La langue 
hebrdique restituee, and was published in two volumes in Paris in 
1815-16. Whorf most probably read an English translation of this 
scarce work published in 1921, for the name of the translator, Nayan 
Louise Redfield, appears in his notes.^ 

According to the Grand dictionnaire universel du XIX^ siecle, Fabre 
d'Olivet died "avec le reputation d'un fou ou d'un visionnaire." A 
rather indifferent dramatist, he retired in his later life to extensive philo- 
logical lucubrations. In La langue hebrdique, his major work in this 
field, he attempted to show that the hidden meanings of the Book of 
Genesis could be elucidated by an analysis au fond of the structure of 
the triliteral Hebrew root. Each letter of the Hebrew alphabet, accord- 
ing to him, contained an inherent meaning; for example, the letter 
Aleph was to him "the sign of the power and stability of ideas, of unity 
and the principle which determines it." The letter Yodh was a sign of 
"manifestation"; thus the partial root Aleph-Yodh "designates," wrote 
Fabre d'Olivet, "the center towards which the will tends, the place 
where it enfixes itself, the sphere of activity in which it operates." Since 
he concluded that the letter Tsadhe denoted "termination," he was not 
surprised to find that the triliteral root Aleph-Yodh(or Waw)-Tsadhe 
meant "any desire tending toward an end." The principle of the root- 
sign was applied to all parts of Hebrew grammar, and to the interpre- 
tation of several hundred Hebrew roots. The whole was offered as 
partly a linguistic study to illuminate the principles of language (he 
claimed having been hard put to choose whether Chinese, Sanskrit, or 
Hebrew would be the basis for his project), and partly as the fulfilment 
of his desire to discover the secret meaning of the cosmogony of Moses. 
In the English translation which Fabre d'Olivet himself obligingly pro- 
vided, the first verse of the Bible conies out as follov^s: 'At-FIRST-IN- 
PRINCIPLE, he-created, yElohim (he caused to be, he brought forth 
in principle, HE-the-Gods, the-Being-of-Beings), the-selfsameness-of- 
heavens, and-the-selfsamencss-of-earth.' He tosses off the comment that 

? Miss Rcdfidd, for some years a resident of Hartford, also translated several 
other works by Fabre d'Olivet. 


this is not a mere result of some system he has estabUshed on the basis 
of "more or less happy conjectures or probabilities," but "the very lan- 
guage of Moses which I have interpreted according to its structural prin- 
ciples, which I have taken pains to develop to a satisfactory' point." 

Despite the dubiety of Fabre d'Olivet's startling results, his book 
seems to have made a strong impression on Whorf, who later character- 
ized him as having been "one of the most powerful linguistic intellects 
of any age." Whorf maintained that, while the Biblical exegesis at- 
tempted by Fabre d'01i\ et could not be taken seriously, his "root-sign" 
was really a foreshadowing of what is nowadays called the phoneme. 
What intrigued Whorf was Fabre d'Olivet's method. For example, in 
arriving at his "meanings" of the letters of the Hebrew alphabet, Fabre 
d'01i\'et had systematically compared and contrasted a wide variety of 
roots in which they occurred, much as one might attempt to obtain a 
"meaning" for the letter M in English by educing the common meaning 
in all English words beginning in M. We may permit ourselves to 
imagine that Fabre d'Olivet could have found a common element even 
in such opposite words as 'mother' and 'murder'! There are limits to 
which such a method can be pushed, which Fabre d'Olivet far exceeded; 
nevertheless, it remains true that such a technique of identifying isolates 
is in essence similar to the procedures of contemporary linguistics in 
identifying phonemes and morphemes. As we shall see, however, 
Whorf's methods in certain spheres of his work closely paralleled those 
of Fabre d'Olivet. Tliis is illustrated in his early efforts to read Maya 
hieroglyphs, as well as in some of his unpublished work on the structure 
of Aztec. Another and perhaps a more profound way in which his 
methods resembled Fabre d'Olivet's is represented in his always bold 
and penetrating search for inner meanings. Just as Fabre d'Olivet 
pushed imagination to the limit in looking for an underlying significance 
in a segment of a Hebrew root, so Whorf persisted in the struggle to 
wrest from the bare linguistic fact its ultimate purport. 

The discovery of the work by Fabre d'Olivet stimulated Whorf to 
read more widely and deeply on the subject of language. He made use 
of the rich collections of the Watkinson Library, a scholarly research 
library in Hartford founded in 1857 under the provisions of the will of 
a wealthy English-born Hartford merchant who wanted the city to have 
a general library of reference. Visited chiefly by an occasional genealo- 
gist or art historian who sought access to its hundred-thousand-odd 


dusty volumes, the library was housed in the upper reaches of a fortress- 
like building known as the Wadsworth Atheneum, which also contained 
the Hartford Public Library and the collections of the Connecticut His- 
torical Society.* At least during the period that Whorf was in the habit 
of frequenting it after his business hours, its extreme stillness and book- 
ish odor were conducive to deep concentration. Its first librarian had 
been James Hammond Trumbull, who was among other things a scholar 
in American Indian lore. During his service as librarian from 1863 to 
1893, Trumbull had built up the library's collections in American Indian 
ethnology, folklore, and language to an extent that would be considered 
unusual except for a large university library. This collection reawakened 
Whorf's interest in Mexican antiquities and lore, and directed his atten- 
tion especially to the Aztec (Nahuatl) language and, later, to Maya hiero- 
glyphs. Just what prompted Whorf to study Aztec, in particular, we do 
not know. Conceivably he chanced upon an account of Nahuatl which 
reminded him of the ideas he had found in La langue hebrdique. Be 
that as it may, Whorf began studying Aztec in 1926; he probably did not 
work seriously on Maya until 1928. He worked not only at the Watkin- 
son Library, but also at any library he could profitably visit on his 
numerous business trips away from Hartford. He made rapid progress, 
and began corresponding with various scholars in Mexican archaeology 
and linguistics, including Herbert J. Spinden of the Brooklyn Museum 
and Alfred M. Tozzer of Harvard University. At Dr. Spinden's sug- 
gestion he addressed himself to an attempt to work out a translation of 
a page of an old Mexican manuscript, a photographic reproduction of 
which was to be found in the Peabody Museum of Harvard University. 
The result was a paper read before the Twenty-Third International 
Congress of Americanists in September 1928 and a corresponding first 
scholarly publication, "An Aztec account of the period of the Toltcc 
decline" (1928),^ which shows an antiquarian's interest in the details of 
Toltec history and chronology as well as a linguist's pride in forcing a 
cranky Aztec word "to yield up its secret," as Whorf himself put it. This 
paper, as read before the Congress, attracted a considerable amount of 
attention and publicity for the young insurance agent, who was hailed 

* In 1952 the Watkinson Library was removed to spacious and modern quarters 
at Trinity College in Hartford. 

5 Sec the bibliography of Whorf's writings, pp. 271-276; this is followed by a 
short list of related writings to which references are made. 


in newspaper reports for his having "unlocked mysteries" which had 
"baffled" other scholars. About the same time there was completed 
another Aztec translation, published in 1929 under the title "The reign 
of Huemac," 

These publications were, howe\er, only the first and easy fruits of a 
period of study in which Whorf was also delving into comparative lin- 
guistics, presumably without any tutoring other than the necessarily brief 
contacts he may have had with such men as Spinden and Tozzer, and, 
in addition, J. Alden Mason of the University of Pennsylvania, whom he 
met during a visit to the first Linguistic Institute, held in the summer 
of 1928. At the International Congress of Americanists, Whorf had 
read another paper besides the one on Toltec history. It attracted much 
less attention, but was closer to his true interests; entitled simply "Aztec 
linguistics," it reported his assertion that Aztec was what he called an 
oligosynthetic language, that is, that all its words were built up out of a 
relatively few elements, perhaps as few as fifty basic monosyllabic roots, 
"each conveying a general notion capable of wide modulation without 
loss of the basic sense" (or so he wrote in the published abstract of the 
paper). On looking into whether the same roots he found in Aztec 
would show up also in languages related to Aztec, he was immediately 
gratified by the results. Toward the end of 1928 the work he had done 
on the familial relationships among Tepecano, Piman, and Aztec— all 
Mexican languages— appeared so promising to Tozzer and Spinden that 
they advised him to seek a research fellowship from the Social Science 
Research Council to enable him to obtain needed materials and work 
more intensi\ely. Whorf countered with the proposal that he use such 
a fellowship to make a field trip to Mexico to locate old Aztec manu- 
scripts for the Watkinson Library, one of the trustees of that library 
having expressed a desire to build up its collection of Aztec materials. 
Tozzer opined, however, that if he wanted to go to Mexico he would be 
better advised to investigate modern Nahuatl, a suggestion to which 
Whorf readily assented. In a letter to Mason dated December 6, 1928, 
Whorf commented, in reference to the fellowship for which he was 
applying, "It is a question whether I get it, because these Fellowships 
are supposed to be for men with Ph.D. degrees, and while they some- 
times make exceptions, these exceptions are rare and hard to get, requir- 
ing very good recommendations." His application to the Social Science 
Research Council was accompanied, first, by a general statement of his 


scholarly plans, and second, by a nearly completed article entitled "Notes 
on the oligosynthetic comparison of Nahuatl and Piman, with special 
reference to Tepecano. " In the first of these documents, Whorf re- 
vealed himself as avowedly visionary, but he may have felt its content 
necessary to win the interest of the committee which reviewed his fellow- 
ship application: 

With the aid of this Research Fellowship if possible I plan to do and 
publish sufficient work on Mexican linguistics to make the principle of oligo- 
synthesis a li\e topic and to interest other investigators in the basic sub- 
stratum of language to which it belongs. 

After I ha\e become better known in this way the next step will be to 
arouse interest in the phenomenon that I call binary grouping in Hebrew 
and the Semitic languages. I am of course still working on this and will 
continue right along to bring it to the attention of Semitic researchers. 

After binary grouping has also become a live topic I will begin to make a 
union between this principle and that of oligosynthesis and thereby reach 
the still deeper principle underlying the Hebrew and Semitic languages. 

The next step will be to use these principles in working out the primitive 
underlying basis of all speech behavior. This will amount to laying the 
foundations of a new science, and although this consummation lies some 
little time in the future I feel that it is quite distinctly in sight. Still further 
ahead are the possible applications of such a science restoring a possible 
original common language of the human race or in perfecting an ideal nat- 
ural tongue constructed of the original psychological significance of sounds, 
perhaps a future common speech into which all our varied languages may 
be assimilated, or, putting it differently, to whose terms they may all be 
reduced. This may seem at the present time very visionary, but it would 
be no more remarkable than what science has already done in other fields 
when it has got hold of sound principles to point the way, and I believe my 
work is tending to unfold such principles. And with the ultimate develop- 
ment of these researches will come manifestation of the deeper psychologi- 
cal, symbolic, and philosophical sense contained in the cosmology of the 
Bible, the starting point and original inspiration of these studies. 

Oligosynthesis is explained further in the first few paragraphs of the 
second document which accompanied Whorf 's fellowship application: 

Oligosynthesis is a name for that t^pe of language structure in which all 
or nearly all of the vocabulary may be reduced to a very small number of 
roots or significant elements, irrespective of whether these roots or elements 
are to be regarded as original, standing anterior to the language as we know 
it, or as never having had independent existence, theirs being an implicit 
existence as parts in words that were always undissociatcd wholes. 


Such a structure was recognized by the writer in the Nahuatl or Aztec 
language of Mexico, in which he has made numerous studies and the term 
ohgosynthesis was thereupon proposed by him. . . . Briefly, the conclusions 
are that nearly all and probably quite all the present known native vocabulary 
of Nahuatl is derived from the varied combination and varied semantic de- 
velopment of NO MORE THAN THIRTY-FIVE ROOTS, for which the wTitcr prcfcrs 
the name of "elements," each of which elements stands for a certain general 
idea, including something of the surrounding field of related ideas into which 
this central idea insensibly shades off. These thirty-five elements (it now 
begins to seem very unlikely that their number will be increased) have been 
obtained by extensi\ e stem-analysis and are listed ... in the appendix of the 
present article. They explain the meaning of thousands of Nahuatl words, 
including great numbers of words learned recently, which have had the 
meaning that would be expected from their elements. Furthermore, it is 
becoming increasingly evident to me that these elements are to be regarded 
as original roots anterior and ancestral to the present language, and my pre- 
vious view that they might be the result of assimilative back-formation is 
becoming less and less of a serious claimant for consideration. Obviously we 
have here a structure, a point-to-point correspondence between the path of 
ideation and the successions of lip, tongue, and glottal acti\ities (i.e. con- 
sonants and vowels) that may be of great linguistic, glottogonic, and psycho- 
logical significance. 

Binary grouping referred to a principle which Whorf believed to in- 
here in the structure of Hebrew roots, as may be seen in the following 
quotation from an unpublished manuscript: "A binary group is a group 
of Semitic roots having in common a certain seciuence of two con- 
sonants, containing all the roots with this sequence in one language, and 
ha\ing these roots with but few exceptions allocated to a few certain 


These quotations are of special interest when \iewed in the light of 
Whorf's early enthusiasm for Fabre d'Olivet. As a further note, it 
should be added that Whorf started to extend the application of the 
oligosynthetic principle to his first work with Maya, concerning which 
he read a paper before the Linguistic Society of America (of which he 
had just become a member) in December 1929, "Stem series in Maya." 
In the abstract of this paper submitted to the program committee of the 
Linguistic Society, we find Whorf pointing out that the majority of 
Maya stems beginning OE- bear the meaning 'turn.' He writes further: 
"So other series, e.g. OI-, radiate, glow, burn, scatter; 00-, 0U-, inward; 
BI-, move; TA-, connect; TZA-, come or bring together; MA-, pass. In 
other words, 'ideology follows phoneticism.' " 


The response of the SSRC to his apphcation for a research fellowship 
being favorable, Whorf set about making arrangements for his field trip 
to Mexico. He obtained a few weeks' leave from his company and left 
for Mexico City with Mrs. Whorf and her mother in January 1930. En 
route he spent a few days researching in the library of the Department 
of Middle American Research at Tulane University of Louisiana. On 
arrival in Mexico City, he sought the assistance of several Mexican 
specialists in Aztec, especially Professor Mariano Rojas of the National 
Museum of Mexico. Partly with their aid, he gained access to several 
excellent informants who spoke a form of Aztec which was believed to 
approximate, as closely as one could expect over the years, the classical 
dialect of Aztec once spoken in Tenochtitlan (now Mexico City) at the 
time of Montezuma. These individuals lived in an outlying suburb in 
the Federal District, known as Milpa Alta, and it is their dialect of 
which Whorf made a detailed linguistic analysis, published posthu- 
mously in 1946 in Hoijer's Linguistic structures of native America. In 
the meantime, Whorf poked around the Mexican countryside in search 
of suggestive archaeological material. In a rmned temple overlooking 
the village of Tepoztlan, where he conducted further linguistic studies, 
he came upon, apparently quite by chance, a band of sculptured figures 
which had previously escaped the close attention of scholars. His sharp 
observation and close familiarity with both Aztec and Maya graphic art 
enabled him to recognize almost immediately that these figures deviated 
from their usual forms as "day signs" of the Aztec calendar and showed 
certain resemblances to Mayan characters. This discovery of "a definite, 
clearly demonstrable rapport between Nahuatl hieroglyphs and early 
Maya ones," as Whorf regarded it, was the basis of one of the papers 
reprinted in the present collection, "A central Mexican inscription com- 
bining Mexican and Maya day signs." It furnishes an excellent example 
of Whorf's methods of work, and is also his earliest publication pointing 
toward his later researches into Maya hieroglyphs. 

For several years after his return from Mexico, Whorf occupied him- 
self with working up the data amassed during the Mexican sojourn. Not 
only was it necessary to sketch out the linguistic analysis of the Milpa 
Alta Nahuatl; it was also urgent to follow up the leads provided by the 
discoveries concerning Mexican and Maya day signs, which had had the 
effect of confirming or modifying certain hunches he had developed pre- 
viously. A major series of publications concerning Maya hieroglyphs 


started about this time, first with a monograph pubhshed by the Pea- 
body Museum of Hansard, "The phonetic value of certain characters in 
Maya writing" (1933) and later with an article, "Maya writing and 
its decipherment" (1935). In the earlier publication, which Professor 
Tozzer of Harvard urged him to prepare, he set forth in detail and with 
his evidence his thesis that Maya writing was at least partly phonetic, 
and he offered a specimen translation of a simple Maya text from one of 
the codices. Since the hypothesis of phoneticism in Maya writing had 
been all but abandoned by Maya scholars at least fifty years previously, 
Whorf's materials must have been exceedingly impressive, at least at 
the time, to have gotten published at all. The later publication, "Maya 
writing and its decipherment," was a reply to a critique published by 
Richard C. E. Long in the journal Maya Research. Besides taking issue 
with many points of detail raised by Mr. Long, Whorf attempted to 
explain why he regarded Long's approach as fundamentally mistaken; 
he also offered another specimen translation of a Maya text. In this 
article, furthermore, he mentioned that he had been working on a manu- 
script, "First steps in the decipherment of Maya writing," which he then 
hoped to publish within a short time. This manuscript, found among 
Whorf's papers, has thus far remained unfinished and unpublished, 
though some parts of it are reflected in a paper which Whorf read be- 
fore a scientific congress in 1940 and which is reprinted in the present 
collection, "Decipherment of the linguistic portion of the Maya hiero- 
glyphs." Whorf was bitterly disappointed by the rather cool reception 
generally accorded his work by Maya scholars after 1933; he was entirely 
confident that his linguistic approach held the key to the interpretation 
of Maya hieroglyphs. The paper he read in 1940 was apparently a last- 
ditch effort to win support for his approach. 

Up to the time of the Mexican trip, Whorf seems to have had very 
little contact, either face-to-face or by correspondence, with any of the 
persons who were later to be his close colleagues in the field of lin- 
guistics. His scholarly relations had been chiefly with a group of special- 
ists in Mexican archaeology, none of whom were particularly well quali- 
fied in, or concerned with, general linguistics. In view of this, the 
competence that Whorf had achieved in general linguistics and lin- 
guistic field methods, purely on the strength of his own untutored 
study, was remarkable. Nevertheless, his talents might never have fully 
matured if he had not eventually met Edward Sapir (1884-1939), a 


foremost authorit}^ not only on American Indian languages but also on 
the general science of language. Whorf had, of course, been aware of 
Sapir's work, and he had doubtless read Sapir's book Language (New 
York, 1921) with intense interest. He first met Sapir, though only 
briefly, at the September 1928 International Congress of Americanists, 
and talked with him further at subsequent meetings of scientific so- 
cieties in 1929 and 1930. He could not, however, make any close con- 
tact with Sapir until the latter came from the University of Chicago in 
the fall of 1931 to take his post at Yale as Professor of Anthropology to 
teach linguistics. Whorf lost no time in enrolling in Sapir's first course 
at Yale in American Indian linguistics; among the Whorf papers can be 
found a manuscript, entitled "The structure of the Athabaskan lan- 
guages," a term paper which Sapir awarded a grade of A and much 
praise. Although Whorf nominally enrolled for a program of studies 
leading to the doctorate, he never sought or obtained any higher degree; 
he pursued his studies for pure intellectual ends. The effects of Whorf's 
first formal studies in linguistics were noticeable and demonstrable. His 
early interests in "oligosynthesis," "binary grouping," and other unusual 
linguistic theories became tempered at least to the extent that he was 
brought to see them in the light of the accumulated experience of men 
like Sapir. (I can find no mention of the idea of oligosynthesis, as such, 
in any of Whorf's writings after 1931.) More important, Whorf was 
put in close touch with the linguistic theories and techniques which 
were most advanced at that time, as well as with the problems which 
were currently considered the most essential to solve. Finally, his 
studies at Yale brought him into contact with a small but earnest band 
of Sapir's students, including such individuals as Morris Swadesh, Stan- 
ley Newman, George Trager, Charles Voegelin, Mar}' Haas, and Walter 
Dyk, all of whom have since made important contributions to linguistics 
or anthropology. In 1937-38, Whorf was a Lecturer in Anthropology at 

Whorf's association with Sapir thus served to intensify his desire to 
develop further the field of American Indian linguistics. In the mono- 
graph about Maya hieroglyphs published in 1933, we find Whorf credit- 
ing Sapir with certain suggestions about the interpretation of the zero 
sign. Sapir was probably more influential, however, in encouraging 
Whorf to expand his \^•ork on the Uto-Aztecan languages (a large stock 
of languages whose relationships Sapir had established), in particular 


to take up the study of the Hopi language, a distant relative of Aztec. 
In December 1932 Whorf read a paper entitled "The characteristics of 
the Uto-Aztecan stock" at the meeting of the Linguistic Society of 
America held at New Haven. His further work on Uto-Aztecan lin- 
guistics (exclusive of Hopi) is represented by a review (1935) of Kroeber's 
Uto-Aztecan languages of Mexico, and several articles, "The compara- 
tive linguistics of Uto-Aztecan" (1935), "The origin of Aztec TL" (1937), 
and (with G. L. Trager) "The relationship of Uto-Aztecan and Tanoan" 
(1937). In these articles, Whorf recognized a superfamily language stock 
which he proposed to call Macro-Penutian, to include Penutian, Uto- 
Aztecan, Mayan, and Mixe-Zoque-Huave. Later, he used this structur- 
ing in preparing a revision of Sapir's classification of American Indian 

In linguistic work as such, Whorf was best known for his studies of 
Hopi. Perhaps through the good offices of Sapir, he made contact with 
a native speaker of Hopi, who then lived, conveniently enough, in New 
York City. Beginning in the spring of 1932 and with the support of a 
small research subvention obtained for him by Sapir, Whorf worked 
intensely on developing a linguistic analysis of Hopi, utilizing field re- 
search methods in which he had received instruction from Sapir. Whorf 
and his informant exchanged visits in New York and W^ethersfield 
(where WHiorf resided); in 1938 Whorf was able to spend a short time 
on the Hopi reservation in Arizona. By 1935 he had prepared a tenta- 
tive grammar and dictionary of Hopi. Except for the brief sketch of 
Hopi grammar in Hoijer's Linguistic structures of native America 
(1946)— a sketch prepared by Whorf in late 1939— the major outcomes of 
these studies remain unpublished. Nevertheless, one can be grateful for 
the two brief but very influential technical articles about Hopi which 
Whorf published during his lifetime: "The punctual and segmentati\e 
aspects of verbs in Hopi" (published in 1936, first read as a paper before 
the Linguistic Society of America in December 1935), and "Some verbal 
categories of Hopi" (1938). In these papers one can see how their 
author was beginning to be fired with the notion, de\cloped more ex- 
tensively in later popularized papers, that the strange grammar of Hopi 
might betoken a different mode of perceixing and concei\ing things on 
the part of the native speaker of Hopi. In the first, he asserted that 
"the Hopi actually ha\e a language better equipped to deal with . . . 
vibratile phenomena than is our latest scientific terminology." lliis was 


followed by "An American Indian model of the universe" (probably 
written in 1936 but not published until 1950), which explores the im- 
plications of the Hopi verb system with regard to the Hopi conception 
of space and time. The work with Flopi must have also influenced the 
writing, about this same time, of the paper entitled "A linguistic con- 
sideration of thinking in primitive communities" (published in the pres- 
ent collection for the first time). "Some verbal categories of Hopi" 
(1938) discusses several interesting distinctions which Hopi makes be- 
tween kinds and modes of happening which English treats as the same, 
and "Linguistic factors in the terminology of Hopi architecture" (written 
in early 1940 and not published until 1953) contains the thesis that the 
Hopi mind automatically separates the "occupancy" or spot of ground 
or floor on which the occupancy occurs from the use to which the occu- 
pancy is put, whereas the speaker of English tends to merge these, as 
where "school" is thought of as both an institution and a building. (In- 
deed, do we not almost instinctively feel that an institution must of 
necessity be housed in some kind of building?) Whorf is probably best 
known for the article "The relation of habitual thought and behavior 
to language," written in 1939, and for the three articles which were pub- 
lished in 1940 and 1941 in the Technology Review— all based to a con- 
siderable extent upon his research in Hopi. What is important to note 
is that, first, these latter papers were grounded upon a solid foundation 
of linguistic analysis done much earlier, and, second, that the ideas of 
linguistic relativity expressed therein were by no means new in Whorf's 
mind; on the contrary, the seeds of these ideas were already apparent in 
materials prepared as early as 1935, if not earlier. 

The three articles written for the Technology Review of M.I.T. and 
the article entitled "Language, mind, and reality" and published in an 
Indian journal of theosophy were addressed to lay audiences. Undoubt- 
edly Whorf had it in mind to bring linguistics before the general public 
in a manner that had scarcely e\er been attempted; in fact, he may be 
credited with being the first popularizer of modern linguistic science. 
He realized, however, that it would be impossible to popularize lin- 
guistics, and there would be little purpose in doing so, unless linguistics 
liekl a message of poj5ular appeal. This message, Whorf believed, was 
that linguistics has much to say about how and what wc think. 

It may be of interest to recount what led to the writing of the articles 
for the Technology Review. As early as 1932 there was an exchange of 


correspondence between Whorf and the editor of the Technology Re- 
view, J. R. Kilhan Jr. (now president of M.I.T.), to whose attention 
Whorfs article on the Mexican day signs had come. Kilhan invited 
Whorf to write an account of his trip to Mexico, and asked whether he 
had unearthed any material relating to "the history of engineering, archi- 
tecture, and the practical sciences." Almost disdainfully Whorf replied 
that this trip had "no bearing on engineering, architecture, or the prac- 
tical sciences" — that "the investigation was conducted thoroughly in the 
spirit of scientific research, but it was in the social sciences, not the 
physical sciences." Nevertheless, he did eventually agree to prepare an 
article on his trip, but for some reason such an article does not seem to 
have been written. The next exchange was between Whorf and Presi- 
dent Karl T. Compton, in late 1939. Whorf initiated this correspond- 
ence as a result of "a slight difficulty" he had had in filling out some sort 
of questionnaire sent him by the Register of Former Students— namely, 
that the questionnaire omitted any mention of insurance, insurance en- 
gineering, fire prevention, or the like, fields which Whorf felt ought to 
receive recognition as engineering professions. He wanted to bring this 
apparent omission to President Compton's attention, and proceeded to 
describe in detail the nature of his own work. In this same letter, he 
took occasion to mention other activities of his which did not quite fit 
the list of rubrics on the questionnaire, to wit, his research on American 
Indian languages. President Compton's reply, after explaining that the 
apparent omission on the questionnaire form was the result of abbrevia- 
tion rather than deliberate exclusion, expressed interest in Whorfs avo- 
cational work and asked permission to submit his letter for publication 
in the Technology Review. Permission was granted, and the letter (in 
much condensed form) is to be seen in the January 1940 issue of the 
Review. In the course of ensuing correspondence, the editor of the 
Review, then F. G. Fassett, Jr., wrote Whorf on November 14, 1939: 
"Your linguistic studies offer a very interesting and provocative possi- 
bility to anyone responsible for a magazine. 'Inasmuch as the analysis 
of reality is a matter of language, and the relativity of such analyses can 
only be appreciated through studies that show the immense range of 
possible diversity in linguistic expression it will be seen that there is a 
connection here with the attempts of science to understand the universe 
and man'— I think it would be very interesting to see the ideas implicit 
in this statement from your October letter expanded in an article aimed 


at the Review group. Is this prospect of interest?" Evidently it was, 
for Whorf was able to submit the first article, "Science and linguistics," 
as early as January 30, 1940, and it was printed in the Review shortly 
thereafter. The exceedingly warm reception this article got, both from 
regular readers of the Review and from the recipients of reprints, pointed 
to the desirability of further articles. The second article, "Linguistics as 
an exact science," was submitted on September 16, 1940, and the third 
and last, "Languages and logic," on February 14, 1941, at a time when 
Whorf's health was failing and his physical weakness was already ap- 
parent in his handwriting. 

Even in the year when he was writing these brilliant articles on lin- 
guistics and at the same time fighting off ill health, Whorf became con- 
sumed with still another interest. On account of a lecture which his 
oldest son attended and described to him, Whorf became acquainted 
with Fritz Kunz, a well-known speaker and writer, at present the execu- 
ti\e vice-president of the Foundation for Integrated Education, Inc. 
Kunz and Whorf had many common interests, especially (as Mr. Kunz 
has written me) in the philosophy and metaphysics of India, and it was 
this that brought them to work together quite closely. One result of 
this friendship was Kunz's suggestion that Whorf write an article about 
linguistics for a theosophical journal published at Madras, India; the 
article entitled "Language, mind, and reality" was the result. Kunz was 
on the point of founding a new journal, Main Currents in Modern 
Thought, and Whorf was of great help in putting out the first few issues, 
in late 1940 and early 1941. The journal (still published today, but in 
a different format) was of an interesting and unusual character; it was 
intended as a clearing house of ideas and information in all sorts of 
fields in natural science, social science, the humanities, mathematics, 
logic, and philosophy; it was to be written chiefly by its subscribers. 
Published in those days in mimeographed form, the varied colors of 
its pages were keyed to subject matter. Whorf wrote literally dozens 
of pages in the first volume, doing highly creative book re\ iews and con- 
tributing little abstracts on such dixersc subjects as "The Hurrians of 
old Chaldea," "Shrinking glass," and "Notes on the demonstration of 
'wetter' water." One of his reviews is based on two books on the eco- 
nomics of primitive societies, and is provocati\ely titled "We may end 
the war that is within all wars that are waged to end all war." "These 
books," Whorf wrote, "are outstanding examples of a type of in\'esti- 


gation that is gradually unsettling the old-style materialistic theory of 
economics. And since both Marxian communism and private capital- 
ism are based on a stereotyped materialistic formulation of economics, 
such irrefutable scientific expositions of the fact that economic behavior 
is conditioned by culture, not by mechanistic reactions, may be the fore- 
runner of a NEW ERA." This quotation is only one of many that could 
be cited to show Whorf's broad humanism and concern for the com- 
monweal. Nor did Whorf neglect to bring the implications of lin- 
guistics to the attention of readers of Main Currents. Reporting several 
interesting facts which had appealed to his interest at a scientific meet- 
ing, he wrote, in a piece entitled "A brotherhood of thought": 

There is no word for 'word' in Chinese. The nearest thing is the element 
tsz, which is translated 'word' but means rather 'syllable' or 'syllabic ele- 
ment.' Many such elements never occur free but only in a few combina- 
tions, like the 'pyr-' in 'pyrometer.' Words in the sense of vocabulary units 
exist as either of one or two syllables, a fact obscured by the traditional 
Chinese system of writing, which keeps every syllable separate. This was 
pointed out by Dr. Y. R. Chao of Yale in a paper "Word conceptions in 
Chinese" at the meeting of the Linguistic Society of America in Providence, 
R. I., 12/50/40. The nature of Chinese grammar is only just beginning to 
be understood; Dr. Chao and others have exploded the idea that Chinese is 
a monosyllabic language. At the same meeting Dr. G. A. Kennedy of Yale, 
analyzing "Complex attributive expressions in Chinese," showed that Chinese 
has no relative clauses, and that a different kind of order-system rules the 
logic of such relationships. If the element te used in this logic be translated 
'-ish,' then "The House that Jack Built" would go in Chinese: 'This is 
Jack-ish build-ish house; this is Jack-ish build-ish house-ish in-ish lie-ish 
malt,' etc. 

It is not sufficiently realized that the ideal of worldwide fraternity and 
cooperation fails if it does not include ability to adjust intellectually as well 
as emotionally to our brethren of other countries. The West has attained 
some emotional understanding of the East through the esthetic and belles- 
lettres type of approach, but this has not bridged the intellectual gulf; we 
are no nearer to understanding the types of logical thinking which are re- 
flected in truly Eastern forms of scientific thought or analysis of nature. 
This requires linguistic research into the logics of native languages, and 
realization that they have equal scientific validity with our own thinking 

After a long and lingering illness, during which he valiantly struggled 
to keep up his study and his writing, Whorf succumbed on July 26, 1941, 
at the age of 44. He had accomplished more than he knew, yet only a 


small part of what he might have done. His passing was taken notice of 
by editorial obituaries not only in the local newspapers but also in such 
papers as the New York Times, and later, of course, in several scholarly 

I cannot close this biographical memoir without a few remarks about 
Whorf's personality and habits of work. Above all he was capable of 
extremely deep and steadfast concentration in everything he did. Noth- 
ing was treated lightly or carelessly. His penciled manuscripts, in beau- 
tifully neat and always legible handwriting, exemplify his meticulous- 
ness; it is also exceedingly rare to find an error in his typescripts (he 
nearly always did his own typing of scholarly manuscripts and corre- 
spondence). He was willing to work almost endlessly; his published 
writings represent but a small fraction of the manuscript material he 
produced, and his notebooks are truly voluminous. Without hesitation, 
he would if necessary go to the trouble of copying out, in longhand or 
script printing, page after page of detailed linguistic texts. In writing, 
he was able to express himself artistically, convincingly, and effortlessly; 
in many cases, his first draft— with a minimum of emendations— was 
final. Yet, he nearly always made a pencil draft preparatory to typing, 
even for correspondence. This tireless devotion to scholarship un- 
doubtedly took its toll of Whorf's strength and health, although he 
never appeared visibly fatigued. He habitually worked late into the 
night, relaxing himself only by taking short catnaps or by pla}ing a few 
rolls of classical music on his mechanical grand piano. He was some- 
what casual about his hours at the office, being dilatory in both arrival 
and departure on many occasions, but he accomplished much while 
there. For exercise, he enjoyed walking, often making the four- or five- 
mile trip from his office to his home in Wethersfield on foot, with 
perhaps a stop at the Watkinson Library en route. 

Social life held little significance for him except when it invoKcd his 
linguistic colleagues, to whom he was always a delight. He maintained 
an air of cheerful curiosity, and continually had interesting and novel 
things to say. As I have recalled elsewhere, "Whorf was a quiet, con- 
templative teacher; he would not stop at remaining silent for a seem- 
ingly interminable time while searching his mind to recall something or 
to think through a problem. Yet, when he became prompted to tell me 
of some new insight he had reached, the smoothness and lucidity of his 
remarks were little short of awesome. His mode of behavior was that 


of neither the scholar nor the businessman— he gave only the impression 
of calm, unhurried, effortless inspiration. Self-seeking was entirely alien 
to him and it is a tribute to him that he was so generous in sharing his 
remarkable perspectives with others." 


The title of this volume, Language, thought, and reality, is the title of 
a book that Whorf hoped to write, and for which a brief outline is to be 
found among the papers left at his death. The book would have been 
dedicated to the memory of Edward Sapir and Antoine Fabre d'Olivet, 
and would have attempted to present the implications of linguistics for 
the clarification of our thinking about the external world of reality-. The 
notes indicate that the book, designed as a college textbook and pro- 
vided with suitable chapter-end quizzes, would have included in its 
appendix language sketches of Latin, Greek, Hebrew, Kota, Aztec, Hopi, 
Shawnee, Russian, Taos, Chinese, and Japanese. This book, of course, 
was never written, but I believe its title is a fitting one for the present 
edition, which includes nearly all of Whorf's writings which are perti- 
nent to what he called the principle of linguistic relativity, which states, 
at least as a hypothesis, that the structure of a human being's language 
influences the manner in which he understands reality and behaves with 
respect to it. This edition also includes what are believed to be the 
most interesting and useful writings of Whorf in Middle American lin- 
guistics and in general linguistics. 

A study of the whole procession of Whorf's writings discloses an 
underlying theme which had its roots in his very early thinking, perhaps 
concurrently with his first steps in linguistic work. We have already 
seen how Whorf tried, as early as 1925, to verify the theories of the 
French mystic and Hebrew scholar, Fabre d'Olivet— theories which 
proposed that certain Hebrew letters and combinations of letters con- 
tained mysterious, fundamental root-ideas. In order to verify these, 
Whorf found himself playing with subtle, below-the-surface similarities 
between seemingly unrelated ideas. This was the first step— to pene- 
trate beneath the veneer of dry, brittle, isolated words which might 
conceal fundamental concepts. We see Whorf's concern with basic 
mental operations, and his seeming discomfort with the straitjacket 
represented by language, in a short essay which I have entitled "On the 


connection of ideas." This essay, printed here for the first time, was 
written in 1927 as a letter to the psychologist Horace B. English, who 
had just authored a dictionary of psychological terms; Whorf was ap- 
pealing to English to supply him with a term for a new kind of asso- 
ciation of ideas. Whorf was groping toward concepts or terms of a more 
general or abstract nature than those provided by any language. None 
of the psychological schools on the contemporary scene were of any real 
help, as Whorf complained in a short unpublished note which, supplied 
with the wholly arbitrary title "On psychology," is also printed here. 
Yet, much of Whorf's work is extremely close to psychology. The 
search for root-ideas led Whorf on many a byway, even in his work with 
Aztec and Maya linguistics. The two papers on Maya hieroglyphs in- 
cluded here give only a glimpse of this fact, but it would have emerged 
very obviously if it had been thought desirable to print here, for 
example, the unpublished paper "Stem series in Maya and certain Maya 
hieroglyphs," to which reference has already been made. Unfortunately, 
one has the suspicion that Whorf occasionally allowed himself too 
many liberties in arraying together ideas which to another would appear 
totally unrelated. For example, at one point in his "Stem series" paper, 
he cites a series of Maya roots which he believed contained "all sorts 
of ideas of dispersal: be dispersed, disappear, spread, radiate, shine." 
Edward Sapir, to whom Whorf at one time submitted the manuscript, 
pinned this comment to it: "I am sorry, but I cannot honestly say I 
feel the cohesiveness of the sa-set as clearly as you do. 'Sand,' 'white,' 
'weave cloth,' 'much,' and 'dislocate,' for instance, on the basis of funda- 
mental 'dispersal' seems to be a purely subjective construction." ^ Whorf 
became aware of the need for objectifying semantic inferences; there is 
to be found among his unpublished papers a brief proposal for an ex- 
periment in which an individual would be presented with a series of 
Aztec words, together with their English meanings, all involving the 
letters ZE-. The subject was to be asked to arrive at some sort of 
semantic grouping for these words; he was to be told that he might 
decide (1) that ZE- has a single meaning throughout, or (2) that there 
are two, three, or more ZE-'s with distinct and quite unconnected mean- 

® The material on the sa- group was considerably reworked before publication in 
the monograph "The phonetic value of certain characters in Maya writing" (1933), 
p. 11. 


ings, or (3) that ZE- has no connection with the meanings whatsoever. 
Apparently, Whorf never performed this experiment. 

It will also be recalled that Whorf early noticed, or believed he 
noticed, that Hebrew, Aztec, and Maya seemed to be built on a dif- 
ferent plan from that of English and other languages which he later 
called "SAE" (standard average European) languages. He called them 
"oligosynthetic" languages, that is, languages whose vocabularies were 
built up out of a very small number of elements. "Each element," he 
stated in a paper concerning Aztec linguistics read in 1928, "is, first, a 
very simple piece of articulation-behavior, and, second, a broad idea or 
complex of related ideas that goes with this piece of behavior." He 
thought he had been able to analyze the Aztec vocabulary into no more 
than thirty-five such roots. "It should now be noted," he continued, 
"that this oligosynthetic phenomenon opens up certain new territories 
in the little-explored field of language-psycholog}'. In it we see the 
whole ideational field of a language partitioned out among or shared 
between some thirty-five elementar>' notions, so as to give us for the 
first time a map or plan of an actual realm of ideas. Previously when 
ideas have been distributed among a set of categories, those categories 
have been the result of some philosopher's introspection, but not so this 
idea-map of a language— we come upon it as we do upon the facts of 
nature, and its as yet dimly seen configuration challenges us to investi- 
gate it by experimental and inductive methods." In these somewhat 
daring ideas we may see, first, a certain appeal to the notion of phonetic 
symbolism, the notion that there may be inherent relations (over and 
above the arbitrary relations established in any given language) between 
sounds and meanings, and, second, the faint suggestion of a theory of 
linguistic relativity. The problem of phonetic symbolism has long chal- 
lenged both linguists and psychologists. Edward Sapir, who was sympa- 
thetic to the notion, performed an experiment which pointed in a 
positive direction,^ and the problem is still alive among contemporary 
psychologists. With reference to the theory of linguistic relativity fore- 
shadowed in Whorf's theory of oligosynthesis, the key is the notion of 
a "broad idea or complex of related ideas" which might be associated 
with a linguistic element, for from this it is but a short step to the notion 
that languages with different collocations of semantic ideas might pro- 

^ Edward Sapir, "A study in phonetic symbolism," /. exp. Psychol, 12:225-239 


vide different "maps" of the realm of possible ideas, or, as Whorf put 
it much later, that different languages might provide different "seg- 
mentations of experience." 

The idea of linguistic relativity did not emerge in a full-fledged form 
until after Whorf had started studying with Sapir. Not until he began 
to analyze Hopi, a language with a grammar much more complex and 
subtle than that of Aztec or even that of Maya, did he begin to appre- 
ciate that the notion of linguistic relativity could be developed in a 
much more telling and effective way by noticing differences not only in 
Vlexation" but also in grammatical structure. The various papers on 
Hopi which are published in this collection will speak for themselves; 
in them Whorf tells of his provocative discoveries in the tense and 
aspect systems of the Hopi verb, in the Hopi treatment of noun classes, 
and so on. 

Whorf's whole outlook in linguistics, apart from his early religious 
concerns, stemmed from his concern with fundamental problems of 
meaning, or, as I like to think, with fundamental intellectual operations. 
In the very interesting and revealing paper written about 1936 and 
printed for the first time in this volume, "A linguistic consideration of 
thinking in primitive communities," Whorf insists that "linguistics is 
essentially the quest for meaning." The "real concern" of linguistics, 
he writes, "is to light up the thick darkness of the language, and thereby 
of much of the thought, the culture, and the outlook upon life of a 
given community with the light of this 'golden something,' as I have 
heard it called, this transmuting principle of meaning." Whorf was 
concerned more with substance than with process. That is, he was more 
interested in what, in some abstract sense, was being thought about 
than with the mental processes by which one might think, and this out- 
look led him to linguistics, full of "content," rather than to psychology, 
relatively "contentless" in its concern with generalized stimulus-response 
mechanisms. Whorf appeared to believe, indeed, that the content of 
thought influences the process of thought, or that differing contents pro- 
duce differing species of process, so that generalization about process 
becomes impossible without content's being taken into account. It was 
his belief that differences in thought content and their corresponding 
effects on thought processes and behavior in general would be spectacu- 
larly revealed by comparison of different language structures. He was 
extremely ingenious in ferreting out both the obvious and the subtle 


differences in language structures, and fully demonstrable differences 
they were, at least on the linguistic level. He did not stop there, how- 
ever; he attempted also to adduce evidence of behavior variations asso- 
ciated with different language phenomena. While this attempt may 
not have been wholly successful, it was at any rate embodied in the 
article written in the summer of 1939 for the Sapir volume, "The rela- 
tion of habitual thought and behavior to language." It was the last 
article he wrote on the subject which was addressed chiefly to his col- 
leagues. The principle of linguistic relativity was stated in the most 
appealing terms, however, in the articles which appear as the last four 
of this collection; they were written primarily for lay audiences. 

Whorf's principle of linguistic relativity, or, more strictly, the Sapir- 
Whorf hypothesis (since Sapir most certainly shared in the development 
of the idea) has, it goes without saying, attracted a great deal of atten- 
tion. Through various reprintings of one or the other of the Technology 
Review articles, beginning as early as 1941 (in Hayakawa's Language in 
action, a Book of the Month Club selection), the material has been 
brought to the notice of a wide public as well as of linguists, anthro- 
pologists, and psychologists. One wonders, indeed, what makes the 
notion of linguistic relativity so fascinating even to the nonspecialist. 
Perhaps it is the suggestion that all one's life one has been tricked, all 
unaware, by the structure of language into a certain way of perceiving 
reality, with the implication that awareness of this trickery will enable 
one to see the world with fresh insight. Surely, at any rate, it would 
have been farthest from Whorf's wishes to condone any easy appeal to 
linguistic relativity as a rationalization for a failure of communication 
between cultures or between nations. Rather, he would hope that a 
full awareness of linguistic relativity might lead to humbler attitudes 
about the supposed superiority of standard average European languages 
and to a greater disposition to accept a "brotherhood of thought" among 
men, as he wrote in the short article of that title quoted above. But, 
even if research in native languages is not for the purpose of helping to 
bridge intellectual gulfs between cultures, Whorf would nevertheless 
aver that the investigation of the "logics" of those languages will con- 
tribute to our understanding of our own thinking habits. 

In truth, the validity of the linguistic relativity principle has thus far 
not been sufficiently demonstrated; neither has it been flatly refuted. It 
seems to be agreed that languages differ in many strange and striking 


vva\'S, but it is a moot point whether such differences in language struc- 
ture are associated with actual differences in ways of perceiving and con- 
ceiving the world. Among the writers who are most impressed with the 
possibilities of such an association are Kluckhohn and Leighton (1946), 
Laura Thompson (1950), Hoijer (1953), and Kluckhohn (1954). Kluck- 
hohn and Leighton, for example, state that the Navaho tongue is so 
radically different from ours that an understanding of Navaho linguistic 
structure is virtually a prerequisite to understanding the Navaho mind; 
they cite the tremendous translation difhculties existing between Navaho 
and English, and imply that the two languages almost literally operate 
in different worlds. Hoijer claims to have found a suggestion of a cor- 
relation between the world view implied by the Navaho verb system 
(that people only "participate" or "get involved" in acts rather than 
initiate them) and the passivity and general restlessness or fatefulness of 
Navaho mythology. 

Two sharp critics of Whorf's methodology and conclusions, on the 
other hand, have been Lenneberg (1953) and Feuer (1953). Lenneberg 
attacks chiefly the methodology. First, he criticizes on several grounds 
the technique of translation which Whorf so often employed to dem- 
onstrate differences in languages; large differences in the linguistic 
handling of an event like cleaning of a gun do not necessarily imply 
corresponding differences in the perception of that event, and may 
merely result from metaphorical developments in the language, of which 
the speakers may not ordinarily be aware (just as we do not ordinarily 
think of "breakfast" as breaking a fast). Second, Lenneberg insists that 
linguistic and nonlinguistic events must be separately observed and de- 
scribed before they can be correlated, and that the usual canons of evi- 
dence must be applied in demonstrating any association between such 
events. Otherwise, the linguistic relativity principle becomes embar- 
rassingly circular, or at least tautological, in that the only evidence for 
differences in "world view" turns out to be the linguistic differences. 
Feuer, a social philosopher, belie\es that on a priori grounds one would 
not expect cultures speaking different languages to have different ways 
of perceiving space, time, causation, and other fundamental elements 
of the physical world, because a correct perception of these elements is 
necessary to survival. 

Since these and other logical, methodological, and psychological dif- 
ficulties have recently received a thorough discussion at the hands of a 


special conference of linguists, anthropologists, psychologists, and phi- 
losophers (Hoijer, 1954), it seems pointless to labor them here. It is 
perhaps desirable to counteract, however, the essential!}' negative, pessi- 
mistic tone which per\aded this conference, by pointing out that ex- 
tremely little research of an appropriate character has thus far been 
conducted on the Sapir-Whorf hypothesis. Except for the experiment 
reported by Brown and Lenneberg (1954), which showed that differences 
in ability to recognize and remember colors were associated with avail- 
ability of specific color names, there have been virtually no researches 
which have adequately tested the existence of correlations between lin- 
guistic structure and nonlinguistic behavior. Numerous suggestions 
pointing toward such research have been made in a monograph edited 
by Osgood and Sebeok (1954). 

There is a further consideration which has not been sufficiently 
stressed in the various discussions of the Sapir-Whorf hypothesis, 
namely, that the principle of linguistic relativity may not be so tauto- 
logical as it has been made to appear. It has been said that one merely 
states a tautology when one appeals to differences in languages as show- 
ing differences in behavior, in "world \'iew." It has also been said that 
it is necessary to find NONlinguistic behaviors which are correlated with 
the linguistic differences. This would doubtless be desirable, but there 
is something to be said for being interested in linguistic differences as 
such, regardless of nonlinguistic behavioral correlates. If we assume 
that there is such a thing as covert, implicit behavior, consisting of 
mental states, sets, attitudes, "mediational processes," and the like, we 
shall have to grant that such beha\ior is largely inaccessible to public 
observation save through the medium of verbal report. Whether or 
not, in fact, we assume any mental processes standing behind them, we 
are led to put a high value on verbal responses in their manifold forms 
as the chief data relating to perception and cognition. 

Suppose, now, we found that, by varying certain environmental con- 
ditions, we could produce corresponding changes in the \erbal reports 
made by speakers of a given language. For example, we might be able 
to find that we could in this way control which of se\eral words (each 
standing for one of several environmental stimuli) was used as the sub- 
ject of a sentence reporting the situation. Suppose further that, upon 
experimenting with speakers of another language, we found it impossible 
to produce changes in sentence structure corresponding to the varying 


environmental conditions— that all speakers invariably used the lin- 
guistic expression for one of the several stimuli as the subject of a sen- 
tence reporting the situations, and that, upon being questioned, the 
speakers of this language stated that it would be "unnatural" or "non- 
sense" to use any other linguistic expression in subject position in the 
sentence. The difference between the language behaviors would then 
be of interest in itself; we would have to conclude that we must take 
language structure into account in describing the verbal behavior of 
speakers in selecting the subject component of sentences. If, further, 
we had some fundamental knowledge about the grammatical meaning 
of sentence subject, we might be able to make some comparisons of the 
cognitive processes of speakers of the two languages. For example, sup- 
pose that the grammatical meaning of sentence subject were to be 
"entity perceived as a potential agent," then we might be permitted to 
conclude that the speakers of the second language in our illustration do 
not readily perceive certain stimuli as "potential agents." 

This illustration has had to be somewhat far-fetched in several re- 
spects: partly that we do not yet know whether differences of the sort 
described could be found between speakers of different languages, and 
partly that we do not know very well how to specify behavioral cor- 
relates of grammatical categories. Yet, it is this kind of language dif- 
ference which Whorf offers on an intuitive basis. He assumes that 
differences bet^^'een languages would be found to correspond to differ- 
ences in ways of reporting e\ents, and that we can feel intuitively the 
grammatical and behavior forces underhing the linguistic phenomena 
which he describes. It must be granted that we have only begun to 
obtain detailed information about language differences and the be- 
havioral pressures exerted by these differences, but, even when we get 
this information, much of it will be strictly in the realm of verbal be- 
havior and the environmental stimulus complexes which evoke this 
verbal behavior. 

One caution that needs to be stated in connection with the linguistic 
relativity principle, regardless of whether it is valid or not, is that the 
interest it has aroused and will continue to arouse should not be allowed 
to distract attention from the importance of language universals. Lan- 
guage universals, phenomena found in all languages, would be of as 
much interest psychologically as language differences. Is it true that 
all languages have subject-predicate construction in sentences? Do all 


languages have some type of noun-verb contrast? What features of 
verb-tense system are common to all languages? Answers to such ques- 
tions would assist in the development of a generalized psychology of 
cognitive functions. 

Because Whorf is chiefly known for his ideas on linguistic relativity, 
this volume features those of his writings which are most relevant to this 
problem. His studies in at least two other areas deserve recognition, 

Whorf's early work in translating obscure Aztec documents was 
undoubtedly brilliant and made a distinctive contribution to a neg- 
lected corner of Middle American research. His interest in translating 
Aztec shortly gave way, however, to the quest for a means of reading the 
noncalendrical portions of Maya hieroglyphic writings. Here the valid- 
ity of Whorf's work is at least highly controversial. There is no gain- 
saying that his observations in this area were exceedingly acute, and his 
"detective work" searching and clever. It is not for a nonspecialist to 
say how much Whorf actually accomplished toward the interpretation 
of the Maya hieroglyphs. There is certainly a great deal of plausibility 
in the translation processes which he presents, as in the paper reprinted 
here ("Decipherment of the linguistic portion of the Maya hieroglyphs"), 
and Whorf was too objective and intelligent a scholar to present mere 
ad hoc constructions which could not be tested and generalized. Yet, 
objections have been raised by several Maya scholars— during Whorf's 
lifetime by Long (1935, 1936) and Andrews (1938), and, most recently 
and also most severely, by J. Eric S. Thompson (1950). Nevertheless, 
the criticisms of Long and Andrews related chiefly to minor points, some 
of which Whorf was subsequently able to adjust, and Thompson does 
not seem to have answered fully the deeply rooted complaint of Whorf 
that the tradition in Maya hieroglyph research has been to ignore lin- 
guistic evidence. As late as 1939, Tozzer (1939) was willing to state his 
belief that there were considerable phonetic components ci Maya hiero- 
glyphs, as Whorf urged, and we are told that the eminent Maya scholar, 
Herbert J. Spinden, was enthusiastic about the paper which Whorf read 
at the Eighth American Scientific Congress in May 1940. This paper is 
included here because so little progress has been made in reading Maya 
hieroglyphs and because it so well presents Whorf's notions about the 
problem and about writing systems in general. 

Finally, Whorf's contributions to general linguistics should not be 


underestimated. His early theories of "oligosynthesis" and "binary 
grouping" were, to be sure, certainly overdrawn, and, except as might be 
implied by his continued admiration for the work of Fabre d'Olivet, he 
ceased to appeal to any such theories after he became a student of Sapir. 
Nevertheless, Whorf de\eloped his theory of oligosynthesis with char- 
acteristic originality and acumen, and it is perhaps unfortunate that he 
never was able to bring himself to publish any full and mature descrip- 
tion of the theor}', for it is at least conceivable that there are languages, 
of which Aztec and Maya are possible examples, in which submorphemic 
elements are more productive throughout the whole of the vocabulary 
than the occasional "phonesthemes" which have been noted in English 
(such as the sp in 'spit, splash, spray, spout, sputter, splatter,' etc., which 
has seemed, to some, to carry the notion of "forceful outward motion"). 

Whorf was, at any rate, a master of straight linguistic description. His 
sketches of Milpa Alta Aztec and Mishongnovi Hopi as published by 
Hoijer (1946) are exemplary; they are characterized not only by the cus- 
tomary minute phonological and morphological descriptions but also by 
an unusual emphasis on seeking to find the meaning of grammatical 
categories. Some of this sort of approach can be seen in papers pub- 
lished in the present volume; for Hopi, in the paper "Some verbal cate- 
gories of Hopi," and for English, or for languages in general, in the 
paper "Grammatical categories," in which he introduced a distinction 
between overt and covert grammatical categories, and first applied the 
term "crj'ptotype." I believe it can fairly be said that contemporary 
linguists have only begun to explore the full implications of Whorf's 
concept of cr\'ptotype. 

E\en when Whorf worked on purely phonetic and phonemic prob- 
lems, he was highly original. He was apparently the first to propose the 
term "allophone," now in common use among linguistic scientists. His 
model of the English monosyllable, as presented in his paper "Lin- 
guistics as an exact science," was at the time an original synthesis of 
facts about English sound clusters. He wrote an interesting paper on 
the phonemics of his own (Boston) dialect of English, published post- 
humously in 1943. 

Whorf was everywhere an exceedingly acute observer of interesting 
and subtle phenomena in language structures. For example, in his 
hands a massive compilation of information about Shawnee, a language 
which he had not previously studied, was able to suggest sexeral novel 


perspectives on figure-ground relationships as exemplified in Shawnee 
word formation; the reader is referred to the paper "Gestalt technique 
of stem composition in Shawnee," printed in this volume. As if to roll 
up in one package all the bases of his insights into language structure, 
Whorf once had occasion to prepare an outline which he thought 
anthropological field workers could use in collecting information about 
new languages. The outline was referred to in a publication by Mur- 
dock, "Outlines of cultural materials" (1938), but it has never before 
been published. In the belief that it will still be useful, even though its 
publication is belated, I have included it in the present collection under 
Whorf 's title "Language: plan and conception of arrangement." It will 
doubtless demand considerable linguistic sophistication on the part of 
the reader to catch the meanings which are often only vaguely suggested 
by its outline form and sketchy phraseology, but at least it may ser\'e in 
this way to stimulate in the reader some of the productive imagination 
which was characteristic of Whorf in whatever he touched. 


The reader is cautioned that the phonetic orthography for Hopi used 
in this volume varies according to the particular system employed in 
each individual paper. To have attempted to regularize this orthog- 
raphy would have required a major linguistic investigation, which the 
editor was not prepared to accomplish. 


Acknowledgment is here made for permission to include in this volume 
various writings of Whorf previously published elsewhere: 

To Dr. Sol Tax, editor, for "A central Mexican inscription combining 
Mexican and Maya day signs" from the American Anthropologist, vol. 34, 
no. 2. 

To Dr. Bernard Bloch, editor, for "The punctual and segmentative aspects 
of verbs in Hopi," "Some verbal categories of Hopi," and "Grammatical 
categories" from Language, vol. 12, no. 2; vol. 14, no. 4; and vol. 21, no. 1, 


To Dr. C. F. Voegelin, editor, for "An American Indian model of the 
universe" and "Linguistic factors in the terminology of Hopi architecture" 
from the International Journal of American Linguistics, vol. 16, no. 2, and 
vol. 19, no. 2. 

To Leslie Spier, editor, for "The relation of habitual thought and behavior 
to language" from Language, Culture, and Personality, pp. 75-93 (Menasha, 
Wis.; Sapir Memorial Publication Fund, 1941). 

To Dr. Leonard Carmichael, secretary, Smithsonian Institution, for "De- 
cipherment of the linguistic portion of the Maya hieroglyphs" from The 
Smithsonian Report for 1941, pp. 479-502. 

To Dr. N. Sri Ram, president, the Theosophical Society, for "Language, 
mind, and reality" from The Theosophist, January 1942. 

The illustrations for several of Whorf's papers in this volume have been 
specially redrawn by J. Martin Rosse, who prepared illustrations for Whorf's 
Technology Review papers in 1940 and 1941 from rough sketches which 
Whorf himself had supplied. 

I am indebted to several persons for furnishing information and docu- 
ments which were necessary to the preparation of this edition. Dr. George 
L. Trager helped me decide which of Whorf's writings should be included. 
Professor C. F. Voegelin supplied numerous scarce reprints, and a manu- 
script copy of "Language: plan and conception of arrangement" was lent 
by Professor Norman McQuown. Professor Herbert Hackett provided sev- 
eral bibliographical items which I might have otherwise missed. 

I am particularly grateful to several members of the Whorf family for 
their kindness in granting me interviews: to Whorf's widow, Mrs. Celia 
Peckham Whorf, and to his brother, John Whorf of Provincetown, Massa- 
chusetts. Special thanks are due to Whorf's son, Robert Peckham Whorf, 
for allowing me to spend several days at his home examining Whorf's papers 
and correspondence, and for permitting me to borrow a number of his manu- 
scripts, some of which are printed in this volume for the first time. 

John B. Carroll 
Arlington, Massachusetts 
June, 1955 


320 Wolcott Hill Road 
Wethersfield, Conn. 
July 12, 1927 
Dear Dr. English: 

I have been intending to write you in regard to your little dictionary 
and especially to ask you for a name by which to denote a certain psy- 
chological concept, but I have not found a chance until the present, and 
I don't know whether this season will find you at your Middletown 
address. I must say that I appreciate that dictionary; it is not only actu- 
ally interesting— a rare thing for a dictionary— but valuable as well. But 
I have not been able to find in it or in any other source a recognized 
term for one of the phenomena in which I am interested and would 
like to know if you know of such a term or could suggest one. 

I have not been able to find a term that I need to denote a kind of 
connection or relation, approximation, closeness, allied character, be- 
tween ideas. The only psychological term I know of that expresses 
connection between ideas is "association," but this has quite a definite 
meaning and one that will not do for the meaning I have in mind. The 

* Tliis unpublished essay was found by me among Whorf's papers as a partly 
typewritten, partly handwritten draft of a letter, dated July 12, 1927, to the psy- 
chologist Dr. Horace B. English, then of Wesleyan University, who had just pub- 
lished a dictionary of psychological terms. There is some question whether the 
letter was ever finished and sent, but Dr. English, now at The Ohio State University, 
has a vague recollection of receiving something like it. I have made minor editorial 
emendations and alterations where necessary. 



"connection" of ideas, as I call it in the absence of any other term, is 
quite another thing from the "association" of ideas. In making experi- 
ments on the connecting of ideas, it is necessary to eliminate the "asso- 
ciations," which have an accidental character not possessed by the "con- 
nections." The subject must not jump at the first idea that comes to 
mind as in a "free association" experiment; hence the experiment might 
be considered a form of "controlled association"; yet it may be quite 
"free" in its own sphere, for any connection may be permitted. 

"Connection" is important from a linguistic standpoint because it is 
bound up with the communication of ideas. One of the necessary 
criteria of a connection is that it be intelligible to others, and therefore 
the individuality of the subject cannot enter to the extent that it does 
in free association, while a correspondingly greater part is played by the 
stock of conceptions common to people. The very existence of such a 
common stock of conceptions, possibly possessing a yet unstudied ar- 
rangement of its own, does not yet seem to be greatly appreciated; yet 
to me it seems to be a necessary concomitant of the communicability 
of ideas by language; it holds the principle of this communicability, 
and is in a sense the universal language, to which the various specific 
languages give entrance. 

For an example of connection, consider first the idea 'down,' and 
then the following ideas: 'set, sink, drag, drop, fall, hollow, depress, 
lie.' I will call these group A. It is clear that there is a "connection" 
between 'down' and each of the ideas in group A. Consider now group 
C, consisting of the ideas: 'upright, heave, hoist, tall, air, uphold, swell.' 
There is a "connection" between these ideas and the idea 'up.' Now in 
a connection experiment the subject, on receiving the idea 'down' would 
be free to connect with any of the ideas in group A or others like them 
but could not give any of the ideas in group C or the like. Yet, if it were 
a question of associations only, he might associate an idea in group C 
with 'down.' He might for instance have had an unpleasant experience 
in a boat when there was a heavy 'swell' on, from which he retained a 
vivid impression of continually going down. But this association would 
not be a connection. It would pertain to his own personal experience 
rather than to the social or collective experience which is embodied in 
the common linguistic stock of concepts, and the reason for the asso- 
ciation would not be intelligible immediately without explanation; it 
would require an explanation bringing in his personal experience. In 


this sense of immediate intelligibility, 'swell' is connected with 'up' or 
the like, and is distinctly removed from 'down.' So, in further defini- 
tion of this concept of connection, it may be said that connections must 
be intelligible without reference to individual experiences and must be 
immediate in their relationship. Mediate connections, i.e., connections 
through the medium of other connections, are to be called rather chains 
or paths of connection, or possibly "communications." 

It is possible to formulate another group of ideas, group B, which 
mediate between A and C, so that we can pass, by means of various 
chains or paths of connective communication, from A to C and hence 
from 'down' to 'up' entirely in a connective way and without the aid of 
association. For instance: 

A set sink drag drop hollow depress lie down 

B stand heavy pull precipice space bear extend 
C upright heave hoist tall air uphold swell up 

Subjects feeling their own way through the congeries of ideas between 
'down' and 'up' do not always follow these paths but often find others. 
For instance, subject M. F. went as follows: 'set— heavy— swell— up.' 
Asked to explain the connection 'set— heavy,' it appeared that 'set' en- 
tailed a strong notion of fixation or fixity, and suggested 'rigidification, 
congelation, stiffening, thickening,' as in the setting of jelly, while 'heavy' 
implied to the subject not merely 'weight' but 'body, density, viscosity,' 
an idea closely similar to the preceding 'set.' This is a true connection, 
although it was not instantly intelligible to the experimenter, but it was 
quickly understood without reference to any personal experience. The 
connection 'heavy— swell' also was not instantly perceived, but it de- 
veloped that 'heavy' conveyed essentially the idea of quantity or mass, 
including 'massiveness, size, increase': hence 'enlarge, expand, swell.' 
This again is a true connection. The same subject starting with 'up' 
traversed the path 'up— hoist— pull— drag— down.' Subject W. W. gave 
'down— drop— heav}'— hoist— up.' Asked to explain the connection 
'heavy— hoist,' it appeared that heavy suggested the feeling or bearing 
of weight, the 'hefting' of a thing, essentially as lifting action. If the 
word 'heave' had been more familiar to the subject, he might have 
chosen it in preference to 'hoist.' 

But a different and nonconnective process appeared when a young 
man having reached an idea 'past' took as the next step 'hiding,' over- 


looking an obvious 'before' in the same group. This still might be a 
connection if it yielded a satisfactory explanation, but the best explana- 
tion he could give was that one's past was usually unpleasant and so one 
would prefer not to remember it; hence it was in hiding. This might 
perhaps be only an awkward way of expressing the connection, but it 
appeared not. He did not respond to the suggestions that 'past' meant 
'receded, withdrawn, retired, concealed,' or that it meant 'gone, vanished, 
invisible, concealed,' or that it meant 'existent (in memory)' but not 
'apparent, stored up, hoarded,' etc., but persisted in this quite extraneous 
idea of the unpleasantness of the past. Hence I concluded either that 
an unpleasant past really had colored his way of thinking, or that he 
wished to pose as somewhat of a misanthrope or cynic, or that he had 
been reading psychoanalysis: that in any case we had to do here with 
something personal, which was indeed an association yet not a con- 
nection. In telling him that I wanted connections that had nothing to 
do with personal experiences, he admitted that this might not apply to 
his association, and then chose 'before.' 

Sometimes a subject will jump to a true connection by association and 
then get the connection later; e.g., W. W., a college freshman with in- 
telligence distinctly above the average, said he thought the connection 
between 'drag' and 'down' was like this: 'drag' meant 'pull' and things 
went 'down' because pulled by the attraction of gravity. He had just 
taken an examination in physics. I asked him whether he could have 
recognized a connection if he had never heard of gravity and he sup- 
posed not. I suggested that gravity might prove to be a compacting 
together due to a kind of external pressure, and then what would 
become of the connection? A mere hint was sufficient to lead him to 
the true connection, which is simply one of linguistic meaning, i.e., 
'drag' = 'trail, dangle'; what is 'dragging' is in general 'down,' not 'up.' 
This is an interesting commentary on the inability to distinguish theory 
from fact in what is learned, even in an exceptionally intelligent student. 
(Or perhaps especially in such a one? That is, if intelligent means quick 
to learn, perhaps it also means receptive and hence too credulous?) 

Can you suggest any better term for this sort of affinity than "connec- 
tion"? I might say that my mental image of the relation is not at all 
one of ideas hitched together by bonds of attachment which they possess 
like miniature hooks and eyes. It is more a concept of continuity, with 
the ideas as relative locations in a continuous medium. Take an idea 


like "up," and say it corresponds to a certain location where we are. 
Now I can conceive that something like motion may happen to us. The 
idea "up" is a sort of neighborhood, and we are leaving that neighbor- 
hood. We cannot tell exactly where any neighborhood leaves oflf. We 
know that the idea up is assuming a different nuance: it is growing to 
be like the idea rise. But, after a certain amount of this change or 
"motion" has taken place, we know that we are in a different spot; the 
idea is now definitely 'rise,' not 'up.' Motion continues, and 'rise' be- 
comes 'left.' 'Left' insensibly becomes 'carry,' and this becomes 'sus- 
tain/ We are now definitely out of the vicinity of 'up.' Any one of 
these ideas might have become something else by var\'ing the "direc- 
tion" of motion. 'Sustain' might become 'nourish,' or it might become 
'continue.' 'Nourish' might become 'feed,' and 'continue' might be- 
come 'long,' 


Psychology has developed a field of research that may no doubt be 
useful or valuable in itself, but it throws little or no light on problems 
of the normal human mind or soul. The person who wishes to under- 
stand more fully the laws and, so to speak, topography, of the inner or 
mental life is as much thrown back on his own difficultly acquired store 
of wisdom and his native judgments, intuitions, sympathies, and com- 
mon sense as though the science of psychology did not exist. Such a 
one, for instance, is the teacher, educator, sociologist, anthropologist, 
trainer, coach, salesman, preacher, manager, diplomat, executive: anyone 
who must deal with human intangibles, especially the man concerned in 
leadership of any sort. If he seeks aid from books, he will get far more 
information about this field from literature not intended to be scientific, 
that is, from the best works of the novelists, playwrights, and poets, than 
he will from any textbook of psychology. There are certain courses that 
psychology has elected to follow that have estranged it, perhaps perma- 
nently, from the truly mental field. 

First, the "old school" of experimental laboratory psychology has 
rather definitely assumed the character of a branch of physiology. Its 

* This hitherto unpublished note was found by me among Whorf s papers as a 
rough handwritten manuscript. The date of its composition is unknown, although I 
would hazard a guess that it was written about the same time as the letter to English, 
that is, about 1927. The latter part of the note is extremely sketchy; perhaps this is 
only the outline of a longer paper which Whorf contemplated writing. I have sup- 
plied a title and made extremely minor editorial changes. 



findings and their value all redound back to physiology. It is un- 
doubtedly valuable to the student of mental phenomena to know the 
mechanisms of the body, but rather in the character of auxiliary infor- 
mation than anything else; and know^ledge about the oxidation of the 
blood and the details of brain and nerve responses, sense perceptions, 
and association times are equally of this character. Moreover, one is 
impressed (and depressed) by the appalling sterility of the vast mass 
of minutiae that this science accumulates, and the dearth of integrating 

Second, the school of behaviorism has begun to appear in its true 
character as simply the old experimental psychology over again in a more 
pick-and-shovel aspect. That it is in many ways an improvement on 
the old school and has enlarged our understanding in certain fields I 
personally believe. It has been of service by teaching us to think more 
in terms of behavior, but, when all is said and done, it can teach us little 
that is new. It has shown us how behavior may be conditioned by 
physical means, but along much the same lines that we already knew 
although they have been more systematically explained. It has become 
apparent that we may "condition" either with or against the cooperation 
of truly psychic considerations. This we already knew, but we are par- 
ticularly interested in "conditioning" with the cooperation of and in 
accordance with the particular laws of the psychic. No doubt the same 
process of stimulus and response "conditions" a man into being a scien- 
tist or a maniac, a leader of men or a nervous wreck, a good workman 
or one who cannot hold a job, an inspiring helper or a resentful cog in 
the machine; but behaviorism does not show us which lines to work 
upon in order to be really in accord with human intangibles, except by 
way of announcing in behavioristic terms things already obvious to com- 
mon sense. 

Gestalt psychology does seem to me to have discovered an important 
truth about mind, the importance of configurations in the mental 
domain. At the same time the Gestalt psychologists have their hands 
full with the manifold mechanical, experimental, and personal data 
required to develop this large subject, most of which data are chiefly 
valid on the animal level. When we attempt to apply the configurative 
principle to the understanding of human life, we immediately strike the 
cultural and the linguistic (part of the cultural), especially the latter, as 
the great field par excellence of the configurative on the human level. 


Here the Gestalt psychologists let the matter drop. They have neither 
the time nor the linguistic training required to penetrate this field; more- 
over their ideas and terminology inherited from the old laboratory psy- 
chology are a liability rather than an asset. 

Psychoanalysis is the one school that really deals with mental ma- 
terial, and it sometimes gets results, but it works only in the sphere of 
the abnormal and the deranged, and it is becoming evident that the 
abnormal is not the key to the normal. Moreover, it is so resolute in 
its determination to deal with intangibles that it shows almost a con- 
tempt for the external world and strays continually into the realms of 
phantasm. It is too heavily stamped with the signature of its founder, 
Freud, an erratic genius with a faculty of apperceiving deep but obscure 
truths, and is notion-obsessed and cluttered with weird dogma. As an 
empirical tool for the clinic it may serve for a while, but I do not see 
how it can possibly be a means for the careful scientific scrutiny of the 
normal mind. 

All the schools then have been surveyed and found wanting, and the 
seeker for knowledge about the human mind is forced to fall back on the 
long-collected mass of empirical observations sometimes called "the 
wisdom of the ages," on the works of keenly intuitive authors, on his 
own insight, and on what few general truths he can cull here and there 
from all the above schools. 

One fact that stands out to a detached viewpoint, but is not stressed 
by any of the schools, is the great and perhaps basic importance of the 
principle we denote by the word "meaning." Meaning will be found to 
be intimately connected with the linguistic: its principle is symbolism, 
but language is the great symbolism from which other symbolisms take 
their cue. 


When in Mexico during the winter of 1930, engaged in Nahuatl 
hnguistic research, I visited the village of Tepoztlan in the state 
of Morelos and while there made the accompanying sketch (Figure 1) of 
a band of sculptured figures in the ruined temple of the Tepoztecatl, the 
ancient tutelary deity, which stands on a great rock pinnacle overlook- 
ing the town. 

The temple has been described by Saville,^ Seler,- and Novelo,^ but 
nowhere do any of them discuss the figures dealt with herein. The 
structure bears indications * of dating from the reign of the Aztec king 
Ahuitzotl, who died in 1502; but, as the figures in question show like- 
nesses to forms known to be over a thousand years more ancient than 
this, it may be that in the building of the temple they were carved in 
obedience to artistic traditions, or copied from older architectural work 
of this region. They form a band extending along the top of a much 
more conspicuous frieze of larger carvings on the inner walls of the 
inner room or court. My sketch and remarks apply only to a clearly 
preserved portion of the band in the southern half of the court. 
Stylistically and on the basis of general probabilities, the figures would, 

* Reprinted from Amer. Anthrop., 34:296-302 (1932). 

1 Bull. Amer. Mus. Nat. Hist. (1896); also Monum. Records, February 1898. 

2 Bull. 28, Bur. Amer. Ethnol, 347; "Die Wandskulpturen im Tempel des Pulque- 
gottes von Tepoztlan," Gesamm. Abh., 3:487. 

3 "Guia para visitar las principales minas arqueologicas del estado de Morelos," 
Publ. Sec. Educacion, Publica, 3 (1929). 

4 See Seler, Bull. 28, Bur. Amer. Ethnol. 




1 . Z. 

Cozcaqufluhtli Quauhtli 



Mai I no 1 1 1 



7 . .. 8. 

Itzcuintli Ail 










" +1- 

M iquiztl I 

Clls/1 I 

gap In stone, about 
space of 2. figures 


02) ^,^ (13.) ,. ,or \Z. (14) , 

(CoatI) (Cuetzpalin) CoatI (Coll.) 


Figure 1. Inscription in tlie Temple of the Tepoztecatl, Tepoztlan, Morelos, 



I think, strike any student as a band of the day signs of the tonalamatl, 
such as one sees continued interminably along with the successions of 
pictures in the Mexican codices. But many of the signs bear little 
resemblance to their Mexican form, the form corresponding to the Aztec 
names Cipactli, Ehecatl, Calli, etc. Nevertheless, as we shall presently 
see, the signs Acatl, Malinalli, and Atl are in practically their regular 
Mexican forms, and to clinch the matter are precisely the correct number 
of signs apart. But moreover, and here the unusual enters, some of the 
signs bear an unmistakable likeness to the quite un-Mexican-looking 
Maya forms, corresponding to the Maya names Imix, Ik, Akbal, etc., 
and these signs, too, are in exactly their proper places. The ruin is of 
course far removed from historical Maya territory, being only about forty 
miles from Mexico City in a region of Toltec and post-Toltec influence. 

As I sketched the figures, their general un-Aztec look quickly struck 
me, but the first clear impression that I was drawing a maya hieroglyph 
came when I began to copy the sign Figure 1, no. 10. Comparison with 
a very common Maya hieroglyph, whose most usual form is that shown 
in Figure 2, no. 10, indicates the similarity. The Maya sign is a hand 
having the distinctive characteristics of a prominent thumb more or less 
opposed toward the fingers, and having in\ariably on the wrist a circle 
usually with a central dot and a small taglike projection from the rim. 
The fingers are usually bent toward the thumb, but there are forms on 
the Maya monuments in which they are extended straight as in the 
Tepoztlan figure. This sign is a hieroglyphic element of wide and varied 
use, but it is especially a grapheme ^ that stands for the day sign Manik, 
In the Mexican day-sign system there is no grapheme even remotely 
resembling a hand. The sign corresponding to Manik is called Mazatl, 
and its grapheme is a deer's head (Figure 2, no. 9). 

Can it be that this Tepoztlan hand figure represents Manik-Mazatl? 
Do the other signs fall into the proper places required by such an as- 
sumption? Let us see. In both day-sign systems the sign before Manik- 
Mazatl is called by a similar name, Cimi in Maya and Miquiztli in Aztec, 
both meaning "death." The grapheme in both systems is a death's head 

s Grapheme is a word formed on the analogy of morpheme, semanteme, to denote 
any written symbol, especially as a linguistic factor, in place of "ideogram," "picto- 
graph," or the ambiguous "character." In discussing hieroglyphs it is desirable to 
liave a term that does not presuppose anything about the nature of the denotative 
process employed. 



Mexican Day Signs 


liwl ^^. Qwouhtii 30" 
3'- CoxcQ^uauhtli f^^N 32^ 

Figure 2. The Mexican and Maya day signs from Coatl-Chicchan to Cozca- 
quauhtli-Cib in their order. BibHographic references: 1, 9, 11, 15, 17, 19, 23, 
and 26 from Sahagun manuscript, 2 and 5 from Zouche Codex; 12 and 31 from 
Fejervary-Maycr manuscript; 14 from Codex Telleriano-Remensis; 21, 24, and 29 
from Seler's Caractere des inscriptions; 3, 4, 6, 13, 16, 18, 20, 22, 25, 27, 28, 30, 
33 from Morley's Introduction to Maya hieroglyphs (3, 18, and 28 inscription forms 
from p. 38, 33 an inscription form from p. 95, the others codex forms from p. 39); 
7, hieroglyph of death god, from Dresden Codex, p. 15; 8, representation of human 
skeleton from Uxmal, from Spinden's Maya art, and 32, Codex Peresianus, from 
Spinden's Maya art, p. 94; 10, from Codex Tro Cortesianus; ISd, Maya hieroglyph 
based on dog's head and related to Oc (18), from Dresden Codex. 


or head of the death god, styhstically different, however, in the two 
systems. Figure 2, no. 5 shows the Aztec type of grapheme, a fieshless 
skull; Figure 2, no. 6, the Maya type; and Figure 2, no. 7, the hieroglyph 
of the Maya death god as it appears in the Maya codices. Now the 
direction of Aztec and Maya writing is like our own, from left to right. 
The figure to the left of the hand (Figure 1, no. 9) bears no resemblance 
to Miquiztli or Cimi. But the figure to the right of the hand (Figure 
1, no. 11), strange to say, shows a most curious likeness to the hieroglyph 
of the Maya death god. The two prefixes attached to the head are 
especially characteristic of this grapheme, indeed are found nowhere else. 
The Tepoztlan glyph however has a suffix that is not found in the death- 
god grapheme, but is an element in other Maya hieroglyphs. It is par- 
ticularly characteristic of the month sign Kankin, and according to Seler 
represents a human skeleton. It seems to be related to the skeletal 
representation shown in Figure 2, no. 8, taken from Figure 115, page 86 
of Spinden's Maya Art, where Spinden treats of the artistic symbolism 
of bones and death among the Maya. 

The fact is, as will soon be proved, that we have here an inscription 
which for some unknown reason is written in reverse order, from right to 
left, and this death-god sign stands for Cimi-Miquiztli. Once this is 
realized, the student of the subject will soon notice another out-and-out 
Maya sign, namely the ninth figure to the left of the hand. Figure I, 
no. I. In the Mexican system the ninth sign after Mazatl is called 
Cozcaquauhtli, and its grapheme is the head of a vulture (Figure 2, no. 
31) or of an eagle wearing a collar. In the Maya system the ninth sign 
after Manik is called Cib, and its grapheme (Figure 2, nos. 32 and 33) 
is a curved line like a question mark or sometimes rather like a letter C 
turned over or turned backward. This last is the form of the Tepoztlan 

The two dots beside the curve are not found in the Maya Cib, but 
they nevertheless confirm the identification. Seler, from the fact that 
the sign Cib was often represented on liquor vessels, connected it with 
a similar sign placed by the Aztec on their drinking vessels and called 
ometoch, from the god of intoxication Ome Tochtli, literally 'Two 
Rabbit.' This god is often represented (e.g., Sahagun Madrid Manu- 
script, under his name Totochtin) carrying a shield with a sign very simi- 
lar to the Tepoztlan figure. An Aztec note on the Sahagun picture says 
that the god bears an ometoch-chimalli, that is, 'shield with the device 


Two Rabbit.' In our present case the two dots are merely the usual 
expression of the number part of such a name as Two Rabbit. Ome 
Tochtli and Tepoztecatl are considered to be the same or related deities, 
so that their especial cult in Tepoztlan might perhaps employ their 
emblem as a day-sign grapheme when it would not be so employed else- 
where. The point is that they should have employed it, not for Tochtli 
or 'Rabbit,' but for Cozcaquauhtli or 'Vulture,' of all signs the one cor- 
responding in position to Maya Cib. 

Let us check the positions of the other signs proceeding from no. 10, 
or Manik, toward Cib. Number 9 is too worn to be distinctly recog- 
nizable as anything; yet, by comparison with the form of the Mexican 
sign Tochtli, shown in Figure 2, no. 12, it will be seen to resemble a 
worn-down carving of this form. 

The next sign, no. 8, is crowded against the following sign no. 7 and 
is placed in an angle where the band turns the corner of the wall. If it 
is compared with the grapheme of the Aztec sign Atl ('Water') shown 
in Figure 2, no. 14, the likeness will be evident. 

Number 7, the next figure, occupies the place of the Aztec Itzcuintli 
('Dog') and the Maya Oc. It shows a head that looks more like a toucan 
or some such bird than a dog. It certainly shows little resemblance to 
the naturalistic dog's head (Figure 2, no. 17), the grapheme for this day 
sign in the Mexican system. Now a toucan-like conventional head, 
shown in Figure 2, no. ISa, is one of the commonest hieroglyphic ele- 
ments in the Maya codices, and Beyer ® has shown that this convention- 
alized head is derived from that of the dog. The Maya day sign cor- 
responding to Itzcuintli is called Oc, and has two distinct forms of 
grapheme. The form of the codices has no resemblance to the 
Tepoztlan. The form of the Maya inscriptions, shown in Figure 2, 
no. 18, may be compared with the Tepoztlan form. The dog hiero- 
ghph may be further compared in that it always bears a suffix contain- 
ing a two-lobed figure, while the Tepoztlan glyph shows suffixed to the 
head a square frame containing a two-lobed figure. 

Sign 6 shows a certain resemblance to the Maya Chuen, Figure 2, 
no. 20, and is quite unlike the naturalistic monkey head of the Aztec 
Ozomatli (Fig. 2, no. 19). 

Sign 5 agrees with the Aztec grapheme for the same position. 

6Amer. Anthrop., 31 (1929). 


Although it is much worn and a cavity in the stone seems to have been 
scooped out of a part of it, the brush of tonguehke streamers character- 
istic of Mahnalh is recognizable (cf. the form of Malinalli shown in Fig. 
2, no. 21). Here there is no trace of the Maya form (Eb, Fig. 2, 
no. 22). 

Sign 4 shows in proper position the distinctive features of certain 
forms of the Mexican Acatl. (Cf. especially the form shown in Fig. 2, 
no. 24.) The Maya Ben, Figure 2, no. 25, is quite different. 

Sign 3 however is a very strange one. It is certainly nothing like the 
Mexican Ocelotl, a jaguar head. Figure 2, no. 26. Nor is there any ex- 
ternal resemblance to the Maya Ix, Figure 2, nos. 27 and 28. And yet it 
contains in a curious way two elements of the Ix grapheme. The really 
essential element of the grapheme is the three dots, and the Tepoztlan 
figure displays very prominently three large dots on the left and three 
smaller dots on the right. Distinctive of the codex form of Ix are the 
two converging dotted lines, and the Tepoztlan sign bears a shield- 
shaped figure on which are two similarly placed lines. 

Sign 2 bears no resemblance to the Maya Men (Fig. 2, no. 30) nor 
to the ordinar}' form of the Aztec Quauhtli, an eagle's head. But Seler ^ 
pictures a form of Quauhtli (Fig. 2, no. 29) in which the eagle wears a 
headdress that compares interestingly in its main features with the 
Tepoztlan figure. This brings us to sign 1, or Cib, which we have already 
discussed. There is a sign beyond this which I have not shown, as it is 
worn and hard to make out, and I did not secure a good drawing. It 
shows no likeness to the unmistakable Aztec Olin or to the Maya Caban. 

What does the band show to the right of sign 11, or Cimi? A blank 
space where the stone has been broken away. Beyond this space appears 
one more sign, no. 12. Of course we do not know whether any signs 
were inscribed in this space, or if so how many, though I should say the 
space would contain only two. Allowing two signs for the space would 
make no. 12 become no. 14, the position of the Aztec Calli or the Maya 
Akbal, neither of which has any resemblance to it. It is a convention- 
alized serpent jaw, a common Middle American art motive. Therefore 
I think that no signs need be allotted to the broken space and that we 
have here the day sign before Cimi-Miquiztli, which is the Maya Chic- 
chan (Fig. 2, nos. 3 and 4), the Aztec Coatl (Serpent) (Fig. 2, nos. 1 and 

■^ Caractere des inscriptions Azteques et Mayas. 


2). The common Aztec form is not conventionalized to the degree of 
this figure. 

It is perhaps these figures to which Novelo refers in his words: 

-hay otros jeroghficos cuya interpretacion no ha sido posible de hacer, 

alguno de los cuales tienen cierta semejanza con los mayas. 
Moreover he refers to the Maya influence in these terms: 
-parcce existir en los relieves de origen tlahuica (Tepoztlan y Xochicalco) 

cierta influencia maya cuya cultura florecio en Mexico, come se sabe, en los 
primeros siglos de la Era Cristiana. 

He also tells us that pilgrims from far-away Chiapas and Guatemala, 
regions once of Maya culture, were accustomed to visit the sanctuary of 

Yet it is certainly unexpected to encounter here, not far from Mexico 
City, definite day signs denoted by graphemes which, like Cib and 
Manik, were being carved on the structures of the old Maya Empire in 
distant Central America a millennium and more before the date of the 
Aztec temple on which they appear interchangeably with ordinary Aztec 
forms. And why was the sign series recorded backward? Here again 
the only comparable thing that I can think of is a Maya one, the fact 
that the Maya inscriptions record a number series in reverse order when 
it is counting back into the past: that is, when its total is to be sub- 
tracted from, and not added to, a beginning date, in order to reach 
a second date. Does this Tepoztlan inscription seek to show the 
tonalamatl receding into the past? 

We have here for the first time evidence of a definite, clearly demon- 
strable rapport between Nahuatl hieroglyphs and early Maya ones. The 
whole subject of the relation of Mexican and Maya graphemes, as it 
reveals itself in other places, is something of which I hope to treat 
extensively and from a linguistic viewpoint at another time. 


Verbs in the Hopi language are noteworthy for their ver}' rich and 
expressive development of verbal aspects and voices. I shall say 
nothing in this paper of the nine voices (intransitive, transitive, re- 
flexive, passive, semipassive, resultative, extended passive, possessive, and 
cessative); and of the nine aspects (punctual, durative, segmentative, 
punctual-segmentative, inceptive, progressional, spatial, projective, and 
continuative) I shall deal with only two. It may be noted that there are 
no perfective and imperfective aspects; in fact Hopi does not in any way 
formalize as such the contrast between completion and incompletion of 
action. Its aspects formalize different varieties of the contrast between 
point-locus and extent-locus of phenomena, indifferently in time or 
space, or in both. Hopi also has three tenses: factual or present-past, 
future, and generalized or usitative. Hopi verbs belong to seven classes 
or conjugations having slightly different inflectional systems. Class 1, 
the largest and most creative class, contains a few categories not found 
in the other classes, among them the segmentative aspect. 

The simplex of the class- 1 verb is a bare root of the form CVCV, and 
is in the third person singular intransitive voice, punctual aspect, and 
present-past tense. The segmentative aspect is formed by final re- 
duplication of this root plus the durative suffix -ta, and produces a 
change in the meaning of the simplex of the following character: the 

* Reprinted from Language 12:127-131 (1936). The paper was read before the 
Linguistic Society of America in December 1935. 



phenomenon denoted by the root, shown in the punctual aspect as 
manifested about a point, becomes manifested as a series of repeated 
interconnected segments of one large phenomenon of a stretched-out 
segmental character, its extension usually being predominantly in one 
dimension, indifferently of space or time or both. The nature of the 
change can best be shown by examples. 

ha'ri it is bent in a rounded angle hari'rita it lies in a meandering 

line, making successive rounded 

angles (applied for instance to 

meander patterns in decoration) 

ho'^ci it forms a sharp acute angle hoci'cita it is zigzag 

pa"ci it is notched paci'cita it is serrated 

pi'va it is gullied out piva'vata it extends in successive 

gullies and gulches (said of 
ca'mi it is slashed inward from the cami'mita it is fringed, it is slashed 
edge into a fringe along the edge 

In these and similar examples, the phenomenon is such that it requires 
a rigid or semirigid substance for its field of manifestation. When this 
is the case the punctual intransitive has somewhat the character of a 
passive,^ and the segmentative shows the phenomenon multiplied along 
one dimension of space, like a candle flame between mirrors. In both 
aspects, the phenomenon shows up as an effect established and there- 
after retained in the rigid substance, so that we are presented with a 
static TABLEAU of tliis effect as it is disposed in space. 

Suppose, however, that the phenomenon denoted by the verb root is 
such as to require a nonrigid or mobile substance for its field of mani- 
festation, for example a liquid or a swarm of mobile particles. In that 
case a deformation of substance such as is denoted by the root will not 
be a permanent deformation but will result in a vibrative or pulsative 
agitation of the substance. The intransitive will no longer seem like a 
passive from our English-speaking viewpoint but will be decidedly active, 
and the punctual will denote one pulse of the deformation or disturb- 
ance, while the segmentative will refer to the entire train or field of the 
vibrations, both as extending in space and as continuing in time. Tlius, 
for instance: 

1 It is not a true passive because it does not imply any external agent; it is not a 
static (at least not in the ordinary sense) because it does not imply duration in time; 
it is not a true active because activity and result are presented as one. 


wa'la it (e.g. a liquid) makes one wala'lata it is tossing in waves, it is 

wave, gives a slosh kicking up a sea 

no'ija several come out (applied to iioija'ijata it is coming out in suc- 
objects or persons) cessive multitudes, it is gushing or 

spraying out (applied e.g. to a 

Note that with mobile-substance phenomena the segmentative is both 
durative in time, in contrast to the momentaneous punctual, and ex- 
tended in space, in contrast to the definitely "spotted" location of the 
punctual. Some phenomena are capable of manifestation in both 
mobile and rigid substances, especially those defined in terms of a cer- 
tain type of contour; e.g. ta'ho 'it exhibits one wavelike curve, or makes 
one undulation'; taho'hota referring to a mobile substance means 'it is 
undulating' (for example a liquid surface, a snake, a shaken rope), to 
a nonmobile substance 'it is scalloped' or 'it forms a wa\e pattern.' 

But suppose again that the phenomenon denoted by the stem is one 
resulting from the type of force known in physics as torque (tendency to 
produce rotation), which in order for any effect to be apparent requires 
that the substance be a body with at least a certain degree of rigidity 
and yet capable of certain degrees of motion relative to other bodies. 
In this case a single deformation or displacement as denoted by the 
punctual will be either a single oscillation or a single turning of this 
body according to the degree of freedom implied in the root-meaning; 
while, if the effect continues, it will continue as a train of oscillations or 
a continued rotation and may or may not involve an advance through 
space at the same time: this, then, will be the meaning of the segmenta- 
tive. Examples of this type of meaning are: 

wa'ya makes a waving shake (like a small tree shaken) 

ija'ya makes a sway from one side to the other 

pi-'ya makes a flap like a pair of wings 

ta'ya makes a racking shake 

yo'ya makes a circuit (axial turning combined with advance in an arc) 

ro'ya makes a turn or twist 

ri'ya makes a quick spin 

It is interesting to note that a great many (though not all) of these 
torque mo\ements are denoted by stems ending in -ya. The segmenta- 
tives of this type correspond to English durative forms denoting vibrative 
or rotative motion, e.g.: 


waya'yata it is shaking p'i-ya'yata it is flapping wings 

ijaya'yata it is swaying ydya'yata it is circling round and 


roya'yata it is rotating riya'yata it is spinning, whirling 

In the case of mi'ma 'rolls over,' where necessarily a lateral motion 
accompanies the turning, we get this phase of the action necessarily 
extended in mima'mata 'it is rolling along.' 

Another type of this general class of phenomena is one which mani- 
fests punctually as a shock, jar, or other sudden disturbance necessarily 
momentary in nature, and is related to a pulsative phenomenon also 
occurring in the natural world as a rapid succession of such shocks. 
Here English generally employs two different stems, but Hopi simply 
uses the punctual and segmentative of the same stem. An example 
from inanimate nature is ti'li 'it receives a slight jar,^ tili'lita 'it is vibrat- 
ing' (like an engine, a wagon, an automobile). But Hopi also discerns 
a grea'L many such phenomena in the animate world, for example: 

ti'r'i he gives a sudden start t'in'nta he is quivering, trembling 

wi'wa he trips over something, or wiwa'wata he is stumbling or hob- 
is suddenly caught by the legs, like bling along 
a lassoed horse 

ya'ro his teeth strike on something yaro'rota he is chewing forcibly on 
hard or gritty, e.g. in the food something hard 

he'ro he (or it) gives out a sudden hi:To'rota he is snoring 
hollow gurgle from within 

Often, again, such verb forms are applied to rhythmical movements of 
the body and limbs: 

\vi"ki he takes a step without mov- wiki'kita he is doing steps, or danc- 
ing from place ing in one place 

h'^i'la he takes a step forward k"ila'lata he does a walk forward 

(not 'is walking forward' which 
English expression is almost a 

yo"ko he gives one nod of the head yoko'kota he is nodding 

Again, the phenomenon may be one of disturbance at a point in a 
subtle medium, that is one that would be scientifically classed as gaseous 
or etheric. Such a medium gives little or no evidence of cither motion 
or extension in space, and the segmentative in these cases denotes only 
pulsation in time: 


ri'pi it gives a flash fipi'pita it is sparkling 

^i'wi it flames up ^iwi'wita it is flaming 

^i'mi it explodes, goes off like a gun ^imi'mita it is thundering 

Finally, there is one class of events to which the segmentative is not 
applied. It is not applied to "mental," "emotional," or other "inner" 
or "psychological" experiences. It concerns only the world of external 

All this has a wider interest than the mere illustration of an aspect- 
form. It is an illustration of how language produces an organization of 
experience. We are inclined to think of language simply as a technique 
of expression, and not to realize that language first of all is a classifica- 
tion and arrangement of the stream of sensory experience which results 
in a certain world-order, a certain segment of the world that is easily 
expressible by the type of symbolic means that language employs. In 
other words, language does in a cruder but also in a broader and more 
versatile way the same thing that science does. We have just seen how 
the Hopi language maps out a certain terrain of what might be termed 
primitive physics. We have observed how, with very thorough con- 
sistency and not a little true scientific precision, all sorts of vibratile 
phenomena in nature are classified by being referred to various elemen- 
tary types of deformation process. The analysis of a certain field of 
nature which results is freely extensible, and all-in-all so harmonious 
with actual physics that such extension could be made with great appro- 
priateness to a multiplicity of phenomena belonging entirely to the 
modern scientific and technical world— movements of machinery and 
mechanism, wave processes and vibrations, electrical and chemical phe- 
nomena—things that the Hopi have never known or imagined, and for 
which we ourselves lack definite names. The Hopi actually have a lan- 
guage better equipped to deal with such vibratile phenomena than is 
our latest scientific terminology. This is simply because their language 
establishes a general contrast between two types of experience, which 
contrast corresponds to a contrast that, as our science has discovered, 
is all-pervading and fundamental in nature. According to the concep- 
tions of modern physics, the contrast of particle and field of vibrations 
is more fundamental in the world of nature than such contrasts as space 
and time, or past, present, and future, which are the sort of contrasts 
that our own language imposes upon us. The Hopi aspect-contrast 



which we have obsen ed, being obligatory upon their verb forms, prac- 
tically forces the Hopi to notice and observe vibratory phenomena, and 
furthermore encourages them to find names for and to classify such 
phenomena. As a matter of fact the language is extraordinarily rich in 
terms for vibrator}- phenomena and for the punctual events to which 
they are related. 

IfTvl ^ l^fli^ 



T find it gratuitous to assume that a Hopi who knows only the Hopi 
-*- language and the cultural ideas of his own society has the same 
notions, often supposed to be intuitions, of time and space that we have, 
and that are generally assumed to be universal. In particular, he has no 
general notion or intuition of time as a smooth flowing continuum in 
which everything in the universe proceeds at an equal rate, out of a 
future, through a present, into a past; or, in which, to reverse the pic- 
ture, the observer is being carried in the stream of duration continuously 
away from a past and into a future. 

After long and careful study and analysis, the Hopi language is seen 
to contain no words, grammatical forms, constructions or expressions 
that refer directly to what we call "time," or to past, present, or future, 
or to enduring or lasting, or to motion as kinematic rather than dynamic 
(i.e. as a continuous translation in space and time rather than as an 
exhibition of dynamic effort in a certain process), or that even refer to 
space in such a way as to exclude that element of extension or existence 
that we call "time," and so by implication leave a residue that could be 

* The manuscript of this article, together with pertinent hnguistic notes, was 
among the papers left by Whorf at his death and turned over to George L. Trager. 
Dr. Trager and Dr. E. A. Kennard edited the manuscript for pubhcation, making no 
substantial changes, and the paper is presented here in the form in which it ap- 
peared in the Int. }. Amer. Linguistics, 16:67-72 (1950). Internal evidence and 
certain comments found in Whorf's correspondence suggest that the paper was writ- 
ten in about 1936. 



referred to as "time." Hence, the Hopi language contains no reference 
to "time," either exphcit or impHcit. 

At the same time, the Hopi language is capable of accounting for and 
describing correctly, in a pragmatic or operational sense, all observable 
phenomena of the universe. Hence, I find it gratuitous to assume that 
Hopi thinking contains any such notion as the supposed intuitively felt 
flowing of "time," or that the intuition of a Hopi gives him this as one 
of its data. Just as it is possible to have any number of geometries other 
than the Euclidean which give an equally perfect account of space con- 
figurations, so it is possible to have descriptions of the universe, all 
equally valid, that do not contain our familiar contrasts of time and 
space. The relativity viewpoint of modern physics is one such view, 
conceived in mathematical terms, and the Hopi Weltanschauung is 
another and quite different one, nonmathematical and linguistic. 

Thus, the Hopi language and culture conceals a metaphysics, such as 
our so-called naive view of space and time does, or as the relativity 
theory does; yet it is a different metaphysics from either. In order to 
describe the structure of the universe according to the Hopi, it is neces- 
sary to attempt— insofar as it is possible— to make explicit this meta- 
physics, properly describable only in the Hopi language, by means of an 
approximation expressed in our own language, somewhat inadequately it 
is true, yet by availing ourselves of such concepts as we have worked up 
into relative consonance with the system underlying the Hopi view of 
the universe. 

In this Hopi view, time disappears and space is altered, so that it is no 
longer the homogeneous and instantaneous timeless space of our sup- 
posed intuition or of classical Newtonian mechanics. At the same time, 
new concepts and abstractions flow into the picture, taking up the task 
of describing the universe without reference to such time or space— ab- 
stractions for which our language lacks adequate terms. These abstrac- 
tions, by approximations of which we attempt to reconstruct for our- 
selves the metaphysics of the Hopi, will undoubtedly appear to us as 
psychological or even mystical in character. They arc ideas which we 
are accustomed to consider as part and parcel either of so-called ani- 
mistic or vitalistic beliefs, or of those transcendental unifications of 
experience and intuitions of things unseen that are felt by the conscious- 
ness of the mystic, or which are given out in mystical and (or) so-called 
occult systems of thought. These abstractions are definitely given either 


explicitly in words— psychological or metaphysical terms— in the Hopi 
language, or, even more, are implicit in the very structure and grammar 
of that language, as well as being observable in Hopi culture and be- 
haxior. They are not, so far as I can consciously avoid it, projections 
of other systems upon the Hopi language and culture made by me in 
my attempt at an objective analysis. Yet, if mystical be perchance a 
term of abuse in the eyes of a modern Western scientist, it must be 
emphasized that these underlying abstractions and postulates of the 
Hopian metaphysics are, from a detached viewpoint, equally (or to the 
Hopi, more) justified pragmatically and experientially, as compared to 
the flowing time and static space of our own metaphysics, which are 
au fond equally mystical. The Hopi postulates equally account for all 
phenomena andTheir interrelations, and lend themselves even better to 
the integration of Hopi culture in all its phases. 

The metaphysics underlying -our own language, thinking, and modern 
culture (I speak not of the recent and quite different relativity meta- 
physics of modern science) imposes upon the universe two grand cosmic 
FORMS, space and time; static three-dimensional infinite space, and 
kinetic one-dimensional uniformly and perpetually flowing time— two 
utterly separate and unconnected aspects of reality (according to this 
familiar way of thinking). The flo^^'ing realm of time is, in turn, the 
subject of a threefold division: past, present, and future. 

The Hopi metaphysics also has its cosmic forms comparable to these 
in scale and scope. What are they? It imposes upon the universe two 
grand cosmic forms, which as a first approximation in terminology we 
may call manifested and manifesting (or, unmanifest) or, again, ob- 
jective and subjective. The objective or manifested comprises all that 
is or has been accessible to the senses, the historical physical universe, 
in fact, with no attempt to distinguish between present and past, but 
excluding everything that we call future. The subjective or manifesting 
comprises all that we call future, but not merely this; it includes 
equally and indistinguishably all that we call mental— everything that 
appears or exists in the mind, or, as the Hopi would prefer to say, in the 
heart, not only the heart of man, but the heart of animals, plants, and 
things, and behind and within all the forms and appearances of nature 
in the heart of nature, and by an implication and extension which has 
been felt by more than one anthropologist, yet would hardly ever be 
spoken of by a Hopi himself, so charged is the idea with religious and 


magical awesomeness, in the very heart of the Cosmos, itself.^ The 
subjective realm (subjective from our viewpoint, but intensely real and 
quivering with life, power, and potency to the Hopi) embraces not only 
our FUTURE, much of w^hich the Hopi regards as more or less predestined 
in essence if not in exact form, but also all mentality, intellection, and 
emotion, the essence and typical form of which is the striving of pur- 
poseful desire, intelligent in character, toward manifestation— a mani- 
festation which is much resisted and delayed, but in some form or other 
is inevitable. It is the realm of expectancy, of desire and purpose, of 
vitalizing life, of efficient causes, of thought thinking itself out from an 
inner realm (the Hopian heart) into manifestation. It is in a dynamic 
state, yet not a state of motion— it is not advancing toward us out of a 
future, but already with us in vital and mental form, and its dynamism 
is at work in the field of eventuating or manifesting, i.e. evolving with- 
out motion from the subjective by degrees to a result which is the ob- 
jective. In translating into English, the Hopi will say that these entities 
in process of causation 'will come' or that they— the Hopi— 'will come 
to' them, but, in their own language, there are no verbs corresponding 
to our 'come' and 'go' that mean simple and abstract motion, our purely 
kinematic concept. The words in this case translated 'come' refer to the 
process of eventuating without calling it motion — they are 'eventuates to 
here' (peMn) or 'e\entuates from it' [angqo] or 'arrived' {pitu, pi. oki) 
which refers only to the terminal manifestation, the actual arrival at a 
given point, not to any motion preceding it. 

This realm of the subjective or of the process of manifestation, as 
distinguished from the objective, the result of this universal process, 
includes also— on its border but still pertaining to its own realm— an 
aspect of existence that we include in our present time. It is that which 
is beginning to emerge into manifestation; that is, something which is 
beginning to be done, like going to sleep or starting to write, but is not 
yet in full operation. This can be and usually is referred to by the same 
verb form (the expective form in my terminology of Hopi grammar) 
that refers to our future, or to wishing, wanting, intending, etc. Thus, 
this nearer edge of the subjective cuts across and includes a part of our 
present time, viz. the moment of inception, but most of our present 

1 This idea is sometimes alluded to as the 'spirit of the Breath' {hikwsu) and as the 
'Mighty Something' {^a^ne h'nnu), although tliese terms may have lower and less 
cosmic though always awesome connotations. 


belongs in the Hopi scheme to the objective realm and so is indistin- 
guishable from our past. There is also a verb form, the inceptive which 
refers to this edge of emergent manifestation in the reverse way— as be- 
longing to the objecti\'e, as the edge at which objectivity is attained; this 
is used to indicate beginning or starting, and in most cases there is no 
difference apparent in the translation from the similar use of the expec- 
tivls. But, at certain crucial points, significant and fundamental dif- 
ferences appear. The inceptive, referring to the objective and result 
side, and not like the expective to the subjective and causal side, implies 
the ending of the work of causation in the same breath that it states the 
beginning of manifestation. If the verb has a suffix which answers 
somewhat to our passive, but really means that causation impinges upon 
a subject to effect a certain result— i.e. 'the food is being eaten,' then 
addition of the inceptive suffix in such a way as to refer to the basic 
action produces a meaning of causal cessation. The basic action is in 
the inceptive state; hence whatever causation is behind it is ceasing; the 
causation explicit]}' referred to by the causal suffix is hence such as we 
would call past time, and the verb includes this and the incepting and 
the decausating of the final state (a state of partial or total eatenness) 
in one statement. The translation is 'it stops getting eaten.' Without 
knowing the underlying Hopian metaphysics, it would be impossible to 
understand how the same suffix may denote starting or stopping. 

If we were to approximate our metaphysical terminology more closely 
to Hopian terms, we should probably speak of the subjective realm as 
the realm of hope or hoping. Ever}- language contains terms that have 
come to attain cosmic scope of reference, that crystallize in themselves 
the basic postulates of an unformulated philosophy, in which is couched 
the thought of a people, a culture, a civilization, e\en of an era. Such 
are our words 'reality, substance, matter, cause,' and as we have seen 
'space, time, past, present, future.' Such a term in Hopi is the word 
most often translated 'ho-pe—tundtya—it is in the action of hoping, it 
hopes, it is hoped for, it thinks or is thought of with hope,' etc. Most 
metaphysical words in Hopi are verbs, not nouns as in European lan- 
guages. The \erb tundtya contains in its idea of hope something of our 
words 'thought,' 'desire,' and 'cause,' which sometimes must be used to 
translate it. The word is really a term which crystallizes the Hopi philos- 
ophy of the universe in respect to its grand dualism of objective and 
subjecti\'e; it is the Hopi term for subjective. It refers to the state of 


the subjective, unmanifest, vital and causal aspect of the Cosmos, and 
the fermenting activity toward fruition and manifestation with which it 
seethes— an action of hoping; i.e. mental-causal activity, which is forever 
pressing upon and into the manifested realm. As anyone acquainted 
with Hopi society knows, the Hopi see this burgeoning activity in the 
growing of plants, the forming of clouds and their condensation in rain, 
the careful planning out of the communal activities of agriculture and 
architecture, and in all human hoping, wishing, striving, and taking 
thought; and as most especially concentrated in prayer, the constant 
hopeful praying of the Hopi community, assisted by their exoteric com- 
munal ceremonies and their secret, esoteric rituals in the underground 
kivas— prayer which conducts the pressure of the collective Hopi thought 
and will out of the subjective into the objective. Tlie inceptive form of 
tundtya, which is tundtyava, does not mean 'begins to hope,' but rather 
'comes true, being hoped for.' Why it must logically have this meaning 
will be clear from what has already been said. The inceptive denotes 
the first appearance of the objective, but the basic meaning of tundtya is 
subjective activity or force; the inceptive is then the terminus of such 
activity. It might then be said that tundtya 'coming true' is the Hopi 
term for objective, as contrasted with subjecti\'e, the two terms being 
simply two different inflectional nuances of the same verbal root, as the 
two cosmic forms are the two aspects of one reality. 

As far as space is concerned, the subjecti\e is a mental realm, a realm 
of no space in the objective sense, but it seems to be symbolically related 
to the vertical dimension and its poles the zenith and the underground, 
as well as to the 'heart' of things, which corresponds, to our word 'inner' 
in the metaphorical sense. Corresponding to each point in the objective 
world is such a vertical and vitally inner axis which is what wc call the 
wellspring of the future. But to the Hopi there is no temporal future; 
there is nothing in the subjective state corresponding to the sequences 
and successions conjoined with distances and changing physical con- 
figurations that we find in the objecti\e state. From each subjective 
axis, which may be thought of as more or less vertical and like the 
growth-axis of a plant, extends the objective realm in every physical 
direction, though these directions are typified more especially by the 
horizontal plane and its four cardinal points. The objective is the great 
cosmic form of extension; it takes in all the strictly extensional aspects 
of existence, and it includes all intervals and distances, all seriations and 


number. Its distance includes what we call time in the sense of the 
temporal relation between events which have already happened. The 
Hopi conceive time and motion in the objective realm in a purely opera- 
tional sense— a matter of the complexity and magnitude of operations 
connecting events— so that the element of time is not separated from 
whatever element of space enters into the operations. Two events in 
the past occurred a long 'time' apart (the Hopi language has no word 
quite equivalent to our 'time') when many periodic physical motions 
have occurred between them in such a way as to traverse much distance 
or accumulate magnitude of physical display in other ways. The Hopi 
metaphysics does not raise the question whether the things in a distant 
village exist at the same present moment as those in one's own village, 
for it is frankly pragmatic on this score and says that any 'events' in the 
distant village can be compared to any events in one's own village only 
by an interval of magnitude that has both time and space forms in it. 
Events at a distance from the observer can only be known objectively 
when they are 'past' (i.e. posited in the objective) and the more distant, 
the more 'past' (the more worked upon from the subjective side). Hopi, 
with its preference for verbs, as contrasted to our own liking for nouns, 
perpetually turns our propositions about things into propositions about 
events. What happens at a distant village, if actual (objective) and not 
a conjecture (subjective) can be known 'here' only later. If it does not 
happen 'at this place,' it does not happen 'at this time'; it happens at 
'that' place and at 'that' time. Both the 'here' happening and the 
'there' happening are in the objective, corresponding in general to our 
past, but the 'there' happening is the more objectively distant, meaning, 
from our standpoint, that it is further away in the past just as it is 
further away from us in space than the 'here' happening. 

As the objective realm displaying its characteristic attribute of exten- 
sion stretches away from the observer toward that unfathomable remote- 
ness which is both far away in space and long past in time, there comes 
a point where extension in detail ceases to be kno\^able and is lost in the 
vast distance, and where the subjective, creeping behind the scenes as 
it were, merges into the objective, so that at this inconceivable distance 
from the observer— from all obseners— there is an all-encircling end and 
beginning of things where it might be said that existence, itself, swallows 
up the objective and the subjective. The borderland of this realm is as 
much subjective as objective. It is the abysm of antiquity, the time and 


place told about m the myths, which is known only subjectively or 
mentally— the Hopi realize and even express in their grammar that the 
things told in myths or stories do not ha\e the same kind of reality or 
validity as things of the present day, the things of practical concern. As 
for the far distances of the sky and stars, what is known and said about 
them is supposititious, inferential— hence, in a way subjective— reached 
more through the inner vertical axis and the pole of the zenith than 
through the objective distances and the objective processes of vision and 
locomotion. So the dim past of myths is that corresponding distance 
on earth (rather than in the heavens) which is reached subjectively as 
myth through the vertical axis of reality via the pole of the nadir— hence 
it is placed below the present surface of the earth, though this does not 
mean that the nadir-land of the origin myths is a hole or cavern as we 
should understand it. It is Paldtkwapi 'At the Red Mountains,' a land 
like our present earth, but to which our earth bears the relation of a 
distant sky— and similarly the sky of our earth is penetrated by the heroes 
of tales, who find another earthlike realm above it. 

It may now be seen how the Hopi do not need to use terms that refer 
to space or time as such. Such terms in our language are recast into 
expressions of extension, operation, and cyclic process provided they 
refer to the solid objective realm. They are recast into expressions of 
subjectivity if they refer to the subjective realm— the future, the psychic- 
mental, the mythical period, and the invisibly distant and conjectural 
generally. Thus, the Hopi language gets along perfectly without tenses 
for its verbs. 




'Tphe ethnologist engaged in studying a living primitive culture must 
-^ often have wondered: "What do these people think? How do they 
think? Are their intellectual and rational processes akin to ours or 
radically different?" But thereupon he has probably dismissed the idea 
as a psychological enigma and has sharply turned his attention back to 
more readily observable matters. And yet the problem of thought and 
thinking in the native community is not purely and simply a psycho- 
logical problem. It is quite largely cultural. It is moreover largely a 
matter of one especially cohesive aggregate of cultural phenomena that 
we call a language. It is approachable through linguistics, and, as I hope 
to show, the approach requires a rather new type of emphasis in lin- 

* This paper was found by me in handwritten manuscript form, undated, among 
the papers left by Whorf to his wife and recently turned over to his son, Robert 
Whorf. The manuscript appeared to be complete (except for certain footnotes), but 
it was generally in a somewhat unfinished state, necessitating some editorial work on 
my part. Notes on the manuscript indicate that Whorf intended to prepare it for 
publication. He even listed individuals to whom he planned to send reprints includ- 
ing Jung, N(ayan) L(ouise) Redfield, Sapir, Carroll, Wayne Dennis, (Claude) Brag- 
don, H. G. Wells, and H. L. Mencken. We may date the writing of this article as 
taking place about late 1936, from two facts: first, it must have occurred after the 
publication, in early 1936, of his article, "The punctual and segmentative aspects of 
verbs in Ilopi," to which he refers, and, second, it probably preceded the writing (in 
late 1937) of his article, "Grammatical categories," which gives a somewhat more 
fully developed notion of cryptotype than occurs in the present paper. 



guistics, now beginning to emerge through the work of Sapir, Leonard 
Bloomfield, and others, though Boas enunciated it decades ago in his 
introduction to the Handbook of American Indian languages. 

One of the clearest characterizations of thinking is that of Carl Jung, 
who distinguishes four basic psychic functions: sensation, feeling 
{Gefiihl), thinking, and intuition.^ It is evident to a linguist that think- 
ing, as defined by Jung, contains a large linguistic element of a strictly 
patterned nature, while feeling is mainly nonlinguistic, though it may 
use the vehicle of language, albeit in a way quite different from thinking. 
Thinking may be said to be language's own ground, whereas feeling deals 
in feeling values which language indeed possesses but which lie rather 
on its borderland. These are Jung's two rational functions, and by con- 
trast his two irrational functions, sensation and intuition, may fairly be 
termed nonlinguistic. They are, it is true, involved in the processes of 
talking, hearing, and understanding, but only in an infinitesimal part of 
their entire range. We are thus able to distinguish thinking as the func- 
tion which is to a large extent linguistic.^ 

The linguistic side of silent thinking, thinking without speaking, is of 
a nature as yet little appreciated. Silent thinking is basically not sup- 

1 To the reader who may not be prepared to accept all of Jung's views, I might say 
that his conception of these functions is essentially that of earlier psychologists such 
as Wundt, to which, however, he adds his own penetrative insight and clarification 
of fundamentals. A distinctive feature in Jung's viewpoint is that his four functions 
are distinguished not merely qualitatively but as separate energy systems of operation 
of an energic principle, the Jungian libido, which feature contrasts them with mere 
processes and complexes. (They are relatively closed systems.) In other words, if I un- 
derstand Jung rightly, none of the libido or energy available for thinking can pass over 
into the form of feeling or sensation and vice versa, except by going into the uncon- 
scious and receding so far therein that it reaches the primitive undifferentiated state. 
This libido concept has proved itself of psychiatric value, and it may also have sig- 
nificance for the "linguistics of thinking" if it is true that the psychic energy available 
for linguistic processes (included in the thinking function) is a differentiated energy, 
entrained in a closed system and not transferable between such systems. However, 
such a Jungian viewpoint is by no means necessary for the linguistic approach to 
thinking which I am here dealing with. [These views of Jung will be found in his 
Psychological types (trans, by Baynes, New York and London, 1923). — JBC] 

2 Some have supposed thinking to be entirely linguistic. Watson, I believe, holds 
or held this view, and the great merit of Watson in this regard is that he was one of 
the first to point out and teach the very large and unrecognized linguistic element in 
silent thinking. His error lies in going the whole hog; also, perhaps, in not realizing 
or at least not emphasizing that the linguistic aspect of thinking is not a biologically 
organized process, "speech" or "language," but a cultural organization, i.e., a lan- 
guage. Some linguists may also hold the idea that thinking is entirely linguistic. 


pressed talking or inaudibly mumbled words or silent lar)'ngeal agitations 
as some have supposed.^ Such an explanation merely appears plausible 
to the linguistically unsophisticated "common sense" view. "Common 
sense" is unaware that talking itself means using a complex cultural 
organization, just as it is unaware of cultural organizations in general. 
Sense or meaning does not result from words or morphemes but from 
patterned relations between words or morphemes. Isolations of a mor- 
pheme, like "John!" or "Come!" are themselves patterns or formulas of 
a highly specialized type, not bare units.* Words and morphemes are 
motor reactions, but the factors of linkage between words and mor- 
phemes, which make the categories and patterns in which linguistic 
meaning dwells, are not motor reactions; they correspond to neural 
processes and linkages of a nonmotor type, silent, invisible, and in- 
dividually unobservable.^ It is not words mumbled, but rapport be- 
tween words, which enables them to work together at all to any semantic 
result. It is this rapport that constitutes the real essence of thought in- 

3 [No text is available for this footnote. Whorf may have intended to refer again 
to Watson, who identiiicd thought with subvocal movements of the speech muscula- 
ture. See his article, "Is thinking merely the action of language mechanisms? (V)," 
Brit. J. Psychol, 11:87-104 (1920).— JBC] 

* Apparent isolations of words m a vocabulary list also derive what meaning they 
have from the patterned "potentials of linkage," which ramify from them and con- 
nect them with complex patterns of Hnguistic formulation. 

5 The pronounced materialist may still be granted leave to regard this matrix of 
relations as consistmg of paths and chains of brain cells or what-not which link and 
relate themselves by physicochemical processes, but no clue to the nature of the 
RAPPORT, the structure of the matrix relations, can be obtained in this way, any more 
than the social organization of a tribe could be worked out from the blood groups of 
its individuals. It can only be determined by a penetrating study of the language 
spoken by the individual whose thinking process we are concerned with, and it will 
be found to be fundamentally different for individuals whose languages are of 
fundamentally different types. Just as cultural facts are only culturally determined, 
not biologically determined, so linguistic facts, which are likewise cultural, and in- 
clude the linguistic element of thought, are only linguistically determined. They are 
determined not merely by language, but by languagES. If the thinkers who are being 
studied speak our own language (let us say English), then the necessary penetrating 
study of the English language which is required can be made only by an investigator 
who has studied and is able to contrast widely diiTering types of language from Eng- 
lish, for only in this way can there be brought into the forefront of consciousness an 
awareness of the existence of mere bare relations that do not correspond to any 
verbalized concepts but nevertheless govern absolutely the linkages of morphemes and 
shape the channels of thinking. [This footnote is extracted from a preliminary draft, 
and appears to represent what Whorf intended at this point. — JBC] 


sofar as it is linguistic, and that in the last resort renders the mumbling, 
laryngeal quiverings, etc., semantically de hop. The nonmotor processes 
that are the essential thing are, of their nature, in a state of linkage ac- 
cording to the structure of a particular language, and activations of these 
processes and linkages in any way, with, without, or aside from laryngeal 
behavior, in the forefront of consciousness, or in what has been called 
"the deep well of unconscious cerebration," are all linguistic patterning 
operations, and all entitled to be called thinking. 

Moreover, an analysis of silent thinking into motor quiverings cor- 
responding to suppressed words and morphemes would no more be a 
real analysis of thinking than the analysis of a language into actual words 
and morphemes would be a real analysis of the language. The crudest 
and most amateurish grammar analyzes more effectively than that, and 
any scientific grammar is necessarily a deep analysis into relations. 

For example, gender in English is a S}stem of relations that has an 
almost minimal outward representation in morphemes. Its only motor 
reactions are the two pronouns 'he' and 'she.' " The motor processes 
which actualize the gender-linked nouns are undifferentiated in gender, 
but the linkage between such a motor process and another motor process 
actualizing the proper pronoun, 'he' or 'she,' is (1) differentiated in 
gender, (2) a nonmotor process, since the two motor processes are dis- 
crete and may even be separated by a prolonged period of rest. The 
gender nouns, such as boy, girl, father, wife, uncle, woman, lady, in- 
cluding thousands of given names like George, Fred, Mary, Charlie, 
Isabel, Isadore, Jane, John, Alice, Aloysius, Esther, Lester bear no dis- 
tinguishing mark of gender like the Latin -us or -a within each motor 
process; but nevertheless each of these thousands of words has an in- 
variable linkage-bond connecting it with absolute precision either to 
the word 'he' or the word 'she,' which however does not come into the 
overt-behavior picture until and unless special situations of discourse 
require it.'^ These thousands of linkage processes rallying around the 
common point of the pronoun and ramifying to all the thousands of 
nouns of one gender form a sort of psychic complex belonging to (1) the 

« Including, of course, their inflections 'his, him, her, hers.' 

"^ [A marginal note in the MS shows that Whorf intended to point out, in a foot- 
note, that use of gender linked nouns is not dependent upon knowing any particular 
individual to which they may refer, although it inevitably classifies such individuals 
as to sex. — JBC] 


iionmotor and nonactualized realm, (2) the thinking function m Jungs 
definition, (3) the hnguistic and cultural order. 

There is no evident reason why such a complex should not enter into 
various functional relations with other material of thought without 
necessarily requiring the activation of any of the individual words or 
class marks with which it is connected. We can be thinking of, say, 
the division of labor between the sexes in a certain culture without 
having to think of the rather bookish words 'female' and 'male' and to 
refer continually to them in our meditations upon such a subject. What 
we more probably do as we run over such a question in our minds is sift 
the facts in terms of a sort of habitual consciousness of two sex classes 
as a standing classificatory fact in our thought-world, something which 
is quite different from sex as a concept or sex as a feeling-value. The 
basis of this shadowy, abstract, and wordless adumbration of a sex classi- 
fication is not a word like 'sex' or 'female' or 'women'; it is a linguistic 
RAPPORT as distinguished from a linguistic utterance. In English it is 
probably a rising toward fuller consciousness of the two great complexes 
of linkage bonds pertaining to the linguistic sex-gender system. It is, 
one might say, the total pronominal-linkage pressure of the George, 
Dick, and William class of words, or of the Jane, Sue, and Betty class, 
that functions in the meditation and not a verbal concept like 'male' 
or 'female.' But in a language without sex gender, like Chinese or Hopi, 
any thinking in terms of a sex classification could not be of this nature; 
it would presumably operate around a word, or a feeling, or a sexual 
image, or a symbol, or something else. 

A linguistic classification like English gender, which has no overt mark 
actualized along with the words of the class but which operates through 
an invisible "central exchange" of linkage bonds in such a way as to 
determine certain other words which mark the class, I call a covert 
class, in contrast to an overt class, such as gender in Latin. Navaho has 
a covert classification of the whole world of objects based partly on 
animation and partly on shape. Inanimate bodies fall into two classes 
which linguists have styled "round objects" and "long objects." * These 
names, of course, misrepresent; they attempt to depict the subtle in 
terms of the gross, and fail. Navaho itself has no terms which ade- 
quately depict the classes. A covert concept like a covert gender is as 

* [Actually, the Navaho verb system provides for more than two classes of inani- 
mate bodies, a fact which makes Whorf's point, if anything, more valid. — JBC] 


definable and in its way as definite as a verbal concept like 'female' or 
feminine, but is of a very different kind; it is not the analog of a word 
but of a rapport-system, and awareness of it has an intuitive quality; we 
say that it is sensed rather than comprehended. It is possibly the kind 
of concept or idea which in Hindu philosophy is called arupa, formless. 
The Navaho so-called ''round" and "long" nouns are not marked in 
themselves nor by any pronouns. They are marked only in the use of 
certain very important verb stems, in that a different stem is required for 
a "round" or a "long" subject or object. Many other verb stems are 
indifferent to the distinction. A new object, for which the Navaho has 
no name, will be put into one or the other class by analog)^ not analogy 
as it would seem to us, but as guided by the contents of the two Navaho 

A covert linguistic class may not deal with any grand dichotomy of 
objects, it may have a very subtle meaning, and it may have no overt 
mark other than certain distinctive "reactances" with certain overtly 
marked forms. It is then what I call a cryptotype. It is a submerged, 
subtle, and elusive meaning, corresponding to no actual word, yet shown 
by linguistic analysis to be functionally important in the grammar. For 
example, the English particle up meaning 'completely, to a finish,' as in 
'break it up, cover it up, eat it up, twist it up, open it up' can be applied 
to any verb of one or two syllables initially accented, excepting verbs 
belonging to four special cryptotypes. One is the cryptotype of dis- 
persion without boundary; hence one does not say 'spread it up, waste 
it up, spend it up, scatter it up, drain it up, or filter it up.' ^ Another is 
the cryptot}'pe of oscillation without agitation of parts; we don't say 
'rock up a cradle, wa\'e up a flag, wiggle up a finger, nod up one's head,' 
etc.^° The third is the cryptotype of nondurative impact which also 
includes psychological reaction: kill, fight, etc., hence we don't say 
'whack it up, tap it up, stab it up, slam it up, wrestle him up, hate him 
up.' ^^ The fourth is the verbs of directed motion, move, lift, pull, push, 

^ 'Burst' belongs to this cryptotype; the colloquial 'bust' does not. 

10 [In a marginal note, Whorf cites 'shake up,' apparently to pouit out that this 
verb implies agitation of parts. The reader should note, incidentally, that this whole 
discussion concerns only transitive verbs, as is made explicit at the end of the para- 
graph.— JBC] 

^^ [In a marginal note, Whorf alludes to such expressions as 'strike up (a band),' 
'hit it up,' but states that they are not true transitives and are not considered. He 
also refers to verbs such as 'sing, shout, cry' in the same way. — JBC] 


put, etc., with which up has the directional sense, 'upward,' or derived 
senses, even though this sense may be contradicted by the verb and 
hence produce an effect of absurdity, as in 'drip it up.' Outside this 
set of cryptotypes, up may be freely used with transitives in the com- 
pletive-intensive sense. 

Another English cr^'ptotype is that of the transitive verbs of a cover- 
ing, enclosing, and surface-attaching meaning, the reactance of which is 
that UN- may be prefixed to denote the opposite. Hence we say 'un- 
cover, uncoil, undress, unfasten, unfold, unlock, unroll, untangle, untie, 
unwind,' but not 'unbreak, undry, unhang, unheat, unlift, unmelt, un- 
open, unpress, unspill.' With the exception of a few words mostly 
semiarchaic, e.g., 'unsay, unthink, unmake,' the use of un- as a reversive 
prefix in true verbs coincides with the centripetal enclosing and attach- 
ing meaning.^^ We have no single word in the language which can give 
us a proper clue to this meaning or into which we can compress this 
meaning; hence the meaning is subtle, intangible, as is typical of crypto- 
typic meanings. Nevertheless this formless idea delimits a quite definite 
class of words and grammatical forms, and may be dredged up from its 
own plane of thought formations and grasped in a semi-intuitive way. 
To do this, one needs only meditate on the meaning of the cr^ptotype, 
e.g. of the typical verbs which take un-, or to use methods of free- 
analogizing akin to the "free-association" methods of Freud and Jung. 
Thus I can imagine a newly coined verb flimniick. If fliimnick means, 
let us say, 'tie a tin can to,' then it falls into the cryptotype and I can 
say, e.g., 'he unflimmicked the dog.' But, if it means 'to take apart,' 
there will be no tendency for anyone to make a form unflimmick mean- 
ing 'put together'; e.g., 'he unflimmicked the set of radio parts.' Such a 
form will appear strange and unacceptable. Similarly a knowledge of 
this cryptotype previous to the adoption of the new words 'camouflage' 
and 'wangle' would have enabled us to predict that it would be possible 
to say 'uncamouflage it,' but not 'unwangle it.' 

12 [From a marginal note, it is evident that Whorf intended to consider the words 
'unstart,' 'unbalance/ and 'undo' in a footnote. Whorf might also have cautioned 
the reader against being misled by participial or adjectival forms such as 'unbroken, 
unheated, unopened,' etc. in which the prefix 'un-' does not denote the reverse of an 
action, but of an adjectivally expressed condition. It is interesting to speculate on the 
possibility that the reason that words such as 'unsay, unthink, unmake' are now ob- 
solete may be precisely the fact that they had to yield to the pressure of the crypto- 
type represented by such words as 'uncover, uncoil, undress,' etc. — JBC] 


In contrast to the cryptotype I give the name phenotype to the hn- 
guistic category with a clearly apparent class meaning and a formal mark 
or morpheme which accompanies it; i.e., the phenotype is the "classical" 
morphological category'. The meanings of 'up' and 'un-' are phenotypes, 
and so are the various tenses, aspects, voices, modes, and other marked 
forms which all grammars study. Grammatical research up to the pres- 
ent time has been concerned chiefly with study of phenotypes. A cer- 
tain type of grammar proceeds as if linguistic meaning dwelt wholly in 
them. The anthropologist should not be satisfied with such a grammar, 
any more than with an ethnology that described only positive behavior 
and ignored the patterning of taboos and avoidances. It can be shown 
that, in some languages at least, linguistic meaning results from the 
interplay of phenotypes and cryptotypes, not from phenotypes alone. 

Thus in Hopi the use of the aspect and tense forms is often governed 
by cryptotypes. They govern, for instance, the way of expressing the 
beginning of an action or state, the English 'begins to do,' or 'begins to 
be' form. First, a different form (phenotype) is used, depending on 
whether the verb is active or inactive (either passive or static), and this 
is a cryptotypic distinction, for the formal apparatus of Hopi grammar 
does not set up any acti\e-versus-inactive contrast. Hopi, moreover, 
classes being 'in, at, over,' or in some other spatial relationship as active, 
but being 'red, long, little, pretty, turned around, shot,' as inactive. 
Causal and incausal are really better terms here than active and passive. 
Next, if the verb is active, the phenotype for beginning depends on 
which of three active cryptotypes is involved. With most verbs one can 
use either the inceptive aspect or the future tense. Analysis seems to 
indicate that Hopi regards the subject of these verbs as working into 
and through the action by a process of dynamic adjustment. The subject 
progressively adjusts himself into the action, and throughout the action 
is maintaining this adjustment either to develop or to stabilize and con- 
tinue the effect. Hopi includes here ^-^ sleeping, dying, laughing, eat- 
ing, as well as most organic functions and most alterative operations, 
e.g., cutting, bending, coxering, placing, and thousands of others. The 
second cryptotxpe uses only future tense to express beginning, and in- 
cludes verbs of straight-line uniform motion, running, fleeing, going, 
coming, being in or at a place or in any spatial relationship, opening, 

13 [Marginally, Whorf notes that this is "strange at first, but illuminating."] 


closing, and certain others. Analysis indicates that here the subject is 
classed as instantly assuming a full-fledged new status, not as dynami- 
cally working into and through a process. The third cryptotype ex- 
presses beginning by means of the projective aspect, a phenotype which 
used elsewhere means 'does with a forward movement.' This crypto- 
type implies that the subject is seized and assimilated by a field of in- 
fluence, carried away by it, as it were; and it consists of gravitational and 
moving-inertia phenomena; 'falling, tumbling, spilling, jumping, whirl- 
ing,' and also, strange though it seems to us, 'going out' and 'going in.' 
According to the logic of Hopi linguistics, a person about to enter a 
house or go outdoors launches off and yields himself to a new influence 
like one who falls or leaps. 

What needs to be clearly seen by anthropologists, who to a large 
extent may have gotten the idea that linguistics is merely a highly 
specialized and tediously technical pigeonhole in a far corner of the 
anthropological workshop, is that linguistics is essentially the quest of 
MEANING. It may seem to the outsider to be inordinately absorbed in 
recording hair-splitting distinctions of sound, performing phonetic gym- 
nastics, and writing complex grammars which onU grammarians read. 
But the simple fact is that its real concern is to light up the thick dark- 
ness of the language, and thereby of much of the thought, the culture, 
and the outlook upon life of a gi\en community, with the light of this 
"golden something," as I have heard it called, this transmuting principle 
of meaning. As I ha\"e tried to show, this amounts to far more than 
learning to speak and understand the language as the practical language 
teacher conceives these ends. The investigator of culture should hold 
an ideal of linguistics as that of a heuristic approach to problems of 
psychology which hitherto he may have shrunk from considering— a glass 
through which, when correctly focused, will appear the true shapes of 
many of those forces which hitherto have been to him but the in- 
scrutable blank of invisible and bodiless thought. 


Awareness of psychological undercurrents is the last thing to arrive in 
the conquest of linguistic understanding, both in the individual and in 
history. The attempt to teach one's language to a foreigner results in 
some awareness of overt formal patterns: paradigms and inflected stems. 


The earliest grammars known are cuneiform wordlists of this kind, giv- 
ing equivalents as between Sumerian and the Semitic Akkadian. A 
further step did not occur until philosophy, in both Greece and India, 
discovered a relation between reasoning and linguistic patterns; this re- 
sulted for philosophy in a formal logic, and for grammar in the discovery 
of at least the more outstanding categories in the classical Indo-European 
tongues. In the Semitic world, grammar remained largely formal, the 
classical Hebrew and Arabic grammars consisting mostly of paradigms, 
known by code names which made no attempt even to characterize, 
much less penetrate, the meanings of these linguistic classes. Even 
Latin grammar, with its terms like indicative, subjunctive, passive, etc., 
was psychological by comparison. The discovery of ancient Hindu 
grammar by Western scholars in the early nineteenth century impressed 
these scholars chiefly by its formal perfection. But it also revealed cer- 
tain psychological subtleties, such as the recognition of different covert 
ideas within word-compounding technique, and the classification of com- 
pounds as tatpurusha, dvandva, bahuvrihi, and so on." 

Even the greatest European grammarians of the nineteenth century 
did not go much beyond formal and overt structures except for riding 
the classical grammatical and philosophical concepts to the limits of 
travel in the languages they studied. To this statement there is one 
grand exception— one of those amazing geniuses who baffle their con- 
temporaries and leave no successors. The real originator of such ideas 
as rapport-systems, covert classes, cryptotypes, psycholinguistic pattern- 
ing, and language as part and parcel of a culture was, so far as I can 
learn, a French grammarian of the early nineteenth century, Antoine 
Fabre d'Olivet (1768-1825),^^ who investigated Semitic languages and 
particularly Hebrew, though his work, like that of Mendel in genetics, 
made no impress whatsoever on the thought of his time. Unfortunately 

!•* [The tatpurusha compounds are those in which one member modifies the other, 
as in such Enghsh words as 'self-made, footsoldier,' etc.; the dvandva compounds are 
those containing coordinate members, the nearest English example being a phrase 
like 'bread and butter,' which would have been rendered in Sanskrit by a single com- 
pound word; the bahuvrihi compounds are adjectival compounds implying a posses- 
sive meaning, and may be exemplified by a Sanskrit word which means 'possessing 
the brightness of the sun.' See William Dwight Whitney, Sanskrit grammar (Har- 
vard University Press, 1931, Chapter XVIII).— JBC] 

1^ [I have supplied the dates, which Whorf intended to fill in. They are as given 
in the Grand dictionnaire universel du XIX^ siecle. Note that the surname is Fabre 
d'Olivet, not d'Olivet.— JBC] 


for its comprehension either then or now, its author was a mystical and 
rehgious metaphysician who mingled this side of his nature with the 
workings of one of the most powerful Hnguistic intellects of any age. 
The result was to produce a mystical and gnostic "translation" of 
Genesis, or rather, an Upanishadic paraphrase that was like some shock- 
ing vision of cosmic space alive with terrific hieroglyphs— that got itself 
promptly placed on the Index. Nor did this repudiation by orthodoxy 
win any encomiums from what was then the radical left, for his Biblical 
views were at the same time too iconoclastic and too transcendental to 
satisfy any possible school of exegesis. But the strictly linguistic part of 
Fabre d'Olivet's work, embodied in La langue hebrdique restituee, which 
appeared in 1815-16,^*^ when separated from his extraordinar}- Upanishad 
upon Genesis, can be seen today to be based on purely linguistic criteria 
and to show great psychological penetration, and ideas far in advance of 
his time. It must be added that, although mystical almost to the point 
of a Jacob Boehme or a William Blake, Fabre d'Olivet steered abso- 
lutely clear of the cabalistic and numerological hocus-pocus with which 
the old Jewish tradition of Hebrew was laden. And, while he threw 
overboard the whole formalistic Hiphil-Hophal conception of grammar, 
he also declined to foist Latin and Greek patterns upon Hebrew. His 
Hebrew stands on its own feet as completely as does Boas's Chinook. 
He reorganized the treatment of verb conjugations on a psycholinguistic 
basis, considered individual prefixes and suffixes from the standpoint of 
their meaning and function, went into the semantics of vowel patterns 
and the semantic coloring of vowels, and showed how many Hebrew 
stems can be resolved into meaningful fractions, as, e.g., such English 
words as 'flash, flicker, clash, click, clack, crack, crash, lick, lash' can be 
so resolved. Refusing to identify the letters of Hebrew writing with the 
actual phonetic elements and yet perceiving that these elements are not 
mere sounds, but stereotyped, codified, and patterned semantic sounds, 
he advanced to a conception of the phoneme, which he called the "sign" 
or the "vocal sign"— struggling with terminology but showing real in- 

1" [I have supplied the dates, which Whorf left blank in the manuscript. The full 
title of this scarce work is La langue hebrdique restituee, et le veritable sens des mots 
hebreux retabli et prouve par leur analyse radicale. Copies are to be seen in the 
Library of Congress, the Cornell University Library, and perhaps a few other libraries 
in the United States. It is probable that Whorf knew the work chiefly from the 
translation into English by Nayan Louise Redfield, The Hebraic tongue restored (New 
York and London, G. P. Putnam's Sons, 1921).— JBC] 


sight into linguistic actualities. He stressed the fact of a complex rap- 
port between signs and between words. A phoneme may assume defi- 
nite semantic duties as part of its rapport. In English the phoneme 6 
(the voiced sound of th) occurs initially only in the cryptotype of de- 
monstrative particles (the, this, there, than, etc.). Hence there is a 
psychic pressure against accepting the voiced sound of th in new or 
imaginary words: thig, thag, thob, thuzzle, etc. not having demonstrative 
meanings. Encountering such a new word (e.g. thoh) on a page, we will 
"instinctively" give it the voiceless sound 9 of th in "think." But it is 
no "instinct." Just our old friend linguistic rapport again. Assign a 
demonstrative meaning, let thag equal 'over the fence,' for instance, and 
we will substitute the voiced phoneme 6 of "there." Fabre d'Olivet 
knew all about such things. 

Moreover, Fabre d'Olivet thought in an anthropological and not 
simply a grammatical way; to him, speech was not a "faculty" exalted 
on its own perch, but something to be understood in the light of human 
behavior and culture, of which it was a part, specialized but involving 
no different principle from the rest. The vocal sign (phoneme) was a 
highly specialized gesture or symbolic act, language a development of 
total somatic behavior becoming symbolic and then diverting its sym- 
bolism more and more into the vocal channel— such is his teaching put 
into the modern idiom. 

No figure so significant for the linguistic approach to thinking again 
appears until we come to the Irish linguist James Byrne (1820-97). His 
studies were based on the exceedingly valuable idea of a worldwide 
survey of grammatical structures in all languages known. His great 
work— it at least deserves to be called great in conception, even though 
perhaps not in execution— in two volumes, called General principles of 
structure of language, appeared in 1885.^^ It had the remarkable feature 
of presenting condensed grammatical sketches of languages all over the 
globe, from Chinese to Hottentot. Almost every linguistic stock outside 
of America is represented, and a good number of American ones. On 
this survey Byrne based his psychological theory. And it seems to me 
at least rather significant that Byrne found, on the basis of language 
structure, a similar contrast of two fundamental psychological types to 

1" [I have supplied the date, as also the dates of Byrne's life. Here and later in the 
manuscript Whorf gives the name as Thomas Byrne, but this must have been due to 
a lapse of memory'. — JBC] 


that which Jung much later found from psychiatry and called the types 
of extraversion and introversion. Jung also showed how, all down 
through history, the irreconcilable opposition of two such types has 
resulted in fundamental controversies and schisms in successive philoso- 
phies and religions. Byrne independently found, or thought he found, 
a correlation betu'cen language structure and two types of mentality, one 
quick-reacting, quick-thinking, and volatile, the other slow-reacting, slow- 
thinking, but more profound and phlegmatic, His slow-thinking men- 
tality, suggestive of Jung's intro\ert, he thought went, on the whole, 
with languages of a synthetic type having a complex overt morphology 
and much derivation and word-building, the extreme of the type being 
polysynthesis. His quick-thinking (extraverted) type went, on the whole, 
with a simpler morphology, lack of synthesis, an analytic or in the ex- 
treme an isolating type of language. 

But, while I am sympathetic to the possibility of such a finding, which 
would indeed be a mighty achie\ement, and also impressed by Byrne's 
anticipation of Jung, I find Byrne's general thesis unconvincing, chiefly 
because I can see how Byrne was working with utterly inadequate ma- 
terials. It is of the greatest importance to man's knowledge of his own 
intellectual makeup, especially in future times, that the really colossal 
task that Byrne so rashly attempted be done as well as possible. This 
would require not only a sur^'ey of many more languages, particularly 
American ones, than Byrne used, but a grammar of each language 
worked out scientifically and on the basis of the language's own patterns 
and classes, and as free as possible from any general presuppositions 
about grammatical logic. Byrne got his materials from old-fashioned 
grammars, formal and even "classical" in cut. These grammars might 
at any juncture quarter a regiment of alien patterns and ideas on the 
unfortunate tongue. Not one of these grammarians, nor Byrne himself, 
could have made a sui generis configurative report on a language as 
Fabre d'Olivet had done; that ability had died. But until it again lives 
as a well-developed scientific technique and is applied to another world 
survey and comparison, man will remain ignorant of the roots of his 
intellectual life. He will be debarred from any consideration of human 
thought on a planetary scale, of what it is in respect to the species. 

That ability began to li\e again with and after the attack made by 
Boas on the American Indian languages, and especially his statement of 
principles, and ideals of method, in his justly celebrated introduction to 


the Handbook}^ And, with Boas, it reappeared in a modern scientific 
form, and in terms of the acceptable science cultus, not as before in 
terms of an exuberant mystically disposed creative imagination. Boas 
showed for the second time in history, but for the first in a scientific 
manner, how a language could be analyzed sui generis and without 
forcing the categories of "classical" tradition upon it. The development 
of an adequate technique for this new outlook had to come haltingly. 
When under Boas the American languages first began to reveal the 
unparalleled complexity and subtlety of their thought categories, the 
phonemic calculus was still unborn. The American field linguist could 
not, like Fabre, intuit the phoneme and morphophoneme in a brilliant 
tour de force of imaginative insight. He had to wait for these concepts 
to be developed by specialized phoneticians, working at first in the mod- 
ern-language field, and at first he lacked in psychological penetration. 

The new era passes into a second phase, into the truly modern lin- 
guistic point of view, with the appearance on the scene of Sapir, and 
particularly with the publication of his Language in 1921.^^ Sapir has 
done more than any other person to inaugurate the linguistic approach 
to thinking and make it of scientific consequence, and moreover to 
demonstrate the importance of linguistics to anthropology and psychol- 
ogy. From this point on it would be a task to mention individual con- 
tributors to this dawning realization and growing idea that linguistics 
is fundamental to the theory of thinking and in the last analysis to all 
HUMAN SCIENCES. The interested reader is referred to the partial and 
very incomplete bibliography appended. 


This linguistic consideration of thinking as applied to primitive com- 
munities is of significance for anthropology in two ways. First, the 
ethnological and the psychological-linguistic insights into the same 
primitive community, especially if made by the same investigator, can 
be reasonably expected to have a very fertilizing effect upon each other. 

^* [Boas, Franz (cd.)- Handbook of American Indian languages (Parts 1 and 2). 
Washington, D. C: Government Printing Office, 1911-22. {Bull. 40, Bur. Amer. 
Ethnol., Smithsonian Institution.) — JBC] 

19 [I have supplied the date. The full reference is: Sapir, Edward. Language; an 
introduction to the study of speech. New York: Harcourt Brace Co., 1921, vii, 
238 pp.— JBC] 


We have the testimony and the enhghtening teaching of Sapir and 
others that this is so. The very essence of hnguistics is the quest for 
meaning, and, as the science refines its procedure, it inevitably becomes, 
as a matter of this quest, more psychological and cultural, while retain- 
ing that almost mathematical precision of statement which it gets from 
the highly systematic nature of the linguistic realm of fact. 

Let us suppose that an ethnologist discovers that the Hopi speak 
about clouds in their rain prayers, etc., as though clouds were alive. He 
would like to know whether this is some metaphor or special religious 
or ceremonial figure of speech, or whether it is the ordinary and usual 
way of thinking about clouds. Here is the sort of problem to which 
language might be able to give a very meaningful answer, and we im- 
mediately turn to it to see if it has a gender system that distinguishes 
living from nonliving things, and, if so, how it classes a cloud. We find 
that Hopi has no gender at all. The traditional grammar of the pre- 
Boas period would stop at this point and think it had given an answer. 
But the correct answer can only be given by a grammar that analyzes 
covert as well as overt structure and meaning. For Hopi does distin- 
guish an animate class of nouns as a cryptotype and only as a cr\'pto- 
type. The crucial reactance is in the way of forming the plural. When 
members of the Flute Society, e.g., are spoken of as Flutes, this (covertly) 
inanimate noun is pluralized in the animate way. But the word 
"^o-'mdw 'cloud,' is always pluralized in the animate way; it has no other 
plural; it definitely belongs to the cryptotype of animateness. And so 
the question whether the animation of clouds is a figure or formality 
of speech or whether it stems from some more deep and subtly per- 
vasive undercurrent of thought is answered, or at the least given a flood 
of new meaning. 

Language thus should be able to analyze some, if probably not all, of 
the differences, real or assumed, between the mentality of so-called 
primitive peoples and modern civilized man. Whether the primitives 
constitute a unit class of mentality over against modern man, apart from 
the differences between their cultures and his, as is implied in Levy- 
Bruhl's concept of participation mystique -^ and in the equation of 

20 [L6vy-Bruhl, Lucien. Les fonctions nientales dans les societes inferieures, Paris 
1912. "Participation mystique" refers to a special kind of psychological relationship 
with the object, in which the individual cannot clearly perceive a separation between 
himself and the object. — JBC] 


"primitive" to "infantile" used by Freud and Jung; or whether (again 
apart from general culture) the civilized modern is the unit class of 
mentality because of the great structural similarity of all the modern 
civilized Western languages, while over against it are many diverse types 
of mentality reflecting a rich diversity of speech structure: This is only 
one of the great psychological world-questions that fall into the domain 
of linguistics and await the impersonal and positive type of answer that 
linguistic research can give. We are accustomed to think of such a 
mentality as is implied by participation mystique as less of a thinking 
mentality, as less rational, than ours. Yet many American Indian and 
African languages abound in finely wrought, beautifully logical dis- 
criminations about causation, action, result, dynamic or energic quality, 
directness of experience, etc., all matters of the function of thinking, 
indeed the quintessence of the rational. In this respect they far out- 
distance the European languages.^^ The most impressively penetrating 
distinctions of this kind often are those revealed by analyzing to the 
covert and even cryptotypic levels. Indeed, covert categories are quite 
apt to be more rational than overt ones. English unmarked gender is 
more rational, closer to natural fact, than the marked genders of Latin 
or German. As outward marks become few, the class tends to crystal- 
lize around an idea— to become more dependent on whatever synthetiz- 
ing principle there may be in the meanings of its members. It may 
even be true that many abstract ideas arise in this way; some rather 
formal and not very meaningful linguistic group, marked by some overt 
feature, may happen to coincide very roughly with some concatenation 
of phenomena in such a way as to suggest a rationalization of this 
parallelism. In the course of phonetic change, the distinguishing mark, 
ending, or what not is lost, and the class passes from a formal to a 

21 See for example the Hopi treatment of repetitive and vibrational phenomena in 
my paper, "Tlie punctnal and segmentative aspects of verbs in Ilopi," or the in- 
stances of [lacuna] in Watkins' Chichewa. [Probably Whorf intended to allude to 
the Chichewa verb system, which is extremely sensitive to the causative aspects of 
acts. For example, there are several past tenses, use of which depends not only on 
the remoteness of the past time being referred to (before or since last night) but also 
on whether the act continues to have an influence on the present. There are also 
seven "voices," which express different kinds of relations among subject, verb, and 
predicate (including object). See pp. 49-57, 72-81 in A grammar of Chichewa, a 
Bantu language of British Central Africa, by Mark Ilanna Watkins, Language Dis- 
sertation no. 24, 1937. See also Whorf's later discussion in his article, "Language, 
mind, and reality" (p. 265 f.). — JBC] 


semantic one. Its reactance is now what distinguishes it as a class, and 
its idea is what unifies it. As time and use go on, it becomes increas- 
ingly organized around a rationale, it attracts semantically suitable words 
and loses former members that now are semantically inappropriate. 
Logic is now what holds it together, and its logic becomes a semantic 
associate of that unity of which the configurative aspect is a bundle of 
nonmotor linkages mooring the whole fleet of words to their common 
reactance. Semantically it has become a deep persuasion of a principle 
behind phenomena, like the ideas of inanimation, of "substance," of 
abstract sex, of abstract personality, of force, of causation— not the o\ert 
concept (lexation) corresponding to the word causation but the co\ert 
idea, the "sensing," or, as it is often called (but wrongly, according to 
Jung), the "feeling" that there must be a principle of causation. Later 
this covert idea may be more or less duplicated in a word and a lexical 
concept invented by a philosopher: e.g., causation. From this point of 
view many preliterate ("primitive") communities, far from being sub- 
rational, may show the human mind functioning on a higher and more 
complex plane of rationality than among ci\ilized men. We do not 
know that civilization is synonymous with rationality. These primitive 
tribes may simply have lacked philosophers, the existence of whom may 
depend on an economic prosperity that few cultures in the course of 
history have reached. Or perhaps too much rationality may defeat it- 
self, or arouse some strong compensatory principle. Tliese are all ques- 
tions, essentially anthropological, to which a liaison between ethnology 
and psychological linguistics would seem to offer the soundest approach. 

The second way in which linguistic consideration of thinking is sig- 
nificant for anthropology has more reference to the future, and perhaps 
most of all to the far distant future of the human species when it will 
have developed into something other, and let us hope far higher, than 
present-day man. Turning first to the nearer future, it is desirable that 
anthropology collaborate in preparation for the time, which cannot be 
too far postponed, when it will be both possible and urgenth- necessary 
to make the cultural and psychological world-survey of languages that 
is envisioned in the work of James Byrne— this time in a way which will 
enrich our science with the prodigal wealth of new truth that lies in that 
field waiting to be discovered. 

As time goes on, the type of knowledge that such a sur\ey would un- 
lock becomes more and more a matter of concern and interest outside 


the world of scholarly pursuits— for it may play a very important part in 
world history that is now in the making. The problems of achieving 
mutual understanding, of language barriers, of propaganda and adver- 
tising, of education, of the technique of managing human affairs with- 
out undue friction, of an intelligence in human relations that can keep 
pace with the changes brought about by the physical sciences, all run 
afoul of this matter of language and thought. Everyone is naturally 
interested in questions of language, although they either do not know 
it, or know it and think they know all about it. There is for example a 
movement for the extended use of Ogden's ingenious artificial language 
called Basic English, which has met with much sympathy among busi- 
nessmen, educators, people interested in international affairs, and social 
prophets like H. G. Wells. There is no use sitting aloof and loftily 
condemning such linguistic movements as unscientific. Unscientific or 
not, they are linguistic phenomena of today, and why should linguistic 
science, which alone can handle the vital underlying principles of such 
movements, stand by in sequestered unconcern and let them blunder 
along, exercising their crude but vast power to change the thinking of 
tomorrow? Basic English appeals to people because it seems simple. 
But those to whom it seems simple either know or think they know 
English— there's the rub! Every language of course seems simple to its 
own speakers because they are unconscious of structure. But English is 
anything but simple— it is a bafflingly complex organization, abounding 
in covert classes, cryptotypes, taxemes of selection, taxemes of order,^- 
significant stress patterns and intonation patterns of considerable in- 
tricacy. English is indeed almost in a class by itself as regards prosodic 
complexity, being one of the most complex languages on earth in this 
respect; on the whole, it is as complicated as most polysynthetic lan- 
guages of America, which fact most of us are blissfully unaware of. The 
complex structure of English is largely covert, which makes it all the 
harder to analyze. Foreigners learning English have to absorb it un- 
consciously—a process requiring years— by dint of constant exposure to 
bombardment by spoken English in large chunks; there exists at this 
moment no grammar that can teach it. As with Basic English, so with 
other artificial languages- underlying structures and categories of a few 

22 [The marginal notation appears: "memberships in covert categories of a certain 
type," and there is a reference to Leonard Bloomfield, Language (New York, 1933) 
where the subject of taxemes is taken up in Chapters 10, 12, and elsewhere. — JBC] 


culturally predominant European tongues are taken for granted; their 
complex web of presuppositions is made the basis of a false simplicity. 
We say 'a large black and white hunting dog' and assume that in Basic 
English one will do the same. How is the speaker of a radically different 
mother tongue supposed to know that he cannot say 'hunting white 
black large a dog'? The English adjecti\es belong to cr)ptotypes having 
definite position assignments, and their formula is a definite and com- 
plex one, but lo, the poor Indian organizes his thinking quite differently. 
The person who would use Basic English must first know or learn the 
immensely intricate covert structure of actual "English as she is spoke." 
We see here the error made by most people who attempt to deal with 
such social questions of language— they naively suppose that speech is 
nothing but a piling up of lexations, and that this is all one needs in 
order to do any and every kind of rational thinking; the far more im- 
portant thought materials provided by structure and configurative rap- 
port are beyond their horizons. It may turn out that the simpler a 
language becomes overtly, the more it becomes dependent upon cr\pto- 
types and other covert formations, the more it conceals unconscious 
presuppositions, and the more its lexations become variable and inde- 
finable. Wouldn't this be a pretty kettle of fish for the would-be advo- 
cates of a "simple" international tongue to have had a hand in stewing 
up! For sound thinking in such fields we greatly need a competent 
world-survey of languages. 


And now, turning to the more distant future, one may perhaps be 
permitted to essay a broader \'iew, to look at the subject of linguistics 
and its bearing upon thinking from the standpoint of the whole human 
species. In order to do this we must not be afraid to begin with a 
platitude. Man is distinguished from other animals by language, and 
by his great de\elopment of thinking. So far as we can en\ision his 
future, we must envision it in terms of mental grow th. We cannot but 
suppose that the future developments of thinking are of priman,' im- 
portance to the human species. They may even determine the duration 
of human existence on the planet earth or in the universe. The possi- 
bilities open to thinking are the possibilities of recognizing relationships 
and the disco\ery of techniques of operating with relationships on the 


mental or intellectual plane, such as will in turn lead to ever wider and 
more penetratingly significant systems of relationships. These possibili- 
ties are inescapably bound up with systems of linguistic expression. The 
story of their evolution in man is the story of man's linguistic develop- 
ment—of the long evolution of thousands of very different systems of 
discerning, selecting, organizing, and operating with relationships. Of 
the early stages of this evolutionary process, the really primitive roots 
of language, we know nothing. What we are at least in a position to 
find out is the results of this evolution as they exist broadcast about 
the planet in our present day. Only the beginnings of such a knowledge 
of worldwide linguistic taxonomy are in evidence. In our armchair gen- 
eralizations about grammar, and the related fields of logic and thought- 
ps}chology, we are in the same position as pre-Linnaean botany. We 
have not yet got anything like a description of existing linguistic species, 
to use a biological metaphor. 

Fortunately for biology, a worldwide systematic taxonomy preceded 
and laid a foundation for the historical and evolutionary approach. In 
linguistics as in other cultural studies, we have had unfortunately the 
reverse situation. The evolutionary concept, having been dumped upon 
modern man while his notions of language and thought were based on 
knowledge of only a few types out of the hundreds of very di\erse lin- 
guistic types existing, has abetted his provincial linguistic prejudices and 
fostered the grandiose hokum that his type of thinking and the few 
European tongues on which it is based represent the culmination and 
flower of the evolution of language! This is as if a pre-Linnaean 
botanist who had conceived the idea of evolution should suppose that 
our cultivated wheat and oats represent a higher evolutionary stage than 
a rare aster restricted to a few sites in the Himalayas. From the stand- 
point of a matured biology, it is precisely the rare aster which has the 
better claim to high evolutionary eminence; the wheat owes its ubiquity 
and prestige merely to human economics and history. 

I'he eminence of our European tongues and thinking habits proceeds 
from nothing more. The relatively few languages of the cultures which 
have attained to modern civilization promise to overspread the globe 
and cause the extinction of the hundreds of diverse exotic linguistic 
species, but it is idle to pretend that they represent any superiority of 
type. On the contrary, it takes but little real scientific study of pre- 
literate languages, especially those of America, to show how much more 
precise and finely elaborated is the system of relationships in many such 


tongues than is ours.-^ By comparison with many American languages, 
the formal systematization of ideas in English, German, French, or 
Italian seems poor and jejune. Why, for instance, do we not, like the 
Hopi, use a different way of expressing the relation of channel of sensa- 
tion (seeing) to result in consciousness, as between 'I see that it is red' 
and 'I see that it is new'? We fuse the two quite different types of rela- 
tionship into a vague sort of connection expressed by 'that,' whereas the 
Hopi indicates that in the first case seeing presents a sensation 'red,' 
and in the second that seeing presents unspecified evidence from which 
is drawn the inference of newness. If we change the form to 'I hear 
that it is red' or 'I hear that it is new,' we European speakers still cling 
to our lame 'that,' but the Hopi now uses still another relater and makes 
no distinction between 'red' and 'new,' since, in either case, the signifi- 
cant presentation to consciousness is that of a \erbal report, and neither 
a sensation per se nor inferential evidence. Does the Hopi language 
show here a higher plane of thinking, a more rational analysis of situa- 
tions, than our vaunted English? Of course it does. In this field and 
in various others, English compared to Hopi is like a bludgeon compared 
to a rapier. We e\en have to think and boggle over the question for 
some time, or ha\e it explained to us, before we can see the difference 
in the relationships expressed by 'that' in the above examples, whereas 
the Hopi discriminates these relationships with effortless ease, for the 
forms of his speech have accustomed him to doing so. 


[At the end of the manuscript appears a section entitled "Bibliography 
(Notes)" which is merely a skeleton of a bibliography; it consists chiefly of 
names. I have already gi\en footnote references for the following names: 
Bloomfield, Boas, Byrne, Fabre d'Olivet, Jung, Sapir, Watkins, and Watson. 
Below I give the citations that Whorf most probably had in mind for the 
other names; in several cases he was explicit. — JBC] 

De Angulo, Jaime. "Tone patterns and verb forms in a dialect of Zapotek." 
Language, 2:238-250 (1926). 

23 [At this point in the manuscript appears a marginal notation: "Conclusion — 
error supposing function of language to be only the communication of thought." 
By emphasizing the word communication, Whorf apparently meant to convey the 
implication that language not only communicates thought but functions in its very 
inception, a conclusion to which we are forced if we accept the main thesis of this 
article.— JBC] 


Flournoy, Theodore. Metaphysique et psychologie. Geneva, 1890. [? This 

may not be the rele\ ant work.] 
Haas, Mary. [Whorf probably referred to unpublished material which he 

had seen. See her sketch of Tunica, an American Indian language, in 

H. Hoijer, Linguistic structures of native America. New York, 1946. — 

Jones, William, and Michelson, Truman. "Algonquian (Fox)." Pp. 75 5- 

873 in Boas, F. (editor). Handbook of American Indian languages, Part 1. 

Washington: Government Printing Office, 1911. 
Koffka, K. Principles of Gestalt psychology. New York: Harcourt, Brace & 

Co., 1935. 
Lowes, John Livingston. Road to Xanadu. Harvard University Press, 1927. 

[Whorf misremembered the author's name as Dickinson. He comments, 

"Interesting for illustrations of the dredging up of linguistic material from 

the unconscious."] 
Murdock, George P. Our primitire contemporaries. New York: Macmillan, 

Newman, Stanley S. A grammar of Yokuts, an American Indian language of 

California. Unpublished Ph.D. dissertation, Yale Univ. 1932. Also, 

Yokuts language of California. New York, 1944. (Viking Fund Publi- 
cation in Anthropology, no. 2.) 
Morice, Adrian G. The Carrier language {Dene family); a grammar and 

dictionary combined. St. Gabriel-Modling near Vienna, Austria, 1932. 
Ogden, Charles K. Basic English: a general introduction with rules and 

grammar. London: K. Paul, Trench, Trubner, 1930. [Whorf's citation 

is to Ogden and Richards, but I believe he meant to refer to this book 

about Basic English.] 
Swadesh, Morris. [Whorf probably referred to unpublished material which 

he had seen. See Swadesh's sketch of South Greenland Eskimo in H. 

Hoijer, Linguistic structures of native America. New York, 1946. — JBC] 
Trager, George L. "The phonemes of Russian." Language, 10:334-344 



' I ^he very natural tendency to use terms derived from traditional gram- 
-■- mar, like verb, noun, adjective, passive voice, in describing languages 
outside of Indo-European is fraught with grave possibilities of misunder- 
standing. At the same time it is desirable to define these terms in such 
a way that we can avail ourselves of their great convenience and, where 
possible, apply them to exotic languages in a scientific and consistent 
way. To do this, we must re-examine the types of grammatical categor}' 
that are found in languages, using a worldwide view of linguistic phe- 
nomena, frame concepts more or less new, and make needed additions 
to terminology. These observations apply pari passu to English, which 
hardly less than some American Indian languages is off the pattern of 
general Indo-European.^ 

In the reaction from conventional grammars of American languages 
based on classical models, there has been a tendency to restrict attention 
to the morphemes by which many grammatical forms are marked. This 
view loses sight of various word-classes that are marked not by mor- 

* Reprinted from Language, 21:1-11 (1945). According to a note supplied by 
the editor of Language, "This paper was written late in 1937 at the request of Franz 
Boas, then editor of the Int. J. Amer. Linguistics. The manuscript was found in 
the Boas collection by C. F. Voegelin and Z. S. Harris." The Boas collection is cata- 
loged in Language Monograph no. 22, 1945. 

1 The author wishes to acknowledge his indebtedness to his colleagues, Dr. George 
L. Trager and Dr. Morris Swadesh, with whom some of these questions of category 
have been discussed. 



phemic tags but by types of patterning: e.g. by the systematic avoidance 
of certain morphemes, by lexical selection, by word-order that is also 
CLASS-ORDER, in general by association with definite linguistic configura- 
tions. At the beginning of investigation of a language, the "functional" 
type of definition, e.g. that a word of a certain class, say a "noun," is "a 
word which does so-and-so," is to be avoided when this is the only test 
of distinction applied; for people's conceptions of what a given word 
"does" in an unfamiliar language may be as diverse as their own native 
languages, linguistic educations, and philosophical predilections. The 
categories studied in grammar are those recognizable through facts of a 
configurational sort, and these facts are the same for all observers. Yet 
I do not share the complete distrust of all functional definitions which 
a few modern grammarians seem to show. After categories have been 
outlined according to configurative facts, it may be desirable to employ 
functional or operational symbolism as the investigation proceeds. 
Linked with configurative data, operational descriptions become valid 
as possible ways of stating the meaning of the forms, "meaning" in 
such cases being a characterization which succinctly accounts for all the 
semantic and configurational facts, known or predictable. 

We may first distinguish between overt categories and covert cate- 

An overt category is a category having a formal mark which is present 
(with only infrequent exceptions) in every sentence containing a member 
of the category. The mark need not be part of the same word to which 
the category may be said to be attached in a paradigmatic sense; i.e. it 
need not be a suffix, prefix, vowel change, or other "inflection," but may 
be a detached word or a certain patterning of the whole sentence. Thus 
in English the plural of nouns is an overt category, marked usually in 
the paradigm word (the noun in question) by the suffix '-s' or a vowel 
change, but in the case of words like 'fish, sheep,' and certain gentilic 
plurals, it is marked by the form of the \erb, the manner of use of the 
articles, etc. In 'fish appeared,' the absence of any article denotes plural; 
in 'the fish will be plentiful,' a pluralizing adjective denotes it; in 'the 
Chinese arrived' and 'the Kwakiutl arrived,' the definite article coupled 
with lack of a singular marker like 'person,' 'Chinaman,' or 'Indian' 
denotes plural. In all these cases plural is overtly marked, and so, with 
few exceptions, are all noun plurals in English, so that noun-plural is an 


overt categoty in English.- In Southern Paiute the subject-person of a 
verb is marked by a sublexical element (or "bound morpheme") that 
cannot stand alone, like English '-s'; but it need not be attached to the 
verb, it may be attached to the first important word of the sentence. In 
English what may be called the potential mode of the verb is an overt 
category marked by the morpheme 'can' or 'could,' a word separate in 
the sentence from the \'erb but appearing in every sentence containmg 
the categor)'. This category is as much a part of the verb system of mor- 
phology as though it were denoted by a bound element in a synthetic 
Algonkian or Sanskrit verb; its morpheme 'can' may replace coordinate 
elements in the same modal system, e.g. 'may, will,' but it may not, like 
a mere lexical item (e.g. 'possibly') be simply added to them. In Hopi 
also there is a rigid system of mutually exclusive "modalities" denoted 
by detached words. 

A covert category is marked, whether morphemically or by sentence- 
pattern, only in certain types of sentence and not in every sentence in 
which a word or element belonging to the category occurs. The class- 
membership of the word is not apparent until there is a question of using 
it or referring to it in one of these special types of sentence, and then 
we find that this word belongs to a class requiring some sort of distinc- 
tive treatment, which may even be the negative treatment of exclud- 
ing that type of sentence. This distinctive treatment we may call the 
REACTANCE of the Category. In English, intransiti^•e \erbs form a covert 
category marked by lack of the passive participle and the passive and 
causative voices; we cannot substitute a verb of this class (e.g. 'go, lie, 
sit, rise, gleam, sleep, arrive, appear, rejoice') into such sentences as 'It 
was cooked, It was being cooked, I had it cooked to order.' An intransi- 
tive thus configuratively defined is quite a different thing from the 
"dummy" intransitive used in traditional English grammar; it is a true 
grammatical class marked by these and other constant grammatical fea- 
tures, such as nonoccurrence of nouns or pronouns after the verb; one 

2 There is of course a minority group of possible or theoretically possible sentences, 
e.g. The fish appeared,' in which plural is not distinguished from singular. But, in 
actual speech, such sentences are embedded in a larger context which has already 
established the plurahty or the singularity of the thing discussed. (Otherwise such a 
sentence is not likely to occur.) Such minority types are not considered in the dis- 
tinction between overt and co\ert; i.e. they do not prevent a category from being 
classed as overt. In covert categories the unmarked forms are relatively numerous, 
often in the majority, and are undistinguished even by context. 


does not say 'I gleamed it, I appeared the table.' Of course compound 
formations involving these same lexemes may be transitive, e.g. 'sleep 
(it) off, go (him) one better.' In the American colloquial forms 'go hay- 
wire, go South Sea Islander,' etc., the word or phrase after the verb is a 
covert adjective, cf. 'go completely haywire.' 

Another type of covert category is represented by English gender. 
Each common noun and personal given name belongs to a certain 
gender class, but a characteristic overt mark appears only when there 
is occasion to refer to the noun by a personal pronoun in the singular 
number— or in the case of the neuter class it may be marked by the 
interrogative and relative pronouns 'what, which.' The grammatical 
alignment is no less strict than in an overt gender s\stem like that of 
Latin, where most nouns bear their gender mark. No doubt for many 
English common nouns a knowledge of actual sex and of scientific 
biological and physical classification of objects could serve a foreigner in 
lieu of knowledge of the grammatical classes themselves, but such knowl- 
edge would be of only limited use after all, for the greater part of the 
masculine and feminine classes consists of thousands of personal names, 
and a foreigner who knows nothing of the cultural background of West- 
ern European Christian names must simply learn, i.e. observe, that 'Jane' 
belongs to the 'she' group and 'John' to the 'he' group. There are plenty 
of names of overt similarity but contrasted gender, e.g. 'Alice : Ellis, 
Alison : Addison, Audrey : Aubrey, Winifred : Wilfred, Myra : Ira, 
Esther : Lester.' ^ Nor would knowledge of any "natural" properties 
tell our observer that the names of biological classes themselves (e.g. 
animal, bird, fish, etc.) are 'it'; that smaller animals usually are 'it'; 
larger animals often 'he'; dogs, eagles, and turke\s usually 'he'; cats and 
wrens usually 'she'; body parts and the whole botanical world 'it'; 
countries and states as Active persons (but not as localities) 'she'; cities, 
societies, and corporations as fictive persons 'it'; the human body 'it'; 
a ghost 'it'; nature 'she'; watercraft with sail or power and named small 
craft 'she'; unnamed rowboats, canoes, rafts 'it,' etc. The mistakes in 
English gender made by learners of the language, including those whose 

3 There are a very few names of indeterminate or double gender: 'Frances 
(Francis),' 'Jessie (Jesse)/ or 'Jess, Jean (Gene), Jocelyn, Sidney, Wynne,' and per 
haps a few others. The number is increased if we include nicknames like 'Bobby, 
Jerry,' etc.; but, all in all, such instances are relatively so few that they in no way 
disturb our alignment of facts. 


own languages are without gender, would alone show that we have here 
covert grammatical categories, and not reflections in speech of natural 
and noncultural differences. 

The classes of nouns based actual])- or ostensibly upon shape, in 
various American languages, may be either overt or covert. In Navaho 
they are covert. Some terms belong to the round (or roundish) class, 
others to the long-object class, others fall into classes not dependent on 
shape. No overt mark designates the class in every sentence. The class 
mark as in English gender is a reactance; not a pronoun, however, but 
a choice between certain verb stems that go definitely with one class and 
no other, although there are ver\' many verb stems indifferent to this 
distinction. I doubt that such distinctions, at least in Navaho, are 
simply linguistic recognitions of nonlinguistic, objective differences that 
would be the same for all obsen-ers, any more than the English genders 
are; they seem rather to be covert grammatical categories. Thus one 
must learn as a part of learning Naxaho that 'sorrow' belongs in the 
"round" class. One's first and "common-sense" impression of covert 
categories like English gender and Navaho shape-class is that they are 
simply distinctions betw^een different kinds of experience or knowledge; 
that we say 'Jane went to her house' because we know that Jane is a 
female. Actually we need not know anything about Jane, Jane may 
be a mere name; yet ha\'ing heard this name, perhaps over the telephone, 
w'e say 'What about her?'. Common sense may then retreat a step 
further and say that we know the name Jane to be given only to females. 
But such experience is linguistic; it is learning English by observation. 
Moreover it is easy to show that the pronoun agrees with the name only, 
not with the experience. I can bestow the name 'Jane' on an automo- 
bile, a skeleton, or a cannon, and it will still require 'she' in pronominal 
references. I have two goldfish; I name one 'Jane' and one 'Dick.' I 
can still say 'Each goldfish likes its food,' but not 'Jane likes its food 
better than Dick.' 1 must say 'Jane likes her food.' The word 'dog' 
belongs to a common gender class with a preference for 'he' and 'it,' 
but the gender-classed gi\en name of a dog determines its own pronoun; 
we do not say 'Tom came out of its kennel,' but 'Tom came out of his 
kennel, Lady came out of her kennel, The female dog came out of its 
{or her) kennel.' "Doggish" names like 'Fido' are of the 'he' class: 
'Towser came out of his kennel.' We say 'See the cat chase her tail,' 
but never 'See Dick chase her tail.' The words 'child, baby, infant' be- 


long to the common class and can take 'it,' but the given names of 
children take either 'he' or 'she.' I can say 'My baby enjoys its food,' 
but it would be linguistically wrong to say 'My baby's name is Helen- 
see how Helen enjoys its food.' Nor can I say 'My little daughter enjoys 
its food/ for 'daughter,' unlike 'baby,' is grammatically in the feminine 

Likewise with various covert categories of exotic languages: where they 
have been thought to be recognitions of objective differences, it may 
rather be that they are grammatical categories that merely accord up to 
a certain point with objective experience. They may represent experi- 
ence, it is true, but experience seen in terms of a definite linguistic 
scheme, not experience that is the same for all obsewers. On the other 
hand, the distinctions between present and absent, visible and invisible, 
made in many American languages, may well represent experiential dif- 
ferences; and again we may have such experiential differences engrafted 
upon purely grammatical classifications, yielding mixed classes such as 
"experiential-present plus grammatical-feminine." 

A covert category may also be termed a cryptotype, a name which 
calls attention to the rather hidden, cryptic nature of such word-groups, 
especially when they are not strongly contrasted in idea, nor marked by 
frequently occurring reactances such as pronouns. They easily escape 
notice and may be hard to define, and yet may have profound influence 
on linguistic behavior. The English intransitive verbs as configuratively 
defined above are a cryptotype. A similar crj'ptotype comprises the 
verbs of "copulative resolution" ('be, become, seem, stay, remain,' etc.), 
which also lack the passive and causative but may be followed by nouns, 
pronouns, and adjectives. Transitives (a cryptotype which includes 'run, 
walk, return,' etc.— indeed most English \erbs) possess the passi\'e and 
causative and may be followed by nouns and pronouns but not by ad- 
jectives alone. Names of countries and cities in English form a cr}'pto- 
type with the reactance that they are not referred to by personal pro- 
nouns as object of the prepositions 'in, at, to, from.' We can say 'I live 
in Boston' but not 'That's Boston— I live in it.' A word of this crypto- 
type is referred to by 'there' or 'here' in place of 'in it, at it, to it,' and 
by 'from there (here)' in place of 'from it.' In various American lan- 
guages such place names constitute a grammatical class; in Hopi they 
lack the nominative and objective cases, occurring only in locational 


cases; in Aztec they bear characteristic endings and exclude the use of 
certain prepositions. 

English adjectixes form two main cr}'ptotypes with subclasses. A 
group referring to "inherent" qualities— including color, material, physi- 
cal state (solid, liquid, porous, hard, etc.), provenience, breed, national- 
ity, function, use— has the reactance of being placed nearer the noun 
than the other group, which we may call one of noninherent qualities, 
though it is rather the residuum outside the first group— including ad- 
jectives of size, shape, position, evaluation (ethical, esthetic, or eco- 
nomic). These come before the inherent group, e.g. 'large red house' 
(not 'red large house'), 'steep rocky hill, nice smooth floor.' The order 
may be reversed to make a balanced contrast, but only by changing the 
normal stress pattern, and the form is at once sensed as being reversed 
and peculiar. The normal pattern has primar\' stress either on the noun 
('steep rocky hi'll') or on the inherent adjective ('pretty Fre'nch girl'). 
We cannot simply re\erse the order of adjectives and say 'French pre'tty 
girl'— the form suggests a contrasted 'French plai'n girl' but the pattern 
of so contrasting adjectives is un-English; the proper contrast is 'plai'n 
French girl.' We can however reverse the adjecti\'es by altering the 
stress pattern and say 'Fre'nch pretty girl,' if in contrast with e.g. 
'Spa'nish pretty girl,' though such forms are clearly exceptional. 

The contrasting term phenotype may be applied to the overt category 
and, when no ambiguity results, to the mark which accompanies the 
overt categor}' in the sentence. 

The distinction between o\ert and covert categories, or phenotypes 
and CRYPTOTYPES, is one of two distinctions of supreme importance in 
the theory of grammatical categories. The other is the distinction be- 
tween what may be called selective categories and modulus cate- 

A selective category is a grammatical class with membership fixed, and 
limited as compared with some larger class. A primary selective cate- 
gory, or LEXEMic category, is one compared to which the next larger class 
is the total lexicon of the language. Certain semantic and grammatical 
properties are assured in the \A'ord by selecting it from a certain class of 
fixed membership not coterminous with the whole \-ocabulary. In order 
that a certain grammatical property may be "in the lexeme," it cannot 
be in all lexemes. The familiar "parts of speech" of most European lan- 
guages, but not of English, are lexemic categories. The situation in 


English is peculiar, and will be touched upon later. Lexemic categories 
may be either overt or covert. Hopi is an example of a language in 
which they are covert. Possibly Maya may be another such case, though 
we lack clear information on that point. In Hopi there is no distinc- 
tion in the simplex (bare-stem) forms between nouns and verbs, and 
sentences are possible in which there is no distinction in the sentence. 
Thus le-'na or pam le-'na means 'it is a flute,' and pe-'na or pam pe-'na 
means 'he writes it.' Hence nouns and verbs may be alike in overt char- 
acteristics. But it is easily possible to make sentences- in which le-'na 
appears with case suffixes and in other forms quite impossible for pe • 'na, 
and vice versa. One has to learn, and cannot always tell from the sen- 
tence, that le-'na and pe-'na belong to different compartments of the 

It is probably more common to find lexemic classes that are overt, as 
in Latin, French, Aztec, Tiibatulabal, Taos, and Navaho. In French, 
ange and mange belong to different compartments of the vocabulary 
(noun and verb) and there is always a feature in the sentence that tells 
which; one does not find such pairs as il mange : il ange, cest un ange : 
c'est un mange. It may be possible to have Ange! versus Mange!, but 
special and abbre\'iated types of sentence like these with their lack of 
formal distinctions do not justify calling the categories covert. In Latin, 
Aztec, Tiibatulabal, and Taos, the distinction is marked not only in the 
sentence, but also usually in the paradigm word itself. Yet this overt 
mark of the noun, verb, or other "part of speech" cannot usually be 
transferred to a lexeme outside of the proper group. The mark that 
goes with a covert lexemic class need not stand for any other category 
such as case, person, or tense, though it does for example in Latin, 
Greek, and Sanskrit. The "absolutive suffixes" found attached to 
lexemic nouns in most Uto-Aztecan languages have basically no other 
character than that of such class-marks, though in Aztec they are also 
tied up with number; and needless ingenuity has been wasted in trying 
to make them out to be "articles" or the like. The absolutive suffixes 
in Taos go with the selective class of nouns but indicate gender and 
number also. In Latin the distinction between the nouns (including in 
this class the adjectives) and the verbs is selective and overt, but that 
between adjectives and substantives is selective and covert; compare est 
gladius and est bonus. As with all covert classes, the distinction is re- 


vealed upon forming the proper type of sentence: est bona occurs, but 
not est gladia. 

Lexemic categories include not only nouns, verbs, adjectives, and other 
"parts of speech," but also "full" words and "empty" ^ words or stems, as 
in Chinese and perhaps the Wakashan languages, and still other types 
of distinction; e.g. in Algonkian the lexemic classes include large groups 
of stems having different combinatory powers and different positions in 
the verbal complex. 

A modulus category is a nonselective category, i.e. it is generally ap- 
plicable and removable at will. Depending on its type, it may be 
applied either to any "major word" (any word excepting small and 
specialized selective classes, e.g. "particles"), or, more often, to any word 
coming within a certain prerequisite larger category, which may be either 
selective or another modulus category'. The cases, tenses, aspects, modes, 
and voices of Indo-European and Azteco-Tanoan ^ languages are modu- 
lus categories, applicable at will to words belonging to the proper larger 
category— cases being moduli of the larger category of nouns; aspects, 
tenses, etc. moduli of the larger category of verbs. Hence the person 
versed only in Indo-European types of grammar poses to himself the 
distinction between selective and modulus classes (or between selectivity 
and modulation) as the distinction between "parts of speech" on the one 
hand and "grammatical forms" of the aspect, tense, and voice type on 
the other. But, in widely different types of speech, these familiar types 
of meaning and function cease to be associated with selectivity and 
modulation in the same way; entirely different alignments there hold 
sway in the grammar, and, until this is recognized, an adequate concep- 
tion of the grammar cannot be obtained. It is not necessary to have 
large categories, such as nouns and verbs, in order to have such modulus 
categories as aspect. In Nitinat ^ (and presumably in the closely related 
Nootka and Kwakiutl) all major words have aspects, such as durative, 

*An "empty" word or stem is probably one that is highly specialized for gram- 
matical or syntactic indication, perhaps in a way that does not admit of being 
assigned a concrete meaning. For example, such a form might have no other mean- 
ing than to serve as the reactance of some other category, or as the signature of a 
modulus category (see the next paragraph). 

5 B. L. Whorf and G. L. Trager, "The relationship of Uto-Aztecan and Tanoan," 
Amer. Anthrop., 39:609-24 (1937). 

6 See Mary Haas Swadesh and Morris Swadesh, "A visit to the other world: a 
Nitinat text," Int. /. Amer. Linguistics, 7:3-4 (1933). 


momentaneous, inceptive, etc.— both the word for 'run' and the word 
for 'house' always bear some element marking this aspect. 

We may use the term "modulus" alone to denote the distinctive class 
meaning and function of the category; thus the present-participle mean- 
ing is a modulus in English. We may also use modulus to mean the 
grammatical operation of producing one such meaning, and hence, 
where no ambiguit}' results, to mean the element or pattern that marks 
the modulus. Thus we can say that in English the present-participle 
modulus is the suffixing of '-ing,' or for short that it is '-ing.' Where 
greater preciseness is desirable, we may call the overt mark the (or a) 
SIGNATURE of the modulus. This distinction is ultimately important; 
sometimes it is necessary to distinguish several signatures of the same 
modulus. In illustrating overt categories we cited the English noun- 
plural, which is a modulus category. The modulus, or plural type of 
meaning, is one and the same thing throughout the various examples, 
but the signatures whereby this plural modulus may be applied to the 
word 'fish' are different from one example to the next. To these signa- 
tures we may add '-s' or '-es,' giving 'fishes.' Since 'sheep, deer, moose, 
caribou,' etc. belong to a cr)'ptotype that excludes '-s,' and "fishermen's 
fish" such as 'trout, bass, salmon, mackerel, cod,' etc. (contrasted with 
"low-grade fish," e.g. 'sharks, skates, eels, sculpins,' etc.) belong to an- 
other such cr}'ptotype, we cannot use this last signature for them. As 
this example shows, it is not necessary to have one-to-one correspondence 
between moduli and signatures. Where a high degree of such one-to- 
one correspondence obtains, it has often been the custom to apply the 
graphic but not very scientific catchword "agglutinative" to the language. 
Languages of the typical "agglutinative" type, such as Turkish, have 
been referred to as if they had such one-to-one correspondence, and 
moreover as if they had no categories but modulus categories. The gram- 
mar of Yana (Hokan stock, California) consists largely of moduli, but 
has also a few selective categories; e.g. a class of stems which must stand 
first in the verbal complex and a class which must stand second. 

A distinction of the same semantic type as that between verbs and 
nouns in selective categories may be handled by modulus categories in- 
stead. That is, the possible moduli include not only voice, aspect, etc., 
but also VERDATiON and stativ.\tion.'^ Whenever, as for example in 

'' Stativation is a term used to denote the modulus of forms which are contrasted 
with verbations in a way similar to that in which nouns, as a selective category', are 


Yana, the mere application of certain distinctive sufExes or other signa- 
tures makes a "verb" out of any stem, then we do not have a class of 
verbs in the same sense as in French, Latin, Greek, Hopi, Aztec, Taos, 
and Navaho: i.e. a selective class. We have verbations instead of such 
verbs. The so-called \erbs and nouns of Semitic are moduli, applicable 
to lexemes in general by signatures consisting largely of \ovvel-consonant 
sequence patterns, though there may be occasional gaps in the uni\er- 
sality of lexical applicability. In Hebrew we have e-e as one of several 
signatures for stativation and d-a as one of several for verbation, e.g. 
herek ^ 'knee' : hdrak 'he kneeled,' derek 'road' : ddrak 'he marched,' 
geher 'man, as virile or strong' : gdhar 'he was strong,' hehel 'cord' : 
hdhal 'he bound,' melek 'king' : mdlak 'he reigned,' qedem 'antece- 
dence' : qddam 'he was before/ regel 'foot' : rdgal 'he went on foot.' 
There are no doubt many Hebrew "nouns" for which we do not know 
the verbation form in texts, but this seems to be so largely because the 
textual Hebrew that we know does not represent the full resources of 
the ancient living language; Arabic shows better the general applicability 
of these moduli to the great majority of lexemes. But verbs and nouns 
which are modulus categories may be found nearer home than Semitic. 
The lexicon of English contains two major selective divisions. One 
division, consisting mostly of long words and words with certain end- 
ings, contains selective verbs like 'reduce, survive, undertake, perplex, 
magnify, reciprocate,' and selective nouns like 'instrument, elephant, 
longevity, altruism.' A limited number of short words belong also to 
the group of selecti\e nouns and verbs, e.g. 'heart, boy, street, road, 
town; sit, see, hear, think.' In this selective vocabulary English is like 
French or Hopi. The other part of the lexicon, mostly the shorter \\ords 
but some long ones, contains bare lexemes to which either verbation or 
stativation may be applied at will, e.g. 'head, hand, stand, walk, ex- 
contrasted with verbs in the languages that have such a contrast. It is used here 
instead of "nomination" or "nominalization," because these terms through past usage 
have come to suggest derivations rather than moduli, while "stativation" helps us to 
think of the form not as a noun derived from a verb, but simply as a lexeme which 
has been affected by a certain meaningful grammatical coloring as a part of certain 

8 Since these Hebrew examples are used only to illustrate vowel patterns, they are 
written in approximate morphophonemic orthography, which does not attempt to 
show the distinction between the stops b, g, k, etc. and the spirants which replace 
them after vowels under regular statable conditions. 


change, sight, skin, weave, dog, surrender, massage,' etc.^ This part of 
the vocabulary is hke Arabic, though the signatures are of a quite dif- 
ferent sort. Those for stativation include the articles, plural signatures, 
position after possessive pronouns and selective adjectives; those for ver- 
bation include position after a nominative pronoun, position before a 
pronoun, noun, or stativation, the tense forms, the verbal auxiliaries and 
modal particles, etc. 

There may be wide variability in the semantic relations between ver- 
bations and stativations in the same language. When contrasted with 
the corresponding stativations, verbations may seem to add in an in- 
constant manner such ideas as 'he engaged in' (hunt, jump, dance), 
'behave like' (mother, carpenter, dog), 'be in' (lodge, hive), 'put in' 
(place, seat, pocket, garage), 'make, add, install' (weave, plant, roof, pipe, 
tin), 'take away' (skin, peel, husk, bone), 'get' (fish, mouse), 'use' (spear, 
hammer, fiddle, bugle); while on the other hand stativations seem to add 
inconstantly such ideas as 'result' (weave, plant, form), 'means' (paint, 
trail), 'action or place' (walk, slide, step, drop), 'instrument' (lift, cover, 
clasp, clip), etc. This inconstancy, or, better, elasticity, in certain aspects 
of the meaning, seen in Semitic as well as in English, is characteristic of 
the simple moduli of verbation and stativation, and it may be contrasted 
with the condition of having a number of different moduli, each a dif- 
ferent specialized type of verbation or stativation, which appears to be 
the situation in Alaskan Eskimo. It merely means that, in a language 
with simple primary types of moduli, the meaning of the individual 
lexeme is more or less under the sway of the entire sentence, and at the 
mercy of the manifold potentialities of connotation and suggestion which 
thereby arise. 

Can there be languages not only without selective nouns and verbs, 
but even without stativations and verbations? Certainly. The power 
of making predications or declarative sentences and of taking on such 
moduli as voice, aspect, and tense, may be a property of every major 
word, without the addition of a preparatory modulus. This seems to be 
the case in Nitinat and the other Wakashan languages. An isolated 
word is a sentence; a sequence of such sentence words is like a compound 
sentence. We might ape such a compound sentence in English, e.g. 

8 Adjectivation in English is another modulus which is applied both to bare lexemes 
and to selective nouns, but there are also selective adjectives, and these are not modu- 
lated into substantives. 


'There is one who is a man who is yonder who does running which 
traverses-it which is a street which elongates,' though the exotic sentence 
consists simply of the predicative lexemes 'one/ 'man,' 'yonder,' 'run,' 
'traverse,' 'street,' and 'long,' and the proper translation is 'A man 
yonder is running down the long street.' Such a structure might or 
might not be found in an isolating language; again it might or might not 
be found in a polysynthetic one like Nitinat. The polysynthetic lan- 
guage might or might not fuse some of the lexemes into long synthetic 
words, but it would doubtless have the power in any case of fusing in a 
great many aspectual, modal, and connective elements (signatures of 
moduli). Of such a polysynthetic tongue it is sometimes said that all 
the words are verbs, or again that all the words are nouns with verb- 
forming elements added. Actually the terms verb and noun in such a 
language are meaningless. The situation therein is radically different 
from for example Hopi, for, though in the latter le-'na 'it is a flute' 
and pe-'na 'he writes it' are both complete sentences, they are words 
which are not equally predicative in all positions of a sentence, and they 
also belong to selective covert classes of noun and verb that in general 
take different inflections, and look alike only in particular types of sen- 
tence. In Hopi the verb-noun distinction is important on a selective 
basis; in English it is important on a modulus basis; in Nitinat it seems 
not to exist. 

So far we have dealt with categories which are distinct both configura- 
tively and semantically, and these are the typical formulations of gram- 
mar. But we also have word groups which are configuratively distinct 
and yet have no difference in meaning; these we may call isosemantic 
or purely formal classes. They in turn are of two sorts corresponding to 
selective and modulus in the semantic categories, but here better styled 
SELECTIVE and ALTERNATIVE. Selective isosemantic classes are typified by 
"declensions" and "conjugations," those very common features of lan- 
guages the world over, richly developed in Latin, Sanskrit, Hopi, and 
Maya, less developed in Semitic, English ("strong" and "weak" verbs), 
and Aztec, and almost lacking in Southern Paiute. They also include 
gender-like classes without semantic difference, as in Bantu and in certain 
of the genders of Taos (all these might be called "declensions" with 
pronominal agreement or tlie like); classes requiring different position in 
a sentence or complex ^^'ithout difference in type of meaning (stem- 


position classes in Algonkian); and classes requiring different signatures 
for the same modulus without difference in type of meaning, e.g. in 
Hebrew the segholate {e-e) "nouns" and parallel stativation groups. 
Alternative isosemantic classes are what their name implies: e.g. the 
English group comprising 'don't, won't, shan't, can't,' etc. and the group 
of 'do not, will not, shall not, cannot.' In this case we could perhaps 
speak of a modulus of brevity, convenience, or colloquial attitude which 
is applied in the former group. Alternative classes sometimes show 
STYLISTIC as opposed to grammatical difference. In other cases there 
seems to be no generalizable difference, as in English 'electrical, cubi- 
cal, cyclical, historical, geometrical' versus 'electric, cubic, cyclic, historic, 

There remains another type of distinction: specific categories and 
GENERIC ones. A specific categor)- is an individual class existing in an 
individual language, e.g. English passive \'oice, Hopi scgmentative aspect. 
A generic category, in the restricted sense of application to a particular 
language, is a hierarchy formed by grouping classes of similar or (and) 
complementary types, e.g. case in Latin, voice in Hopi. Here much de- 
pends on both the insight and the predilections of the systematizer or 
grammarian, for it may be easy to build up specific categories into very 
logical schemes; yet what is rather desired is that such generic categories 
should represent systems which the language itself contains. We do well 
to be skeptical of a grammarian's systematization when it is full of 
ENANTiOMORPHiSM, the pairing with e\ er}' category of an opposite which 
is merely the lack of it. Specific categories of seemingly opposite mean- 
ing such as passive voice and active voice (when this term "active" means 
merely "nonpassive") should be brought into one generic category 
("voice") only when they are more than two, or when, if there are only 
two, taken together they contrast as a unit with some other system of 

Finally, in a still wider sense generic categories may be so formulated 
as to become equivalent to the concepts of a general science of grammar. 
Such categories are made by grouping what seem to us to be similar 


we speak of a category of "passive voice" which would embrace the 
forms called by that name in English, Latin, Aztec, and other tongues. 
Such categories or concepts we may call taxonomic categories, as op- 


posed to DESCRIPTIVE categories. Taxonomic categories may be of the 
first degree, e.g. passive voice, objective case; or of the second degree, 
e.g. voice, case. Perhaps those of the second degree are the more im- 
portant and ultimately the more valuable as linguistic concepts, as gen- 
eralizations of the largest systemic formations and outlines found in 
language when language is considered and described in terms of the 
whole human species. 


320 Wolcott Hill Road, 
Wethersfield, Conn. 
Dear John: 

You will be interested to hear that I ha\e been appointed to a part- 
time lectureship at Yale for the term January to June 1938, Department 
of Anthropology, to give one 2-hour lecture a week on Problems of 
American Linguistics. During the fall term my colleague George L. 
Trager will have the same group in Phonetics, so that I do not plan to 
devote much of any time to phonetic or phonemic problems per se. 
I am going to orient my lectures largely toward a psychological direc- 
tion, and the problems of meaning, thought, and idea in so-called primi- 
tive cultures. Methods of in\'estigating language which reveal some- 
thing of the psychic factors or constants of the American Indians in the 
given linguistic community will be emphasized. I say psychic instead 
of mental, since affect as well as thought, insofar as it is linguistic, will 
be treated. I expect to give a good deal of attention to the subject of 
the organization of raw experience into a consistent and readily com- 
municable universe of ideas through the medium of linguistic patterns. 
Altogether I hope to present some conceptions rather exciting to the 

* The following, heretofore unpublished, was a draft of a letter addressed to nie 
while I was a student of psychology at the University of Minnesota in the fall of 
1937. Even though part of the handwritten draft was copied out on the typewriter, 
with an original and one carbon, it was apparently never sent, for I never received 
it. The draft manuscript was found among family i:)ai5ers. 



cultural anthropologist and to the psychologist, and to have some bud- 
ding exponents of both disciplines among my students. 

In preparing material for the course I should be very glad to have any 
assistance that you and the University can render, for I know you and 
they are interested along this linguistic-psychological line. I might out- 
line here certain concepts that I am now working on, based on analysis 
of the Hopi language, which naturally will be one I shall draw on con- 
siderably for examples, though I expect also to devote some time to 
Aztec and Maya. An introduction to the general problem may be found 
in the analysis of Hopi categories of verb morphology, especially those 
which for purposes of convenience may be called aspect and tense, 
although exactly the same meaning that these terms have in classical 
European linguistics cannot be taken for Hopi. We have however two 
distinct morphological categories, the suffixes of which are differently 
treated and have different positions, the tense suffix coming after the 
voice suffix. There are three tenses: past (i.e., past up to and including 
present), future, and generalized (that which is generally, universally, or 
timelessly true), all of which are mutually exclusive. Of these, the only 
one to be considered here is the future (suffix -ni). A first approxima- 
tion to its meaning is the English future. There are nine aspects, of 
which 1 shall consider chiefly the inceptive {-va) and projective {-to). 
The punctual aspect is the aspect of the simplex (stem without suffixes 
or modification); the past tense is the tense of the simplex. A first ap- 
proximation to the meaning of the incepti\e is 'begins doing' (I shall 
translate the Hopi past tense by the English present), and to that of the 
projective is 'does with a forward movement.' Later I shall refer to the 
progressional resultati\e or "crescentive" form in -iwma. There is an- 
other inceptive-like progressional form in -^yma but I am not including 
it as it has a noticeably different meaning ('is well on the way to getting 
it done'). 

In getting translations of the English 'begins doing it' form with a 
large number of different verbs, we eventually find that, whereas the 
Hopi uses the inceptive for the majority of the verbs, for a certain num- 
ber it uses the projective, and for a considerable number it jumps out 
of the "aspect" categor}' entirely and uses the future tense (in the 
punctual or nonspecified aspect). The usage is consistent and does not 
depend on the formal type ("conjugation") of the verb. Analysis shows 
that it follows in a curious way the lexical meaning of the verb. The 


question at once strikes one: Why should a pattern ('begins doing it') 
which appears to us to be perfectly uniform and of the same nature 
in all cases present itself to the mind of the bilingual English-speaking 
Hopi informant as a meaning which switches back and forth betA'een 
two (or more) fundamental meaning categories of his own language? 
It is also to be noted that, in almost all the cases where the inceptive is 
used for 'begins doing,' the Hopi uses not only the suffix -va but also a 
reduplication. The meaning of reduplication is a durativizing of the 
punctual to denote a more extended process and hence would seem to 
be a logical prerequisite to a form denoting the beginning of a process, 
but nevertheless, where the projective or the future tense translates the 
'begins doing,' the reduplication is not employed. This tends to con- 
firm one's impression that the Hopi observer conceives the events in a 
different manner from the one whose natiNC language is English. 

Thus in the expressions for 'begins writing, breathing, sleeping, flying, 
rolling (over and over), laughing, fighting, smoking, singing, swimming, 
dying, looking, bumping it,^ turning it, digging it, eating it, breaking it, 
tearing it, killing it, tying it, gathering it, lifting it, bending it, putting 
it in, putting it down, picking it up,' and many more, the inceptive {-va 
on reduplicated base) is used. In those for 'begins running, moving, 
fleeing, going homeward, going away, going to (a place), going up (or 
down), talking, opening it, closing it, shooting it (arrow), driving it (car), 
being in or at (a place),' the future is used. Thus the forms are the 
same as for 'will run, move, flee,' etc. In the expressions for 'begins 
going out, going in, coming (= 'arriving' in Hopi), falling down, fall- 
ing (through space, a different verb), going in a circuit, turning, rotat- 
ing, splitting open (intransitive), tipping over, spilling,' and others, the 
projective aspect is used. 'Does with a forward movement' seems ap- 
propriate to many of these, but why it is not also appropriate to many 
of those in the other two groups is not clear, nor why it should be the 
necessary translation of English 'begins doing' in this third group. The 
informant himself can give no explanation. 

From phenomena of this sort, which are not confined to the inceptive 
problem but pervade all Hopi grammar, I conclude that there must be 
to the Hopi speaker a dimly felt relation of similarity between the verb 
usages in each group having to do with some inobvious facet of their 

1 "It" denotes a transitive verb, requiring an object expressed or implied. 


meaning, and therefore itself a meaning, but one so nearly at or below 
the threshold of conscious thinking that it cannot be put into words by 
the user and eludes translation. To isolate, characterize, and understand 
the operation of these dimly felt, barely conscious (or even unconscious) 
meanings is the object of my further analysis. Such an elusive, hidden, 
but functionally important meaning I call a cryptotype. 

Thus I would say that the meanings of the Hopi stems translated 'be 
writing, breathing, sleeping, breaking it,' etc. are similar in that they all 
contain "cryptotype A," while 'running, moving, going homeward, open- 
ing it,' etc. contain "crjptotype B," and 'going in, falling, turning,' etc. 
"cryptotype C." The mode of lettering is simply a varying and pro- 
visional one for each problem. In contrast with the cryptotype, which 
has no formal mark and whose meaning is not clearly evident, but is 
rather a submerged meaning shown as an influence, I give the name 
PHENOTYPE to the Categories inceptive, future, projective, etc.; i.e., the 
phenotype is the "classical" morphological category', having a formal 
mark and a clearl}' apparent class-meaning. For schematic purposes, we 
may refer to inceptive, future, projecti\e, as phenotypes 1, 2, 3, with 
reference to the present problem. It will be seen that, in a language 
such as Hopi, the meanings of grammatical forms result from the inter- 
play of phenotype and cryptotype, and not from phenotype alone. This 
concept is of course extensible to many other languages than Hopi. Lin- 
guistics up to now has studied almost entirely the phenotypes. Study 
of cr}ptotypes opens up a more psychological phase of linguistics. 

So far we have three types of 'begin' forms, Al, B2, C3. If we equate 

'begins ing' to 'begins to be ing' (which is rendered the same in 

Hopi) and then pass to 'begins to be ,' where is not an -ing form 

but either (a) an adjecti\e or (b) a past participle, we get an interesting 
difference. With an adjecti\e, e.g., 'begins to be sweet, red, white, blue, 
hot, big, short, good, round,' etc. a new phenotype (4), i.e. another 
inceptive-like morpheme, appears. A form in -iwnia is now used. This 
form is a combination of voice and aspect: the resultative voice ('as a 
result is sweet') plus the progressional aspect -yna ('does in the course of 
motion, secondarily "passes" or "goes" into or toward a state or condi- 
tion'). In the case (b), e.g. 'begins to be torn, bumped, dug, cut, writ- 
ten,' etc., it makes a difference whether the condition is to be considered 
as resultative or passive, i.e., whether what is beginning is a torn condi- 
tion, which uses phenotype 4 {-hvma), or what is beginning is a tearing 


act happening to a passive subject, which uses (2) (future tense, -ni) plus 
the suffix of the passive voice. If, however, cryptotype C is present, the 
form is C3 (projective, -to) just as in the previous case without passive 
meaning, i.e., 'begins to be split open, rotated, spilled,' etc. is the same 
as 'begins splitting open, rotating, spilling,' although it is possible to use 
the -iwnia form here to give a definitely resultative meaning. In all 

these 'begin to be ' forms we see another cryptotype, D, which is 

evidently the passive-resultative side of the meanings of both A and B, 
besides also including the ordinary descriptive adjective. D can use two 
inceptive phenotypes, 4 or 2. Phenotype 4 is really the progressional 
aspect -ma, -iw- being the sign of the resultative voice corresponding to 
the sign of the passive used with 2. We have added the types D4, D2. 

It may now be noted that D is contrasted with A, B, C as inactive 
with active. A, B, and C are alike in being active: i.e., actual movement 
and change is shown going on, or else it is a vital state, a life process, that 
is shown as in 'sleeping'; the only exception (from our own viewpoint) 
is 'being in a place,' which is very likely not a real exception but a phase 
of the same idea. The "role" of the subject is that of actor, even if the 
actor does nothing but be in a place; this seems to be a common idea to 
A, B, and C. In D the "role" of the subject is the nonacting substance 
that serves to display some condition or quality. Perhaps it might be 
said that, in the first case, the subject is regarded as the causal agent, 
in the second as not the causal agent, of what I call the 'verbation,' 
i.e., the manifestation (action, operation, condition, state, status, rela- 
tion, etc.) announced by the verb. We still cannot state the differences 
among the A, B, and C cryptotypes, but one factor of cryptotypic mean- 
ing has been shown up. The contrast active-inactive or causal-incausal 
nowhere appears as a phenotype in Hopi. But as a cr}'ptotypic contrast 
it is decisive in governing the outward form. 

If we get renderings for 'begins to do,' distinct differences appear as 
compared to 'begins doing.' The Hopi now omits the reduplication 
from the inceptive aspect forms of Al. Evidently the Hopi feels that 
our -ing form denotes a more extended process than the infinitive. The 
B cryptotype is however rendered just the same: i.e., -ni for both 'begins 
moving' and 'begins to move.' The C is also the same: i.e., -to for 
'begins going out' and 'begins to go out.' But, moreover, the A crypto- 
type can now use phenotype 2 {-ni) as well as 1; both forms are possible 
for the same verb with the same translation. In certain cases a slight 


difference in meaning creeps out. We have added one more type, A2, 
which has a shghtly different translation from the reduplicated Al, and 
an even slighter difference in meaning from the unreduplicated Al. 
This difference is hard to perceive, but is shown in certain examples and 
is presumably connected with the difference in the phenotypes -ni and 
-va (and is not cr}ptotypic). 

Returning to cn,ptotypes A, B, and C: A and B are related by the fact 
that A uses both -va and -ni and B uses -ni only. There is then pre- 
sumably something about the B type of ideas that makes the character- 
istic inceptiveness of -va inappropriate. And there is presumably some- 
thing about C that makes both -va and -ni ineligible and demands -to. 
D is characteristically inactive or incausal, and, in contrast. A, B, and C 
represent three different types of activity or causality. Omitting the 
space-relational forms and a few others, causality becomes the same as 
activity. After a prolonged scrutiny and analysis, aided by what grasp 
I have of Hopi linguistic ideas and viewpoints, I arrive at the following 
characterization of these cr>'ptotypes A, B, and C. B is an activity, the 
starting of which implies that there will be a certain amount of actual 
maintenance thereof, springing from the initial impulse supplied by the 
subject or actor. B represents acts springing from a subject-initiated 
impulse, but not necessarily all such acts— onh' those in which the first 
initiation of a phenomenon by the impulse immediately shows the 
activity in full-fledged form, a form for which continuing has simply 
the meaning of adding time, but not of further developing nor of con- 
tinuously adapting the form. In nearly all cases we may read for "im- 
pulse," "volitional impulse" or "will," either actual in an animate actor 
or "felt into," the rarely occurring inanimate subject. Hence the type 
includes the subjectively determined kinds of uniform motion like run- 
ning, fleeing, and "going" of various sorts. Thus, in order to say 'he 
begins to run,' the activity must already have shown itself in the effec- 
tive form of running, \\'hich will not thereafter be developed and stabil- 
ized, but merely continued "as is." "Being," i.e., in Hopi a verbation of 
spatial relationship like 'in, at, on, along, under, with,' etc., is classed as 
of the same type as "running." A ver)' few transitive verbs are included, 
in which the actor transfers his impulse at once to the object in such a 
way that its form of activity is at once completely displayed. Thus, 'he 
opens it' (no matter how little, it is open and no longer closed); closing 
it is classed as a converse operation of the same type (the moving of any 


barrier over the opening is a partial closing which, so far as it goes, has 
been maintained and will not be changed in form by continuance); so 
also 'shooting an arrow.' 

A however is an activity, the start of which does not mean even a 
brief maintenance of itself as a result of one first impulse or tendency. 
If the very first impulse should not be reinforced the laugh would not 
be a laugh, the writing not a writing but at most a mark, or a pen-grasp- 
ing, the break not a break but only a strain. The action is felt to con- 
sist of a developing train of events or a more than momentary applica- 
tion of the will to action, a following-up sort of participation of the 
subject necessary to even the briefest establishment of the action in its 
representative form. Some of the verb meanings of A will at first seem 
strangely selected from this standpoint, but a little meditation on the 
matter will often show a peculiar insight revealed in the Hopi crypto- 
typic meaning. Thus "sleep" is classed here by Hopi as though it felt 
sleep to be a state which the subject developed into by a continuous 
readjustment, not one which he launched himself into; while "running" 
and "talking" are regarded as states launched into, not progressed or 
adjusted into. All transitive verbs except the few special ones in B are 
regarded as of this A type— the affecting of an object requires an ad- 
justment to it. Cr\"ptotype A is thus an essentially dynamic or sym- 
bolically dynamic concept, even though it may refer to a resting state, 
and the dynamism emanates from the subject. Cryptotype A is active 
or symbolically active (actorial) but not dynamic; the subject is simply 
introjected into a state of uniform motion, or its symbolic equi\alent, 
and left there. 

Cryptotype C is an activity (always intransiti^'e), into which the sub- 
ject needs only to be placed in the initial stage in order to be seized by 
a natural tendency and carried on beyond the initial state in spite of 
itself. It is delivered over, as it were, to a realm of tendency, and hence- 
forth is no longer master but must submit to an inevitable development 
and change of the initial state. Thus the initial state may be losing a 
support or losing balance, whereupon the subject is delivered over to 
gravity and 'falls' or 'tips over' or 'spills'; once entered upon this state, 
it has to "see the finish of it." Or, as in the case of turning, rotating, 
and other mechanical actions, it may be inertia, momentum, elasticity, 
or simply some indefinite automatic tendency that seizes the subject, 
once the first projection of action has, like a trigger, released it. The 


symbol for this is "being thrown," which is the symbohc image behind 
the projective aspect -to 'does with a forward movement.' It may at 
first be thought curious that Hopi puts 'going out' and 'going in' in this 
category. Actually no external force seizes the subject and moves him, 
but he does pass over a definite dividing line into another realm, the 
realm of the exterior, whereas he has been a part of the interior, or vice 
versa, and once he has made the change he is subject to the laws and 
nature of the new realm, those of the initial state being absolutely left 
behind. Thus in a way the verb 'go in' or 'go out' might serve as the 
symbol of all the other ideas in this cryptotype. Cr^ptotype C is thus 
dynamic, or symbolically d}namic, but the dynamism does not emanate 
from the subject but from the external field. Finally, cryptot\pe D, as 
we saw, is not dynamic or e\en acti\e, but inactive, i.e. either static or 

The meaning of a phenotype, though ostensibly plain, can really not 
be completely understood in all its subtlety until the cryptotypes that 
go with it have been dredged up from their submerged state and their 
effective meanings to some extent brought into consciousness. There- 
upon the different effects produced by the same phenotype with different 
cr}'ptotypes, and vice versa, result in a more pronounced consciousness 
and clearer understanding of the phenotype itself. We are now in a 
better position to stud}' the subtle meaning of -va and -ni. In the case 
of the latter w^e of course ha\e to begin with a somewhat intuitive study 
of the tense system as a whole, and then of the "future" tense, denoted 
by -ni. Tlie "future" tense asserts that expectancy of the event is pres- 
ent, that the subject's will to the event, if it is a \'oluntary act, is present, 
and that the ver\' first initial point of the e\ent ma\' ha\e been reached 
(here context governs), but that all beyond this is not present but 
future: i.e. the event as a \\'hole is future. In order that the event as 
a whole may be future, the tense cannot imply that the subject's tend- 
ency will persevere beyond the initial point; otherwise the sense of 
futurity would be greatly weakened or completely lost; it would mean 
only 'he is starting his doing of it' or 'he is starting to do it.' On the 
other hand -va means 'starts with perse\erance or gathering impetus to 
do it'; the initial point is present, the next point swiftly impending, and 
the follow-through promised. With the A cryptotype, unreduplicated, 
either -va or -ni may be used with a slight difference in nuance, which is 
shown by some of the examples to mean 'begins to do it.' The -ni says 


the activity has gone to the initial point, but the dynamic, perseverating 
meaning of the cryptotype itself promises that the nondurative condition 
denoted by the \'erb will be fairly embarked into. It is the same with 
'Va, except that it makes the embarkation even more positive. But, with 
the reduplicated or durative A verb, we need -va because the dim crypto- 
typic perseveration in the A verb is merely sufficient to embark, and, in 
order to promise a follow-up through a durative action, we need to assert 
the persevering will or tendency of the subject himself or itself, which is 
done by means of -va. Thus, to say 'he begins to chop it,' meaning that 
he swings the axe to make at least one chop (whether he makes any 
more or decides to quit then and there we don't know), we may use 
either -ni or -va. But, to say 'he begins chopping it,' meaning that his 
first chop will be followed by a second and a third and so on, we need 
the reduplication and -va, not -ni. It is the same if the action is not 
repetitive but merely continuous: i.e. 'he begins sleeping.' 

On the other hand, with cryptotype B, the cryptotypic meaning im- 
plies that, once the initial point is reached (for which -ni is sufficient), 
the event is manifested in its typical form and will maintain itself for 
a time at least in that form; hence the whole question of perseverance 
by the subject is thrown out of the window, so to say. This means that 
-va would be redundant and inappropriate, perhaps suggesting more 
purposive dri\e than is needed in the bare statement of a B-t\pe activity. 
There is about the B-cryptotype words a certain bareness and abstract- 
ness; they announce a type of motion or position and let it go at that. 
All the 'going' and 'coming' words are particularly abstract, having no 
really verbal roots but being merely verbalized postpositions or adverbs, 
'to, from, away,' etc. It is perhaps a matter of the deep layers of Hopi 
thought process that this cold bareness should not be spoiled or falsified 
by the use of an element like -va. 

This illustration will show how the meaning of a form in a language 
like Hopi is capable of being more deeply analyzed by the cryptotype 
concept, and how the totality of meaning is a joint product of crypto- 
typic and phenotypic factors. In many languages the cryptotype con- 
cept would be of little use, but there are languages like Hopi in which 
much of the influential material of paradigm production lies in this 
heavily veiled state, just as there are people whose mental life is much 
less accessible than that of others. Cryptotypes play a much larger part 
in Hopi than this rather minor problem of inceptive forms, which how- 


ever yields a neat illustration. I believe I am the first to point out the 
existence of this submerged layer of meaning, which in spite of its sub- 
mergence functions regularly in the general linguistic whole. 

I am very curious to know what you as a psychologist think of this 
general idea. How does it bear on the problems you have in hand? The 
resemblance of the crj'ptotypes to the concept of the unconscious of 
Freud, and still more perhaps of Jung, will no. doubt strike you, although 
the parallel should perhaps not be carried too far. 


Tn the earlier stages of work on the Hopi language, I had the pleasant 
■*• feeling of being in familiar linguistic territory. Here, wondrous to 
relate, was an exotic language cut very much on the pattern of Indo- 
European: a language with clearly distinct nouns, verbs, and adjectives, 
with voices, aspects, tense-moods, and no outre categories, no gender- 
like classes based on shape of objects, no pronouns referring to tribal 
status, presence, absence, visibility, or invisibility. 

But, in course of time, I found it was not all such plain sailing. The 
sentences I made up and submitted to my Hopi informant were usually 
wrong. At first the language seemed merely to be irregular. Later I 
found it was quite regular, in terms of its own patterns. After long study 
and continual scrapping of my preconceived ideas, the true patterning 
emerged at last. I found the experience highly illuminating, not only in 
regard to Hopi but as bearing on the whole subject of grammatical cate- 
gories and concepts. It happens that Hopi categories are just enough 
like Indo-European ones to give at first a deceptive impression of iden- 
tity marred with distressing irregularities, and just enough different to 
afford, after they have been correctly determined, a new viewpoint 
toward the, on the whole, similar distinctions made in many modern 
and ancient Indo-European tongues. It was to me almost as enlighten- 
ing to see English from the entirely new angle necessitated in order to 
translate it into Hopi as it was to discover the meanings of the Hopi 

* Reprinted from Language, 14:275-286 (1938). 



forms themselves. This was notably true for the four types of verbal 
category herein discussed. 

It will be well to outline first the following general distinctions: 

1. Overt category: one marked by a morpheme which appears in 
every sentence containing the category, vs. covert category: not 
marked in sentences in general, but requiring a distinctive treatment 
in certain types of sentence, e.g. English genders. 

2. Word category: a category (overt or covert or mixed) which 
delimits one of a primary hierarchy of word classes, each of limited 
membership (not coterminous with entire vocabulary), e.g. the familiar 
"parts of speech" of Indo-European and many other languages, vs. 
MODULUS category: one which modifies either any word of the vocabu- 
lar}' or any word already allocated to a delimited class, e.g. voices, aspects, 

3. Specific category: an individual class of any of the abo\e types, 
e.g. passive voice, durative aspect, vs. generic category: a higher hier- 
archy formed by grouping classes of similar or complementary types, 
e.g. voice, aspect. 

The categories treated in this article are all of the overt and the modu- 
lus types, but it should be stated that covert categories and word cate- 
gories are also of great importance in Hopi grammar. Failure to define 
such classes would at once gi\e the grammar a \'ery irregular appearance. 
The generic categories here treated belong to the verb system, and have 
been designated assertion, mode, status, and modality. 


Hopi verbs have three assertions: reportive (zero form), expective 
(suffix -77/'), NOMic (suffix -ifi). Thcsc translate, more or less, the Eng- 
lish tenses.^ But they do not refer to time or duration. They distin- 
guish three different kinds of information. Assertion, in other words, is 
a classification that refers the statement to one of three distinct realms 
of validity. The reporti\e is simply a reporting statement, telling of the 

' In ""riie piiiictual and scgmcntativc aspects of verbs in Hopi," Language, 12 
(1936), I referred to the assertions as tenses and called them factual or present-past, 
future, and generalized or usitativc. 


historical actuality of a certain situation: e.g. Eng. 'he ran, he is running, 
I see it.' 2 

The expecti\'e declares an expectation or anticipation of a situation. 
It is translated by the English future, or by 'is going to,' or by 'begins 
to,' for an attitude may still be one of expectant anticipation, rather 
than one of reporting a fait accompli, while the action is already begin- 
ning. Since the assertions have nothing to do with time as such, an 
expectant attitude may be projected into an account of past events, 
giving the translations 'was going to,' 'began to.' A clue to meanings 
otherwise obscure resides in the rendering 'his doing it is (or was) ex- 
pected.' Thus, addition of the concursive mode suffix -kai] 'while' forms 
an expective concursive in -nikai], but to translate here by the English 
future 'while he will do it' would be quite incorrect. The form means 
'before he does it': i.e. 'while his doing it is expected.' ^ 

The nomic does not declare any particular situation, but offers the 
statement as a general truth, e.g. English: 'she writes poetry, he smokes 

2 There is no distinction in the reportive between past and present, for both are 
equally accomplished fact. What we call present tense (not counting our present 
form which corresponds to the nomic) is from the Hopi standpoint simply a report 
to others concerning a situation shared with them, this report being either redundant 
information, or used to call attention to, or tell about some fragment of the situation 
not fully shared. Thus to the Hopi 'he is running' need not be different from 'he 
was running,' for, if both the speaker and listener can see the runner, then the 'is' 
of the former sentence means merely that the listener can see for himself what he is 
being told; he is being given redundant information, and this is the only difference 
from the latter sentence. Now the Hopi listener senses no lack in Hopi grammar 
for its not telling him that the information is redundant, when he can see for him- 
self that it is. If the speaker can see the runner but the listener cannot, then the 
information is not redundant, but the situation in that case is one of rapidly relaying 
the information, which rules out the distinctive past meaning of 'was,' and again the 
Hopi find our tense distinction irrelevant. 

8 The orthography used for Hopi is phonemic and employs the American An- 
thropological Association symbols generally used for American Indian languages, with 
the following minor variations: k is somewhat fronted and before a and s sounds like 
kv; c is the affricate ts; ^ is glottal stop; v is bilabial and when syllable-final unvoiced, 
r is retroflcx, untrilled, and slightly si^irantal, and when syllable- final unvoiced, small 
capitals denote voiceless consonants which are separate phonemes, a dot under a 
vowel denotes the short, and lack of a mark the medium grade of the Hopi three- 
quantity vowel system, ' ("high accent") denotes raised pitch and fairly strong stress 
and on monosyllabic words is not marked but to be understood, ^ ("low accent") 
denotes a lower level of pitch and weak stress and is marked where it occurs on 
monosyllabic words. The mode sufhxes and most of the particles have special pausal 
forms when they stand at the end of sentences; these are not given except for the 
suffix -qo^, which is used in an example. 


only cigars, rain comes from the clouds, certain dinosaurs laid eggs in 
sand.' The three assertion categories are mutually exclusive. 


Mode, in Hopi, is the generic category of the system by which is de- 
noted the nature of the mingled discreteness and connection between a 
sentence (clause) and the sentence (clause) which follows or precedes it. 
The INDEPENDENT modc (zero form) implies that the sentence is de- 
tached from others, though it is possible to relate such sentences by 
paratactic connectives like our 'and.' But the Hopi show a great liking 
for hypotactic constructions. 1 hese employ six mutually exclusive de- 
pendent MODES, denoted by sufExes placed after the assertion sufhx. 
Each mode denotes a basic type of relatedness invohing both linkage 
and discreteness, or disparity. With the further addition of qualifying 
particles, these modes can distinguish a great number of possible rela- 
tionships, much as in Greek the basic system of three oblique cases can 
be further developed by a large number of prepositions. However, the 
mode relationships are not case relations, nor are the modes defective 
forms like IE infinitives and gerunds, but full verb paradigms. 

The names, suffixes, and t}pes of discreteness-connection are as fol- 
lows: CONDITIONAL {-£^ cclipsiug final vowel of base), condition needed to 
justify a nonreporti\e (expective or nomic) assertion in the other clause 
(Eng. 'when, if) (the mode clause is logically in this same assertion, 
though it does not bear any assertion suffix); correlative {-qaY), ex- 
planatory justification of the statement of the other clause (Eng. 'be- 
cause, since, as, for,' gerund construction); concursive {kaij, -kakaij), 
parallel contemporaneous occurrence (Eng. 'while, as, and'); sequential 
(-t), sequence in time (Eng. 'after, . . . and then . . .'); agentive {-qa), 
qualification of a person or thing in one clause as the subject in the 
other, the mode clause (Eng. 'who, which,' though not Eng. 'whom'); 
1 ransrelative {-q, -qo^), general relatedness bridging a difference of sub- 
ject in the two clauses (no Eng. equivalent). Each mode refers to a 
certain kind of disparity or contrast, as well as of connection, between 
the two clauses; and a separateness of subjects or performers in the two 
clauses is itself one basic type of disparity on a level with the other basic 
types, and necessitates the transrelative mode in one of the clauses. 


Hence all the other modes refer to conditions in which the subject of 
both clauses is the same: i.e. clause-contrast is based on other factors 
than disparity of subjects. If it is desired to add to the general notion 
of the transrelative mode an expression of factors like those which are 
basic in the other modes, this can be done by detached particles. The 
subject of the transrelati\c in many constructions and of the agentive 
in some is in the objective case. That of all other modes is in the nomi- 
native. There follow some illustrations of the uses of the modes. In 
these examples and henceforth throughout this paper, the mode sufEx 
is hyphenated to assist in identifying the dependent verb. 

Conditional: ni'm-e^ mi-'nat tiwa'ni 'when he goes home he will see 
the river' (expective); ni'm-e^ mi-'nat tiwq'}fi 'when he goes home he 
sees the ri\er' (nomic). 

Correlative: ni'ma-qa^Y mi-'nat t'i'wa 'because he went home he saw 
the river'; ni'ma mi-'nat t'iwa'ni-qa^Y 'he went home in order to see the 
river': i.e. 'he went home because his seeing the river was expected, 
looked forward to (by him)'; pa'?/qa^W-qaY ya\v wii-'nat ti'wa 'he said 
that he saw the river' (by his sa\ing, to quote, 'he saw the river'). 

Concursive: ni'ma-ka^ij nii-'nat t'i'wa 'as (or while) he went home he 
saw the ri\er'; wini'ma-ka^jj td-'w/ciVz 'he danced and sang' (at the same 

Sequential: ki-y ^aw pit'i'-t mi-'nat t'i'wa 'after (or when) he arrived at 
his home he saw the river'; w'ini'ma'-t pf^ ta - 'wla^w'i 'he danced and then 
he sang.' 

Agentive: ta-'qa ni'ma^-qa nii-'nat ti'wa 'the man who went home saw 
the river'; 7iP t'i'wa ta-'qat ni'ma-qa^-t 'I saw the man who went home' 
(the agentive takes noun cases and here is in objective case, as likewise 
its subject).'* 

Transrelative: ni'ma-q ta-'qa '\iw pit'i'ni 'when he goes home a man 
will meet ('arrive to') him'; ni'ma-q mo'ifi ^aw pit'i'if'i 'when he goes 
home the chief meets him' (nomic); ti'yo w'ini'ma-q ^o-'viy ma-'na 
ta-'wla\n 'since the boy danced, therefore the girl sang'; pa'P/qaW-q 
ya'w ma-'na ni'ma 'he said that the girl went home'; na-t ta-'wlaW-q 
ma-'na w'ini'ma 'while he sang the girl danced'; pa'nis w'ini'ma'-q pp 
ma-'na ta-'wla'w'i 'right after he (or, as soon as he had) danced the girl 

4 This objective-case agentive clause is the one exception to the rule that disparity 
of subject requires the transrelative. 


sang'; "^a'son m'nia-q mo'ij"'i '■'aw piti'ni 'after he goes home the chief will 
meet him.' 

Our relative clause with relative pronoun object is transrelative m 
Hopi, since there is disparity of subject; e.g. ta-'qat riP tiwa'-q ni'ma 
'the man whom I saw went home' ('man,' objective dependent on 'my 
seeing,' 'he went home'). Hopi treats 'man' as the object of the seeing 
clause, while the subject of the going-home clause is 'he' expressed m 
the verb. English makes very httle distinction between this construc- 
tion and the one expressed in Hopi by the agentive, often using the 
same connecti\e, 'that' or 'which,' for both, so that bewildering changes 
of construction may result, in translating a number of outwardl}' similar 
relative clauses into Hopi. These changes however are perfectly trans- 
parent to a Hopi; even a bilingual Hopi, when given the two proposi- 
tions dressed in the same garb of English form, 'the man that I saw 
went home' and 'the man that saw me went home,' will iristantly react 
with two completely dissimilar patterns: ta-'qat nP tiwa'-q ni'ma, and 
ta-'qa riiy tiwa"qa ni'ma. Hopi also produces transrelative forms that 
translate our relati\e pronoun object of a preposition; and here the re- 
moteness from our pattern is extreme: e.g. yama'kpit^qij wa-'yma-q W'pe 
'the bridge on which he was walking collapsed' ('bridge [objectixc] on-it 
being his walking, it collapsed'). Most bewildering of all to the usual 
Indo-European view is the manner in which Hopi combines modes with 
modalities according to a systematic logic of its own, of which more 

Lest the omission of "imperati\e" from the modes seem peculiar, I 
may say that it belongs to a category of injunctives (impcrati\c, semi- 
imperative, optati\e. \etati\e), defective paradigms that arc neither 
modes nor modalities. 


The status categories have been named affirmative (zero form, 
declarative sentence), negative, interrogative, indefinitive. The 
negative is formed in the reportive and nomic by the particle qa^ 'not' 
before the verb, in the expective by so'^on before the \crb. The inter- 
rogative is the status of asking a question for a yes-or-no answer. It is 
formed by the particle pT as the first word in the sentence, e.g. pT ma ■ 'na 
ti'yot t'i'wa 'did the girl see the boy?' There is no different intonation 


from any other type of sentence. The reply-forms are ^owi-^y 'yes,' 
qa'^e or qa?£-^y 'no,' pi-'hi-^y, an unanalyzable word meaning 'I don't 
know.' The indefinitive is the status of the sentence containing an 
"interrogative" (better, indefinitive) pronoun, adverb, or verb (e.g. the 
verb 'do what'). It need not be necessarily a true question, because such 
interrogative words are also indefinites (i.e. the words 'what?' and 'some- 
thing' are the same); or, better said, the meaning of the word is an in- 
definite suggestion that implies also a more or less inquisitive attitude 
'something— I wonder what?' The Hopi sentence hi -'mi pe'wP 'some- 
thing is coming' could be rendered psychologically as an 'I-wonder-what 
is coming.' Whether this is to be interpreted as a request for informa- 
tion or merely as an interesting remark will depend on the context, which 
may include the general behavior of the speaker. 


JModality hi Hopi is m rough terms the sort of thing that is referred to 
by the subjunctive and some other moods of IE languages. But of course 
we ought not to rest content with this "sort of" sort of definition. I 
would say that the modalities of Hopi are moduli of moduli, that is, they 
are methods of further modifying and amplifying the three-assertion 
system that distinguishes three basic realms of validity, so that in effect 
many more than three realms and subrealms of validity are distin- 
guished—again, much as prepositional relationships in Greek amplify a 
basic system of case relationships. Modalities are to assertions what 
the particles na-t, "^a'son, etc. in our mode examples are to modes. They 
are denoted by particles designated modalizers. The word-category of 
modalizers is more abundant in morphemes than the category of modal- 
ity strictly considered. While there are many particles of the modalizcr 
type, it seems necessar>' to distinguish eight (perhaps more) as mutually 
coordinated to the point of forming a system of eight modalities, to 
which on schematic grounds we may add as ninth the indicative modal- 
ity or zero form. The line between the modalities and the lexical use of 
other modalizers is not sharp; yet on the whole the modalities are a set 
of mutually cxclusi\c forms (with certain exceptions), while the other 
modalizers are used more freely along with each other or with the modal- 
ity modalizers, occur in less frequent and less formally patterned uses, 


and are perhaps most conveniently treated as lexical items (see examples 
at end of this paper). The modalities have been named: indicative, 


Quotative: modahzer yd'w. The assertion becomes the assertion of a 
linguistic report. In telHng a folk-tale )'dV is used like the English 'so' 
or 'and so' at the beginning of almost ever\' long sentence; it there means 
'so' in the sense of 'according to the story.' In the simple independent 
sentence it adds the idea 'they say,' 'on dit que.' It is placed imme- 
diately after a direct quotation, the last word in which then receives 
high stress on the last syllable, this combination being equivalent to 
spoken quotation marks. It is used in indirect discourse, the verb of 
saying, hearing, etc., being in correlative or transrelative mode (see 
examples under these modes). Contrast n'P navo't-q ya'w rrii'ni 'I heard 
that he fell down' (by \erbal report) with nP navo't-q m'i'ni 'I heard him 
fall down' (heard the sound of his fall) . However the quotative cannot 
imply the confirmation or concede the truth of the report, which is the 
function of the concessive modality (Hr); hence nP navo't-q k'ir m'i'ni 'I 
heard (of the fact) that he fell down' (which is conceded to have actually 

Inhibitive: modalizer kirhi'n. This means that the subject is blocked 
or prevented from producing the effect specified by the verb, with com- 
plete lack of implication about the cause of this condition, as to whether 
it lie in the ability of the subject or externally, etc. It is translated 
simply by 'cannot.' 

Potential: This formulation strikes the English speaker at first as 
topsy-turvy, if not positively weird. It is translated by 'can,' but is 
simply the negati\e of the inhibitive, donated by kirhi'n qa\ Yet 
analysis shows the form to be remarkably logical. By this means the 
Hopi produce a perfectly neutral, potential 'can' that does not merely 
refer to personal ability, but denotes that the way is entirely open for the 
subject to turn potentiality into action if he chooses. (For the 'can' of 
personal or technical ability, 'knowing how,' there is a definite \erb, used 
with the expective correlative of the action verb.) But how could such 
a neutral potential 'can' be better expressed than by a negati\e form 
which declares simply the absence of all inhibitive or frustrative checks 
between the subject and action? 


Indeterminate: modalizer sen. This indicates uncertainty, corre- 
sponding to 'perhaps, possibly, maybe,' or in the expective to 'may': 
e.g. ni'm-s^ sen nio'if'it ^aw pit'i'ni 'when he goes home he may meet the 
chief (and again he may not). The uncertainty is like that of a balancing 
between about equal positive and negative probabilities; hence, e.g. nP 
^aw ti'viijtd'-q sen ni'mg'ni 'I asked him if (whether) he were about to 
go home' (transrelative construction). Here at first sen seems to play 
the role of Eng. 'if,' but it is not so. It answers only to the indetermi- 
nacy expressed by the 'if,' whereas the linking and relating function of 
the 'if is performed by the transrelative mode relation; sen itself is quite 
unable to effect any linkage. 

Advisory: modalizer ke. It denotes an uncertainty like sen, but 
stresses slightly the possibility of the positive rather than the negative 
outcome. If this positive possibility is being asserted in the presence 
of a somewhat opposing attitude, or a dread of such an outcome, it 
gi\es the feeling of our 'might,' 'may nevertheless,' or in the reportive 
'might have done so.' Thus, ta-'qa fiw-e'"^ ke wa-'ya'ni 'when the man 
sees it he may run away' (his possible running away is the thing to be 
kept in mind). Hence the sentence containing ke has an advisory char- 
acter, since it does not merely adumbrate an uncertainty, but calls atten- 
tion to one possible outcome thereof. Logically enough, our 'may not' 
is ke qa\ not sen qd", e.g. ni'm-e^ ke qa" tiwa'ni 'when he goes home he 
may not see it/ 

Concessive: modalizer kir. It denotes that the assertion is given valid- 
ity as a concept rather than validity as an objective experience: ^ e.g. 'it is 

5 We may not here read "senson,'" for "objective," for experiences which psychol- 
ogy would place at the level of percepts rather than sensations, or even at the level 
of simple concepts, do not require the concessive, though they are treated differently 
from sensations. They are the percepts (or simple concepts) of seeing an action or 
phenomenon of a type having a lexical name (verb), and are put in the indicative 
transrelative, while it is now the verb of seeing that is in the independent mode, 
e.g., np tfwa wa-'ya'-qo^ 'I saw him run away,' nP tfwa ci'rot mf^a-qo^? '1 saw him 
shoot the bird.' It is a remarkable fact that the Hopi seem to recognize, in their lan- 
guage, a distinction of four types of received information which correspond roughly 
to gradings made by psychology: (1) sensory, e.g. 'I see that it is red,' 'I hear him fall,' 
sensing verb in transrelative, information verb in indicative; (2) perceptual conceptual, 
e.g. 'I see him fall,' sensing verb in independent, information verb in transrelative; 
(3) overt linguistic, e.g. 'I heard that he fell' (unconceded), sensing verb in transrela- 
tive, information verb in quotative; (4) pure conceptual, e.g. 'I see that it is new,' 
'I heard that he fell' (a conceded truth), sensing verb in transrelative, information 
verb in concessive. 


conceded, granted, niferred from the evidence at hand, assumed, con- 
sidered as known,' etc. In the mdependent sentence it may be rendered 
'it seems that, evidently, apparently,' or merely 'so': e.g. kir mg'ifi 
ni'tna 'so the chief went home' (I gather). It is m complex sentences 
that its subtle importance stands out. Consider the transrelative pat- 
tern; rip tiwa'-q pa'la 'I see that it is red' (by my seeing it is red), riP 
thva'-q fipi'pita' 'I see that it sparkles.' I'he Ilopi refuse to use this 
pattern as it stands for, e.g. 'I see that it is new,' which demands the 
concessive in the clause expressing newness, i.e. riP tiwa'-q kir pi- 'hi (by 
my seeing it is inferentially new). In other words newness is not a 
visual sensation like redness or light; it is not seen directly, but is in- 
ferred or assumed, kir, from seen data. To us this seems like psycho- 
logical analysis, but to the Hopi it is a clear and practical distinction. 
The ordinary English conditional 'if construction requires in Hopi thai 
the conditional or transrelative mode which represents the linking func- 
tion of our 'if be also concessi\'e to represent the hypothecating function 
of the 'if; otherwise the mode would be translated 'when' and not 'if.' 
Thus, kir iii'm-s^ rrii-'nat tiwa'ni 'if he goes home he will see the river' 
(assuming that he goes home, etc.). The contrary-to-fact 'if is another 
matter {^as, see below). 

Necessitative: modalizer so''^on qd'. It means 'necessarily, naturally, 
inevitably,' and seems queer from the IE standpoint in being simply a 
double negative, which in Hopi always makes a positive. It is a com- 
bination of qd' 'not' and sp'^on 'expective not' and thus means that there 
can be no expectancy of a negative. It often translates English 'must' 
and 'have to' but is not tinged with any notions of compulsion, duty, or 
obligation, being entirely neutral and abstract. It is often used in the 
conclusion of conditional statements to indicate a necessary conse- 
quence: e.g. kir ni'ni-£^ so'^on qd mi-'nat ftwa'ni 'if he goes home he 
will see the rivef (as a necessary^ consequence). 

Impotential: modalizer ^as. This modality is very difficult to express 
in terms of our ways of thinking. It indicates what I might call teleo- 
logical incffectixeness. Wz go part way along the road with the Hopi 
by recognizing forms of assertion, like 'may' and 'can,' that are on a 
different plane from the bare 'does' and 'does not,' and as compared with 
these more rudimentary assertions have a status that combines the quali- 
ties of affirmative and negative, of reality and unreality. Tliey are on a 
middle ground between these opposites, even though formally cast in an 


affirmative pattern. But all our middle-ground expressions refer to the 
realm of latency; the reason for the statement's dual positive-negative 
character is that its truth is that of a latency, of which the manifestation 
belongs to the future. The Hopi also have middle-ground expressions 
of this sort, as we have seen. But they have, further, a middle-ground 
expression in which the dual positive-negative character is not a matter 
of latency, but is ascribed to events that have already happened. More- 
over an expression referring to something that never happened at all 
can be assigned to this same realm of quasireality, along with references 
to actual happenings. The criterion that fastens this particular stigma, 
as it were, of quasireality upon the subject matter of discourse is ineffec- 
tiveness in terms of the purpose, goal, drive, need, function, etc. (a 
variety of concepts in our own ideology are here applicable) that origi- 
nally formed the grounds for the action. If a Hopi is reporting a train 
of events in which a man ran away from his pursuers but was eventually 
captured by them, he will use the impotential, and say ta-'qa ^as wa-'ya 
'the man ran away' (implying that 'ran away' cannot here be held to 
mean 'escaped'). If the man ran away and escaped, the statement would 
be simply ta-'qa wa-'ya. N'P ma'qto is 'I went hunting'; riP ^qs ma'qto 
is the same, except that it implies that I came back empty-handed or 
practically so. We could convey such information by saying 'well, I 
went hunting!' in a disgusted tone, but the Hopi sentence is not really 
comparable to this. It is a quite unemotional statement; ^qs is not an 
expression of affect but is intellectual, and would be used whether the 
speaker be unmoved, displeased, or pleased; e.g. co-'viw ^qs wa-'ya 'the 
deer fled' (but I caught him just the same). My name for this modality, 
"impotential," refers to the connotation of impotence that it gives to 
the statement of actions ajpd attempts. In the expective, it changes 
the translation from 'will' to 'tries to': e.g. ma-'na ^qs wini'mq''ni 'the 
girl tries to dance' (but does not for the present succeed). However, 
expective impotential does not imply that later attempts will not suc- 
ceed. When the expective refers to the past of narration, its impoten- 
tial refers to a frustrated attempt at some event that did not actually 
happen: e.g. ^qs wa-'ya'ni 'he tried to escape,' of a prisoner who failed 
to escape. When the impotential is expective in a dependent mode 
(i.e. when it is conditional or transrelative and the other clause is expec- 
tive), the reality is further attenuated to a ne\er-rcalized theoretical past 


possibility. Thus our contrary-to-fact condition is impotential in Hopi: 
e.g. ^gs ni'm-s'' so' ''on qd" mi-'nat tiwg'ni 'if he had gone home he would 
have seen the river' ('when in not-to-be-realized capacity of going home 
he was in necessitous expectancy of seeing the river'). Here the conces- 
sive may be added to the impotential, especially if the mode is trans- 
relative: e.g. kir ^gs ni'ma-q sg'^on qa" mo'ij'^i ^aw pit'fni 'if he had gone 
home the chief would have met him.' Our 'although, but, yet,' etc. 
indicate a tension of some sort between two conflicting tendencies. 
The Hopi unerringly discriminate whether one of these tendencies has 
actually aborted the other (the impotential meaning) or whether the 
opposition is of the sort to be indicated by ks, sen, or some other means. 
Thus, in concursive, nP ^gs qati'-kat] 7na-'}ji'^i 'although I was sitting I 
felt tired'; in transrelative, ^gs wa-'ya'-q nP y'Pa 'although he fled I 
caught him.' On the other hand, ^gs would be wrong in 'although he 
was running he was singing,' since neither action has aborted the pur- 
pose of the other; here a Hopi would say simply wa'riki'w-kaij ta-'wla'wi 
'while he was running he was singing,' or perhaps add to this an element 
denoting mild surprise. In the sequential form ^gs pit'i'-t qd" wini'ma 
'although he had arrived he did not dance,' ^gs implies that the function 
of coming was to dance; he might as well not have come.® 

The disparity of pattern from Indo-European appears in that, while 
all these modalities resemble the IE subjunctive, not one aligns with it. 
The Hopi interpret our subjunctive in various ways according to a recog- 
nition of relationships of which we are not linguistically conscious. 
Thus, in 'if I were king,' 'were' from the Hopi viewpoint is impotential; 
in 'to see if he were brave,' 'were' is indeterminate; in 'though he be 
stubborn/ 'be' is advisory; in 'if he be right,' 'be' is concessive. Or is 
the pattern so very different from Indo-European after all? It remains 
a fact that the Uto-Aztecan languages in general, and Hopi especially, 
are for American languages unusually reminiscent of IE in their type 
of grammar. Could it be possible that in ancient forms of IE, perhaps 
in Hittite, patterns of syntactic construction may exist that would lend 
themselves to an analysis following somewhat the Hopi outline? 

•^ The realm of the mightaswell-not-havebeen is, in a nutshell, that middle ground 
between positive and negative which the impotential represents. What we call the 
might-have-been is to the Hopi simply a part of this realm. It is the expectancy 
(potency, tendency, possibility, wish) that might as well not have been. 



There are a number of modalizers having uses less definitely formal 
than those of the modahty system. Their wide range of expression is 
shown by the following samples: 

^i'ra: memory, recollection — 'according to recollection it is so.' 

H'sninti'q: probability, reasonable expectation, justifiable assumption 
or hope— 'supposed!}'— and in the expective is like our 'if all goes well.' 

na'w'is: obligation to voluntary action without compulsion; differs from 
our 'ought to, should' by greater promise that the action will be per- 
formed; may sometimes be rendered 'has to,' but, according to my in- 
formant, corresponds to our expression 'can't \ery well refuse to.' 

pi- -v. 'almost, nearly,' ps-v ks 'possibly almost,' 'almost may/ 

pi": acceptance of conditions as they are or must be, undeniable fact, 
inevitability; sometimes corresponds to English word stress, e.g. pam 
pit'i'ni 'he will arrive,' pi" pgm pit'i'ni 'he will arrive,' 'he, at least, will 
arri\e.' It also corresponds to 'after all,' and still more to our shrug of 
the shoulders, while pay pi" 'already,' pi" indicates philosophic resigna- 
tion to unchangeable reality— pen' pi" wa-'ya 'he ran away, and that's 
that.' Also common are the combinations with negative {pi" qa") and 
necessitative {so'^on pi" qa"). 

ta'tam: necessity to which one resigns oneself with a self-sacrificing 
feeling; 'must,' or 'may as well' imbued with this feeling; implies that 
the subject is sacrificing his own interests or preferences. 

t'ir: intention without clear resolve, vaguer than 'wants to' or 'intends 
to'; more like 'is (was) thinking of ing,' 'would,' 'would like to.' 

The extent to which "modal feeling" is di\crsificd and the finesse of 
its application differ markedly in different languages, but probably few 
languages have gone so far into these fields as Hopi. 


Editor's note: In 1938, Whorf circulated this table and accomp>anying 
outline in manuscript form among selected colleagues. It was written 
as a supplement to the Outline of cultural materials prepared by George 
P. Murdock and his colleagues at the Department of Anthropology at 
Yale University as a guide to ethnological field workers, and is referred 
to in the brief "Language" section of that outline. 

In se\eral places in his writings Whorf mentions the desirability of a 
"world-survey" of languages; this outline was doubtless intended by him 
as a suggested standard framework for collecting the information on 
particular languages which would be needed for such a surxey. 

The reader's attention should be directed first to the table on page 126, 
which displays the whole scheme of language as conceived by Whorf. 
The subsequent outline, which presents an expansion of the semasiology 
section of the table, is thus only an appendage to the table, even though 
it contains most of the meat. The material is printed with only minor 
alterations and corrections of the original manuscript, which was fur- 
nished by Professor Norman McOuown of the Universit)' of Chicago. 

[k.xp.\nsion of Semasiology section of the table] 
A. The sentence 

1. Sentence-end marking: by 

intonation (one pattern, several patterns) 
pause forms 




Si:, --n 

.2 l^ 

to ^ 'Elh j; 


w J 


^ 1- 

^ E 

O ^ .t; O 3 

Zj , rt 3 

-^ ■•-■ 

u tj > cr 

S I E 


U i: 

















^ o 

- u'5 

"? CL, c;5 h5 < cc; w 

.2i M 

o t; t" 

•. rt g t" 


bD u 

C C 


t3 <u 

w 5 .2 
)i Jjt V. -^ oj rt i; rt rt ^ j;^ 

E c E t-^V.-^^.S.^ 

C X 

''O — 

2 5 o a 

§y„:2-5 ,, E-£:$>^ 

u b c/j^- 

.P g R « 

u "-^ u JJ .— 


bO 5 

c/i rt 

E 8 ^ = ^--s 

c vj [- n *: c!! -r. 

5 o 

^0 u 


g-S « bO <u 

■= o 

language: plan and conception of arrangement 127 

special marking elements 

word order 

affective marking. Intonations or other elements marking sentence end 
and also denoting, e.g., emphasis, interest, surprise, doubt, interroga- 
tion, force of conviction, affective diminution, or augmentation, etc. 

2. Sentence-initial marking: by 

word order, e.g., verb ahvays first, etc. 

sentence introducers 




3. Intrasentential linkage (syntax) (integrating principle within thesentence) 
word order. Expansion of word order, subject-predicate order, adjacent 

order (modifier before, modifier after, mixed), interrupted order 

(tmesis), e.g., English split infinitive, 
functional categories (e.g., nouns, verbs, predicators, etc.) variously 

marked. See categories under B, The word, 
marking of relationships, e.g., by cases, pre- or postpositions, action direc- 
tors (i.e., different markings of action-goal or cause-effect, transitives, 

applicatives) et al. 
subsumption (reference within verb or other keyword to the syntax) 

pronominal incorporation or reference, noun incorporation 

directive and instrumental elements, and body-part elements, etc. 

verb a nuclear sentence (e.g. Navaho) 

holophrastic sentence (form of polysynthesis in which verb takes in 
most of the sentence) 
sentence harmony, i.e., agreement of formal classes, e.g., agreement in 

gender, number, etc. (Bantu is an extreme instance) 

4. Intersentential linkage (external syntax) (addition of sentence to sentence) 
Paratactic (coordinating) 

juxtaposition of sentences 
coordinating elements 

prosodic (intonations, etc.) 

sublexical (suffixes, etc.) 

particles (enclitics or words, e.g., 'and') 
Hypotactic (subordinating) (use of dependent clauses) by 

prosodic means (intonation, etc., e.g., the English comma intonation) 
sublexical marking, including special verb forms, gerundials 
subordinating particles, conjunctions 

5. Predication 

word order or sentence pattern (the isolating type) 

sublexical predicators (i.e., suffixes, etc., which make a "verb" out of 


anything to which they are added. Here the suffix, not the whole 
"verb" is the prcdicator) 
predicating word-classes (verbs and various types of quasiverbs; the 
predicative force is fused with a lexical meaning, e.g., 'eat, kill, 
stand') (See categories under B, The word) 
verbal sentence only 

verbal and nominal sentences — nominal sentences designated by 
order pattern, sublexically, etc. 
operators (words specialized for predication, otherwise lexical meaning 
blank ['be, become, cause, do'] or vague ['make, turn, get,' etc.]). 
auxiliary verbs, 
mixtures of techniques 
Categories of predication 
copulative (be) 

inherent (Spanish ser), or subjective 
objective (Spanish estar) 
general verbal, including all the following: 


intransitixe — mediopassive 
others (the above may be stated differently according to the pattern 
of language, e.g., instrumental verb, etc.) 
6. Minimal and abbreviated sentences 
abbreviated sentences 

retrospective forms, e.g., 'I did.' 'Will you?' 
elliptic forms, e.g., 'But tomorrow — !' 
social formulas ('thank you, hello, please,' etc.) 
minimal sentences 

vocative type ('John! Mother!') 
imperative type ('Come!') 
'yes' and 'no' 

ordinary ('oh, ouch, alas,' etc.) 


"swear words" 

other h'pes with special cultural force 

B. The word {the word as part of the sentence) 

1. Morpholog>' (overt structure) 

Techniques (of overt marking) (if possible, state degree, e.g., slight, mod- 
erate, abundant, profuse) 




vocalic ablaut 

change of 1 \owel of stem, of 2 or more 
ablaut elision, length-change, quavering 

consonantal ablaut 

accent and (or) tone changes 

reduplication (total, partial, initial-syllabic, final-syllabic, vocalic, re- 
duplication with interpolation) 

extrusion (commonly called reduplication) 
vocalic, initial (tak — atak), final [lein — leme) 
consonantal, initial [lem — leml), final {lem — mlem) 
mixed (e.g., lem — lemel) 

vowel-harmony accompanying other techniques 
Categories, morphological. See 3. Categories 

2. Covert structure and relationship 

selection * (e.g., difference between 'John, come, dog, kill') 

suppletion (e.g., 'go, went') 

order (in phrase or sentence) 

pronominal reference (e.g., used in English to mark gender-class) 

reference by key-word not a pronoun 

reactance (word with the covert relation go\erns choice of certain other 

words, e.g., round- vs. long-object terms in Navaho, governs choice 

of verb stem) 
Categories, covert. See 3. Categories 

3. Categories 

May be either o\ert or co\crt. If possible, say which and how marked, 
or mixed, e.g., verbs, suffixing; nouns, pure selection; nouns, absence 
of other marking.! 

* Selection, i.e., pure selection. Selection also accompanies all other marking of 
word-categories. Pure selection requires existence of covert word-categories (q.v.). 

t Where possible, state avoidance of commonly found categories, e.g.. no plural, 
no gender. 


I. Word categories 
a. Functional-lexical 





none (it must be remembered that any or all of these types may 
not exist; e.g., it may be impossible to have a "verb" without a 
suffix of verbation which can be appHed to any stem. There is 
then no class of verbs, but only the "modulus category" (q.v.) 
of verbation.) 
h. Functional 

pronouns (personal, state kinds of — demonstrative, directive — 
state ideas connoted — interrogative, negative, indefinite, rela- 
tive, etc.) 

particles (words used in marking or linking the sentence, q.v., 
end-markers, initial-markers (inceptors), pro- and postpositions, 
conjunctions, modal particles, predicators, operators, etc.) 

articles. See ll,c. definite-indefinite 

c. Reference categories (which imply a classification of experience, 

different kinds of things, state, or action) 

genders (many different kinds, masculine, feminine, animate, 
inanimate, personal, rational, irrational, integral, etc.) 

multireference categories (many gender-like classes, often with in- 
distinguishable meaning, e.g., Bantu) 

social status and rank classes 

shape classes (e.g., Navaho, Haida) 



locus and extension classes 

d. Purely formal classes, e.g., conjugations and declensions 

e. Personal name classes, gender, age, respect, etc. 

II. Modulus categories. (These do not delimit word-classes in them- 
selves; they MODIFY, cither any class, or classes already delimited 
by other means) 

a. Generally applicable 

verbation (predication other than 'be') 
"nomination" or noun-designation 

absolutive suffixes, nominative articles, zero-marking, etc. 

b. Mixed application — sometimes generally applied, sometimes 

specially applied 
number (kinds 1, 2, 3, several, many, plural) 

language: plan and conception of arrangement 131 

collectivity and distribution ^ 


tension (extension-duration) 
time or tense 

comparison, e.g., of adjectives 

see also reference categories; the same ideas may be applied as 
, Special application 
Applied usually to verbs (or along with verbation) 
categories of predication, q.v. A. The sentence 
voice: active, passive, etc. 

resolution: transitive, intransitive, passive (voice and resolu- 
tion merge) 
aspect (duration, extension, etc., e.g., punctual, durative, per- 
fective, imperfective, inceptive, continuative, progressive, fre- 
quentative, iterative, usitative, etc.) 
tense systems 

mode (mood), e.g., indicative, subjunctive, inferential, dubita- 
tive, optative, potential, permissive, concessive, adversative, 
et al. 
status, e.g., interrogative, negative, quotative, emphatic 
exclamative and other affective forms 
address-forms, e.g., imperative, vetative (negative imperative), 

hortative, etc. 
gerundials or subordinators (cross-refer to A4) 
Applied usually to nouns (or along with nomination) 

state (i.e., possessed, unpossessed, pronominally possessed 

case (various cases) 

adjectivization (also given under general) 
definite-indefinite (articles, etc.) 
partitives (some or any) 
generality, e.g., 'man, woman, canis,' as opposed to 'a man, a 

woman, a dog' 
continual, e.g., 'wood, metal,' as opposed to 'stick, piece of 

individuative, e.g., 'stick' 

others — See Reference categories, into which these merge 
d. Affective modulus categories (express speaker's feelings rather than 
an idea) 
affective diminution (diminutives) 
affective augmentation 
respect forms 


emphasis, exclamative, and other forms — what classes of words 

is the affective modulus applied to? 

III. Cryptotypes. Co\ert word categories with subtle meaning marked 

only by reactances. Skip this in the survey except for obvious 

cases, as determination of cryptotypes usually requires deep study 

of a language. 

C. The lexeme {the word or stem as an item of the vocabulary, and as a 
part analyzed or abstracted from sentence words) 

1 . Lexical hierarchies 

a. The language possesses distinctions between roots and derivative for- 

mations (stems, bases, themes). In this case the minimal irreducible 
base is called root. 

b. The language may have only one sort, or one main sort, of lexical 

element. In this case such an element is usually called stem — 
e.g., stems in Algonkin, Yana. Characterize language in this way 
if possible. 

c. Lexeme may be identical with word (word in sentence). 
Lexeme always different from word. Lexeme occurs in sentence — 

1. with morphological elements 

2. in polysynthetic composition 

2. Root and stem types 

polymorphous (no particular form for root or stem — however, this is 
rare; apt to be appearance based on insufficient analysis, e.g., as re- 
gards roots, English is not polymorphous) 

monomorphous. One characteristic root type, or 1 or 2 related types, 
e.g., CV, CVC, etc. 

restricted. Type with considerable freedom of form with certain restric- 
tions, e.g., limitation on the kind and position of consonant clusters 
within root or within stem, e.g., English. Indicate the restrictions on 
clusters, etc. if possible. 

3. Derivation (formation of secondary lexemes, i.e., word bases, from roots) 
Techniques. Overt. These are similar to morphological techniques; e.g., 

prefixing, suffixing, etc. 
Covert techniques — transfer to a different covert class and change of 

meaning with covert marking, e.g., '(to) stand, (a) stand (position), 

(a) stand (pedestal)' 
Degree of derivation — none, slight, moderate, great, cumulative (piling of 

derivative upon derivative, e.g., the mock-learned 'honorificabilitudi- 

nity'; this is found in Aztec, less so in Sanskrit and Greek, possibly in 

Mag\ar and Turkish) 
Derivational types t 

I These may merge into or become identical with morphological categories, and 
in some languages this section is to be transferred from the lexeme to the word: 

language: plan and conception of arrangement 133 

Noun types from verb-like bases 

action and state nouns, abstract nouns 
agentives — nouns of the doer 
instrunicntives — nouns of instrument 
place nouns 

. I noun of one affected 
nomen patientis { ^^^^^ ^^ ^^^^^ produced 

adjectival nouns 

others — there are many possibilities 
Verb types from nounlike bases 

verbs of activation, of possession, etc. 

Compounding (essentially binary complexes; the 2 main parts may also be 
modifier types. Does modifier come before or after? 

types: noun-noun, verb-noun, noun-verb, etc. 
coordinate types, e.g., 'space-time' in 'space-time relationships' 
Polysynthetic composition 

composition of many stems with rules of order, e.g., Algonkin 

a further possibility is: no distinction between stems (lexemes) and 
markers of modulus categories. In this case cross-refer to modulus 
Mixed types, e.g., "interrupted synthesis" 

in Athabascan — may be skipped as often difficult to analyze 
Nonisolatable lexemes — few, many, or all lexemes thus 
Semantic root-structure 
root analyzable into more or less vague parts and meanings, e.g., 'tread, 

track, trip.' Root-nucleus (e.g., tr) and root-determinative 
phonemic symbolism (correspondence between sound and sense) 

recurrence of the same phoneme or phoneme-group with a type of 

overt manipulation of phonemes for semantic and affective results (e.g., 

childish forms in some NW coast languages) 
roots susceptible of considerable intra-radical analysis. 


Human beings do not live in the objective world alone, nor alone in 
the world of social activity as ordinarily understood, but are very much 
at the mercy of the particular language which has become the medium 
of expression for their society; It is quite" an illusion \o imagine that one 
adjusts to reality essentially without the use of language and that lan- 
guage is merely an incideutal means of sohing specific problems of com- 
munication or reflection. (The fact of the matter is that the "real world" / 
f^s to a-Jarge extent unconsciously built up on the language habits of the 
group. ). . y.\ Ve see and hear and otherwise experience very largely as 
. ,we do 'because the langua ge Imfa its o f our communit}^ predispo se certain^ 
choices of interpretation.*^ 

-Edward Sapir 

'T^here will probably be general assent to the proposition that an ac- 
-L cepted pattern of using words is often prior to certain lines of think- 
ing anH^fdrms orT)ehaviorf*l3ut he who assents often sees in such a 
statement nothing more than a platitudinous recognition of the hypnotic 
power of philosophical and learned terminology on the one hand or of 
catchwords, slogans, and rallying cries on the other. To see only thus 
far is to miss the point of one of the important interconnections which 
Sapir saw between language, culturc2_anjl_4i5icholog;y, and succinctly 
expressed in the introcltiHory~quotation. It is not so much in these 

* Reprinted from pp. 75-93, Language, culture, and personality, essays in memory 
of Edward Sapir, edited b.y Leslie Spier (Menasha, Wis.: Sapir Memorial I'ublication 
Fund, 1941). The article was written in the summer of 1939. 



special uses of language as iiv^its constant ways of arranging data and 
its most ordinary everyday analVsis of phenomena that we need to recog- 
nize the influence it has on other activities, cultural and personal, j 



I came in touch with an aspect of this problem before 1 had studied 
under Dr. Sapir, and in a field usi^ally considered remote from linguistics. 
It was in the course of my professional work for a fire insurance com- 
pany, in which I undertook the task of analyzing many hundreds of 
reports of circumstances surrounding the start of fires, and in some cases, 
of explosions. My analysis was directed toward purely physical condi- 
tions, such as defective wiring, presence or lack of air spaces between 
metal flues and woodwork, etc., and the results were presented in these 
terms. Indeed it was undertaken with no thought that any other sig- 
nificances would or could be revealed. But in due course it became 
evident that not only a physical situation t/ucz physic s, but the meaning 
of that situation to people, was sometimes a factor, through the be- 
havior of the people, in the start of the fire. And this factor of meaning 
was clearest when it was a ling uistic meaning, residing in the name or 
the linguistic description commonly applied to the situation. Thus, 
around a storage.of what are called "gasoline drums," behavior will tend 
to a certain type, that is, great care will be exercised; while around a 
storage of what are called "empty gasoline drums," it will tend to be 
different— careless, with little repression of smoking or of tossing ciga- 
rette stubs about. Yet the "empty" drums are perhaps the more dan- 
gerous, since they contain explosive vapor. Physically the situation is 
hazardous,/but the linguistic analysis according to regular analogy must 
employ the word 'e mpty,' which inevitably _i uggests lack of hazard. The 
word 'empty' is used in two linguistic patterns: (1) as a virtual synonym 
for 'null and void, negative, inert,' (2) applied in analysis of physical sit- 
uations without regard to, e.g., vapor, liquid vestiges, or stray rubbish, in 
the container. The situation is named in one pattern (2) and the name 
is then "acted out" or "lived up to" in another (1), this being a generaK 
formula for the linguistic conditioning of behavior into hazardous forms. 1 

In a wood distillation plant the metal stills were insulated with a com- 
position prepared from limestone and called at the plant "spun lime- 


stone." No attempt was made to protect this covering from excessive 
heat or the contact of fllame. After a period of use, the fire below one 
of the stills spread to the "limestone," which to everyone's great surprise 
burned vigorously. Exposure to acetic acid fumes from the stills had 
converted part of the limestone (calcium carbonate) to calcium acetate. 
This when heated in a fire decomposes, forming inflammable acetone. 
Behavior that tolerated fire close to the covering was induced by use of 
the name "limestone," which because it ends in "-stone" implies non- 

A huge iron kettle of boiling varnish was obser\ed to be overheated, 
nearing the temperature at which it would ignite. The operator moved 
it off the fire and ran it on its wheels to a distance, but did not cover it. 
In a minute or so the varnish ignited. Here the linguistic influence is 
more complex; it is due to the metaphorical objectifying (of which more 
later) of "cause" as contact or the spatial juxtaposition of "things"— to 
analyzing the situation as 'on' versus 'off' the fire. In reality, the stage 
when the external fire was the main factor had passed; the overheating 
was now an internal process of con\ection in the varnish from the in- 
tensely heated kettle, and still continued when 'off' the fire. 

An electric glow heater on the wall was little used, and for one work- 
man had the meaning of a convenient coathangcr. At night a watch- 
man entered and snapped a switch, which action he verbalized as 'turn- 
ing on the light.' No light appeared, and this result he verbalized as 
'light is burned out.' He could not see the glow of the heater because 
of the old coat hung on it. Soon the heater ignited the coat, which set 
fire to the building. 

A tanner}' discharged waste water containing animal matter into an 
outdoor settling basin partly roofed with wood and partly open. This 
situation is one that ordinarily would be verbalized as 'pool of water.' 
A workman' had occasion to light a blowtorch near by, and threw his 
match into the water. But the decomposing waste matter was evolving 
gas under the wood cover, so that the setup was the reverse of 'waten,'.' 
An instant flare of flame ignited the woodwork, and the fire quickly 
.spread into the adjoining building. 

A drying room for hides was arranged with a blower at one end to 
make a current of air along the room and thence outdoors through a 
vent at the other end. Fire started at a hot bearing on the blower, 
which blew the flames directly into the hides and fanned them along 


the room, destroying the entire stock. This hazardous setup followed 
naturally from the term 'blower' with its linguistic equivalence to 'that 
which blows/ implying that its function necessarily is to 'blow.' Also its 
function is verbalized as 'blowing air for drying.' overlooking that it can 
blow other things, e.g., flames and sparks. In reality, a blower simply 
njalees-a-etuTCiit of ai4-aiui.canexhaust as we ll as blow; It should have 
been installed at the vent end to draw the air over the hides, then 
through the Hazard (its own casing and bearings), and thence outdoors. 

Beside a coal-fired melting pot for lead reclaiming was dumped a pile 
of "scrap lead"— a misleading verbalization, for it consisted of the lead 
sheets of old radio condensers, which still had paraffin paper between 
them. 'Soon the paraffin blazed up and fired the roof, half of which was 
burned off. ^^- ,^. i ,^ , , , r^- ^ d 'y n^-^.^ c' -7 ^ ^ < -^-" " ^ 

Such examples, which could be greatly multiplied, will suffice to show 
I how the cue to a certain line of beha\ior is often given by the analogies 
of the linguistic formula in which the situation is spoken of, and by 
which to some degree it is analyzed, classified, and allotted its place in 
that world which is "to a large extent unconsciously built up on the lan- 
guage habits of the group." And we always assume that the linguistic v 
analysis made by our group reflects reality better than it does. 


The linguistic material in the above examples is limited to single 
words, phrases, and patterns of limited range. One cannot study the 
behavioral compulsiveness of such material without suspecting a much 
more far-reaching compulsion from large-scale patterning of grammati- 
cal categories, such as plurality, gender and similar classifications (ani- 
mate, inanimate, etc.), tenses, voices, and other verb forms, classifica- 
tions of the type of "parts of speech," and the matter of whether a given 
experience is denoted by a unit morpheme, an inflected word, or a syn- 
tactical combination. A categor}' such as number (singular vs. plural) is 
an attempted interpretation of a whole large order of experience, vir- 
tually of the world or of nature; it attempts to say how experience is to 
be segmented, what experience is to be called "one" and what "several." 
But the difficulty of appraising such a far-reaching influence is great be- 


cause of its background character, because of the difficulty of standing 
aside from our own language, which is a habit and a cultural non est 
disputandum, and scrutinizing it objectively. And if we take a very dis- 
similar language, this language becomes a part of nature, and we even 
do to it what we have already done to nature. We tend to think in our 
own language in order to examine the exotic language. Or we find the 
task of unraveling the purely morphological intricacies so gigantic that 
it seems to absorb all else. Yet the problem, though difficult, is feasible; 
and the best approach is through an exotic language, for in its study we 
are at long last pushed willy-nilly out of our ruts. Then we find that the 
exotic language is a mirror held up to our own. 

In my study of the Hopi language, what L-now see as an opportunity 
to \\ork on this problem was first thrust upon me before I was clearly 
aware of the problem. The seemingly endless task of describing the 
morphology did finally end. Yet it was evident, especially in the light 
of Sapir's lectures on Navaho, that the description of the language was 
far from complete. I knew for example the morphological formation of 
plurals, but not how to use plurals. It was evident that the category 
of plural in Hopi was not the same thing as in English, French, or 
German. Certain things that were plural in these languages were singu- 
lar in Hopi. The phase of investigation which now began consumed 
nearly two more years. . -_' ■ 

The work began to assume the character of a comparison between 
Hopi and western European languages. It also became evident that 
even the grammar of Hopi bore a relation to Hopi culture, and the 
grammar of European tongues to our own "Western" or "European" 
culture. And it appeared that the interrelation brought in those large 
subsummations of experience by language, such as our own terms 'time,' 
'space,' 'substance,' and 'matter.' Since, with respect to the traits com- 
pared, there is little difference between English, French, German, or 
other European languages with the possible (but doubtful) exception 
of Balto-Slavic and non-Indo-European, I have lumped these languages 
into one group called SAE, or .".Standard Average European." 

That portion of the whole investigation here to be reported may be 
summed up in two questions: (1) Are our own concepts of 'time,' 'space,' 
and 'matter' gi\en in substantially the same form by experience to all 
/^ ^ men, or are they in part conditioned by the structure of particular lan- 
guages? (2) Are there traceable affinities between (a) cultural and be- 
havioral norms and {b) large-scale linguistic patterns? (I should be the 


last to pretend that there is anything so definite as "a correlation" be- 
tween culture and language, and especially between ethnological rubrics 
such as 'agricultural, hunting,' etc., and linguistic ones like 'inflected,' 
'synthetic,' or 'isolating. -J^' When I began the study, the problem was 
by no means so clearly formulated, and I had little notion that the 
answers would turn out as they did. 


In our language, that is SAE, plurality and cardinal numbers are 
applied in two ways: to real plurals and imaginary plurals. -/Or more 
exactly if less tersely: perceptible spatial aggregates and metaphorical 
aggregates. We say 'ten men' and also 'ten days.' Ten men either are 
or could be objectively perceived as ten, ten in one group perception ^— 
ten men on a street corner, for instance. . But-~lt£a_d ays' can not be 
objecti vely exp erienced. We experience only one day, today; the other 
nine (or e\'en all ten) are something conjured up from memory or.^ 
imagination. If 'ten days' be regarded as a group it must be as an// 
"imaginary," mentally constructed group. Whence comes this mental 
pattern? Just as in the case of the fire-causing errors, from the fact that 
our l anguage confuses the two different situations, has but one pattern 
f or, both. When we speak of 'ten steps forward, ten strokes on a bell,' ' 
or any similarly described cyclic sequence, "times" of any sort, we are 
doing the same thing as with 'days.' cyclicity brings the response of$C^^^^ 
imaginar}' plurals. But a likeness of cyclicity to aggregates is not un- 
mistakably given by experience prior to language, or it would be found 
in all languages, and it is not. 

Our AWARENESS of time and cyclicity does contain something imme- 
diate and subjective— the basic sense of "becoming later and later." But, 
in the habitual thought of us SAE people, this is covered under some- , 

thing quite different, which though mental should not be called §ub;_ //_ 
jective. I call it object ified, or imaginan,^, because it is patterned on 

1 We have plenty of evidence that this is not the case. Consider only the Hopi 
and the Ute, with languages that on the overt morphological and lexical level are as 
similar as, say, English and German. The idea of "correlation" between languae e 
a nd culture, in the generally accepted sense of correlation, is certainly a mistaken one . 

2 As we say, 'ten at the same time, showing that in our language and thought we 
restate the fact of group perception in terms of a concept 'time,' the large linguistic • 
component of which will appear in the course of this paper. 


the OUTER world. It is this tha t reflects our lingu istiC-Jisagef, Our tongue 
makes no distinction between numbers counted on discrete entities and 
numbers that are simply "counting itself." Habitual thought then as- 
sumes that in the latter the numbers are just as much counted on "some- 
thing" as in the former. This is objectification. Concepts of time lose 
contact with the subjective experience of "becoming later" and are ob- 
jectified as counted quantities, especially as lengths, made up of units 
as a length can be visibly marked off into inches. A 'length of time' is 
envisioned as a row of similar units, like a row of bottles. 

InHojgi there is a different linguistic situation. Plurals and cardinals 
are used only for entities that form or can form an objective group. 
There are n o imaginary plurals, but instead ordinals used with singu-^ 
lars. Such an expression as 'ten days' is not used. The equivalent state- 
ment is an operational one that reaches one day by a suitable count. 
'They stayed ten days' becomes 'they stayed until the eleventh day' or 
'they left after the tenth day.' 'Ten days is greater than nine days' 
becomes 'the tenth day is later than the ninth.' Our "length of time" 
is not regarded as a length but as a relation between two events in late- 
ness. Instead of our linguistically promoted o^jecTiEcatiori of~-that 
datum of consciousness we call 'time,' the Hopi language has not laid 
down any pattern that would cloak the subjective "becoming later" that 
is the essence of time. y>^. ^Jr^^^ ^ pie.ause <U. ^l leS* 


, We have two kinds of nouns denoting physical things: individuaK-^ 
nouns, and mass nouns, e.g., 'water, milk, wood, granite, sand, flour,jP^''^'' 
meat' Individual nouns denote bodies with definite outlines: 'a tree, 
a stick, a man, a hill.' Mass nouns denote homogeneous continua with- 
out implied boundaries. The distinction is marked by linguistic form; 
e.g., mass nounslackplurals,^, in English drop articles, and in French 
take the partiti\c article du, de la, des. 71ie distinction is more wide- 

3 It is no exception to this rule of lacking a plural that a mass noun may sometimes 
coincide in lexeme with an individual noun that of course has a plural; e.g., 'stone' 
(no pi.) with 'a stone' (pi. 'stones'). The plural form denoting varieties, e.g., 'wines' 
is of course a different sort of thing from the true plural; it is a curious outgrowth 
from the SAE mass nouns, leading to still another sort of imaginary aggregates, which 
will have to be omitted from this paper. 


spread in language than in the observable appearance of things. Rather 
few natural occurrences present themselves as unbounded extents; 'air' 
of course, and often 'water, rain, snow, sand, rock, dirt, grass.' We do 
not encounter 'butter, meat, cloth, iron, glass,' or most "materials" in 
such kind of manifestation, but in bodies small or large with definite 
outlines. The distinction is somewhat forced upon our description of 
events by an unavoidable pattern in language. It is so inconvenient in 
a great many cases that we need some way of indi\'idualizing the mass 
noun by further linguistic -devices. This is partly done by names of 
body-types: 'stick of wood, piece of cloth, pane of glass, cake of soap'; 
also, and e\'en more, by introducing names of conta i ners though their 
contents be the real issue: 'glass of water, cup of coffee, dish of food, bag 
of flour, bottle of beer.' These \er\' common container formulas, in 
which 'of has an obvious, visually perceptible meaning ("contents"), 
influence our feeling about the less obvious type-body formulas: 'stick 
of wood, lump of dough,' etc. The formulas are very similar: individual 
noun plus a similar relator (English 'of'). In the obvious case this » j 
relator denotes contents. In the inobvious one it "suggests" contents. ^J^\^ 
Hence the 'lumps, chunks, blocks, pieces,' etc., seem to contain some- 
thing, a "stuff," "substance," or "matter" that answers to the 'water,' 
'coffee,' or 'flour' in the container formulas. So with SAE people the 
philosophic "substance" and "matter" are also the naive idea; they are 
instantly acceptable, "common sense." It is so through linguistic habit. 
Our language patter ns ofte n require us to nam.e a physical thing j^y-a 
binomial that splits the ref erence into a formless jtem plus a f orm. 

Hopi is again different. It has a formally distinguished class of nouns. 
But this class contains no formal subclass of mass nouns. All nouns 
have an individual sense and both singular and plural forms. Nouns 
translating most nearly our mass nouns still refer to vague bodies or 
\'aguely bounded extents. They imply indefiniteness, but not lack, of 
outline and size. In specific statements, 'water' means one certain mass 
or quantity of water, not what we call "the substance water." Gen- 
era lity of statement is conveyed through the verb or predicator, notjhe 
n^iun. Since no uns are individual already, they are not individualized 
by either type-bodies or names of containers, if there is no special need 
to emphasize shape or container. The noun itself implies a suitable 
type-body or container. One says, not 'a glass of water' but kj-yi 'a 

(A.\ CO^ ^^^ r^diy^ Mai. 'vv^^^'Uf (TT- (d^^r^C/W <?^ Sfjv 


water/ not 'a pool of water' but pa- ha,* not 'a dish of cornflour' but 
ijamni 'a (quantity of) cornflour/ not 'a piece of meat' but sik"H 'a meat.' 
The language has neither need for nor analogies on which to build the 
concept of existence as a duality of formless item and form. It deals 
with formlessness through other symbols than nouns. 


Such terms as 'summer, winter, September, morning, noon, sunset' 
are with us nouns, and have little formal linguistic difference from other 
nouns. They can be subjects or objects, and we say 'at sunset' or 'in 
winter' just as we say 'at a corner' or 'in an orchard.' ^ They are plural- 
ized and numerated like nouns of physical objects, as we have seen. Our 
thought about the referents of su'ch words hence becomes objectified. 
Without objectification, it would be a subjective experience of real time, 
i.e. of the consciousness of "becoming later and later"— simply a cyclic 
phase similar to an earlier phase in that ever-later-becoming duration. 
Only by imagination can such a cyclic phase be set beside another and 
another in the manner of a spatial (i.e. visually perceived) configuration. 
But su ch is the power of linguistic analogy that we do so objectify cyclic 
Rasing. We do it even by saving 'a phase' and 'phases' instead of, 
e ^ 'phasing .' And the pattern of individual and mass nouns, with the 
resulting binomial formula of formless item plus form, is so general that 
it is implicit for all nouns, and hence our very generalized formless items 
like 'substance, matter,' by which we can fill out the binomial for an 
enormously wide range of nouns. But even these are not quite gen- 
eralized enough to take in our phase nouns. So for the phase nouns we 
have made a formless item, 'time.' We have made it by using 'a time,' 
i.e. an occasion or a phase, in the pattern of a mass noun, just as from 
'a summer' we make 'summer' in the pattern of a mass noun. Thus with 
our binomial formula we can say and think 'a moment of time, a second 

* Hopi has two words for water quantities; ko-yi and pa-hj. I'he difference is 
something hke that between 'stone' and 'rock' in English, pa- ha implying greater size 
and "wildness"; flowing water, whether or not outdoors or in nature, is pa- ha; so 
is 'moisture.' But, unlike 'stone' and 'rock,' the difference is essentia l, not pertaining 
to a co nnotative ma rgin^ and the two can hardly ever be interchanged? " ^"~' 

5 To be sure, there are a few minor differences from other nouns, in English for 
instance in the use of the articles. . 


of time, a year of time.' Let me again point out that the pattern is 
simply that of 'a. bottle of milk' or 'a piece of cheese.' Thus_we are 

assisted to imagine that 'a summer' actually contains or consists of such - 
and-such a quan tity of 'tirn e.' 

^In JHlopi however all phase terms, like 'summer, morning,' etc., are not 
nouns but a kind of adverb, to use the nearest SAE analogy. They are 
a formal part of speech by themselves, distinct from nouns, verbs, and 
even other Hopi "adverbs." Such a word is not a case form or a locative 
pattern, like 'des Abends' or 'in the morning.' It contains no morpheme 
like one of 'in the house' or 'at the tree.' '^ It means 'when it is morn- 
ing' or 'while morning-phase is occurring.' These "temporals" are not 
used as subjects or objects, or at all like nouns. One does not say 'it's 
a hot summer' or 'summer is hot'; summer is not hot, summer is only 
WHEN conditions are hot, when heat occurs. One does not say 'this 
summer,' but 'summer now' or 'summer recently.' There is no objecti- 

fication, as a region, an extent, a quantity, of the subjective duratio n- 
feeling! NothTngTs suggested about time except the perpetual "getting 
later" of it. And so there is no basis here for a formless item answering 
to our 'time.' 


The three-tense system of SAE verbs colors all our thinking about 
time. This system is amalgamated with that larger scheme of objecti- 
fication of the subjective experience of duration already noted in other 
patterns— in the binomial formula applicable to nouns in general, in 
temporal nouns, in plurality and numeration. This objectifica tion 
enables us i n imagination to "stand t ime units in a row." Imagination 
of time as like a row harmonizes with a system of three tenses; whereas 
a system of two, an earlier and a later, would seem to correspond better 
to the feeling of duration as it is experienced. For if we inspect con- 
sciousness we find no past, present, future, but a unity embracing com- 
plexity. Everything is in consciousness, and everything in conscious- 

^ 'Year' and certain combinations of 'year' with name of season, rarely season names 
alone, can occur with a locative morpheme 'at,' but this is exceptional. It appears 
like historical detritus of an earher different patterning, or the effect of English 
analogy, or both. 


ne^__iSj,_andL^.LQ^cther./ There is in it a sensuous and a nonsensuous. 
We may call the sensuous— what we are seeing, hearing, touching— the 
'present' w^hile in the nonsensuous the vast image-world of memory is 
being labeled 'the past' and another realm of belief, intuition, and un- 
certainty 'the future'; yet sensation, memory, foresight, all are in con- 
sciousness together— one is not "yet to be" nor another "once but no 
more." Where real time comes in is that all this in consciousness is 
"getting later," changing certain relations in an irreversible manner. In 
this "latering" or "durating" there seems to me to be a paramount con- 
trast between the newest, latest instant at the focus of attention and 
the rest— the earlier. Languages by the score get along well with two 
tenselike forms answering to this paramount relation of "later" to 
"earlier." We can of course construct and contemplate in thought 
a system of past, present, future, in the objectifie dconfiguration of pomts 
on a line. This is what our general objectification tendency leads us to 
do_a nd our tense system confirmj . 

In English the present tense seems the one least in harmony with the 
paramount temporal relation. It is as if pressed into various and not 
wholly congruous duties. One duty is to stand as objectified middle 
term between objectified past and objectified future, in narration, dis- 
cussion, argument, logic, philosophy. Another is to denote inclusion in 
the sensuous field: 'I see him.' Another is for nomic, i.e. customarily 
or generally valid, statements: 'We see with our eyes.' These varied uses 
introduce confusions of thought, of which for the most part we are 

Hopi, as we might expect, is different here too. Verbs ha\'e no 
^ "tenses" like ours, but have validity-forms ("assertions"), aspects, and 
clause-linkage forms (modes), that yield even greater precision of speech. 
The validity-forms denote that the speaker (not the subject) reports the 
situation (answering to our past and present) or that he expects it (an- 
swering to our future) '^ or that he makes a nomic statement (answering 

■^ The expective and reportive assertions contrast according to the "paramount rela- 
tion." The expective expresses anticipation existing earlier than objective fact, and 
coinciding with objective fact later than the status quo of the speaker, this status 
quo, including all the subsummation of the past therein, being expressed by the 
reportive. Our notion "future" seems to represent at once the earlier (anticipation) 
and the later (afterwards, what will be), as Hopi shows. This paradox may hint of 
how elusive the mystery of real time is, and how artificially it is expressed by a linear 
relation of past-prescnt-future. 

Le+ ijJb How 5ee(< l^^urs4.t WipU^^^c^i ^ 


to our nomic present). The aspects denote different degrees of duration 
and different kinds of tendency "during duration." As yet we have 
noted nothing to indicate whether an event is sooner or later than an- 
other when both are reported. But need for this does not arise until 
we have two verbs: i.e. two clauses. In that case the "modes" denote 
relations between the clauses, including relations of later to earlier and 
of simultaneity. Then there are many detached words that express 
similar relations, supplementing the modes and aspects. The duties of 
our three-tense system and its tripartite linear objectified "time" are 
distributed among various verb categories, all different from our tenses; 
and there i s no more basis for an objectified time in Hopi verb s than in 
other Hopi patterns; although this does not in the least hinder the verb 
forms and other patterns from being closely adjusted to the pertinent 
realities of actual situations. 


To fit discourse to manifold actual situations, all languages need to 
express durations, intensities, and tendencies. It is characteristic of 
SAE and perhaps of many other language types to express them meta- 
phorically. The metaphors are those of spatial extension, i.e. of size, 
number (plurality), position, shape, and motion. We express duration 
by 'long, short, great, much, quick, slow,' etc.; intensity by 'large, great, 
much, heavy, light, high, low, sharp, faint,' etc.; tendency by 'more, in- 
crease, grow, turn, get, approach, go, come, rise, fall, stop, smooth, even, 
rapid, slow'; and so on through an almost inexhaustible list of metaphors 
that we hardly recognize as such, since they are virtually the only lin- 
guistic media available. The nonmetaphorical terms in this field, like 
'early, late, soon, lasting, intense, very, tending,' are a mere handful, 
quite inadequate to the needs. U/^^ I do allr/^ht K^^,'M/^ ^^M^- iMi ^^^i^^t > ^ 

It is clear how this condition "fits in." It is part of our whole scheme '^ ± 
o f OBJECTIFYING— imaginatively spatializing*qual ities and pot entials that ^pg, 
are quite nonspatial (so far as any spatially perceptive senses can tell pL 6 
us). Noun-meaning (with us) proceeds from physical bodies to referents 
of far other sort. Since physical bodies and their outlines in perceived 
SPACE are denoted by size and shape terms and reckoned by cardinal 
numbers and plurals, these patterns of denotation and reckoning extend 


to the symbols of nonspatial meanings, and so suggest an imaginary 
SPACE. Physical shapes 'move, stop, rise, sink, approach,' etc., in per- 
ceived space; why not these other referents in their imaginary space? 
This has gone so far that we can hardly refer to the simplest nonspatial 
situation without constant resort to physical metaphors. I "grasp" the 
"thread" of another's arguments, but if its "level" is^^jivef my head" my 
attention may "wander" and "lose touch" with the "drift" of it, so that 
when he "comes" to his "point" we differ "widely," our "views" being 
indeed so "far apart" that the "things" he says "appear" "much" too 
arbitrary, or even "a lot" of nonsense! [/KuiJl rAa^/i U vififM dC^^oS (vji-re. 
The absence of such metaphor from Hopi speech is striking. Use of 
space terms wh en there is no space involved is not there— as if on it 
had been laid the taboo teetotal! The reason is clear when we know 
that Hopi has abundant conjugational and lexical means of expressing 
duration, intensity, and tendency directly as such, and that major gram- 
matical patterns do not, as with us, provide analogies for an imaginary 
space. The many verb "aspects" express duration and tendency of mani- 
festations, while some of the "voices" express intensity, tendency, and 
duration of causes or forces producing manifestations. Then a special 
part of speech, the "tensors," a huge class of words, denotes only in- 
tensity, tendency, duration, and sequence. The function of the tensors 
is to express intensities, "strengths," and how they continue or vary, their 
rate of change; so that the broad concept of intensity, when considered 
as necessarily always varying and/or continuing, includes also tendency 
and duration. Tensors convey distinctions of degree, rate, constancy, 
repetition, increase and decrease of intensity, immediate sequence, in- 
terruption or sequence after an interval, etc., also qualities of strengths, 
such as we should express metaphorically as smooth, even, hard, rough. 
A striking feature is their lack of resemblance tq the ^rms of real space 
and movement that to us "mean the same.'^ There is not even more 

than a trace of apparent derivation from spacexterms.^ So, while Hopi • . 

8 One such trace is that the tensor 'long in duration,' while quite different from 

'large' ^ 

ans 'at y 
-«.S -1 

ig in duration, while qi 
the adjective 'long' of space, seems to contain the same root as the adjective 
of space. Another is that 'somewhere' of space used with certain tensors mear 
some indefinite time.' Pcssibly however this is not the case and it is only the tensor J." , 
that gives the time element, so that 'somewhere' still refers to space and that under C ; 
these conditions indefinite space means simply general applicability, regardless of^^^. 
either time or space. Another trace is that in the temporal (cycle word) 'afternoon' 3- J 
the element meaning 'after' is derived from the verb 'to separate.' There are other / < 
such traces, but they are few and exceptional, and obviously not like our own spatial ^p-j 
metaphorizing. O', Obw/au.sCu Ykat^ . ^ ^ 


in its nouns seems highly concrete, here in the tensors it becomes ab- V' 
stract almost beyond our power to follow. S pex-t- ^ c^^^^yrse (( ^ uyU^rr-^ 


The comparison now to be made between the habitual thought worlds 
of SAE and Hopi speakers is of course incomplete. It is possible only to 
touch upon certain dominant contrasts that appear to stem from the 
linguistic differences already noted. By "habitual thought" and "thought<_^^ 
world" I mean more than simply language, i7e. than the linguistic pat- 
ferns themselves^_Iin£lud£^^ and suggesfive value of 
th e patterns (e.g., our ''imag inary space" and its distant implications), 
and all the give-and-take between language and the culture as a whole, 
wherein is a vast amount that is not linguistic but yet shows the shaping 
Influence of language^ in brief, this " thou p; ht wbrTJ" is the microc osm 
that each man carries about with in hirnsel f, by which h e measures and 
understands what he can of the macrocosm. 

The SAE microcosm has anahzcd reality largely in terms of what it \/ 
calls "things" (bodies and quasibodies) plus modes of extensional but 
formless existence that it calls "substances" or "matter." It tends to see 
existence through a binomial formula that expresses any existent as a 
spatial form plus a spatial formless continuum related to the form, as 
contents is related to the outlines of its container. Nonspatial existents 
are imaginatively spatialized and charged with similar implications of 
form and continuum. 

T he Hopi microcosjsr seems to have analyzed reality largely in terms / 
qf_EyEKi:&-(or ^tt^ "eventing"), referred to in two ways, objective and 
subjective. Objectivel}', and only if perceptible physical experience, 
events are expressed mainly as outlines, colors, movements, and other 
perceptive reports. Subjectively, for both the physical and nonphysical, 
events are considered the expression of invisible intensity factors, on 
which depend their stability and persistence, or their fugitiveness and 
proclivities. It implies that existents do not "become later and later" 
all in the same way; but some do so by growing like plants, some b\ 
diffusing and vanishing, some by a procession of metamorphoses, some 
by enduring in one shape till affected by violent forces. In the nature 
of each existent able to manifest as a definite whole is the power of its 
own mode of duration: its growth, decline, stability, cyclicity, or crca- 


tiveness. Everything is thus already "prepared" for the way it now mani- 
fests by earlier phases, and what it will be later, partly has been, and 
partly is in act of being so "prepared." An emphasis and importance 
rests on this preparing or being pr epared aspect of the world that may 
to the Hopi correspond to that "quality of reality" that 'matter' or 'stuff' 
has for us. 


\ Our behavior, and that of Hopi, can be seen t© be .coordinated in 

L^ many ways to the Hnguistically conditioned microcosm. As in my fire 
^ casebook, people act about situations in ways which are like the ways ^ 
^-^irfThey talk about them. A characteristic of Hopi behavior is the em- -^ 
\Va/ • ' P^^^sis on preparation. This includes announcing and getting ready for "^ 
■^ ^ cvents"well beforehand, elaborate precautions to insure persistence of v 
1^ desired conditions, and stress on good will as the preparer of right 
jJ\^ results^ Consider the analogies of the day-counting pattern alone. Time 

> ■ : 

f'S^ words are not nouns but tensors, the first formed on a root "hght, day," 

is mainly reckoned "by day" {tOLk, -tala) or "by night" [tok), which ^ V 

the second on a root "sleep.' The count is by ordinals. This is not N '^ 
the pattern of counting a number of different men or things, even '^ \ 
though they appear successively, for, even then, they could gather into ^ ^ 
an assemblage. It is the pattern of counting successive reappearances 
of the SAME man or thing, incapable of forming an assen^blage. The ;^ ^ 
analogy is not to behave about dav-c\clicit\' as to several men ("several t 
days"), which is what we tend to do, but to behave as to the successive ^ ^ 
visits of the same man. One does not alter several men by working S 
upon just one, but one can prepare and so alter the later visits of the ^'^ •^ 
same man by working to affect the visit he is making now. Tliis is the J "i^ 
way the Hopi deal with the future— by working within a present situa- .^ , 
tion which is expected to carry impresses, both obvious and occult, for-^^.^ 
ward into the future event of interest. One might say that Hopi society^ 
understands our proverb 'Well begun is half done,' but not our "1 o-a 
morrow is another day.' This may explain much in Hopi character. ^ 

This Hopi preparing behavior may be roughly divided into announc- 
ing, outer preparing, inner preparing, co\crt participation, and persist- 
ence. Annovnicing, or preparative publicity, is an important function 



in the hands of a special official, the Crier Chief. Outer preparing is 
preparation in\olvhig much visible activity, not all necessarily directly 
useful within our understanding. It includes ordinary practicing, re- 
hearsing, getting ready, introductory formalities, preparing of special 
food, "etc. {all of these to a degree that may seem overelaborate to us), 
intensive sustained muscular activity like running, racing, dancing, which 
'is thought to increase the intensity of development of events (such as 
growth of crops), mimetic and other magic, preparations based on eso- 
teric theor}- in\ol\ing perhaps occult instruments like prayer sticks, 
prater feathers, and prayer meal, and finally the great cyclic ceremonies 
and dances, which have the significance of preparing rain and crops. 
From one of the \erbs meaning "prepare" is derived the noun for 
"harvest" or "crop": na'twani 'the prepared' or the 'in preparation.' ^ 

~ Inner preparing is use of prayer and meditation, and at lesser intensity 
good wishes and good will, to further desired results. Hopi attitudes 
stress the power of desire and thought. With their "microcosm" it is 
utterly natural that they should. Desire and thought are the earliest, 
and therefore the most important, most critical and crucial, stage of 
preparing. Moreo\er, to the Ilopi, one's desires and thoughts influence 
not only his own actions, but all nature as well. This too is wholly 
natural. Consciousness itself is aware of work, of the feel of effort and 
energy, in desire and thinking. Experience more basic than language 
tells us that, if energy is expended, effects are produced. We tend to 
believe that our bodies can stop up this energy, prevent it from affecting 
other things until we will our bodies to overt action. But this may be 
so only because \\c ha\e our own linguistic basis for a thcon' that form- 
less items like "matter" are things in themsehes, malleable only by simi- 
lar things, by more matter, and hence insulated from the powers of life 
and thought. It is no more unnatural to think that thought contacts 
everything and pcr\'ades the universe than to think, as we all do, that 
light kindled outdoors does this. And it is not unnatural to suppose 
that thought, like any other force, leaves everywhere traces of effect. 
Now, when we think of a certain actual rosebush, we do not suppose 
that our thought goes to that actual bush, and engages with it, like a 
searchlight turned upon it. What then do we suppose our conscious- 

3 The Hopi verbs of preparing naturally do not correspond neatly to our "prepare"; 
so that na'twani could also be rendered 'the practiced-upon, the tried-for,' and other 


ness is dealing with when we are thinking of that rosebush? Probably 
we think it is dealing with a "mental image" which is not the rosebush 
but a mental surrogate of it. But why should it be natural to think 
that our thought deals with a surrogate and not with the real rosebush? 
Quite possibly because we are dimly aware that we carry about with us 
a whole imaginary space, full of mental surrogates. To us, mental sur- 
rogates are old familiar fare. Along with the images of imaginary space, 
which we perhaps secretly know to be only imaginary, we tuck the 
thought-of actually existing rosebush, which may be quite another story, 
perhaps just because we have that very con\enient "place" for it. The 
Hopi thought-world has no imaginary space. The corollary to this is 
that it may not locate thought dealing with real space anywhere but in 
real space, nor insulate real space from the effects of thought. A Hopi 
would naturally suppose that his thought (or he himself) traffics with 
the actual rosebush— or more likely, corn plant— that he is thinking 
about. The thought then should leave some trace of itself with the 
plant in the field. If it is a good thought, one aboift health and growth, 
it is good for the plant; if a bad thought, the reverse. 

The Hopi emphasize the intensity-factor of thought. Thought to be 
most effective should be vivid in consciousness, definite, steady, sus- 
tained, charged with strongly felt good intentions. They render the 
idea in English as 'concentrating, holding it in your heart, putting your 
mind on it, earnestly hoping.' Thought power is the force behind 
ceremonies, prayer sticks, ritual smoking, etc. The prayer pipe is re- 
garded as an aid to "concentrating" (so said my informant). Its name, 
na'twanpi, means 'instrument of preparing.' 

Covert participation is mental collaboration from people who do not 
take part in the actual affair, be it a job of work, hunt, race, or cere- 
mony, but direct their thought and good will toward the affair's success. 
Announcements often seek to enlist the support of such mental helpers 
as well as of overt participants, and contain exhortations to the people 
to aid with their active good will.^° A similarity to our concepts of a 
sympathetic audience or the cheering section at a football game sliould 

'0 See, e.g., Ernest Beagleliolc, Notes on Hopi economic life (Yale University Pub- 
lications in Anthropology, no. 15, 1937), especially the reference to the announce- 
ment of a rabbit hunt, and on p. 30, description of the activities in connection with 
the cleaning of Toreva Spring — announcing, various preparing activities, and finally, 
preparing the continnitv of the good results already obtained and the continued flow 
of the spring. 


not obscure the fact that it is primarily the power of directed thought, 
and not merely sympathy or encouragement, that is expected of covert 
participants. In fact these latter get in their deadliest work before, not 
during, the game! A corollary to the power of thought is the power of 
wrong thought for evil; hence one purpose of covert participation is to 
obtain the mass force of many good wishers to offset the harmful thought 
of ill wishers. Such attitudes greatly favor cooperation and community 
spirit. Not that the Hopi community is not full of rivalries and collid- 
ing interests. Against the tendency to social disintegration in such a 
small, isolated group, the theory of "preparing" by the power of thought, 
logically leading to the great power of the combined, intensified, and 
harmonized thought of the whole community, must help vastly toward 
the rather remarkable degree of cooperation that, in spite of much pri- 
vate bickering, the Hopi village displays in all the important cultural 

Hopi "preparing" activities again show a result of their linguistic 
thought background in an emphasis on persistence and constant in- 
sistent repetition. A sense of the cumulative value of innumerable small 
momenta is dulled by an objectified, spatialized view of time like ours, 
enhanced by a way of thinking close to the subjective awareness of dura- 
tion, of the ceaseless "latering" of events. To us, for whom time is a 
motion on a space, unvarying repetition seems to scatter its force along 
a row of units of that space, and be wasted. To the Hopi, for whom 
time is not a motion but a "getting later" of everything that has ever 
been done, unvarying repetition is not wasted but accumulated. It is 
storing up an invisible change that holds over into later events.^^ As we 
have seen, it is as if the return of the day were felt as the return of the 
same person, a little older but with all the impresses of yesterday, not as 
"another day," i.e. like an entirely different person. This principle 

11 This notion of storing up power, which seems imphed by much Hopi behavior, 
has an analog in physics: acceleration. It might be said that the linguistic back- 
ground of Hopi thought equips it to recognize naturally that force manifests not as 
motion or velocity, but as cumulation or acceleration. Our linguistic background 
tends to hinder in us this same recognition, for having legitimately conceived force to 
be that which produces change, we then think of change by our linguistic metaphori- 
cal analog, motion, instead of by a pure motionless changingness concept, i.e. ac- 
cumulation or acceleration. Hence it comes to our naive feeling as a shock to find 
from physical experiments that it is not possible to define force by motion, that 
motion and speed, as also "being at rest," are wholly relative, and that force can be 
measured only by acceleration. 


joined with that of thought-power and with traits of general Pueblo 
culture is expressed in the theory of the Hopi ceremonial dance for 
furthering rain and crops, as well as in its short, piston-like tread, re- 
peated thousands of times, hour after hour. 


It is harder to do justice in few words to the linguistically conditioned 
features of our own culture than in the case of the Hopi, because of both 
vast scope and difficulty of objectivity— because of our deeply ingrained 
familiarity with the attitudes to be analyzed. I wish merely to sketch 
certain characteristics adjusted to our linguistic binomialism of form 
plus formless item or "substance," to our metaphoricalness, our imagi- 
nary space, and our objectified time. These, as we have seen, are 

From the form-plus-substance dichotomy the philosophical views most 
traditionally characteristic of the "Western world" have derived huge 
support. Here belong materialism, psychophysical parallelism, physics— 
at least in its traditional Newtonian form— and dualistic views of the 
universe in general. Indeed here belongs almost everything that is 
"hard, practical common sense." Monistic, holistic, and relativistic 
views of reality appeal to philosophers and some scientists, but they are 
badly handicapped in appealing to the "common sense" of the Western 
average man— not because nature herself refutes them (if she did, phi- 
losophers could have discovered this much), but because they must be 
talked about in what amounts to a new language. "Common sense," 
as its name shows, and "practicality" as its name does not show, are 
largely matters of talking so that one is readily understood. It is some- 
times stated that Newtonian space, time, and matter are sensed by 
everyone intuitively, whereupon relativity is cited as showing how mathe- 
matical analysis can prove intuition wrong. This, besides being unfair 
to intuition, is an attempt to answer offhand question (1) put at the 
outset of this paper, to answer which this research was undertaken. 
Presentation of the findings now nears its end, and I think the answer 
is clear. The offliand answer, laying the blame upon intuition for our 
slowness in discovering mysteries of the CosmoS, such as relativity, is 


the wrong one. The right answer is: Newtonian space, time, and matter 
are no intuitions. They are recepts from culture and language. That 
is where Newton got them. 

Our. objectified view of ti me is, however, favorable to hi storicity antl 
to. everything connected with the kee ping o f records, while the Hopi 
view is'unfavorable thereto. The latter is too subtle, complex, and ever- 
developing, supplying no ready-made answer to the question of when 
"one" event ends^ and "another" begins. When it is implicit that 
everything that e\ er happened still is, but is in a necessarily different 
form from what memory or record reports, there is less incentive to 
study the past. As for the present, the incentive would be not to record 
it but to treat it as "preparing." But our objectified time puts before 
imagination something like a ribbon or scroll marked off into equal 
blank spaces, suggesting that each be filled with an entry. Writing has 
no doubt helped toward our linguistic treatment of time, even as the 
linguistic treatment has guided the uses of writing. Through this give- 
and-take between language and the whole culture we get, for instance: 

1. Records, diaries, bookkeeping, accounting, mathematics stimulated- x 
by accounting. 

2. Interest in exact sequence, dating, calendars, chronology, clocks, 
time wages, time graphs, time as used in physics. 

3. Annals, histories, the historical attitude, interest in the past, ar- 
chaeology, attitudes of introjection toward past periods, e.g., classicism, 

Just as we concei\e our objectified time as extending in the future in 
the same way that it extends in the past, so we set down our estimates 
of the future in the same shape as our records of the past, producing 
programs, schedules, budgets. Tlie formal equality of the spacelike 
units by which we measure and conceive time leads us to consider the 
"formless item" or "substance" of time to be homogeneous and in 
ratio to the number of units. Hence our prorata allocation of value to 
time, lending itself to the building up of a commercial structure based 
on time-prorata values: time wages (time work constantly supersedes 
piece work), rent, credit, interest, depreciation charges, and insurance 
premiums. No doubt this \ast system, once built, would continue to 
run under any sort of linguistic treatment of time: but that it should 
have been built at all, reaching the magnitude and particular form it 


has in the Western world, is a fact decidedly in consonance with the 
patterns of the SAE languages. Whether such a civilization as ours 
would be possible with widely different Hnguistic handling of time is a 
large question— in our civilization, our linguistic patterns and the fitting 
of our behavior to the temporal order are what they are, and they are 
in accord. We are of course stimulated to use calendars, clocks, and 
watches, and to try to measure time ever more precisely; this aids science, 
and science in turn, following these well-worn cultural grooves, gives- 
back to culture an ever-growing store of applications, habits, and values, 
with which culture again directs science. But what lies outside this 
spiral? Science is beginning to find that there is something in the 
Cosmos that is not in accord with the concepts we have formed in 
mounting the spiral. It is trying to frame a new language by which 
to adjust itself to a wider universe. 

It is clear how the emphasis on "saving time'^jwhich goes with all the 
above and is very obvious objectification of time, leads to a high valua- 
tion of "speed," which shows itself a great deal in our behavior. 

Still another behavioral effect is that the character of monotony and 

regLilarit}' possessed by our image of time as an evenly scaled limitless 

tape measure persuades us to behave as if that monotony were more 

true of events than it really is. That is, it helps to routinize us. We 

tend to select and favor whatever bears out this view, to "play up to" 

the routine aspects of existence. One phase of this is behavior evincing 

a false sense of security or an assumption that all will always go smoothly, 

and a lack in foreseeing and protecting ourselves against hazards. Our 

technique of harnessing energ}' does well in routine performance, and it 

is along routine lines that we chiefly strive to improve it— we are, for 

'example, relatively uninterested in stopping the energy from causing 

accidents, fires, and explosions, which it is doing constantly and on i 

J wide scale. Such indifference to the unexpectedness of life would be 

\ disastrous to a society as small, isolated, and precariously poised as the 

( Hopi society is, or rather once was. 

I Thus our linguistically determined thought world not only collabo- 
^ rates with our cultural idols and ideals, but engages even our uncon- 
scious personal reactions in its patterns and gives them certain typical 
characters. One such character, as we have seen, is carelessness, as in 
reckless driving or throwing cigarette stubs into waste paper. Another 
of different sort is gesturing when we talk. Very many of the gestures 


made by English-speaking people at least, and probably by all SAE 
speakers, serve to illustrate, by a movement in space, not a real spatial 
reference but one of the nonspatial references that our language handles 
by metaphors of imaginary space. That is, we are more apt to make a 
grasping gesture when we speak of grasping an elusive idea than when 
we speak of grasping a doorknob. The gesture seeks to make a meta- 
phorical and hence somewhat unclear reference more clear. But, if a 
language refers to nonspatials without implying a spatial analog}', the 
reference is not made any clearer by gesture. The Hopi gesture very 
little, perhaps not at all in the sense we understand as gesture. 

It would seem as if kinesthesia, or the sensing of muscular movement, 
though arising before language, should be made more highly conscious 
by. linguistic use of imaginary space and metaphorical images of motion. 
Kinesthesia is marked in two facets of European culture: ar^and^ sport. 
European sculpture, an art in which Europe excels, is strongly kines- 
thetic, conveying great sense of the body's motions; European painting 
likewise. The dance in our culture expresses delight in motion rather 
than symbolism or ceremonial, and our music is greatly influenced by 
our dance forms. Our sports are strongly imbued with this element of 
the "poetry of motion." Hopi races and games seem to emphasize 
rather the virtues of endurance and sustained intensity. Hopi dancing 
is highly symbolic and is performed with great intensity and earnest- 
ness, but has not much movement or swing. 

Synesthesia, or suggestion by certain sense receptions of characters 
belonging to another sense, as of light and color by sounds and vice 
versa, should be made more conscious by a linguistic metaphorical 
system that refers to nonspatial experiences by terms for spatial ones, 
though undoubtedly it arises from a deeper source. Probably in the first 
instance metaphor arises from synesthesia and not the reverse; yet meta- 
phor need not become firmly rooted in linguistic pattern, as Hopi shows. 
Nonspatial experience has one well-organized sense, hearing— for smell 
and taste are but little organized. Nonspatial consciousness is a realm 
chiefly of thought, feeling, and sound. Spatial consciousness is a realm 
of light, color, sight, and touch, and presents shapes and dimensions. 
Our metaphorical system, by naming nonspatial experiences after spatial 
ones, imputes to sounds, smells, tastes, emotions, and thoughts qualities 
like the colors, luminosities, shapes, angles, textures, and motions of 
spatial experience. And to some extent the re\'erse transference occurs; 


for, after much talking about tones as high, low, sharp, dull, heavy, bril- 
liant, slow, the talker finds it easy to think of some factors in spatial 
experience as like factors of tone. Thus we speak of "tones" of color, 
a gray "monotone," a "loud" necktie, a "taste" in dress: all spatial meta- 
phor in reverse. Now European art is distinctive in the way it seeks 
deliberately to play with synesthesia. Music tries to suggest scenes, 
color, movement, geometric design; painting and sculpture are often 
consciously guided by the analogies of music's rhythm; colors are con- 
joined with feeling for the analogy to concords and discords. The Euro- 
pean theater and opera seek a synthesis of many arts. It may be that in 
this way our metaphorical language that is in some sense a confusion of 
thought is producing, through art, a result of far-reaching value— a 
deeper esthetic sense leading toward a more direct apprehension of 
underlying unity behind the phenomena so variously reported by our 
sense channels. 


I low does such a network of language, culture, and behavior come 

about historically? Which was first: tlie language patterns or the cul- 

^tural norms? In main they ha\c grown up together, constantly in- 


flue ncmg each other. But in this partn ership the nature of the language 
is the factor that limits fre e plasticity a nd rigidities channels of develop- 
nient in the more autocratic way. This is so because a language is a 
system, not just an assemblage of norms. Large systematic outlines can 
change to something really new only very slowly, while many other cul- 
tural innovations are made with comparative quickness. Language thus 
represents the mass mind; it is affected by inventions and innovations, 
but affected little and slowly, whereas to inventors and innovators it 
legislates with the decree immediate. 

The growth of the SAE language-culture complex dates from ancient 
times. Much of its metaphorical reference to the nonspatial by the 
spatial was already fixed in the ancient tongues, and more especially in 
1 ,alin. It is indeed a marked trait of Latin. If we compare, say Hebrew, 
we find that, while Hebrew has some allusion to not-space as space, 
Latin has more. Latin terms for nonspatials, like educo, religio, prin- 
cipia, comprehendo, are usually metaphorized physical references: lead 


out, tying back, etc. This is not true of all languages— it is quite untrue 
of Hopi. The fact that in Latin the direction of development happened 
to be from spatial to nonspatial (partly because of secondar)- stimulation 
to abstract thinking when the intellectually crude Romans encountered 
Greek culture) and that later tongues were strongly stimulated to mimic 
Latin, seems a likely reason for a belief, which still lingers on among 
linguists, that this is the natural direction of semantic change in all 
languages, and for the persistent notion in Western learned circles (in 
strong contrast to Eastern ones) that objective experience is prior to sub- 
jective. Philosophies make out a weighty case for the reverse, and cer- 
tainly the direction of development is sometimes the reverse. Thus the 
Hopi word for 'heart" can be shown to be a late formation within Hopi 
from a root meaning think or remejnber. Or consider what has hap- 
pened to the word "radio" in such a sentence as "he bought a new 
radio," as compared to its prior meaning "science, of wireless-telephony." 

In the Middle Ages the patterns already formed in Latin "began to 
interM'ea\'e with the increased mechanical invention, industr)-, trade, and 
scholastic and scientific thought. Tlie need for measurement in industry 
and trade, the stores and bulks of "stuffs" in various containers, the type- 
bodies in which various goods were handled, standardizing of measure 
and weight units, in\ ention of clocks and measurement of "time," keep- 
ing of records, accounts, chronicles, histories, growth of mathematics and 
the partnership of mathematics and science, all cooperated to bring our 
thought and language world into its present form. 

In Hopi histon.-, could we read it, we should find a different type of 
language and a different set of cultural and environmental influences 
working together. A peaceful agricultural society isolated by geographic 
features and nomad enemies in a land of scanty rainfall, arid agriculture 
that could be made successful only by the utmost perseverance (hence 
the value of persiitence and xepetition), necessity for collaboration 
(hence emphasis on the psychology of teamwork and on mental factors 
in general), corn and rain as primary criteria of \'alue, need of extensive 
^^REPARATIONS and precautions to assure crops in the poor soil and pre- 
carious climate, keen realization of dependence upon nature favoring 
prayer and a religious attitude toward the forces of nature, especially 
prayer and religion directed toward the ever-needed blessing, rain— these 
things interacted with Hopi linguistic patterns to mold them, to be 


molded again by them, and so little by little to shape the Hopi world- 
ouilook. , 

To ^n rr] up fhe matt er, our first question asked in the beginning (p. 
138) is answered thus: Concepts of "time" and "matter" are not given 
in substantiall y the same form by e xperience to all men but depend 
upon the nature of the language or languages through the use of which 
they have been developed. They do not depend so much upon any one 
SYSTEM (e.g., tense, or nouns) within the grammar/as upon the ways of 
analyzing and reporting experience which have become fixed in the lan- 
guage as integrated "fashions of speaking" and which cut across the 
typical grammatical classifications, so that such a "fashion" may include 
lexical, morphological, syntactic, and otherwise systemically diverse 
means coordinated in a certain frame of consistency. Our own "time" 
differs markedly from Hopi "duration." It is conceived as like a space 
of strictly limited dimensions, or sometimes as like a motion upon such 
a space, and employed as an intellectual tool accordingly. Hopi "dura- 
tion" seems to be inconceivable in terms of space or motion, being the 
mode in which life differs from form, and consciousness in toto from 
the spatial elements of consciousness. Certain ideas born of our own 
time-concept, such as that of absolute simultaneity, would be either 
very difficult to express or impossible and devoid of meaning under the 
Hopi conception, and would be replaced by operational concepts. Our 
"matter" is the physical subtype of "substance" or "stuff," which is con- 
ceived as the formless extensional item that must be joined with form 
before there can be real existence. In Hopi there seems to be nothing 
corresponding to it; there are no formless extensional items; existence 
may or may not have form, but what it also has, with or without form, 
is intensity and duration, these being nonextensional and at bottom the 

But what about our concept of "space," which was also included in 
our first question? There is no such striking difference between Hopi 
and SAE about space as about time, and probably the apprehension of 
space is given in substantially the same form by experience irrespective 
of language. The experiments of the Gestalt psychologists with visual 
perceptior) appear to establish this as a fact. But the concept of space 
will vary somewhat with language , becau se, as an intellectual tool,^^ it 

12 Here belong "Newtonian" and "Euclidean" space, etc. 


is SO closely linked with the concomitant employment of other intellec- 
tual tools, of the order of "time" and "matter," which are linguistically 
conditioned. We see things with our eyes in the same space forms as 
the Hopi, but our idea of space has also the property of acting as a 
surrogate of nonspatial relationships like time, intensity, tendency, and 
as a void to be filled with imagined formless items, one of which may 
even be called 'space.' Space as sensed by the Hopi would not be con- 
nected mentally with such surrogates, but would be comparatively 
"pure," unmixed with extraneous notions. 

As for our second question (p. 138): There are conne ctions but n ot 
orrelatio ns or diagnostic correspondences between cultural norms an d 
l inguistic pattern s. Although it would be impossible to infer the exist- 
ence of Crier Chiefs from the lack of tenses in Hopi, or vice versa, there 
is a relation between a language and the rest of the culture of the so -^ 
ciety which uses it. There are cases where the "fashions of speaking" 
are closely integrated with the whole general culture, whether or not 
this be universally true, and there are connections within this integra- 
tion, between the kind of linguistic analyses employed and various be- 
havioral reactions and also the shapes taken by various cultural develop- 
ments. Thus the importance of Crier Chiefs does have a connection, 
not with tenselessness itself, but with a system of thought in which cate- 
gories different from our tenses are natural. These connections are to 
be found not so much by focusing attention on the typical rubrics of 
linguistic, ethnographic, or sociological description as by examining the 
culture and the language (always and only when the two ha\e been to- 
gether historically for a considerable time) as a whole in which concate- 
nations that run across these departmental lines may be expected to 
exist, and, if they do exist, eventually to be discoverable by study. 


C. F. Voegelin Has accomplished the difficult and signal work of analyz- 
ing an immense number of baffling stem compounds of Shawnee into 
their component lexemes (stems) and other morphemes (formatives), 
classifying them according to formal categories of Shawnee grammar, 
and discovering an important native semantic relation, that of the oc- 
current, a lexeme that has some pervasive semantic influence that in- 
duces the native to cling to translation of the occurrent even when he 
neglects specific translation of the other lexemes in the compound. 

Voegelin has asked me to illustrate the application of a different 
aspect of linguistic method, which can be applied only after a formal 
grammatical analysis has been made, but which then can sometimes 
show the principles by which lexemes of differing meaning are placed 
in certain sequences to produce semantic effects, whether in compounds 
or in syntactic constructions. 

Linguists have studied the Indo-European languages so long that they 
have been able to generalize their most typical sequences and resultant 
semantic effects into such general formulas as subject and predicate, 
actor, action, and goal, attribute and head, and exocentric versus endo- 
centric; also to tag and handle relations that have a superficial similarity 
in languages that may otherwise differ greatly from Indo-European. But 

* Reprinted from the appendix, pp. 393-406, to C. F. Voegelin, Sliawnee stems 
and the Jacob P. Dunn Miami dictionary. Indianapolis: Indiana Historical Society, 
1940. {Prehistory Research Series, vol. I, no. 9, April 1940). 



this last ability must turn out in many cases to be only a happj, or at 
times perhaps an unhappy accident. When the principles of composi- 
tion are themselves widely different, these schematizations break down 
and cannot account for either the rules of sequence or the resulting 
semantic effects. Let me give a simple instance from a language not 
wideh^ removed from the syntactic type of Indo-European— Aztec, be- 
longing to the Uto-Aztecan stock. Here an apparent attribute-head re- 
lationship is very definite, and the attribute term or modifier always 
precedes the head or modified term (why this is a necessary conclusion 
would require some length to explain). Yet many expressions follow the 
type of 'narrow road,' o''^-picak-tli, in which 'narrow' has to be expressed 
by a sort of \erbal passi\e participle, 'narrowed' {-picak-), and such a 
participle is placed last, hence after 'road' (-0^-). The complete cor- 
relation of attribute and head with word order in this language forces 
us to conclude that 'narrow' is the head and 'road' is the attribute, as 
it is in English 'roadside.' Yet if one wished to say 'new road, good 
road, brick road,' in these 'road' would be 'head' and would come last. 
Of what use then, to one v\'ishing to compose in Aztec, is the categorv 
of attribute and head, when it cannot say whether such a simple notion 
as 'road' is attribute or head in semantic effects that seem so closelv 
parallel as 'narrow road' and 'good road'? One concludes that such 
categories are but linguistic kinship systems, and like social kinship 
systems do not follow any uni\ersal norm. 

It is the same with the schematizations of subject-predicate, actor- 
action, and action-goal. Even in English, the description of such a 
sentence as 'the tree stood here' as 'actor-action' is rather forced, even 
if it is formally parallel to 'the boy ran.' A hypothetical American lan- 
guage X might use three or more lexemes instead of two for the latter; 
perhaps (1) mo\'ement-of-foot (2) over-a-surface (3) manifestation-of- 
boy-occurs-quickly. Perhaps (3) might bear formatives that make it 
formally a verb, or an 'action,' but again such formatives might be 
'operators' apph ing to the whole sentence, no more to one lexeme than 
another. Such a sentence cannot really be broken into a subject and a 
predicate, not even when it consists of just two formal words. Never- 
theless it has an analysis, and the parts correspond to certain essentials 
that have been segregated out of the situation reported: i.e., the situa- 
tion does contain something that might be called a surface and some- 
thing that can be called moving feet, besides something that can be 


called a boy. Our problem is to determine how different languages 
segregate different essentials out of the same situation. This is often 
a crucial question in the description of a language, and it must not be 
supposed that it has been answered by an account of the formal rules 
for combination into sentences of the lexemes and other morphemes 
that represent the language's segregation of essentials out of situations. 
Our hypothetical language X might express the sentence (l)-(2)-(3) by 
a polysynthetic compounding of stems and formatives into one formal 
word, as often in Shawnee, or by a number of words arranged into a 
sentence as analytic as one of English; yet in either case the really im- 
portant difference from English is the same, viz., that it has isolated 
the peculiar group of essentials (1), (2), (3), and ignored our own isola- 
tion of 'boy (as actor)' and 'ran.' So, where we speak of 'cleaning (a 
gun) with a ramrod,' Shawnee does not isolate any rod or action of 
cleaning, but directs a hollow moving dry spot by movement of tool 
{Shawnee stems, part III, 157). This is what makes Shawnee so strange 
and baffling from the standpoint of English, and not at all the mere 
fact that it is polysynthetic. A language can be polysynthetic and still 
say 'clean with a ramrod' polysynthetically, thereby remaining quite 
transparent from the standpoint of English. 

To compare ways in which different languages differently "segment" 
the same situation or experience, it is desirable to be able to analyze or 
"segment" the experience first in a way independent of any one lan- 
guage or linguistic stock, a way which will be the same for all observers. 
This cannot be done by describing the situation in terms of subject- 
predicate, actor-action, attribute-head, etc., for any scientific use of 
such terms contemplates that they shall have a variable meaning as 
defined for each particular language, including the possibility that for 
some languages their meaning shall be nil. Neither can it be done 
wholly by familiar terms ranging from the common-sense type to the 
quasiscientific, as by tr\'ing to break up the situation into 'things, ob- 
jects, actions, substances, entities, events.' Cautious use of such terms 
may be helpful, perhaps unavoidable, but it must be remembered that 
in their ranges of meaning they are but the creatures of modern Indo- 
European languages and their subsidiary jargons, and reflect the typical 
modes of segmenting experience in these tongues. They do not become 
scientific for linguistics because they may happen to be used in physics 
or chemistry. When they refer to psychological experience, like the 


terms 'thoughts, ideas, concepts,' they require no less caution in use, but 
they are under no specially strong taboo for being "mentalistic" or 
"mystical." Mystical in the proper sense they certainly are not, but 
are merely "lexations," no better and no worse than 'gravitation' or 

There is one thing on which all observers of the appearance of a 
running boy will agree, at least after questioning or experimental test- 
ing—that it can be divided into parts— and they will all make the 
division in the same way. They will all divide it into (1) a figure or 
outline having more or less of motion (the boy) and (2) some kind of 
background or field against which, or in which, the figure is seen (that 
is, if we define observation in its common visual sense and so leave out 
the blind "observers"). 

A discovery made by modern configurative or Gestalt psychology 
gives us a canon of reference for all observers, irrespective of their 
languages or scientific jargons, by which to break down and describe 
all visually observable situations, and many other situations, also. This 
is the discover}' that visual perception is basically the same for all 
normal persons past infancy and conforms to definite laws, a large 
number of which are fairly well known. It is impossible here to do 
more than touch on these laws, but they bring out clearly that the basal 
fact of visual perception is the relation of figure and ground, that per- 
ceptions are largely in the nature of outlines, contrasted more or less 
with the grounds, fields, and fillings of outlines, and that perception of 
motion or action is figural in type, or connected with the perception of 
at least a vague outline quality. 

To say that the facts are essentially the same for all obseners is not 
to deny that they have their fringe of aberrations and individual dif- 
ferences, but these are relatively minor. Brain lesions and eye defects 
produce distortions; special skills or mental effort can rearrange em- 
phases and sometimes change the figure-ground roles of certain items, 
as when one "wills" the drawing of a cube seen edgewise to look like 
a hexagon with three radii. Color blindness and unequal sensitivity to 
colors are such marginal variations; impression of size, too, has marginal 
variation, as when the moon looks to one person the size of a nickel, 
to another as big as a house, yet always subtends on the retina less than 
a pencil at arm's length. When it comes to shape, the variations arc 
still more marginal and slight. All these variations operate within the 


frame of known laws, and so do not hinder a normative account of per- 
ceived data. The facts may differ slightly; the laws are the same for 
all. If the perceptual influences are such as to cause one normal person 
to see a definite outline, they will cause all other normal persons to see 
the same outline. For example, all people see the constellation Ursa 
Major as the outline which we call dipper-shaped, though they may not 
call it a dipper or have such a utensil in their culture, and though there 
are, of course, no lines connecting the stars into this or any other outline. 

But how do these laws of vision gi\e any canon of reference for non- 
\isual experience? By process of elimination. Everything that "takes 
up space" can be shown to be known directly or indirectly through 
vision. E\er\thing unvisual is unspatial in character (and vice versa) 
and is felt as immediate to the experiencer. Touch alone is somewhat 
fused with visual material, and, when it tells us form, contour, and tex- 
ture, it is indirectly visual. Visual experience is projected and consti- 
tutes space, or what we shall call the external field of the observer; un- 
^'isual experience is introjected and makes up what we shall call, follow- 
ing some Gestalt psychologists, the ego field, or egoic field, because the 
observer or ego feels himself, as it were, alone with these sensations and 
awarenesses. Hence in referring a certain experience to the egoic field, 
because it is not in the visual field, or to the ambivalent borderland, as 
when a sensation is known by both modes as within the observer's body, 
we are classing it as all observers class it, regardless of their language, 
once they understand the nature of the distinction. Moreover, the 
egoic field has its own Gestalt laws, of sense quality, rhythm, etc., which 
are universal. We can unhesitatingly class the referent of a lexeme of 
hearing, tasting, or smelling along with those of thinking, emotions, 
etc., in the egoic realm and apart from any lexeme referring to an 
experience having outline or motion. Tlie difference between light and 
darkness, and the referent of seeing, not of what is seen, also, is either 
borderland or of the egoic field, because the sensation quality is intro- 
jected though the figure-ground quality is projected; the referent of 
saying something is also egoic, because the observer introjects both his 
own and other people's speech, equating an essential from it to his egoic 
field of hearing or sound; and the referent of possessing or ha\ing is also 

This principle of classifying referents is nonlinguistic and nonsemantic 
in the ordinary sense of semantic. An isolate of experience in either the 
external or the egoic field, e.g., a shape or a noise, is not a meaning. 


Ne\ertheless a language may sometimes have a principle of classifying 
groups of morphemes and their semantic effects which is coordinated 
with this universal principle. Thus, in English, verbs referring to the 
subject's ego-field experience use the simple present tense for momen- 
tary' present fact, and not the present progressive. Other verbs employ 
the present progressive tense for either momentary or continued present 
fact, and the simple present (except in special locutions like 'here he 
comes') for the nomic or customary tense aspect. Foreigners learning 
English do not know this, and hence may say 'I am hearing you, he is 
seeing it.' English speakers say 'I hear you, he sees it, he feels sick, I 
say that—, I think that—'; but, on the contrar}', 'I am working' (not 'I 
work'), 'the boy is running' (not 'the boy runs,' which is nomic, e.g., 'the 
boy runs whenever — ')} 

I have found this Gestalt method of describing referents and situa- 
tions of so much service toward understanding puzzling points of lan- 
guages, as different in viewpoints as English, Hopi, Aztec, and Maya, 
that I decided to try it on Shawnee, though I know nothing of Shawnee 
or any other Algonkian language except from what Voegelin has pub- 
lished in the present scries and his manuscript for the remainder of the 
series, which will complete his lexicon of Shawnee and Miami. The 
results are as follows, and it is for Algonkianists to say whether they 
have any significance or utility. 

A fairly simple rough general rule applies to formation of a Shawnee 
stem compound. It may be likened in applicability to the rough gen- 
eral rule for noun-phrase composition in English: modifier precedes 
modified. Neither rule is absolute; e.g., in English 'brick buildings' 
represents the typical case where the general rule applies, 'buildings 
brick except for frame porches' one of the overriding special rules. The 
English rule is a good rough guide to a modern European, learning to 
compose English, because his own language is sufficiently similar so 
that he understands what is meant by modifier and modified: his lan- 
guage makes a similar classification of experience, and at most, as in 
French, merely reverses the order.- The phraseology of modifier and 

1 Cf. 'he is feeling (outlining) it' — external field visual touch, and 'he feels it' — 
egoic field sensation. 

2 Yet sometimes the non English speaker makes a Gestalt error, as when a Mexican 
translated for me desierto de los leones as 'lion desert.' Normal English does not say 
'lion wilderness' or ' ocean' because of an overriding rule that a small compact 
figure does not modify a total external-field ground, or in vernacular 'a little object 
can't niodifv all rmt of doors.' 


modified does not work for Shawnee stem compounds that result in a 
verb, as they usually do. The rough general rule (aside from the over- 
riding rules) for Shawnee is: figure precedes external field, the more 
figural precedes the less figural, but the egoic field generally precedes all 
these. The chief overriding rules are: (1) a group of stems of vague 
figure (vague movement, texture, size, etc.) precedes everything; (2) non- 
initial stems must be preceded somehow, even in contravention of the 
general principle, although they generally comply in being less figural 
than what precedes; (3) when the result is- a noun (but not a bahuvrihi 
or sentence used like a noun) the rule is reversed, and ground or field 
characters precede the figural; (4) two themes each compounded ac- 
cording to all the foregoing may be placed together, sometimes result- 
ing in irregular sequences within the total formation; (5) such a theme 
is sometimes used like a stem. 
The descriptions of the reference of stems will be: 

svf special stem of vague figure (often of vague motion, direction, texture 
or plasticity of surface or mass, size, etc.). 

ef egoic field reference. 

f figure — this group expresses outline and space distribution more than 

the others; it does not necessarily imply movement but movement may 
be present, and an f stem may be used after some more figural f stem 
to denote relative external field, ground, or filling quality of the latter. 

frg figure as relative ground, as in preceding description, often a body part. 

mf movement figure, the 'idea' or image of a certain outline of movement. 

fcm figure containing movement, a vaguely outlined field which is relatively 
stationary but has movement or "coming to rest" within it. 

xf external field or ground, with a minimum of figural or outline quality. 

i instrumental, a special small group of elements. 

This is the normal order of position, i.e., svf, ef, f, frg, mf, fcm, i— 
although there is nothing very rigid about the relative order of frg, mf, 
fcm, with respect to each other. Formatives will be denoted: 

s (formative of) subject. 

o object and/or transitive. 

t transitive element, 

m miscellaneous formatives. 

In a broad sense the group f, frg, mf, fcm, xf, is one, and sometimes the 
same stem can fill any one of these positions if it is preceded or followed 
by stems in such a way that the progression of decreasing figural quality 
and increasing ground or field quality is carried out. 


Examples of svf stems are: pa^- 'going, moving thither'; paak- 'hard, 
firm, staccato movement'; tep- 'acquiring'; kt- 'big, superlative'; ka^k- 
'rough, dry'; laakeet- 'lightly, easily, quickly'; laasiwe- 'down, oflf from, 
above'; lo^Oee- 'go out, off, through'; liil- 'diversity.' 

Examples of ef stems are: pa^pa- 'tapping sound' (ef because a sound); 
pedkw- 'aversion, repugnance'; pet9ak{y)- 'trouble, nuisance, intricate, 
difficult, confused, excited'; pt- 'accidental, unintentional, erroneous'; 
teepwe- 'truth'; tepaat- 'satisfactory'; cP9- 'fear'; ki^- 'warm, hot'; 
-kii^kwe- 'consciousness'; katawi- 'ability'; katow- 'ask, beg'; lalalwee- 
'rattling noise'; miim^kaw- 'disco\'ered, remembered'; wiyakowee- 'anger'; 
waa^i- 'intentional'; halan- 'notifying'; -eele- 'thought.' 

Examples of the (f xf) group of the first degree, operating chiefly 

as f, are: pap- 'roomy configuration'; pat- 'moist spot or mass'; PpeH- 
'weakly propping configuration'; petekw- 'rounded, around, roll'; petakw- 
'covering, top, abo\'e'; piit- 'interior outline, inside, hole'; pPtaw- 'in- 
between figure'; pakw- 'plantlike, leafshaped'; peekw- 'dry spot'; pe^kw- 
'cluster, bunched together'; po^k{y)- 'broken, smashed condition'; 
palk{y)- 'emergence from opening'; posk-w- 'irregular fraction, hah'ed, 
broken'; tepilahi 'straight (outline)'; tepet{w)- 'together, in a group'; 
cee- 'matched pair or combination, equal, even'; kip- 'co\'ered, closed up'; 
kotekwi 'turning, winding'; kakaanwi 'long (long outline)'; kooky- 'im- 
mersed in water'; saapw- 'in and out, through'; §kote 'fire (fier\' figure)'; 
laa- 'midst of area'; leep- 'tapering at base'; liipiik- 'settled liquid'; —Ppw- 
'contracted, pounded down'; laka^kwa 'ribbed contour, like washboard 
or palate'; lekw- 'covered with ground or ashes'; liiky- 'dismantled, apart'; 
le^Oawaa- 'forked'; laal- 'hanging down, away from'; lePky- 'torn outline, 
tear cloth, etc.'; \vaaMnyaa- 'circle.' 

Some f stems often operating as frg are: {-)lec- 'finger, hand, in the 
fingers, on the hand'; -ece- 'belly, body' (in fact, all body and body-part 
terms are usually frg); -a^kwi- '(mass of) vegetation, flora, wood'; -aalaka 
'hole, hollowness'; -kamekwi 'house, in the house'; -^ee- 'cloth, clothing'; 
-wale 'back pack'; -api 'sitting configuration, sitting.' 

Some mf stems are: peteki 'back in time or space' (movement or 
path); -pho 'picking up while running'; ptoo- 'running'; clip- 'conveying 
in, conveying secretly'; cPciip- 'shaking'; -^tan- 'flowing, floating'; -ke- 
'general bodily mo\ement'; -e^ka- 'initiating bodily mo\emcnt'; -eka 
'dancing'; -kawi 'dripping'; -^9a- 'flying'; loop- 'swinging'; lek- 'dissolving, 


melting'; miil- 'giving'; hee- 'going'; -'''Oen 'breaking off from moving 

Some fcm stems are: ppte- 'foaming'; pootawe 'burning wood'; 
-e^ekwi 'stream'; kapee- 'crossing stream'; kalawi- 'talking (person talk- 
ing)'; kon- 'swallowing'; kwaap- 'lifting from water'; kwaskw- 'recoiling'; 
kwke- 'movement of hook in water'; -a- 'teeth movement'; -^si-, -^sin- 
'coming to rest'; -laa- 'boiling.' 

Examples of stems chiefly operating as xf are: -piiwe 'hair, feathers'; 
-aapo 'liquid'; -pki 'scattering over level surface, level surface'; tepki 
'swampy, marshy terrain'; tepe^ki 'night'; -taskwi 'flora'; -la 'color'; -kami 
'expanse of water'; -''ki 'expanse, abundance'; -'^kwatwi 'sky'; -swaa- 'space, 
room'; -^skw-atwi 'herbage'; -aam- 'soil, ground'; -^ho- 'water, wetness'; 
-^sfe(y)- 'softness, sliminess.' 

Examples are hardly needed of the small group of very common i 
stems (instrumentals) indicating operation by hand, foot, tool, heat, etc. 

A few examples of composition may be explained in detail. In 
Shawnee Stems, part III, 289, under kip-, kipw-, stem of f type, an out- 
line of closure or of something covered up is placed or 'pictured' on a 
ground or in a setting of: {a) buckskin, (b) a path, (c) the eye region, 
(d) the eye region with movement of hand {-kip-iikwee-n-, -f-frg-i-), also 
in the anus region, the mouth, the ear, etc. Or consider the way of 
saying 'among the swamps.' Our own manner of lexation is to isolate 
from the experience an essential that we call 'swamp,' in the form of a 
typical English noun. As such a noun, it slides in the grammatical 
grooves prepared for all nouns, is treated as a typical "thing," referred 
to as having individual separation, singularity, plurality, suitability for 
article and preposition treatment. There is little difference in linguistic 
treatment between a swamp and a butterfly, in spite of the enormous 
difference in the perceptual experience. In Shawnee we have to forget 
the English type of lexation and fall back on the perceptual situation. 
The referent of our preposition 'among' becomes actually the part of 
the picture with the most quality of outline— a limited, defined spot in 
the midst of an indefinite field— which is a field of swampiness. TTie 
picture is, as it were, rough sketched by placing first the figural element 
laa- 'midst of area' followed by its ground or setting tepki 'swampy ter- 
rain,' > laa-teepki (f-xf) '(spot) among the swamps, in the swamp' (part 
IT, 137). 

In part II, 1 57, we have for 'I clean or dry gun by running ramrod in 


it,' ni-peekw-aalak-h-a, s-f-frg-i-o. The figural center of the composition 
is a dry or clean spot {peekw) which is placed in a setting of hoUowness 
or 'hole' by the stem -aalak-, frg, a figure serving as relative ground or 
field for the first figure; the figural center is then activated or given 
motion by the instrumental -h- 'by movement of tool,' and denoted as 
transiti\'e with inanimate object by the formative -a. In part II, 143, the 
f-stem cee- sets up a basic outline for the composition, the configuration 
of a matched pair, or matching units. Essentials of filling quality for 
the matched pair are given by xf stems signifying 'certain kind or t}pe, 
general appearance, color,' or by frg stem 'person-s,' or by xf-xf, e.g., 
'color plus body hair,' and 'color plus water.' Thus the words mean 'of 
the same type, looking alike, of the same color,' etc. In ni-PpeO-k-a, 
s-f-mf-o, part I, 69, 'I lean against it to hold it up,' the basic outline of 
propping, \'isually often something like a rough T or a lower-case lambda, 
is given by Ppe9 (< Ppe^t), and the vague figural quality of a moving 
animate body is imparted by the mf stem -k- 'bodily mo\ement.' In 
ni-Ppec-si-m-a, s-f-fcm-t-o, 'I put him there (on some support) to keep 
him from falling,' the second figure -si-, a vague outline of motion com- 
ing to rest within an area, comes to rest in or on (as filling of) the basic 
prop outline, and is made transitive with animate object. 

Many further examples of stem compounding are briefly analyzed in 
the list at the end of this appendix. It remains to speak of noun com- 
pounding and theme compounding, which are here regarded as over- 
riding the basic rule of stem compounding. When the result of com- 
pounding is a noun, the rule is reversed: field or ground precedes figure, 
the less figural precedes the more figural. Since this is also commonly 
the case in the usual type of attribute-head relation in English, the 
Shawnee noun (unlike the verb) can usually be understood in terms of 
such a relation; e.g., part III, 290, kopeleko-miyeewn, xf-f, 'iron road 
(railroad).' The thing mentally associated with the head term or figure 
precedes it, again as in English; in this t}pe of analysis it is regarded as 
a datum of the speaker's egoic field due to memor\', and hence is de- 
noted efm. Thus, part II, 139, takhwaan-ekaawe, efm-f, 'bread dance'; 
p. 141, taamin-aapo 'corn liquid (whisky)'; p. 143, ciipa-yeemo, efm-f, 
'spirit bee'; p. 145, caki-y^k\vee9a, frg-f, 'small woman,' where the vague 
figure, awareness of degree of size, precedes the more definite outline. 

Two compound themes may be compounded with each other, but 
what rule if any governs the order I am unable to say. The examples 


seem few compared to the huge number of ordinary stem compounds. 
Thus in part I, 67, is a form that can perhaps be analyzed as waaUnitasi- 
paHenaweewi-ci, theme-theme-s. The first theme would be waa^i-nitasi, 
ef-f 'intentionally at a spot there,' the second paHenaweewi, svf-ef 
'going thither, living.' On the other hand, perhaps we have here a 
theme preceded by two stems. 

It may be that there are two kinds of lexemes in Shawnee, stems and 
themes, and two kinds of compounding techniques, one for stems and 
one for themes. According to such a theory, stem compounds would 
use the principle of figure before ground and would result in a verbal 
theme, which if ending the word makes it a verb or normal sentence, or 
else a nomic sentence used as a noun (a bahuvrihi form). Theme com- 
pounds would use the principle of ground before figure, the result being 
whatever the last theme is, verb or noun. It would then appear that 
certain lexemes although not analyzable are always themes, e.g., funda- 
mentally nominal lexemes and the svf and ef lexemes. These svf and ef 

elements, being themes, precede the f xf verbal theme, as less figural. 

This is, of course, very tentative. 

Nothing treated in this appendix explains which stem the native re- 
gards as most pervasive: i.e., as the occurrent. This may be a question 
quite apart from that of the method of compounding. I might hazard 
a guess that it may depend on the degree of analogic pressure behind 
the various stems in a combination. Some stems in the nature of the 
case would be more productive of combinations than others. The stem 
with the greatest number of close parallels to the combination in ques- 
tion might be felt by the native as nuclear. 

There follows a list of analyzed combinations from parts I, II, III, 
each preceded by page number and followed by formula and translation, 
often reworked from Voegelin's to illustrate the technique, with oc- 
casional comments on the semantic effect. 

Part 1 : 67 pa-kwke, svf-fcm 'he went to condition of water with hook 
moving in it, he went fishing.' 67 ni-pa^-pem-^9e-Ho, s-svf-f-mf-o 'I 
pass it (the resolution) along (around).' 67 ye^-pa^-nekot-Oee-Oi-ya, m-svf- 
f-mf-m-s 'when I go there alone.' 68 papi-^w^aa-m, f-xf-s 'figure of roomy 
occupancy in field of general space occurs, it has plenty of room.' 
69 ni-pat-sk-a-m-a, s-f-xf-fcm-t-o 'I kissed him' (f moist spot, xf general 
softness, fcm lip mvt., stem -a- can be called fcm but is perhaps also i). 
69 ni-pat-Sk-a^h-w-a, s-f-xf-i-t-o 'I made him wet with mud' (i, mvt. of 


tool or means). 70 ni-peteko-n-a, s-f-i-o '1 roll it.' 70 peteko-ce-§ka, 
f-frg-mf 'he doubled up his body' (rolled outline in relative ground of 
body or belly, general bodily mvt.). 70 ta'''petekisi-miim^kaweele-ta-m- 
akwe, theme-theme-t-o-s < ta?-peteki-si, svf-mf-m 'back to that place,' 
miim^kaw-eele, ef-ef 'remember.' 71 ni-petskw-eele-m-a, s-ef-ef-t-o '1 hate 
him.' 71 pet9aki-lee-6a, ef-f-m '(person who) is a nuisance.' 71 ni- 
petako-l-aw-a, s-f-mf-t-o '1 shot above him' (outline of superposition, pro- 
jectile mvt.). 71 ni-pt-a-m-a, s-ef-fcm-t-o '1 accidentally bit him.' 72 piic- 
Oe-^Oen-wi, f-mf-fcm-s 'it breaks off and flies in.' 72 piici-lece-^sin-wa, 
f-frg-fcm-s 'he laid his hand inside.' 72 ni-piici-miil-a, s-f-mf-o '1 gave it 
to him through a hole.' 72 piit-alwa, f-frg (interiority figure, bullet fill- 
ing) = 'bullet sack.' 72 piHeewi-laate, fcm-fcm as relative ground 'it 
foams when boiling.' 73 ni-pPtawise-^9e-to, s-theme-fcm-o '1 joined 
pieces of cloth,' pPtawi-se, f-xf 'juncture (f) in cloth (xf).' 73 piimi- 
pooteO-wa, ef-f-s 'he makes error in smoking.' 74 paak-aame^ki, svf-xf 'it 
is a hard spot (svf) of ground (xf). 74-75 ni-paak-eele-m-a, s-svf-ef-t-o 'I 
think of his strength.' 76 ni-paki-kaw-Ho, s-svf-mf-o 'I make it drip.' 
77 ni-pkw-e^ko-ta, s-f-fcm-o 1 cut a piece off it.' 83 leelawi-piikwa, f-xf 
'central point (f) of undergrowth (xf) occurs.' 83 kinwi-piikwa, f-xf 'long 
narrow area (f) of brush (xf) occurs, brush extends along.' 83 ni-po^ki- 
ce-el-aw-a, s-f-frg-mf-t-o 'I caused breech (f) in body (frg) with projectile 
mvt. (mf) to him, 1 shot him in the body.' 83 po^k-iikwe, f-frg 'he has 
breech (f) in one-side-of-face (frg), he has one eye out.' 87 ni-poskwi- 
piye-en-a, s-f-frg-i-o '1 caused irregular fraction (f) of long extension (frg) 
by hand mvt. (i) of it (tree), I broke a limb of the tree.' 87 ni-poskwi- 
n?ke-'^U-m-a, s-f-frg-fcm-t-o 'I broke (f) his arm (frg) by motion coming 
to rest (fcm), by throwing him against something.' 91 ni-paalaci-\ve-l-a, 
s-f-mf-t-o 'I carry (mf) him downward (f).' 91 ni-pele-se-en-a, s-f-frg-i-o 
'I ripped the seam of it.' 92 meelawaaci-paam-'^Oe, ef-f-mf 'he was tired 
(ef) of running (mf) around (f).' 99 piyet-aalak-9en-wi, svf-f-xf-s 'it lies 
(is in xf) with hole (f) this way (svf).' 

Part II: 135 ni-tephikan-'^Oe-to, s-f-fcm-o '1 put it in jar-s.' 135 ni-tepi- 
kii^kwe, s-svf-ef 'I returned to consciousness.' 136 pa?-tepowee-ki, svf- 
ef-s 'they went to council (counseling).' 137 ni-tepeto-kalawi-pe, s-f- 
fcm-m 'they were talking (fcm) in a group, together (f).' 137 tepeto- 
ptoo-ki, f-mf-s 'they were running in a group.' 138 tetep-a^kwi, f-frg 
'rolling outline of flora' = 'grape\'ine.' 138 ni-waawiyaa-tap-^k-a, s-f- 
mf-i-o 'I rolled (mf) it in a circle (f) by kicking (i).' 141 ni-me^ci-tehe. 


s-svf-ef 'I ha\e thought.' 149 ni-Oaki-caalee-pi-l-a, s-f-frg-fcm-t-o '1 cause 
holding-outhne (f) on nose (frg) with tying (fcm), 1 put halter on him.' 
Part III: 289 kape-ho-kwi, fcm-xf-s 'he crosses floating.' 293 kotekw- 
aakami, f-xf 'water (xf) is in a winding (f) channel.' 295 ni-kakaanwi- 
lece, s-f-frg '1 have long hands.' 297 ni-kooki-tepe-en-a, s-f-frg-i-o 'I 
dipped his head (frg, filling of the immersion-figure) in the water.' 
298 ni-Kki-lecee-pi-l-a, s-f-frg-fcm-t-o 'I put (binding action, fcm) a ring 
(circular outline, f) on his finger (frg). 300 ka§ko-He, f-frg 'he has sharp 
ears.' 300 ni-kisw-eele-m-a, s-ef-ef-t-o 'I regard him as worthy, permit 
him.' 301 kiisoo'kwaam-wa, ef-ef-s or svf-ef-s he 'sleeps warmly.' 
303 ni-kilek-a-m-a, s-f-fcm-t-o 'I caused mixed configuration (f) with mvt. 
in mouth (fcm), I mix it in my mouth.' 304 kolep-sin-wa, f-fcm-s 'he 
turned over lying down.' 306-307 ni-kaawat-eele-m-a, s-theme-ef-t-o '1 
think of him as round' [kaaw-at- theme, hence can precede ef?). 308 ni- 
kw^asko-l-aw-a, s-fcm-mf-t-o 'I knocked him down by shooting him.' 
308 ni-kwaskwi'tepe-en-a, s-fcm-frg-i-o (irregular order but fcm could be 
considered f) 'I pushed his head away.' 310 saapot-aalakat-wi, f-frg-s 'it 
has a hole through to the other side.' 319 9aak-ho-Wen-wi, f-xf-xf-s 'it 
is partly immersed sticking out of the water.' 320 ni-Oak-aalow-een-a, 
s-f-frg-i-o 'I caught him by his tail.' 


The Maya were the only fully literate people of the aboriginal Amer- 
ican world. The buildings and monuments of stone that they left 
are covered with their writings— writings of which little has yet been 
read except the dates with which they begin. Moreover, they wrote 
many books and manuscripts, and three such books of fairly late period 
have been preserved. These are the famous three Maya codices, and 
I propose, before the end of this paper, to read a very brief extract from 
one of them, and to show, in a very plain and simple way, what the 
Maya writing system was like, and how its signs were put together. 

Included in this writing system is a group of signs and combinations 
of signs referring to a special kind of subject matter. These are signs 
denoting numerals, periods of time, and terms of the calendar, between 
which mathematical relations exist and the use of which constitutes a 
system of mathematics. The mathematical references of these signs 
have been determined from these mathematical relations that are ob- 
served to exist between them, and thus we can read the dates and the 
positions of the solar-lunar calendar that are recorded at the beginning 
of most inscriptions. Besides this mathematical record, there is the 
purely linguistic portion of the writings, between the parts of which we 
can observe grammatical or linguistic relations, but no mathematical 

* Reprinted from pp. 479-502 of the Smithsonian Report for 1941 (Washington: 
Government Printing Office, 1942). This paper was read before the Section on 
Anthropological Sciences of the Eighth American Scientific Congress, Washington, 
D. C, May 10-18, 1940. 



relations. These purely linguistic portions are those with which I shall 
deal. I shall deal, moreover, with the writing in the codices, not that of 
the inscriptions, though the inscriptional writing is generally similar to 
that of the codices. It may surprise many to know that, in the codices, 
the nonmathematical, linguistic signs outnumber the mathematical ones 
by more than a hundred to one (not counting repetitions of the same 
sign). So much for the belief that the Maya writings are mainly 

When Champollion began the decipherment of Egyptian writing, he 
was in the relatively fortunate position of not having to oppose an 
extensive body of established doctrine holding that the markings were 
not writing but a nonlinguistic symbolism. To be sure there were the 
fantastic speculations of Athanasius Kircher, concerned wholly with the 
religious and mystical symbolism which he read into the hieroglyphs, 
but these were upheld by none of the scholarly disciplines and quickly 
went down before Champollion's irrefutable logic. At that time the 
philologist and literary scholar reigned supreme in the study of ancient 
cultures. Champollion therefore had only to prove the linguistic logic 
of his results to philologists; he needed not to advocate his methods to 
archaeologists, for there were none, except philologists. There was not 
then the specialized separation of disciplines which prevails now. At 
that time philology led the way, read inscriptions, and stimulated 

It is popularly supposed that the success of Champollion's effort was 
wholly due to the discovery of the Rosetta Stone with its bilingual 
inscription and that there is nothing corresponding to the Rosetta 
Stone in Maya hieroglyphs. Both suppositions are wrong. Champol- 
lion would have ultimately succeeded without a Rosetta Stone, for the 
inscriptions happened to be in a language that he knew. He knew 
Egyptian, that is Coptic, the late form of the language and still essen- 
tially the same tongue, which the ancient Egyptians spoke and wrote. 
Just so the Central American writings happen to be in a language that 
it is possible to know. They might have been in a dead language, and 
then the case would indeed have been difficult, but fortunately they are 
in Maya, which is still spoken and can be studied from many sources. 
But how do we know they are in Maya? This will be quite clear to a 
linguistic scholar, who appreciates that, if texts in an unknown char- 
acter are in a language that he knows, it is likely that he can detect that 


fact from the nature and frequency of repeated collocations of signs. 
In addition, the meaning of various clusters of signs in the Maya system 
is known from tradition (e.g., the glyphs of the months) and others from 
pictures that accompany them in the codices. The hieroglyphs record 
a language in which the writings for a certain month and for 'sitting 
position' begin with the same sign, which is the image of a feather. 
This condition is satisfied only by the Maya language, in which the roots 
of these particular words and the root of the word 'feather' all begin 
with the same syllable. Again, it is a language in which the writings 
for 'snake, fish/ and a certain time period all begin with the same or 
with mutually interchangeable signs, a condition also satisfied by Maya. 
It is a language in which the writings for 'honeybee, earth,' and the 
name of a day begin the same, in which 'hold in the hand' and 'nothing' 
begin the same, in which 'spear' and 'noose' begin with the same sign, 
which is also found in the clusters that mean 'jaguar,' 'nine,' and 'lunar 
month,' and so on. The evidence mounts and becomes at last over- 
whelming. Not even Cholti or Tzeltal, the languages closest to Maya, 
can satisfy the requirements; only Maya can do so. 

There exists also a lesser equivalent of the Rosetta Stone, i.e., the pre- 
served names of the ancient months and other calendar terms with the 
sign clusters for writing them, the ways of writing the numerals, the 27 
characters recorded by Bishop Landa, the sign clusters for the cardinal 
directions, the colors, quite a number of animals, and various gods— a 
collection of odd bits that, when gathered together, make a not incon- 
siderable total. Finally there are many texts in the codices in which the 
meaning is almost as plain as though a translation ran beside it, because 
of the detailed pictures that run parallel with the text and illustrate it. 
Thus we really do have a Maya Rosetta Stone, as well as a knowledge 
of the language of the texts, so that, given linguistic scholarship like that 
of Champollion, it is perfectly feasible to decipher and translate some of 
the texts now, and eventually all of them. 

But, on the other hand, the linguistic decipherer today has to contend 
with the chasm that now exists between American archaeology and 
philology. The philological viewpoint, with its scholarly interest in 
texts simply as texts, has become rather strange and incomprehensible 
to modern American archaeology, with its high development, along the 
scientific side, of the logical correlating of strictly material evidence, the 
while its popular side and its financing is largely connected with the 


esthetic interest, and with the interest that attaches to concrete human 
subject matter, particularly that of an exotic kind. Now the linguistic 
and philological interest is to be distinguished both from the materially 
and physically scientific interest and from the estheticohuman one; for, 
while it is not entirely divorced from either, and it cannot live in a 
vacuum, yet it finds its main concern upon a different level, a level of 
its own. The linguistic scholar is interested in a text as the monument 
of a language arrested and preserved at a certain point of time. He is 
not primarily interested in the subject matter of the text, as either his- 
tory, folklore, religion, astronomy, or whatnot, but in its linguistic form, 
which to him is the supreme interest of interests. From this proceeds 
his type of objectivity, an earnest that his reading will not be affected 
by theories concerned with the content of the writing. He puts aside 
content to concentrate on linguistic form. He aims to reconstruct the 
language as it actually was, with its consonants and vowels in their actual 
places in words, its paradigms of declension and conjugation and its 
patterns of syntax, thereby adding a new body of facts to the whole 
domain of linguistic taxonomy. A by-product of his research is the 
reading of history and culture, but it may be questioned if his discovery 
of strictly linguistic fact in a time perspective is not the more important. 
The decipherment of Hittite has proved to be far more important for 
the light it has thrown on the development of the Indo-European lan- 
guages than for all the accounts of Hittite reigns and conquests. The 
battles and politics of the Hittites are as dead as a nail in Hector's coffin, 
but their verb forms and pronouns and common words are matters of 
live interest in American universities at this moment, since the accurate 
facts of the Hittite language revealed by careful decipherment are com- 
pletely revolutionizing our concepts of Indo-European linguistics. This 
authoritative knowledge of Hittite could not have come about if the 
deciphering scholars had not been linguists who had slowly and care- 
fully ascertained, by scholarly methods, with profound respect for the 
text as a text, the exact words and grammar, conceiving this as their 
paramount duty. It could not have come about if they had conceived 
their duty as that of reading off a sweeping survey of Hittite history and 
culture, or even of clothing the dry bones of archaeology with the flesh 
of human narrative, important as these things are. 

lire desiderata for Maya decipherment are no different. Reading 
Maya texts must be a slow, careful investigation of linguistic forms, 


regardless of the interest or lack of interest of their subject matter. We 
must not conceive it our task to read off sweepingly the Maya literature 
for the sake of the information on histor\', culture, religion, or whatever 
else may be contained in it. The annals of this subject are cumbered 
by such attempts to read off or "interpret" the whole corpus of the Maya 
codices at one fell swoop, from Brasseur de Bourbourg to one very 
recent such an attempt. Such amusements proceed from a longing for 
glamour and quick results, misconceiving what is the most valuable 
thing to be obtained from the results. On the other hand, much of the 
work of Cyrus Thomas and various bits of linguistic data pointed out 
by Morley and others have been at least in the right direction— they 
seem to have understood what the problem really is. 

The Maya writing system was a complex but very natural way— natu- 
ral to minds just beginning to exploit the idea of fixing language in 
visual symbols— of using small picture-like signs to represent the sounds 
of fractions of utterances (usually of a syllable or less in extent), com- 
bining these signs so that the combined fractions of utterance outlined 
the total utterance of a word or a sentence. Past study of this system 
has been considerably retarded by needless and sterile logomachy over 
whether the system, or whether any particular sign, should be called 
phonetic or ideographic. From a configurative linguistic standpoint, 
there is no difference. "Ideographic" is an example of the so-called 
mentalistic terminology, which tells us nothing from a linguistic point 
of view. No kind of writing, no matter how crude or primitive, sym- 
bolizes ideas di\orced from linguistic forms of expression. A symbol 
when standing alone may symbolize a "pure idea," but, in order to rep- 
resent an idea as one in a definite sequence of ideas, it must become the 
symbol for a linguistic form or some fraction of a linguistic form. All 
writing systems, including the Chinese, symbolize simply linguistic ut- 
terances. As soon as enough symbols for utterances have been as- 
sembled to correspond uniquely to a plainly meaningful sequence 
(phrase or sentence, e.g.) in the language being written, that assembly 
of signs will ine\'itably convey the meaning of that linguistic sequence 
to the reader native to that language, no matter what each sign may 
symbolize in isolation. Meaning enters into writing, writing of any 
kind, only in this way, and in no other. The meaning of any linear or 
temporal succession of symbols is not the sum of any symbolisms or 
denotations that the symbols may have in isolation, but is the meaning 



of the total linguistic form which that succession suggests. Hence the 
fact that some individual signs look like pictures of the things or ideas 
denoted by the words of the utterance plays no real part in the reading; 
those signs are just as much symbolic, learned, and at bottom arbitrary 
signs for fractions of utterance as any other characters or letters. On 
the other hand, resemblance to an object or picture may be really im- 

Number sound Symbol FVoboble 


2 b 

3. e 

4 h 

5. hw 

6 -u 


8 ka 

a kok 
' ka 

10 k<a 

11 kum 


rTODooie ftu/c* Name 

Object -io^rce o« Object Source 
scraper hcjob 

'c«p./lo,o qual<^uteT- 

instromento poro Tospor ' 

perfovotions bis 

'pe^u«»ioA o^ujereodos* 

pol-nts, dots e 


ope»MBg,doer he 

'obrtr como pvertai * 

(<oce of) chief ahaw 

letter, book hwun 

'corta o libro' 

VI pplesCoi animal) im 
'teta de -mujer y 
de Cfual^u'tor anivnol ' 

pan kot 


Tia 8 exIorqCcJ o-nd 
doubled ( ?) 

^li'd tied oT> Kal 

*<er rtjT" con cemaduna, 
obT"ochaT; atvancaT*' 

fcother KuKum 

'p'umo de ovc' 

Number Sound 
U lie 

13 1. le 

K l,I".lo 
IS'"! wa 





la s, so 


22 f 00 


fVoboble. Moyo Name 

Object-Source of ObjectSou-rv 

loop.Ttooie ]g 

* loz.0 pora color* 

double loot> )e 

(No. 12 dovbled; 

. drinkin^-cupC&loop) lot 
'voso psro. beb€T- ' 

grpsp of the hand mal 

* o3ir, tovnor co_n loij 
Tnopos, o emp«noT^ * 


woven work "Ti loom sakol 
'telex e-n el telot-' 

stretched strings sin 
ettender poHai e cueroj^ 'itrct<h 

cel^or ct(«iidt«ndo J er^trin^* 

ormor- logos' 

rood-crost ings iay 

'encTvcijoda de tamino* 

floTties C?) took 

'qooTnaT"' 'born* 

face of deq f?) j«,*^, 
'perro liso iitt pelo ' •^oirlMl ^09' 

Craitant Woon ? U 

•luna* 'moon* 

Figure 3. Examples of Maya symbols having phonetic values. 

portant in decipherment, as a clue to how the sign came to be invented, 
to the logic of its original use, and hence to the fraction of utterance, 
i.e., sound, which answers to it in reading— a clue to be tested by how 
well that proposed fraction, or sound, fits into each proposed reading. 
Figure 3 shows 23 symbols selected out of the several hundreds found 
in the whole Maya literature. These particular ones have been chosen 
because they enter into the written words and the codex sentence used 
as examples of decipherment in this paper. The fractions of utterance 
to which these signs regularly correspond have been identified by com- 
parative evidence— running back ultimately to that body of evidence 
which I have called the Maya Rosetta Stone. Signs 1, 2, 3, 7, 8, 12, 


17, 22, are also given by Landa with the same values (1, 7, 12, 17 being 
shghtly altered in form) in his book Relacion de las cosas de Yucatan, a 
first-hand account of the Maya shortly after the Conquest. The left- 
hand column shows in alphabetic order the fraction of utterance, i.e., 
sound which regularly ^ corresponds to the appearance of the sign in a 
written form. The next column to the right shows the usual appearance 
of the written sign, with common variants added in some cases. The 
list includes less than a third of all the signs the phonetic values of 
which I consider fairly well established. The column headed "Probable 
Object Source" names the thing or condition of which the written sign 
was probabl)' at one time a picture. However, these theories of pictorial 
origins, while they seem probable and have a substantiating value, are 
not the evidence for the phonetic values, and their being proved wrong 
would not invalidate the latter nor alter the readings, but would merely 
mean that the origin of the sign was other than 1 have supposed. There 
are several signs for which I am unable to offer any explanation (e.g., 
no. 16), yet for which the phonetic value is reasonably certain. I did 
not guess the probable object source of no. 6 until after I had known its 
phonetic value for several years. 

The extreme right-hand column shows the Maya word, as given in the 
Motul dictionary',- for the thing or condition postulated as the object 

1 Regularly but not always in the case of all these signs, for polyphony is a prevalent 
trait in Maya writing, as it is also in Sumerian and Akkadian (Babylonian) cuneiform. 
That is, various signs are polyphones, with two or more contrasting sets of sound 
values, besides the slightly differing values within a set, such as either ha or h with 
vowel lacking or indefinite, which slight differences are on another level than the 
polyphonic contrasts. The native reader, able to grasp words as wholes, is not con 
fused by these polyphonic values; he knows from the other signs assembled with the 
one in question just which of the polyphonic values applies in a given case, just as 
the reader of English is not confused by the 'o' in 'women' or the 'olo' in 'colonel,' 
but is governed by the total collocation so that he reacts with fractions of utterance 
entirely unlike those regularly associated with the written forms o and olo. Polyphony 
is therefore the same type of thing as irregular spelling under an alphabetic system 
of writing. Thus the Maya sign no. 5 of Fig. 3 has also the value la, I, as in the 
writing of the word lak'in, lik'in 'east'; this value may very likely derive from the 
word Mail 'the largest, greatest, principal, chief — a near synonym of d/idw. Sign 
no. 15 occasionally has the value c, as in the writing of cik'in 'west'; this value prob 
ably derives from cuk 'catch or seize with the hand,' a near synonym of mac. 

2 The Motul dictionary is an anonymous sixteenth centur\' work ascribed to Fray 
Antonio de Ciudad Real, and is the most voluminous and authentic source of infor 
mation on the Maya language at the time of the Conquest. Actually it is not only 
a dictionary' but a grammar and a chrestoniath\- ns well, for ninst of the word citations 


source. It will be observed that the initial sound of this Maya name of 
the object (i.e., the first consonant and/or the first consonant and 
vowel) is the sound which the sign represents in writing, as shown in 
the left-hand column, except in the case of no. 1, in which the initial h 
is either lost or transposed, yielding a or ah. The Spanish entry under 
the English name of the object source is the way in which the Motul 
dictionary defines the Maya word in the extreme right-hand column. 

Figure 3 then should be self-explanatory. The following supplemen- 
tary remarks may be added: number 1 does not occur initially in a word. 
Primary word-initial h in Maya, in becoming secondarily word-internal, 
as when it begins the second member of a compound word, tends to be 
weakened or lost. This explains why a syllable originally denoting ha 
would denote a when used only to write noninitial fractions of words. 
Number 6 is especially interesting. Maya has simple, unanalyzable 
words for 'write' or 'book,' not connected with 'paint' or 'draw' as in 
Aztec and many other American languages. This fact, ceteris paribus, 
argues for the greater antiquity of writing in the Maya culture than in 
these other cultures. Maya missi\es and books (e.g., the codices) were 
written on an elongated strip of tissue which was then folded up, and, 
when tied or clasped, would have an appearance not unlike a modern 
letter sealed in its envelope, or like no. 6.^ The nipple (im) sign for i 
appears in the codices usually with three nipples, which leads me to 
think that the teats of a deer or other animal may have been one of the 

are accompanied by copious examples of phrases and sentences. The technique of 
stem composition in Maya of this period is beautifully brought out in these examples; 
the same is true of syntax. The Maya words in Fig. 3 are not cited in the conven- 
tional Maya orthography used in the Motul dictionary, but in the phonetic alphabet 
used by most present-day linguists for American Indian languages (the revised Ameri- 
can Anthropological Association system), except that g is used instead of c for the 
alveolar affricate (a sound like ts). The cedilla has been added to the c to avoid con- 
fusion with the c of Maya orthography which represents k. Symbol 22 is cited by 
Landa with the value c; it is unquestionable that he meant Spanish g or the soft 
sound of c, as in the name of the letter "ce," which is very likely what he asked his 
Maya informant to write. This soft sound of c was close to ts in old Spanish, which 
is why it was equated to the Maya sign for ts, no. 22. The sounds c and s are Eng- 
lish ch and sh, k' is a glottalized k; the language has a scries of such glottalized 
sounds: p', t', c', c', k'. Through some curious omission, the Motul dictionary does 
not actually cite the word ne, 'tail,' but this is, of course, a well known Maya word. 

3 As may be inferred from this, I regard the previous theories about what no. 6 
represents, one of which calls it a kernel of maize (to which it has no resemblance), 
as fanciful. The fact that in some Maya pictures corn plants may sprout from char- 
acters of writing, and characters may take part in the scenes like persons or objects, 


original forms; sometimes it appears with two; Landa shows it with two, 
and the sign of the day Ik [ik') may be based on an original human 
breast form with only one. Number 8 probably represents a kat, an 
earthen, basketry, or wooden pan, tray, or low fiat tub, often boat- 
shaped; it was also called a cem or 'boat' (see Motul, chem licil ppo and 
chem che), and conversely a boat may have been called a kat. The 
combhke lines may be the conventionalization of a fluted rim or of 
projecting basketr>' withes, or may represent people in a kat in the 
meaning 'boat.' Number 10 is an example of the many perspective 
drawings found in both Maya art and the writing symbols— a rounded, 
flattened pot, basket, or calabash with a k'al, a tied-on or clasped lid or 
cover. The Maya, as is well known, drew in perspective from very early 
times. Number 11 is a k'uk'um, 'feather' or 'plume,' and in this word 
k'um was probably felt to be the true initial form of the stem and fe'u- 
a reduphcation, which may not have been historically the case, but 
which would be felt analogically in a language like Maya in which 
initial reduplication is a derivational process in wide use. Nothing as 
yet is postulated as to the object source of 16, a profile head with a sort 
of parrotlike beak; a suggestion here would be the parrotlike bird called 
moan or muan. Tlie sign corresponds to the consonantal sequence nin, 
with any or no vowel inter\ening, and as a day sign denotes the day 
Men. Number 23 looks very much like a form of no. 1, but it is always 
upright and placed in front of a sign cluster with its concave side 

is secondar)' symbolism, not the original logic from which the character arose. All 
this elaborate secondan. symbolism, perhaps religious and magical in large degree, has 
nothing whatever to do with the reading of the characters in their capacity of symbols 
of writing, any more than the elaborate symbolism and numerology that grew up 
around the Hebrew letters in rabbinical tradition affects the reading of the Hebrew 
text by one jot. This secondary symbolism may eventually become a matter of philo- 
logical literar}' study, wherein it will very likely prove important. At present, and 
from a linguistic standpoint, clearing away all this sort of symbolism is essential to 
understanding the proper symbolism and function of the Maya signs in writing. The 
use of no. 6 to denote the day Kan is a writing of the original name of the day Hu — 
i.e., 'lizard, iguana' (cf. Aztec Cuetzpalin, 'lizard,' for the same day). All the original 
names of the days, except for Ik, Cimi, Caban, and perhaps Manik, Cauac, and 
Eznab, and one or two more, became changed under the Maya culture after the 
establishment of the writing system. Some of the days continued to be represented 
by the initial letter or character of their original names, much as we write 'lb.' for 
'libra,' but read it 'pound.' The voluminous speculations of Seler concerning the day 
symbols are to be taken with a great deal of caution, if they are not indeed stumbling 
blocks of the worst kind. 



toward the cluster, while no. 1 is not placed in front of a cluster and 
is usually horizontal. Number 23 corresponds to initial u of a word or 
to u as a separate word or as a prefix. 

Figure 4 shows the writing of six words occurring in the codices.* The 
sign clusters or glyphs of various animals, originally determined by 
Schellhas from their concurrence with pictures, have long been known.^ 

kaT) "snake" 






I Q (J a 7? a 


Tia-me of a -montpi 


kuTM - hu 


"stabbed , speared " 

/ e ' -rioose. ', an d 
"catch IT) noose -'trap' 

les iri cf h "catcU 

i-n st-rvng noose -trap'' 

'siqno de CQceyi^a 
P o-r -medio d e 
f \e cha y la-nz-Q " 




5iq-no de CQcevio 
po>" xVa'mpa 

le "cogev por /a-zo" 
sm "arina-r Iflzos" 

Figure 4. Maya sign clusters representing words. 

Number 1 is cited by Schellhas as the glyph meaning 'snake.' It will 
be noted that it consists of no. 8 of Figure 3, ka, and no. 17, n, and a 
third symbol. This third symbol and the iguana figure in the next glyph 
of Figure 4 are the only symbols cited in this paper which are not found 
in Figure 3. The first two symbols spell kaii, which is the Maya word 
for 'snake' The third s\nibol is probably derived from a picture of a 

■* In an unpublished paper read before the annual meeting of the American An- 
tliropological Association at Washington, D. C, in December 1936, entitled "A com- 
parative decipherment of forty-six Maya written words," I exhibited 46 word-writings 
similarly analyzed, including hu and kumhu of the present six. 

5 Paul Schellhas, Gottergestalten der Mayahandschriften, 1897. 


rattlesnake's rattles, intended to evoke the linguistic response "snake," 
i.e., kan, and has itself the value kan. However, it is apparently insuf- 
ficient by itself to write the word kan. It was not usual in the Maya 
system to write a word of one syllable simply by one sign having the 
value of that syllable, probably because that sign often was polyphonic, 
having other values. Instead, the Maya method was to suggest the 
syllable by a combination of signs that, to Maya speakers acquainted 
with the conventions of the writing, was probably unambiguous. This 
combination of signs could be made according to two principles: 
(1) synthetically, building the syllable from signs to be understood as 
fractions of the syllable, which together made the whole syllable; (2) by 
repeated affirmation, that is, by combining, in the sense of repeating, 
different ways of denoting the whole syllable. A word of one syllable, 
or often a syllable within a longer word, could be written by either 
method, or by both together, as in the case of this writing of the word 
kan. The signs ka and n build the word synthetically, the sign kan re- 
peats it; we have double writing, but only single reading. It is as if 
the writing said "my first is ka, my second is n, my whole is one of the 
values of the snake-rattle sign, and so must be kan." The combina- 
tion is, by sum of all its parts, ka-n-kan, but we may use the convention 
of transliteration ka-n'^"" to show that the final kan is a doubling in the 
writing only, not in the reading. 

Number 2, Figure 4, is the sign cluster meaning 'iguana,' or 'large 
lizard,' a meaning which is quite obvious, since it accompanies plain 
pictures of that animal, besides containing such a picture itself. But 
this one picture-like sign, no matter how much it may look like the 
animal, is not sufficient by itself to write the word meaning 'iguana.' 
The Maya system, as already noted, requires combination with at least 
one other sign before we can have a unit of writing, capable of standing 
alone. The exceptions to this rule form a very restricted list indeed, 
the most important ones being the 20-day signs, which are single ele- 
ments enlarged to the size of a full cluster and capable of standing alone. 
The month glyphs and calendric and mathematical glyphs, in general, 
conform quite to the rule, being clusters of signs. Number 2 writes the 
one-syllable word hu 'iguana' entirely by the method of repeated af- 
firmation, using the ordinary sign for /zu, no. 6 of Figure 3, topped by 
an iguana figure, which of course has the linguistic value of the animal's 


name. Here the formula which we use in transliterating is hu-hu, to be 
read or pronounced, of course, as "/lu." 

Number 3 writes the word kumhu, the name of a Maya month, en- 
tirely by the synthetic method. It is the well-known glyph of this month 
Cumhu as found in the codices. It uses the feather sign kum, no. 11 of 
Figure 3, plus hu, no. 6; so we transliterate kum-hu. Some other words 
of the codices using the sign kum, no. 11 of Figure 3, are kumah, the 
stem "sit" with transitive suffix meaning 'seats' or 'carries seated/ and 
kumag, another word meaning snake (cf. Quiche kumag 'snake'). Al- 
though we are still somewhat in doubt concerning the values of the 
vowels in these words, the general phonetic contour is interestingly con- 
firmed by the fact that the codices write kumah not only as kum-ma 
(with 1 1 and 1 5) but also as kw-m-a, while Landa cites a way of writing 
the month Cumhu which is the cluster of kw-m-hw; in both of which 
writings kw and m are signs not included in Figure 3 (but confirmed by 
other evidence) while a is \ and hw is 5 of Figure 3. 

Number 4 of Figure 4 occurs in texts of the Codex Tro-Cortesianus 
dealing with hunting and illustrated with hunting pictures. It is ob- 
viously a sign cluster or word referring to animals killed by spears or 
arrows, and the commentary in the Villacorta edition ^ of the Tro- 
Cortesianus calls it "signo de caceria por medio de flecha y lanza." It 
is a writing composed synthetically with doubling of one subsyllabic 
sign. At the top is the cup-and-loop sign lu, lo, no. 14 of Figure 3, 
written within the outlines of no. 15, m, ma, which is doubled, the 
lower member of the doubled pair enclosing the tail sign n, no. 17 of 
Figure 3. When we find doubled a sign which according to the total 
setup is probably to be interpreted as a syllabic confirmed by a sub- 
syllabic, we may transliterate without the convention of writing a super- 
script, using instead a convention that permits of possible interpretation 
as a long consonant or vowel, e.g., in this case not ma-ma but m-ma. 
Number 4 is then transliterated lu-m-ma-n or lo-m-ma-n, which is a word 
meaning exactly what the accompaniment of pictured scenes tells us. It 
is the passive participial inflection in -an of the stem lorn, which means 
a 'spearing or stabbing thrust or blow,' and by extension a 'spear,' while 
with the verbal inflection it denotes the occurrence of a spearing 
action. The Motul dictionary' gives "lom : tiro de lanza, o dardo, y 

^ J. Antonio Villacorta C. and Carlos A. Villacorta, Codices Mayas, published in 
Arqueologia Guatemalteca, 1932. 


cosas assi, y estocada, o puiialada." This stem with the transitive verbal 
inflection is given by the Motul as "lomah, oh : fisgar, o harponear, dar 
estocada o puiialada, alancear y aguijonear," this citation being followed 
by that of the passive participial form, ''Ionian : cosa que esta assi fis- 
gada." Hence this word loman written in the hieroglyphs of the Maya 
text means 'speared, stabbed; pierced, wounded or killed by a spear, 
arrow,' etc. 

Number 5 of Figure 4 is synthetic with doubling of the inherent vowel 
of one sign. It is common in the hunting section of the Codex Tro- 
Cortesianus, and is obviously the word denoting catching of animals 
by a noose or lasso, or in a noose snare— a trap consisting of a noose 
set to spring by a stretched rope triggered and attached to a small 
bent-down tree so that, when the animal steps in the noose and re- 
leases the trigger, the tree springs back, drawing the slipknot of the 
noose and catching the animal. The glyph or sign cluster no. 5 accom- 
panies pictures of this operation, e.g., Tro-Cortesianus 42c. Villacorta 
calls it "signo de caceria por trampa." It consists of the double loop or 
knot sign I, le, no. 13 of Figure 3, and the dot sign e, no. 3 of Figure 3, 
and is to be transliterated le-e and read le 'loop, noose, slipknot, noose 
trap or snare,' Motul ''le : lazo para cazar y pescar, y pescar con lazo," 
with the verbal inflection, e.g., leah meaning 'catch or trap with noose 
snare,' for which the Motul gi\'es the participial ''lean : cosa enlazada 
o cogida en lazo." Here again we see the principle that a sign is inade- 
quate by itself, in that no. 13, though itself derived from the picture of 
a slipknot or noose le and denoting the sound fraction le, is not suf- 
ficient alone to write the monosyllabic word having this sound, i.e., le 
'noose,' but is subject to the rule that a sign must be combined with 
another and cannot stand alone. Here it has its inherent vowel re- 
aflfirmed by attachment of the sign e. Hence there is a mixture of the 
synthetic and the repeated-affirmation principles in sign clusters or 
glyphs of this type. We also find the verbally inflected form leah 'catch 
with noose,' written le-e-a, with no. 1 of Figure 3 for a. Cyrus Thomas 
correctly analyzed the le-e 'cluster,' I believe, though 1 worked it out 
without referring to his work. A number of Thomas' readings are un- 
doubtedly correct. 

In no. 6, Figure 4, we have one of the polysynthetic words common in 
Maya, in which two stems are compounded and suffixes attached. It is 
illustrated in Tro-Cortesianus, page 46, by three pictures showing vividly 


in successive stages of action a deer caught and jerked upward by the 
spring of the bent tree to which the noose of the trap is attached. It is 
written le-e-sin-a (or -ah), with signs 12, 3, 19, and 1 of Figure 3, and is 
to be read lesinah. This word is typical of a common kind of Maya 
compound, consisting of two stems with the verbal inflection sufExed 
after the second. The stems are le, already defined, and sin 'stretch or 
string tightly (as cloth, hides, or cords are stretched on a frame), draw 
taut, string with stretched cords, string up, string or rig a noose trap 
or the like to spring when released,' etc. The Motul gives "zin (i.e., 
sin) : estender panos o cueros y colgar estendiendo o tender desarru- 
gando; armar lazos; armar arco o ballesta." Such a compound usually 
has the following type of meaning: designating the two stems as X and 
Y, a compound X-Y-ah or X-Y-t-ah ^ means do X by means of Y, transi- 
tively, or to an object. Thus, since le-ah means catch in a noose, we can 
form freely words such as le-k'ab-ah (or more modern le-k' ab-t-ah) 'catch 
in a noose by action of the hand' {k'ab 'hand'), le-k'as-ah 'catch in a 
noose by a tying action,' and so on. Our word le-sin-ah then means 
'catch in a noose by the action sin or catch in a noose by tight stretch- 
ing, catch by the spring of a tautly strung noose trap.' ^ 

Now, having noted the reading of a few individual words, let us read 
a short sentence written in Maya hieroglyphs. Figure 5 shows page 38 
of the Codex Tro-Cortesianus, and the sentence thereon to be examined 
in particular is that made by the four sign clusters or glyphs over the 
second seated figure in section b, the middle of the three horizontal 
divisions of the page. Figure 6 shows this sentence written on one line, 
analyzed, transliterated, and translated. As can easily be observed from 
Figure 5, the texts which comment on the pictures, or, to put it the 
other way, which are illustrated by the pictures, are placed over the pic- 
tures, reading from right to left across the width of the picture, and then 
on the line below similarly; or they run vertically downward in the cases 
where there are no pictures. This order is easily demonstrated from the 

7 ITie form with the suffix -t- before the suffix -ah is the common form in Maya of 
the Motul dictionary for binary compounds of this type. 

8 We find in the codices other compounds of this type, including some others with 
sin as second member; thus in the Tro-Cortesianus (e.g., 41d) the picture of a deer 
trussed up in a bundle, legs folded up, with cords lashed around it, is accompanied 
by the sign cluster ma-sin-a (with Landa's ma sign), to be read probably massinah, 
assimilated from macinah (compound of stems mac and sin), meaning 'clasp together 
(like a clasped fist) by pulling and tension, by tight stringing, by tightly drawn cords.' 

Figure 5. Page 38 of the Codex Tro-Cortesianus. 


parallelism of the writing; we have here plainly a repetition of very 
similar short sentences or clauses. Thus, if we give a letter to each 
cluster or glyph which is the same, the middle-section text over the first 
or left-hand picture runs A-B, and then on the line below C-D, next to 
the right running straight downward we have A-B-E-F, then over the 
next picture A-B-C-D again, then downward again A-B-G-H. The 
texts of the top and bottom sections can be seen to run in the same 
manner, which indeed is general throughout the codices. The texts 
would seem to be in a style which is common enough in aboriginal 
American songs, chants, and ceremonies: sets of phrases containing a 
constant element repeated throughout a set, as when each line of a song 
stanza begins the same way but then introduces a certain difference. 
Thus the text which we have just examined consists of lines each be- 
ginning A-B and then becoming different. Navaho chants are of 
course typical cases of this sort of thing. In the top section, dealing, as 
the pictures show, with hunting by means of the spear, each clause 
begins with the word loman 'speared' that we have already studied. We 
shall not pause however to analyze this top section in detail, since the 
limits of this paper do not allow it. 

The middle and bottom sections are verj' similar to each other, 
though not identical, and deal with drilling, as can be seen from the 
pictures. The pictures of the middle section show the using of the 
drill to make fire; the bottom set shows the drilling of an object which 
appears to be a stone. Each clause in each section begins with the word 
for drilling or drill, as is evident not only from comparison with these 
pictures, but also from one of the other Maya books, the Dresden 
Codex, in which the same sign cluster accompanies pictures of drill- 
ing. This cluster, A, occupies first position, which is the regular position 
of the predicating word of a clause in Maya of the sixteenth century (if 
not also today) as shown by the hundreds of short simple sentences in 
the Motul dictionary. This predicator need not be a formal verb in 
Maya grammar (though it most often is), but it is what corresponds to 
the predicate in an English translation. The final two words of each 
clause, C, D, • • • etc., are the well-known name glyphs of the Maya 
gods. They are the names of the persons shown in the pictures, as has 
long been known, and consequently they are undoubtedly the gram- 
matical subjects of the clauses. The second cluster of each clause may 
be called Bi in the middle section, Bo in the bottom section, to indicate 



that it is the same throughout each section but differs between the two 
sections. By eHmination and by position after the predicator, it should 
indicate the grammatical object and/or result of the verb action, which 


5cript ) 



U-to-kak i-§-Tn-n-a ka-haw 

\J "to k - ka k i § a TD n a ka ahaw 
tronslation [Causes byl his burning-fire Itzamna our lord 

reconstruction hasesoh 

haxezah u tooc kak Itzanrna ca ahau 


stcTns: hai 'drill* 'taladrar o 
vocabular/ aejvjeTsay taladTamdo ' 
of the haS UaU 'e-ncendei- luTnbre 

text J f-rota-ndo un palo con otro' 

"tok 'buT-n' 'cjuerrar' 
Mok 'fire' 'fuego o luwibre 
ijaTTmO. Tia-me of a god 

'^ ahaW lord r"ey, ogronsenor' 
u 'his' ka 'out-' -es- causative 

-oh transitive, rion-futore 


i I lust rat I on 

Figure 6. Analysis of a Maya sentence taken from page 38 of the Codex Tro- 


agrees with the fact that the drilHng is pictured with different objects 
and results in the two sections. Thus we have, as a first schematization: 

A, predicator or verb (driHing) 

Bi, B2, object and/or result (fire, stone) 

C, D, ' • • etc., subject (names of gods or persons) 

Figure 6 is a detailed exposition of the sentence over the second pic- 
ture of the middle section, which shows the Roman-nosed god of the 
codices, or god D, making fire with a drill. The top line is a copy of 
the text, arranged from left to right on one line, instead of on two lines 
as in the original. This line, like the original text, is in glyphic script, 
the form of writing used in the codices. It closely resembles the monu- 
mental glyphic style of the stone inscriptions, but is less ornate and has 


more rounded outlines. In both these styles the signs in a cluster are 
gathered into a tight bunch or cartouche, in which they are grouped 
in two dimensions, and there is only a \estige of linear order in that the 
front or extreme left-hand part of a cluster never stands for the last part 
of a word, and similarly the rear or right-hand part never stands for the 
beginning of a word. The signs in a cluster are usually in contact and 
often fused together or enveloped in the same flowing outline; they may 
be attached to the top or bottom of a central sign, or they may be one 
within another; i.e., one sign may serve as the frame or ground of 
another. In short, the putting together of signs is more like a heraldic 
device than like our kind of writing.^ But the reading of the signs is 
exactly as if they were written in linear order, although this order must 
be learned separately for each glyph and hence requires a separate and 
often prolonged study of each by the decipherer. 

Tlie second line from the top in Figure 6 shows the signs which com- 
pose each cluster regrouped in one-dimensional linear order. Such an 
arrangement I call open transcription or linear script, and there is some 
evidence that the Maya actually used such a form of script, though not 
in the inscriptions or in the codices that happen to have been preserved. 
Landa cites instances of the utterances ma in k'ati and elele written by 
a native informant in this manner," the signs delineated consecuti\ely 
from left to right and either close together or actually touching one 
another. It seems not unlikely that such a linear script may have been 
used by the later Maya for convenience in ordinary purposes, as the 
Egyptians used demotic, while the glyphic script would have been re- 
garded as more hieratic and ornamental and used for important books, 

3 It should be pointed out that, even in our kind of writing, i.e., the alphabetic 
kind, linear order of signs is not quite absolute in many systems, which contain ves- 
tiges of an older two-dimensional way of grouping. Thus, in the writing of pointed 
Arabic, pointed Hebrew, and Pitman shorthand, the vowel points are grouped two- 
dimensionally with the consonantal signs, not written consecutively with them in the 
order of actual utterance. In the Devanagari alphabet the vowel signs are fused two- 
dimensionally with the consonant signs, and the vowel z to be uttered after a con- 
sonant is actually attached in front of that consonant. Our own 'wh' is similarly 
written backward, being actually 'hw' — a special cluster of signs that retains an 
unusual order of positions. Some monograms and modern advertising placards also 
use two-dimensional groupings of letters. 

10 Diego de Landa, Relacion de las cosas de Yucatan. The first phrase means 'I 
do not want.' The second utterance is gibberish from the Maya standpoint, but, 
judging from the context, it evidently represents the informant's attempt to comply 
with a request to "write L-E, 'le.' " 


priestly writings, and inscriptions. Be that as it may, conversion of a 
passage of glypliic into open transcription is a device which is often 
helpful to the decipherer. It will be noted that all the signs in this 
passage are gi\en in Figure 3, so that, from this line of open transcrip- 
tion, the whole utterance can be read off in rough outline, as shown in 
the third line or transliteration. Since many of the signs can be in- 
definite as to vocalic timbre, even when they imply a preferred inherent 
vowel, the vowels of the utterance are here and there doubtful, although 
the indication of definite vowels is generally much better than in Egyp- 
tian or unpointed Hebrew. To a certain extent, but by no means wholly, 
the transliteration of vowels is based on sixteenth century Maya, which 
can hardly have changed radically in this respect since the period of the 
codices, probably not very many centuries earlier; and it is also based 
partly on comparative evidence from other Mayan dialects, a field of 
research which must of course go hand in hand with scholarly and philo- 
logical reading of the codices. But it must also be emphasized that the 
text itself contains unmistakable reference to many of its vowels; thus 
the signs a, e, i, u of Figure 3 are unambiguous in their indication of 
vowels, though the position of the vowel in the word may not always 
be clear. Thus we arrive at the transliteration, namely: 

h-s-e-sa u-to-kak i-g-mn-a k-ka-haw 

The position of the e in the first word is not wholly clear, since this e 
is written inside both the h and the s signs; and another possible trans- 
literation is h-e-s-sa or he-e-s-sa, to be read either hesesah or hessah, 
which would indicate that the stem which means 'drilling,' which is has 
in sixteenth century Maya, was pronounced more nearly hes in the 
dialect of the codices. At present more evidence would be needed to 
confirm this, and the reading hasesah seems preferable, the vowel a not 
being indicated in the writing but a reasonable reconstruction from 
Maya linguistic evidence. 

Under the transliteration is a reconstruction of the original sentence 
in the light of Maya linguistics, written in the usual Americanist pho- 
netic system, and below the translation of this is a repetition of the 
reconstruction written in the traditional Maya orthography. This is 
included in order that Maya students may see the sentence written in 
the way most familiar to them, though the use of this traditional spelling 
for linguistic purposes is not to be recommended and imposes a handi- 


cap, indeed may breed quite misleading notions in the minds of students. 
Thus we have for the reconstruction: 

Phonetic hasesah u-to • k-k'ak' igamna ka-ahaw 

Traditional haxezah u tooc kak Itzamna ca ahau 

Under the phonetic transcription is the literal translation: 'makes (or 
made) by drilling his burning-fire Itzamna our lord/ or in smoother 
English: 'Our lord Itzamna kindles (kindled) his fire with a drill.' 

The first word is a deri\ative of the stem has meaning 'twisting or 
rolling between the palms, drilling,' and with the verbal inflection, 'twist 
between the palms, work a drill, bore, drill.' The Motul has "hax, ah, 
ah (i.e. has, hasah, hasab) : torcer con la palma o palmas de las manos 
y hazer tomiza, o cordel assi, y lo assi torcido" and again "haxs : taladrar 
o agujerar taladrando y la cosa taladrada o agujerada assi." Tliis stem is 
the only word for drilling in Maya that 1 know of, so the case is particu- 
larly convincing. The word for 'a drill, the instrument,' is hasab; we do 
not have it in this codex, but rather the verbal inflection. The suffix -es, 
-s (followed by -ah) of the verbal inflection is causative, similar in mean- 
ing to the suffix -bes; X-es-ah means 'puts (put) it (grammatical object) 
into the condition X,' or else 'cause (caused) it to exist by the condition 
or action X, makes (made) it by X-ing, by doing X.' The second type of 
causati\e meaning is that which fits the present case. The suffix -ah 
denotes transiti\e action already accomplished, in contrast with -ik, 
transitive action not accomplished or not finished, either future or con- 
tinuing in the present. Thus hasesah means 'makes (made) it by 

Makes what by drilling? According to our scheme above, that which 
is denoted by the next sign cluster, Bi. In the bottom section of the 
same page, the corresponding cluster B2 denotes the stone or stone ob- 
ject being drilled. In that case 'makes by drilling' of course does not 
mean create the object wholly by drilling, but rather perform that step 
in the manufacture of the object that requires drilling. Hence in that 
case there is merely a subtle shade of difference between hasesah and 
hasah 'drills it.' To digress a little, cluster Bo is probably to be read 
e-i-l-l: e, dots, here many instead of three, i of three nipples, and a form 
of double-loop I doubled by scratches {lac) between the loops. The 
word eil could mean 'edge tool,' i.e., 'weapon point, knife,' etc. Such 


points or knives were of course predominantly of stone among the Maya, 
and were no doubt sometimes drilled. 

Returning to the middle-section text; here hasesah Bi means 'makes 
Bi by drilling,' actually in the sense of 'causes' or 'creates,' since Bi 
evidently denotes 'fire.' This fits in well with the expression cited by 
the Motul for 'making fire with the firedrill': hasah k'ak' {k'ak' 'fire'), 
which uses the simpler or less inflected form hasah rather than hasesah. 
The Motul gives "hax kak (i.e., has-[ah] k'ak') : encender lumbre 
casando fuego frotando un palo con otro, " also "haxab kak [hasab 
k'ak' 'drill for fire') : artificio o recaudo con que sacan fuego los indios." 

The cluster By is analyzed as u-to-kak, consisting of sign 23 of Figure 
3, u; sign 21, to, tu (to be read here to); and 9 of Figure 3, which if it is 
a doubled and enlarged ka (no. 8) might be read kaka, kak, or simply ka. 
Here the reading kak fits exactly. The initial u here would denote the 
preposed third-person pronominal reference u. For our present purposes 
it is immaterial whether this be regarded as a prefix or a separate word 
always occurring immediately before nonpronominal stems. Owing 
wholly to the grammatical patterns of English (and other European 
languages), it must be translated as 'he (she, it, they)' if the following 
stem is translated as an English verb, but as 'his (her, its, their)' if that 
stem is translated as an English noun. From the Maya standpoint it 
denotes the same relationship at all times; Maya stems are neither nouns 
nor verbs in the English sense, but a single class delimited on a quite 
diflFerent basis from our parts of speech. The stem with which this u 
is in construction is what is written as to-kak in the rest of the cluster. 

The writing to-kak however is only approximately phonetic, as with 
Maya writing in general; it suggests only in rough outline the sound of 
the utterance, from which suggestion the reader is expected to infer the 
right Maya word; the Maya application of phonetics in writing had 
progressed no farther than this, as we have already seen. Now the word 
that is apparently indicated is not what a modern Americanist phoneti- 
cian understands by the transcription tokak, but rather what he would 
transcribe as to-kk'ak'. This is a compound word, to-k-k'ak', consist- 
ing of the stems to-k 'burn, burning, ignition' (o- denotes long o) and 
k'ak' 'fire.' The Motul gives these as "tooc (i.e., to-k) : quemar, 
abrazar, y cosa quemada" and "kak (i.e., k'ak') : fuego, o lumbre." 
Note that the Maya way of writing to-kk'ak' does not distinguish the 
glottalized palatal stop k' at the end of k'ak' from the corresponding 


unglottalized stop k at the end of to-k, nor does it distinguish the 
sequence of the two, kk' from either one singly nor the long vowel o 
from a short o. This is all part and parcel of the approximate and out- 
line-like character of the phoneticism, implicit rather than clearly con- 
scious phoneticism, which Maya scribes employed. There is a pho- 
nemic difference between the simple and the glottalized stops in Maya, 
but it is a minimal difference. The writing used the same symbol for 
both a simple stop and the homorganic glottalized stop; instances of 
this are numerous. This does not mean that these were not distinct 
sounds in the Maya dialect of the codices. It is almost a certainty that 
they were distinct, just as they are in all modern dialects of Maya. They 
were not distinguished in writing probably in the same way that mini- 
mally differing phonemes (e.g., the long and short vowels of Latin) are 
often not distinguished in a writing system, because the native reader 
can always tell from the context w^hich sound to supply, x^nd this con- 
dition is no more than we meet, to varying degree, in all systems of 
writing other than those devised by linguistic scientists for the express 
purpose of an accuracy going beyond the needs of simple communication. 

The expression u-to-k-k'ak' may be translated 'his burning fire,' or 
probably better 'his kindling fire, his igniting of fire.' It follows a type 
of Maya two-stem compound, probably the same type as already ex- 
plained, though the idea of "by means of" here need not be injected into 
the translation. We now have attained to translation of the whole 
predicate: '(he) causes by drilling his ignition of fire'; and it is exident 
that this expression hasesah u-to-k-k'ak is but a more elaborate form of 
the hasah k'ak' cited by the Motul dictionary as the way of saying that 
one starts fire with a fire drill; it follows the same basic pattern. 

I might here digress briefly, anticipating a misconceived objection that 
might be raised, to say that the sign cluster to-kak sometimes occurs in 
the codices where there is no pictured reference to fire, and seems in 
these cases to refer to an animal in a hunting scene. An instance of this 
is seen in Figure 5, top section, over the second picture, where occurs 
the cluster to-kak-a, with -a of no. 1, Figure 3, and without preceding u-, 
forming part of a sentence roughly analyzable as loman u-NORTH 
tokaka X 'speared (in) his north (is) (grammat. object)-X.' I shall sug- 
gest first, but not in seriousness, a type of explanation that overstresses 
the mentalistic approach. 1 shall suggest that the reason why this glyph 
accompanies both pictures of fire and pictures of a hunted animal is that 


it is a glyph which denotes sacrifice or a sacrifice: hence either a sacri- 
ficial fire or a sacrificial animal. Now apparently just this sort of ex- 
planation, with its thin veneer of ethnological allusion, sounds plausible 
to some minds that have engaged themselves with Maya hieroglyphs, 
and it is necessary to warn against it. This is the reason why no people 
but linguists should touch the hieroglyphs. In the present case of course, 
the explanation is an out-and-out concoction of my own, cooked up in 
a few seconds merely to illustrate a point. A trained linguist would, 
I believe, be inclined to ask: "Have you searched for an explanation in 
the configurations of utterances and in the data of the vocabulary, be- 
fore adopting this quite speculative hypothesis?" The real reason no 
doubt is that, besides the stem to-k 'burn,' Maya has the similar sound- 
ing stem tok (with nonlong o) 'take away, take by force, capture, carry 
off,' etc. The Motul has "toe, ah, oh (i.e., tok) : quitar, tomar por 
fuerza, privar, arrebatar, robar y usurpar casas, y cosas muebles." The 
sign cluster to-kak in this case is not being used to write the compound 
word to-k-k'ali but to write some similarly-sounding derivative or in- 
flection of the stem tok, and the word probably means 'prey, animal 
taken or carried off, catch, game.' Possibly the word contains tok and 
the repetitive plural suffix -ak; hence '(successive) catches of game.' The 
context is enough to distinguish this word from the similarly-written 
word pertaining to fire. 

The next sign cluster, i-g-mn-a, writing the word igamna "Itzamna, 
name of the leading Maya god, the Roman-nosed god of the codices," 
is very important because it is the first proper name written in Maya 
hieroglyphs to be deciphered. Proper names and especially personal 
names have a peculiar convincingness in the decipherment of any script. 
They are ideal tools for decipherment when they can be had. When a 
decipherer can with the aid of his system spell out some well-known 
proper name which should occur in his text, he knows that he is on the 
right track. It will be remembered that it was the names of Ptolemy 
and Cleopatra in an inscription that gave Champollion his most effec- 
tive clues, and similarly it was the names of Xerxes and Darius in the 
Behistun inscription that afforded Rawlinson his starting point for the 
decipherment of cuneiform. It has long been agreed that the Roman- 
nosed god of the codex pictures, or god D, corresponds in characters to 
the one traditionally known as Itzamna. His glyph is always written in 
this way. If we knew more of the ancient names of the gods, our 


progress in decipherment would be materially aided. Unfortunately 
the god Kukulcan, who appears so frequently in the codices, evidently 
is not called by that name in the codices, or else, if he is called by that 
name, it is written by a unitary word sign. 

The next cluster, k-ka-haw, representing the pronunciation kahaw, is 
to be reconstructed ka-ahaw 'our lord, our master, our king.' This was 
the characteristic epithet of Itzamna as the Maya Zeus. In the Chilam 
Balam of Chumayel and also that of Tizimin, this god is referred to 
and called Itzamna kavil. Here kavil equals in the Americanist phonetic 
system, k'awil, from kahawil (glottalization arising from loss of -ah-) 
from ka-ahawil, which has the same meaning as ka-ahaw. Thus this 
decipherment may be likened to Rawlinson's recognition of 'king, great 
king, king of kings' after the name of Xerxes. The Motul defines ahaw 
as "ahau {ahaw) : rey o emperador, monarca, principe o gran seiior." 
The preposed pronominal ka (traditional spelling ca) is the second- 
person plural governing the following word, the translation of the rela- 
tionship being possessive when that word is translated as a noun, sub- 
ject when it is translated by a verb. Here of course the translation is 
'our.' The cluster k-ka-haw 'our lord' is an almost invariable accom- 
paniment of the name Itzamna in the codices; rarely it is omitted, and 
rarely it occurs with the names of other gods. Occasionally also with 
names of gods we find the simple epithet ahaw 'lord,' written a-hw, with 
an a sign not listed in this paper but cited in slightly \ariant form by 
Landa, and with no. 6 of Figure 3 for hw. In accordance with the 
general principle of Maya writing that signs may not be used in isola- 
tion, except as day signs, the word ahaw is not written with sign 5 {haw) 
alone, except when it means the day Ahau. 

Thus we arrive at our final translation: 'Our lord Itzamna kindles his 
fire by drilling.' 

The importance of this decipherment and translation is quite inde- 
pendent of the interest or lack of interest of the subject matter. As far 
as concerns the information which this translation gives us about the 
Maya, or about its own subject matter, it is quite trivial; it is no more 
than we could ha\e gathered from the pictures alone. Its importance is 
linguistic and philological— linguistic because it gixcs information about 
the structure of a language, as far as the writing can express it, at a 
certain period of past time; philological because it is precedent to the 
study of a literature and of culture as reflected in this literature, at a 


period of past time and in a historical context and perspective. From 
this one short sentence can be gathered a host of Knguistic and philo- 
logical data, only a small fraction of which has been discussed in this 
paper, data which can be tested and correlated, and employed heuristi- 
cally in further investigations, of progressive difficulty. A very few of 
these further ramifications of this sentence are barely hinted at in the 
footnotes, which the exigencies of space have kept relatively brief. Each 
such footnote actually represents an extensive study. In this way the 
decipherment establishes itself upon a constantly growing enlacement of 
sentences, their translations controlled by sets of pictures, which sen- 
tences mutually give rise to a growing grammar, syntax, vocabulary, and 
sign list. 

There are two main wrong ways of trying to read the Maya codices. 
One wrong way is to attempt a clean sweep of the job— to retire into 
seclusion and eventually emerge with a book— a book which "tells all," 
which reels off, interprets, explains, epitomizes, and comments on every- 
thing from page 1 of the Tro-Cortesianus to the last page of the Dresden. 
There have been several such books in the past hundred years. Usually 
such books proclaim the discovery of a key. This key is then applied at 
the author's sweet will, and the trick is turned as easily as a magician 
lifts a rabbit out of a hat. Often, moreover, such an author has exposed 
his slight acquaintance with the Maya language and with linguistic pro- 
cedures in general. Historical writings are not to be read with keys; 
there is never any key but research. The amateur decipherer is prone 
to make a false analogy between straightforward writing and a cipher. 
Actually the very word "decipher" which I have employed so profusely 
in this essay, embodies a misconception. Why have I used it? I sup- 
pose because it is simple and vivid, it has been generally used for this 
sort of research, and I have succumbed to usage. But really one does 
not decipher a literature; one deciphers only a cipher. A cipher is a 
method of writing with deliberate intent to conceal the content from 
those who do not possess the key. It is deciphered with a key because 
it has first been enciphered with a key. A straightforward writing, not 
intended to conceal its tenor from all but a select iew, is not really 
deciphered; it is analyzed and translated. The methods of such analysis 
and translation are quite different from the methods of message de- 
coders; they are the methods of Champollion and Young with Egyptian, 


of Rawlinson and Grotefend with Babylonian, of Hrozny and Sturte\ant 
with Hittite; they are the methods of Hnguistics and philology. 

The other wrong way of attacking the linguistic portion of the Maya 
codices is the Sitzenfleisch approach. It concentrates for long periods 
upon isolated glyphs or words, having conveniently forgotten that such 
things as sentences exist. Suppose that, in this method, one succeeds 
in deciphering or partly deciphering the glyph of Itzamna. Then one 
next spends years scrutinizing every glyph of Itzamna in the literature, 
noting the most minute differences, to the pen quirk, and linking it up 
first with e\er}' scrap of information that can be gleaned about Itzamna, 
then with every god in the Middle American area that can be connected 
with Itzamna. The mere glyph disappears from view, having served 
as the springboard into a sea of mythology, religion, and folklore, from 
which one may perhaps emerge at last with a monograph entitled "The 
Concept of Itzamna." This method, through concentrating entirely on 
word study, wanders so far from the specific incidences of the word in 
the texts that it finally ceases to be linguistic altogether, and becomes 
something else. Words are nothing without sentences. What a word 
is depends on what it does: i.e., on its position and function in the sen- 
tence. This is even more important than how it is written. In Maya 
as in English there are many homonyms, and also words which though 
not homonyms are written alike, as in English are 'lead' (the metal) and 
'lead' (go in front). Hence the determination of the sounds of signs 
and of their glyphic combinations is only half the battle. 

There is only one road to decipherment of the Maya hieroglyphs and 
reading of the Maya literature. It is through a growing concatenation 
of sentences, proceeding from the less to the more diflftcult, beginning 
with sentences whose meaning can be understood from pictures, with 
the linguistic interest and linguistic findings kept constantly foremost, 
and conclusions relative to subject matter resolutely submerged. The 
linguistic findings must eventually bear the scrutiny of, and become the 
ground of, collaboration for various linguistic scholars. One man can- 
not be the medium for interpreting a literature; such a task requires the 
mutual contributions of many scholars who are able to proceed in gen- 
eral agreement as to basic principles. Linguistic principles alone earn- 
the conviction necessary to such scientific agreement. 

As the research progresses and expands and grows more sure, it be- 
comes able to read with some confidence sentences which lack pictures 


to control the translation. We shall thus begin to read cautiously por- 
tions of the inscriptions, and the long pictureless texts of the Peresianus 
codex whose meaning is now utterly mysterious. As the major linguistic 
difficulties are conquered, the study becomes more and more philologi- 
cal; that is to say, subject matter, cultural data, and history play an in- 
creasing role— it becomes a matter not only of reading but also of under- 
standing as much as possible the allusions, the references, the nonlin- 
guistic contexts, the cultural patterns which are seen by glimpses, as it 
were, through the bare words and grammar of the translations. This is 
philology. But as the base of philology we must have linguistics. Only 
in this way can we ever hope to understand the history and culture of 
the Maya. 


The common material of Hopi buildings is stone. Adobe, the usual 
building material of the Rio Grande country, is rarely used. The 
stone is quarried and roughly dressed by the Hopi themselves, and set 
up without mortar. Walls are stone, roofs and floors above ground are 
tamped earth or clay several inches thick on a layer of close-set poles laid 
across cylindrical timbers or beams ledged in the walls. Interior surfaces 
of walls and ceilings are usually finished with a clay plaster or stucco, 
and then whitewashed with a fine white clay; exteriors are sometimes 
stuccoed, usually left in bare masonry. One-story buildings are the most 
frequent, but two-stor\' ones are not uncommon, and in Walpi even 
occasional third stories are to be seen. Stairways and ladders, both of 
which are used, are external to the buildings. The pit house or kiva, 
used for ceremonial purposes, is essentially similar except that it is buried 
and wholly or partially hollowed out of the ground, its upper portion 
projecting like a well-curb and bearing a roof with a hatchway for the 
entrance ladder. 

Hopi has a fairly considerable number of terms for what might be 
called structural elements or component parts of a building, including 
essential appurtenances to a building such as ladders, stairs, and windows. 

* Reprinted from Int. J. Amer. Linguistics, 19:141-145 (1953). This article was 
prepared by Whorf for presentation to the New York and Yale informal linguistic 
group at its meeting on February 25, 1940. The manuscript was among the papers 
left by Whorf to George L. Trager. It was checked by Edward A. Kennard. 



Such terms are grammatically all nouns. The following is a representa- 
tive list; it follows the order of construction, from the foundations up. 

kiya 'foundation/ or 'foundations'; te'k^'a 'an erection of masonry,' 
not a part of a finished building, 'unfinished wall, unroofed wall, or 
standing portion of a ruin'; tek"'dnmere 'encircling or enclosing wall'; 
te-wi 'ledge, shelf, or setback,' applied both to a natural ledge and to the 
architectural form; tek"'ni 'wall,' chiefly a roofed wall, but also applied 
to a finished stone fence or rampart; ^e'ci 'partition,' or 'closure' of any 
kind; ^e^ecpi, ^ecpi 'door,' i.e. closing piece or door proper; hociwa 
'door opening, doorway'; pokso 'vent hole, unglazed window, chimney'; 
pandvca 'piece of glass, glazed window'; naydve 'adobe brick'; palwi, 
pdlwicoqa 'plastering clay,' used as stucco whitewash, and 'floor clay'; 
ki-qolo 'lowest stor\' of a building of more than one story, or a sunken 
room like a basement; story with a floor above it'; ki-vela, ks-vii^pi 
'ceiling'; wuna piece of timber of any kind, 'board, plank, post, log, pole,' 
whether placed in structure or not; Ustavi 'beam or joist of roof or upper 
floor, timber or log' for this purpose, usually when in place in the struc- 
ture; kiqdlmo 'eaves,' or 'cornice'; ki-'^ami 'roof (not however the term 
in such expressions as 'on the roof); kiska 'tunnel or covered way, 
roofed passage.' 

Of these 19 terms, which are some of the most common ones, 8 are 
unanalyzable stems, or in a few cases partially and conjecturally analyz- 
able; the other 1 1 are transparent derivatives or compounds. 

Terms of this class, all denoting structural elements or parts, are 
nouns, having the noun declension of two cases, nominative and ob- 
jective, and a set of construct-state or possessed forms. Tliey all denote 
three-dimensional solids in the geometrical sense, solid and rigid masses, 
or definitely bounded areas on or perforations through such solids. 

As we look over the grammatical class of nouns we are struck by the 
absence of terms for interior three-dimensional spaces, such as our words 
'room, chamber, hall, passage = interior passage, cell, crypt, cellar, attic, 
loft, vault, storeroom,' etc., in spite of the fact that Hopi buildings are 
frequently divided into several rooms, sometimes specialized for different 
occupancies. We should never notice this state of affairs unless we 
approached it from the grammatical viewpoint first, for, if we simply 
ask the informant for the word for 'room' we shall certainly get a reply— 
a word which to him is the equivalent and translation of our word 'room.' 
Nevertheless, this word, and a few other words used to denote interior 


spaces, on examination will be found to have different grammatical or 
paradigmatic properties from the words for architectural elements or 
structural members which we have just noted. They do not seem to be 
nouns in the strict sense at least. The word for 'room' ^d'pave^, has no 
nominative or objective case and no construct state; one cannot say, 
based on this stem, 'my room,' whereas one can say 'my door' or 'my 
ceiling' even though these expressions have no socially functioning mean- 
ing, for Hopi society does not reveal any individual proprietorship or 
retainership of rooms, doors, or ceilings. Here we see the difference 
between a purely linguistic or formulaic meaning, which could be said 
but probably would not be said, like 'my ceiling,' and the case of a cul- 
tural and practically recognizable meaning which also coincides with 
linguistic meaning, like 'my house.' On the contrary, an expression 
formally equivalent to English 'my room' does not exist, or have even 
a formulaic meaning; there is a gap here in the language as compared 
to ours. If the Hopi should borrow from us the custom of having in- 
dividual "own" rooms, or should rent individual rooms when they visited 
other Hopi villages, they would still be unable to say 'my room.' What 
they would probably do would be to coin a new expression for this need. 
There are many ways they could do this. They could say for instance 
'my ceiling,' 'my door,' or 'my floor,' and in time the word 'ceiling,' 
'door,' or 'floor,' as the case might be, would acquire the extended mean- 
ing of an individual person's own room, like French foyer 'hearth' mean- 
ing one's home. This slight digression from the main topic will serve 
to illustrate the conser\ativism of grammatical patterns and their re- 
sistance to change as compared to simple lexical items. 

Returning to the word for 'room' ^d'pave^, let us examine its case 
properties. Though it has not the noun cases nominati\e and objec- 
tive, it has the cases locative, illative, and ablative ^d'pave^, ^d'phmiq, 
'-^d'paijk, case-relations which are found among pronouns, along with a 
number of others of similar nature called the locational cases. But pro- 
nouns have the nominative and objective cases also, as well as other 
peculiar properties all their own. It will be found that ^d'pavc^ belongs 
to a part of speech called locators, which include such words as 'here, 
there, above, below, in front, in back, north, south, east, west,' and a 
good many others, among them the Hopi geographical names, such as 
Oraibi, Walpi, Shipaulovi. These all have a paradigm of locational 
cases, and ever)' form ends in suffix; there is no bare-stem form such as 



exists in nouns and therein forms the nominative. In saying 'north' it 
is necessary to say 'in the north, from the north, to the north,' etc. 
These are forms that all belong in the predicate, while something else 
serves as the subject, or else there is no subject. In other words they 
are adverbial forms. The word translated 'room' means 'in a house, 
room, or other enclosed chamber' or, more precisely, 'in an architectural 
interior,' or 'into' such an interior, or 'from' such an interior, etc., ac- 
cording to the case sufEx. The -vs^ sufEx in ^d'pavs^ is the locative. 
There is also a quasilocative sufEx -vi or -pi which cannot be applied to 
the stem ^d'pa-, nor to most of these locationals, but can be applied to 
geographical place names and one or two other special words for 'room.' 


-► 4 


Figure 7. 

Its locative sense is so weak that it can be used as a nominative or ob- 
jective noun, though this use is rare. The chief other word for interior 
space in a building is locative ye-mokvi, illative ye-mok, often trans- 
lated 'the other room' or 'the next room'— but it is also used equivalent 
to 'back room, closet, recess, spare room, storeroom.' Perhaps the closest 
semantic effect in English would be 'inner room,' though this must be 
construed 'inner' in the sense of the Hopi illative, which would include 
'further room' or 'adjoining room.' It is any chamber one goes into 
from another interior which is the point of reference. One of the most 
common room plans of a Hopi house is shown in Figure 7. 

This is in line with the way Hopi and, in fact, most or all Uto-Aztecan 
languages represent location in space, or regions of space. They are not 
set up as entities that can function in a sentence like terms for people, 
animals, or masses of matter having characteristic form, or, again, human 
groups and human relations, but are treated as purely relational con- 
cepts, of an adverbial type. Thus hollow spaces like room, chamber, 
hall, are not really named as objects are, but are rather located; i.e. posi- 
tions of other things are specified so as to show their location in such 
hollow spaces. Contrasted with the considerable number of terms for 
solid architectural members, there seems at first a remarkable paucity of 


terms for architectural hollow spaces, with only two stems of any prac- 
tical importance. At first one might be led to ascribe this to something 
culturally peculiar in Hopi architecture— their concepts of building con- 
struction are limited and one-sided as a given cultural fact, we might 
suppose. On more intimate knowledge of the language we see that it 
has nothing to do with the architecture that there is this paucity of 
terms where we have a rich array of terms; it is a matter of the structure 
of the language. It is not the two stems that determine paucity or 
richness of expression, but the large array of suffixes with locational 
case-endings which may be used on these stems, because they belong to 

k i-go I o 



Figure 8. 

the locator group. In the locator group the number of initial stems is 
not an important criterion of vocabulary richness; it is rather the pro- 
fusion of suffixes, which in this group are in effect noninitial stems. 

Now there is still a third class of architectural terms to be considered- 
terms for different types of buildings. There are certainly different 
structural types of building among the Hopi. The three main types may 
be illustrated by diagrams (Figure 8). 

These buildings are put to various specialized uses. Most are dwell- 
ings, but the so-called piki-houses are used only as bakeries for baking 
piki or Hopi corn wafers; others are used only for storehouses, the kivas 
only for ceremonies. Since white influence there are buildings occupied 
solely as stores, churches, and schools. Now we, and many peoples 
much less sophisticated architecturally than the Hopi, have a \ocabu- 
lary of different terms for buildings: we have the terms 'house, building, 
cottage, castle, fort, temple, church, chapel, palace, theater, school, store, 
inn, hotel, barn, shed, garage, stable, hut, shack, shanty, prison, jail, 
tower, station, depot' and so on. Many of the terms denote occupancy- 
types; others structural types. It might be noted, from a detached view- 
point, that this English list is quite a miscellany and has practically no 


system to it. However, it seems to us the natural thing, for a people 
who have a building technology at least as diversified as the Hopi. 

Yet the fact is that, except for a few marginal terms of extremely 
restricted application (below), the Hopi language has only one word for 
a building; and it may be said, without any qualifications, that the lan- 
guage has no architectural terminology that classifies buildings into 
types— in spite of the fact that it does have a considerable architectural 
terminology serving another purpose. There is only the word ki-he 
'house' (as usually translated), which really means 'building' of any kind. 
This word is a noun, but the only noun of its kind in the language. 
This word 'house,' though not compounds ending in '-house,' can ser\'e 
as base or initial stem for suffixing the vocabulary of locational suffixes 
terminating in locational case endings, as if it were a pronoun or a 
locator. Certain other nouns can take a few of these suffixes, but 'house' 
is the only one that can take the entire set. In this respect it is like a 
place pronoun, but it is not a pronoun, for it has the construct state 
forms which only nouns have, so that one can say 'my house,' 'your 
house,' etc. 

The marginal cases which might possibly be considered terms for 
building other than ki-he may be cited, and it will be seen that these 
hardly denote true buildings in the usual sense: mecdvki 'tent,' lit. 'cloth 
house'— a foreign object to the Hopi, and denoted by a compound of 
'house'; te-teska 'shrine'— a small, crude enclosure of stones, covered 
o\er and located outdoors; kiska 'tunnel'— also a covered outdoor passage 
with walls and roof, usually connecting different buildings. 

One reason for the great paucity of Hopi building terms is that the 
Hopi either do not use occupancy terms as synonymous with the term 
for the building housing the occupancy, or, if they do this, they have 
begun to do it only recently, and few terms of this sort have accumu- 
lated. They do not have, at least not firmly rooted, the pattern which 
is so natural to us, in which 'a church,' i.e. an institution, is a term that 
merges quite imperceptibly into 'a church' meaning a type of building 
used as a meeting place for this institution, with the distinction hardly 
felt until attention is called to it; or in which 'a school,' the institution, 
is hardly distinguished from 'a school,' i.e. a schoolhouse, or 'garage' 
denoting a kind of occupancy from 'garage' the building housing this 
occupancy, or 'hospital' the occupancy from 'hospital' the building, or 
'the theater' in the sense of 'the dramatic art' from 'the theater,' a build- 


ing. The Hopi language does not have this imperceptible fusion, but 
rather a distinction between the two. The occupancy and the spot of 
ground or floor on which the occupancy occurs is called simply 'the 
building,' a ki-he. This is not a matter of stylistics, for it is not variable 
by the artistic modulations of the speaker, but of linguistics, for it is a 
form which the native speaker must follow willy-nilly, just as much as 
he must follow matters of grammar. The occupancy of a piki-house is 
called by a term which means 'place where the griddle is set up,' but it 
would be called that if set up outdoors, and there is no term for the 
piki-house itself, except in English, although the piki-house is a rather 
distinct architectural type. 

The fact that occupancy terms can be used in conjunction with the 
separate and uncombined word 'building,' with its array of inflected 
forms, to specify all sorts of places both outside and inside of given 
buildings, makes up for the lack of building terms, so far as fluency of 
expression goes. Howe\er, it seems rather queer that there should be 
no terms for such distinctly different shapes of buildings as e.g. the one- 
story building, the two-stor\' setback building, and the kiva; this fact has 
to be recorded as a peculiar datum of the language not explainable either 
from other patterns in the language or from anything in the architec- 
ture or anything else in the culture. 

It seems especially queer from the standpoint of our ways of thinking 
that there should be no name for the kiva, that structure so highly typi- 
cal of pueblo culture and so intimately connected with their religion. 

Many people know that our word kiva is taken from Hopi, but they 
think that it is the Hopi word for a kiva, which it is not. 


Unanalyzable stems: ^d'pa- 'interior,' ^a'pays"^ 'at the inside'; "^e'ci 
'door,' ^&cpi, ^e^e'cpi 'at the door'; ^eci 'closure, partition'; sa'qa 'ladder'; 
te'kwa 'something built of stones, but not finished building, erection 
of stones, unfinished wall, standing portion of a ruin'; te-wi 'ledge, 
shelf (natural or architectural); wena 'a timber, board, plank.' 

ki-he 'building, house.' 

Compounds with -ki: occupancy terms: he-ya-Nld 'store, trading post'; 
te-teqayki 'school.' 


Locator sufExes: kico^o- 'roof,' kico^ovi 'from the roof,' kico^dmiq 'to 
the roof; ye-mok 'into another enclosure, room, closet,' etc., ye-mokvi 
'far inside.' 

Compounds and miscellaneous architectural terms: khja, ki-het ija^at 
'foundation (root of the house)'; kiska 'tunnel, covered way'; ki-coki 
'village'; ki-sonvi 'plaza'; hociwa 'opening, doorway'; ki-vda 'ceiling'; 
kiqdlmo 'eaves, cornice'; ki-qolo 'lowest story of a house, cave'; Isstavi 
'viga, beam, roof support'; mecdvki 'tent'; naydvs 'adobe'; pdlwicoqa 
'plastering clay'; pandvca 'window, glass, mirror'; pokso 'vent hole, chim- 
ney, window hole'; tek^'ni 'roofed wall, stone fence'; tek^"dnmere 'enclos- 
ing wall, corral.' 


Tn* very normal person in the world, past infancy in years, can and does 
■*~^ talk. By virtue of that fact, every person— civilized or uncivilized- 
carries through life certain naive but deeply rooted ideas about talking 
and its relation to thinking. Because of their firm connection with 
speech habits that have become unconscious and automatic, these 
notions tend to be rather intolerant of opposition. They are by no 
means entirely personal and haphazard; their basis is definitely syste- 
matic, so that we are justified in calling them a system of natural logic— 
a term that seems to me preferable to the term common sense, often 
used for the same thing. 

According to natural logic, the fact that every person has talked 
fluently since infancy makes every man his own authority on the process 
by which he formulates and communicates. He has merely to consult 
a common substratum of logic or reason which he and e\eryone else are 
supposed to possess. Natural logic says that talking is merely an inci- 
dental process concerned strictly with communication, not with formu- 
lation of ideas. Talking, or the use of language, is supposed only to 
"express" what is essentially already formulated nonlinguistically. For- 
mulation is an independent process, called thought or thinking, and is 
supposed to be largely indifferent to the nature of particular languages. 
Languages have grammars, which are assumed to be merely norms of 
conventional and social correctness, but the use of language is supposed 

* Reprinted from Technol. Rev., 42:229-231, 247-248, no. 6 (April 1940). 



to be guided not so much by them as by correct, rational, or intelHgent 


Thought, in this view, does not depend on grammar but on laws of 
logic or reason which are supposed to be the same for all observers of 
the universe— to represent a rationale in the universe that can be "found" 
independently by all intelligent observers, whether they speak Chinese 













shawnee to say 


Figure 9. Languages dissect nature differently. The different isolates of meaning 
(thoughts) used by Enghsh and Shawnee in reporting the same experience, that of 
cleaning a gun by running the ramrod through it. The pronouns T and 'it' arc not 
shown by symbols, as they have the same meaning in each language. In Shawnee 
ni- equals T; -a equals 'it.' 

or Choctaw. In our own culture, the formulations of mathematics and 
of formal logic have acquired the reputation of dealing with this order 
of things: i.e., with the realm and laws of pure thought. Natural logic 
holds that different languages are essentially parallel methods for ex- 
pressing this one-and-the-same rationale of thought and, hence, differ 
really in but minor ways which may seem important only because they 
are seen at close range. It holds that mathematics, symbolic logic, 
philosophy, and so on are systems contrasted with language which- deal 
directly with this realm of thought, not that they are themselves special- 
ized extensions of language. The attitude of natural logic is well shown 
in an old quip about a German grammarian who devoted his whole life 


to the study of the dative case. From the point of view of natural logic, 
the dative case and grammar in general are an extremely minor issue. 
A different attitude is said to ha\e been held by the ancient Arabians: 
Two princes, so the story goes, quarreled over the honor of putting on 
the shoes of the most learned grammarian of the realm; whereupon their 
father, the caliph, is said to ha\e remarked that it was the glory- of his 
kingdom that great grammarians were honored even above kings. 

The familiar saying that the exception pro\es the rule contains a good 
deal of wisdom, though from the standpoint of formal logic it became 
an absurdity as soon as "prove" no longer meant "put on trial." The 
old saw began to be profound psychology from the time it ceased to 
have standing in logic. What it might well suggest to us today is that, 
if a rule has absolutely no exceptions, it is not recognized as a rule or as 
anything else; it is then part of the background of experience of which 
we tend to remain unconscious. Never having experienced anything 
in contrast to it, we cannot isolate it and formulate it as a rule until we 
so enlarge our experience and expand our base of reference that we 
encounter an interruption of its regularity. The situation is somewhat 
analogous to that of not missing the water till the well runs dry, or not 
realizing that we need air till we are choking. 

For instance, if a race of people had the physiological defect of being 
able to see only the color blue, they would hardly be able to formulate 
the rule that they saw only blue. The term blue would convey no mean- 
ing to them, their language would lack color terms, and their words 
denoting their various sensations of blue would answer to, and translate, 
our words "light, dark, white, black," and so on, not our word "blue." 
In order to formulate the rule or norm of seeing onh^ blue, they would 
need exceptional moments in which they saw other colors. The phe- 
nomenon of gravitation forms a rule without exceptions; needless to 
say, the untutored person is utterly unaware of any law of gravitation, 
for it would never enter his head to conceive of a universe in which 
bodies behaved otherwise than they do at the earth's surface. Like the 
color blue with our hypothetical race, the law of gravitation is a part 
of the untutored indiNidual's background, not something he isolates 
from that background. The law could not be formulated until bodies 
that always fell were seen in terms of a wider astronomical world in 
which bodies mo\ed in orbits or went this way and that. 

Similarly, whene\er we turn our heads, the image of the scene passes 


across our retinas exactly as it would if the scene turned around us. But 
this effect is background, and we do not recognize it; we do not see a 
room turn around us but are conscious only of having turned our heads 
in a stationary room. If we observe critically while turning the head or 
eyes quickly, we shall see, no motion it is true, yet a blurring of the 



HOPI - PAHE -^^-=^1^-^^ HOPI - KEYI 


Figure 10. Languages classify items of experience differently. The class corre- 
sponding to one word and one thought in language A may be regarded by language 
B as two or more classes corresponding to two or more words and thoughts. 

scene between two clear views. Normally we are quite unconscious of 
this continual blurring but seem to be looking about in an unblurred 
world. Whenever we walk past a tree or house, its image on the retina 
changes just as if the tree or house were turning on an axis; yet we do 
not see trees or houses turn as we travel about at ordinary speeds. Some- 
times ill-fitting glasses will reveal queer movements in the scene as we 
look about, but normally we do not see the relative motion of the en- 
vironment when we move; our psychic makeup is somehow adjusted ^to 
^i ^regard whol e r ealms of phenomena t hat are s o all-pcrvasivc as t o be 
'irrelevant to our daily lives and needs. 


Natural logic contains two fallacies: First, it does not see that the 
phenomena of a language are to its own speakers largely of a background 
character and so are outside the critical consciousness and control of the 
speaker who is expounding natural logic. Hence, when anyone, as a 
natural logician, is talking about reason, logic, and the laws of correct 
thinking, he is apt to be simply marching in step with purely gram- 
jTiatical f acts that h a\c somewhat of a background character m his o wn 
language or tamUv ot langna^ j ^es bnt pr^ bv n^ mponc nni\prtj^:^1^jr. oil 
languages ^]y] in it^ <:pngp ^ rommop substratum nf YP^son SccOnd, 

natural logic confuses agreement about subject matter, attained through 
use of language, with know^ledge of the linguistic process by which agree- 
ment is attained: i.e., with the province of the despised (and to its 
notion superfluous) grammarian. Two fluent speakers, of English let us 
say, quickly reach a point of assent about the subject matter of their 
speech; they agree about what their language refers to. One of them. A, 
can give directions that will be carried out by the other, B, to A's com- 
plete satisfaction. Because they thus understand each other so perfectly, 
A and B, as natural logicians, suppose they must of course know^ how it 
is all done. They think, e.g., that it is simply a matter of choosing 
words to express thoughts. If you ask A to explain how he got B's 
agreement so readily, he will simply repeat to }0u, with more or less 
elaboration or abbreviation, what he said to B. He has no notion of the 
process involved. The amazingly complex system of linguistic patterns 
and classifications, which A and B must haxe in common before they 
can adjust to each other at all, is all background to A and B. 

These backgrou nd phpr ^nmena are the province of the gran iyfinr^" — 

or of the linguist, tn piyP ln 'm Vik mnrp mnrlprn ngp-|P as a Qripnficf . Xhc 

word linguist in common, and especially newspaper, parlance means 
something entirely different, namely, a person who can quickly attain 
agreement about subject matter with different people speaking a num- 
ber of different languages. Such a person is better termed a polvglot or 
a multilingual. Scientific linguists have long understood that ability to 
speak a language fluently does not necessarily confer a linguistic knowl- 
edge of it, i.e., understanding of its background phenomena and its 
systematic processes and structure, any more than abilitv to play a good 
game of billiards confers or requires any knowledge of the laws of me- 
chanics that operate upon the billiard table. 
The situation here is not unlike that in any other field of science. All 


real scientists have their eyes primarily on background phenomena that 
cut very little ice, as such, in our daily lives; and yet their studies have a 
way of bringing out a close relation between these unsuspected realms of 
fact and such decidedly foreground activities as transporting goods, pre- 
paring food, treating the sick, or growing potatoes, which in time may 
become very much modified, simply because of pure scientific investi- 
gation in no way concerned with these brute matters themselves. Lin- 
guistics pr esents_a quite similar case; the background phenomena with 
which it deals are involved in all onr foreground arhvi>ips nf t-a1ki'nP| and 
of reaching agreement, in all reasoning and arguing of cases, in all la w, 
arbitration, conciliation, contracts, treaties, public opinion, weighing of 
scientific theories, formulation of scientific results. Whenever agree- 
ment or assent is arrived at in human affairs, and whether or not mathe- 
matics or other specialized symbolisms are made part of the procedure, 
v^is agreement is reached by linguistic processes, or else it is not 

As we have seen, an overt knowledge of the linguistic processes by 
which agreement is attained is not necessary to reaching some sort of 
agreement, but it is certainly no bar thereto; t he more comp li cated and 
diffic ult the matter, the more such knowledge is a distinct aid, till the 
point may be reached — I suspect the modern world has about arrived 
at it— when the knowledge becomes not only an aid but a necessity. 
The situation may be likened to that of navigation. Every boat that 
sails is in the lap of planetary foj^ces; yet a boy can pilot his small craft 
around a harbor without benefit of geography, astronomy, mathematics, 
or international politics. To the captain of an ocean liner, however, 
some knowledge of all these subjects is essential. 

When linguists became able to examine critically and scientifically a 
large number of languages of widely different patterns, their base of 
reference was expanded; they experienced an interruption of phenomena 
hitherto held universal, and a whole new order of significances came into 
their ken. It was found that the background linguistic system (in other 
words, the grammar) of each language is not merely a reproducing in- 
strument for voicing ideas but rather is itself the shaper of ideas, the 
program and guide for the individual's mental activity, for his analysis 
of impressions, for his synthesis of his mental stock in trade. Formula- 
tion of ideas is not an independent process, strictly rational in the old 
sense, but is part of a particular grammar, and differs, from slightly to 



great!}', between different granmiars. We disseet nature along lines laid 
down by our nati\c languages. The eategories and types that we isolate 
from the world of phenomena we do not find there because they stare 
every observer in the face; on the contrary, the world is presented in a 







HOPI . . , "WARl" (RUNNING. 





HOPI . . . "WARl" (RUNNING, 




HOPI . . . 'WARl" (RUNNING, 





HOPI . . . "era WARl" (RUNNING. 







ENGLISH. .."he runs" (E.G. ON 


I'igurc 11. Contrast between a "temporal" language (English) and a "timeless" 
language (Hopij. What are to English differences of time are to Hopi differences 

in the kind of .validity. 

kaleidoscopic flux of impressions which has to be organized by our 
minds— and this means largely by the linguistic systems in our minds. 
We cut nature up, organize it into concepts, and ascribe significances 
as we do, largely because we are parties to an agreement to organize it 
in this way— an agreement that holds throughout our speech community 
and is codified in the patterns of our language. The agreement is, of 
course, an implicit and unstated one, but its terms are absolutely 


OBLIGATORY; wc cannot talk at all except by subscribing to the organi- 
zation and classification of data which the agreement decrees. 

This fact is very significant for modern science, for it means that no 
individual is free to describe nature with absolute impartiality but is 
constrained to certain modes of interpretation even while he thinks him- 
self most free. The person most nearly free in such respects would be 
a linguist familiar with very many widely different linguistic systems. 
As yet no linguist is in any such position. We are thus introduced to 
a new principle of relativity, ^hichJiolds that alL observers are not led 
by the same physical evidence to the same picture of the universe, unless 
their linguistic backgrounds are similar, or can in some way be calibrated. 

This rather startling conclusion is not so apparent if we compare only 
our modern European languages, with perhaps Latin and Greek thrown 
in for good measure. Among these tongues there is a unanimity of 
major pattern which at first seems to bear out natural logic. But this 
unanimity exists only because these tongues are all Indo-European dia- 
lects cut to the same basic plan, being historically transmitted from what 
was long ago one speech community; because the modern dialects have 
long shared in building up a common culture; and because much of this 
culture, on the more intellectual side, is derived from the linguistic back- 
grounds of Latin and Greek. Thus this group of languages satisfies the 
special case of the clause beginning "unless" in the statement of the 
linguistic relativity principle at the end of the preceding paragraph. 
From this condition follows the unanimity of description of the world 
in the community of modern scientists. But it must be emphasized that 
"all modern Indo-European-speaking observers" is not the same thing as 
"all observers." TTiat modern Chinese or Turkish scientists describe the 
world in the same terms as Western scientists means, of course, only 
that they have taken over bodily the entire Western system of rationali- 
zations, not that they have corroborated that system from their native 
posts of observation. 

When Semitic, Chinese, Tibetan, or African languages are contrasted 
with our own, the divergence in analysis of the world becomes _jii&f€- 
apparent; and, when we onng m the native languages of the Americas, 
where speech communities for many millenniums have gone their ways 
independently of each other and of the Old World, the fact that lan- 
guages dissect nature in many different wavs becomes patent. The rela- 
tivity of all conceptual systems, ours included, and their dependence 


upon language stand revealed. That American Indians speaking only 
their native tongues are never called upon to act as scientific obser\'ers 
is in no wise to the point. To exclude the evidence which their lan- 
guages offer as to what the human mind can do is like expecting bota- 
nists to study nothing but food plants and hothouse roses and then tell 
us what the plant world is like! 

Let us consider a few examples. In English we divide most of our 
words into two classes, which have different grammatical and logical 
properties. Class 1 we call nouns, e.g., 'house, man'; class 2, verbs, e.g., 
'hit, run.' Many words of one class can act secondarily as of the other 
class, e.g., 'a hit, a run,' or 'to man (the boat),' but, on the primary level, 
the division between the classes is absolute. Our language thus gives us 
a bipolar division of nature. But nature herself is not thus polarized. If 
it be said that 'strike, turn, run,' are verbs because they denote tempo- 
rary or short-lasting events, i.e., actions, why then is 'fist' a noun? It 
also is a temporary e\ent. Why are 'lightning, spark, wave, eddy, pulsa- 
tion, flame, storm, phase, cycle, spasm, noise, emotion' nouns? They are 
temporary events. If 'man' and 'house' are nouns because they are long- 
lasting and stable events, i.e., things, what then are 'keep, adhere, extend, 
project, continue, persist, grow, dwell,' and so on doing among the verbs? 
If it be objected that 'possess, adhere' are verbs "because they are stable 
relationships rather than stable percepts, why then should 'equilibrium, 
pressure, current, peace, group, nation, society, tribe, sister,' or any kin- 
ship term be among the nouns? It will be found that an "event" to us 
means "what our language classes as a verb" or something analogized 
therefrom. And it will be found that it is not possible to define 'event, 
thing, object, relationship,' and so on, from nature, but that to define 
them alwa\s in\ol\es a circuitous return to the grammatical categories 
of the definer's language. 

In the Hopi language, 'lightning, wave, flame, meteor, puff of smoke, 
pulsation' are verbs— e\'ents of necessarily brief duration cannot be any- 
thing but verbs. 'Cloud' and 'storm' are at about the lower limit of dura- 
tion for nouns. Hopi, you see, actually has a classification of events 
(or linguistic isolates) by duration type, something strange to our modes 
of thought. On the other hand, in Nootka, a language of Vancouver 
Island, all words seem to us to be \erbs, but really there are no classes 1 
and 2; we have, as it were, a monistic view of nature that gives us only 
one class of word for all kinds of events. 'A house occurs' or 'it houses' 


is the way of saying 'house,' exactly like 'a flame occurs' or 'it burns.' 
These terms seem to us like verbs because they are inflected for dura- 
tional and temporal nuances, so that the suExes of the word for house 
e\ent make it mean long-lasting house, temporary house, future house, 
house that used to be, what started out to be a house, and so on. 

Hopi has one noun that covers every thing or being that flies, with 
the exception of birds, which class is denoted by another noun. The 
former noun may be said to denote the class (FC-B)— flying class minus 
bird. The Hopi actually call insect, airplane, and aviator all by the same 
word, and feel no difficulty about it. The situation, of course, decides 
any possible confusion among very disparate members of a broad lin- 
guistic class, such as this class (FC-B). This class seems to us too large 
and inclusive, but so would our class 'snow' to an Eskimo. We have the 
same word for falling snow, snow on the ground, snow packed hard like 
ice, slushy snow, wind-driven flying snow— whatever the situation may 
be. To an Eskimo, this all-inclusive word would be almost unthink- 
able; he would say that falling snow, slushy snow, and so on, are sensu- 
ously and operationally diflferent, different things to contend with; he 
uses different words for them and for other kinds of snow. The Aztecs 
go even farther than we in the opposite direction, with 'cold,' 'ice,' and 
'snow' all represented by the same basic word with different termina- 
tions; 'ice' is the noun form; 'cold,' the adjectival form; and for 'snow,' 
"ice mist." 
'v^What surprises most is to find that various grand generalizations of 
the Western world, such as time, velocity, and matter, are not essential 
to the construction of a consistent picture of the universe. The psychic 
experiences that we class under these headings are, of course, not de- 
stroyed; rather, categories derived from other kinds of experiences take 
over the rulership of the cosmology and seem to function just as well. 
Hopi may be called a timeless language. It recognizes psychological 
time, which is much like Bergson's "duration," but this "time" is quite 
unlike the mathematical time, T, used by our physicists. Among the 
peculiar properties of Hopi time are that it varies with each observer, 
does not permit of simultaneity, and has zero dimensions; i.e., it cannot 
be given a number greater than one. The Hopi do not say, "I stayed 
five days," but "I left on the fifth day." A word referring to this kind 
of time, like the word day, can have no plural. The puzzle picture (Fig. 


11, page 213) will give mental exercise to anyone who would like to fig- 
ure out how the Hopi verb gets along without tenses. Actually, the only 
practical use of our tenses, in one-verb sentences, is to distinguish among 
five typical situations, which are symbohzed in the picture. The time- 
less Hopi verb does not distinguish between the present, past, and future 
of the event itself but must always indicate what type of validity the 
SPEAKER intends the statement to have: (a) report of an event (situations 
1, 2, 3 in the picture); (b) expectation of an event (situation 4); (c) gen- 
eralization or law about events (situation 5). Situation 1, where the 
speaker and listener are in contact with the same objective field, is di- 
\ided by our language into the two conditions, Id and lb, which it calls 
present and past, respectively. This division is unnecessary for a lan- 
guage which assures one that the statement is a report. 

Hopi grammar, by means of its forms called aspects and modes, also 
makes it easy to distinguish among momentary, continued, and repeated 
occurrences, and to indicate the actual sequence of reported events. 
Thus the universe can be described without recourse to a concept of 
dimensional time. How would a physics constructed along these lines 
work, with no T (time) in its equations? Perfectly, as far as I can see, 
though of course it would require different ideology and perhaps dif- 
ferent mathematics. Of course V (velocity) would have to go too. The 
Hopi language has no word really equivalent to our 'speed' or 'rapid.' 
What translates these terms is usually a word meaning intense or \ery, 
accompanying any verb of motion. Here is a clue to the nature of our 
new physics. We may have to introduce a new term I, intensity. Every 
thing and event will have an I, whether we regard the thing or event as 
moving or as just enduring or being. Perhaps the J of an electric charge 
will turn out to be its voltage, or potential. We shall use clocks to 
measure some intensities, or, rather, some relative intensities, for the 
absolute intensity of anything will be meaningless. Our old friend ac- 
celeration will still be there but doubtless under a new name. We shall 
perhaps call it V, meaning not velocity but variation. Perhaps all 
growths and accumulations will be regarded as Vs. We should not 
have the concept of rate in the temporal sense, since, like velocity, rate 
introduces a mathematical and linguistic time. Of course we know that 
all measurements are ratios, but the measurements of intensities made 
by comparison with the standard intensity of a clock or a planet we do 


not treat as ratios, any more than we so treat a distance made by com- 
parison with a yardstick. 

A scientist from another culture that used time and velocity would 
have great difficulty in getting us to understand these concepts. We 
should talk about the intensity of a chemical reaction; he would speak 
of its velocity or its rate, which words we should at first think were 
simply words for intensity in his language. Likewise, he at first would 
think that intensity was simply our own word for velocity. At first we 
should agree, later we should begin to disagree, and it might dawn upon 
both sides that different systems of rationalization were being used. He 
would find it very hard to make us understand what he really meant by 
velocity of a chemical reaction. We should have no words that would 
fit. He would try to explain it by likening it to a running horse, to the 
difference between a good horse and a lazy horse. We should try to 
show him, with a superior laugh, that his analogy also was a matter of 
different intensities, aside from which there was little similarity between 
a horse and a chemical reaction in a beaker. We should point out that 
a running horse is moving relative to the ground, whereas the material 
in the beaker is at rest. 

One significant contribution to science from the linguistic point of 
view may be the greater development of our sense of perspective. We 
shall no longer be able to see a few recent dialects of the Indo-European 
family, and the rationalizing techniques elaborated from their patterns, 
as the apex of the evolution of the human mind, nor their present wide 
spread as due to any survival from fitness or to anything but a few events 
of history— events that could be called fortunate only from the parochial 
point of view of the fa\'ored parties. I'hey, and our own thought 
processes with them, can no longer be envisioned as spanning the gamut 
of reason and knowledge but only as one constellation in a galactic 
expanse. A fair realization of the incredible degree of diversity of lin- 
guistic system that ranges over the globe leaves one with an inescapable 
feeling that the human spirit is inconceivably old; that the few thousand 
years of history coNcred by our written records are no more than the 
thickness of a pencil mark on the scale that measures our past experi- 
ence on this planet; that the events of these recent millenniums spell 
nothing in any e\olutionary wise, that the race has taken no sudden 
spurt, achieved no commanding synthesis during recent millenniums. 


but has only played a little with a few of the linguistic formulations and 
views of nature bequeathed from an inexpressibly longer past. Yet 
neither this feeling nor the sense of precarious dependence of all we 
know upon linguistic tools which themselves are largely unknown need 
be discouraging to science but should, rather, foster that humility which 
accompanies the true scientific spirit, and thus forbid that arrogance of 
the mind \\'hich hinders real scientific curiosity and detachment. 



The revolutionary changes that have occurred since 1890 in the world 
of science — especially in physics but also in chemistry, biology, and 
the sciences of man— ha\e been due not so much to new facts as to new 
ways of thinking about facts. The new facts themselves of course have 
been many and weighty; but, more important still, the realms of research 
where they appear— relativity, quantum theory, electronics, catalysis, 
colloid chemistry, theory of the gene, Gestalt psychology, psychoanalysis, 
unbiased cultural anthropology, and so on— have been marked to an 
unprecedented degree by radically new concepts, by a failure to fit the 
world view that passed unchallenged in the great classical period of 
science, and by a groping for explanations, reconciliations, and restate- 

I say new ways of thinking about facts, but a more nearly accurate 
statement would say new ways of talking about facts. It is this use of 
LANGUAGE UPON DATA that is Central to scientific progress. Of course, 
we have to free ourselves from that vague innuendo of inferiority which 
clings about the word 'talk,' as in the phrase 'just talk'; that false op- 
position which the English-speaking world likes to fancy between talk 
and action. There is no need to apologize for speech, the most human 
of all actions. The beasts may think, but they do not talk. 'Talk' 
OUGHT TO BE a morc noble and dignified word than 'think.' Also we 
must face the fact that science begins and ends in talk; this is the re- 

* Reprinted from Technol. Rev., 43:61-63, 80-83 (December 1940). 



verse of anything ignoble. Such words as 'analyze, compare, deduce, 
reason, infer, postulate, theorize, test, demonstrate' mean that, whenever 
a scientist does something, he talks about this thing that he does. As 
Leonard Bloomficld has shown, scientific research begins with a set of 
sentences which point the way to certain obser\ations and experiments, 
the results of which do not become fully scientific until they ha\e been 
turned back into language, yielding again a set of sentences which then 
become the basis of further exploration into the unknown. This scien- 
tific use of language is subject to the principles or the laws of the science 
that studies all speech— linguistics. 

As I was concerned to point out in a pre\"ious article, "Science and 
linguistics," in the Review for April, we all hold an illusion about talk- 
ing, an illusion that talking is quite untrammeled and spontaneous and 
merely "expresses" whatever we wish to have it express. This illusory 
appearance results from the fact that the obligatory phenomena within 
the apparently free flow of talk are so completely autocratic that speaker 
and listener are bound unconsciously as though in the grip of a law of 
nature. The phenomena of language are background phenomena, of 
which the talkers are unaware or, at the most, ver\' dimly aware— as they 
are of the motes of dust in the air of a room, though the linguistic phe- 
nomena govern the talkers more as gravitation than as dust would. 
These automatic, involuntary patterns of language are not the same for 
all men but are specific for each language and constitute the formalized 
side of the language, or its "grammar"— a term that includes much more 
than the grammar we learned in the textbooks of our school days. 

From this fact proceeds what I have called the "linguistic relativity 
principle," which means, in informal terms, that users of markedly dif- 
ferent grammars are pointed by their grammars toward different types 
of observations and different evaluations of externally similar acts of 
observation, and hence are not equi\alent as obser\crs but must arrive 
at somewhat different \'iews of the world. (A more formal statement 
of this point appears in my article of last April.) From each such un- 
formulated and naive world view, an explicit scientific world view may 
arise by a higher specialization of the same basic grammatical patterns 
that fathered the naive and implicit \iew. Thus the world \icw of 
modern science arises by higher specialization of the basic grammar of 
the Western Indo-European languages. Science of course was not 
CAUSED by this grammar; it was simply colored by it. It appeared in this 


group of languages because of a train of historical events that stimulated 
commerce, measurement, manufacture, and technical invention in a 
quarter of the world where these languages were dominant. 

The participants in a given world view are not aware of the idiomatic 
nature of the channels in which their talking and thinking run, and are 
perfectly satisfied with them, regarding them as logical inevitables. But 
take an outsider, a person accustomed to widely different language and 
culture, or even a scientist of a later era using somewhat different lan- 
guage of the same basic type, and not all that seems logical and in- 
evitable to the participants in the given world view seems so to him. 
The reasons that officially pass current may strike him as consisting 
chiefly of highly idiomatic "famous de parler." Consider the answers 
that were at one time given even by learned men to questions about 
nature: Why does water rise in a pump? Because nature abhors a 
vacuum. Why does water quench fire? Because water is wet or because 
the fiery principle and the watery principle are antithetical. Why do 
flames rise? Because of the lightness of the element fire. Why can one 
lift a stone with a leather sucker? Because the suction draws the stone 
up. Why does a moth fly toward a light? Because the moth is curious 
or because light attracts it. If once these sentences seemed satisfying 
logic, but today seem idiosyncrasies of a peculiar jargon, the change did 
not come about because science has discovered new facts. Science has 
adopted new linguistic formulations of the old facts, and, now that we 
have become at home in the new dialect, certain traits of the old one 
are no longer binding upon us. 

We moderns are not yet in a position to poke fun at the wiseacres 
of old who explained various properties of water by its wetness. The 
terminology which we apply to language and cultural phenomena is 
often of a piece with the wetness of water and nature's abhorrence of a 
vacuum. The researches of linguists into the ways of languages many 
and diverse are needed if we are to think straight and escape the errors 
which unconscious acceptance of our language background otherwise 
engenders. An increasing contribution from linguistics to the general 
philosophy of science is demanded by the new ways of thinking implied 
by those new realms of science cited at the beginning of this essay. It 
is needed for science's next great march into the unknown. 

The situation is not likely to be aided by the philosophical and mathe- 
matical analyst who may try to exploit the field of higher linguistic 



symbolism with little knowledge of linguistics itself. Unfortunately the ' 
essays of most modern writers in this field suffer from this lack of ap- 
prenticeship training. To strive at higher mathematical formulas for 
linguistic meaning while knowing nothing correctly of the shirt-sleeve 
rudiments of language is to court disaster. Physics does not begin with 
atomic structures and cosmic rays, but with motions of ordinary' gross 
physical objects and symbolic (mathematical) expressions for these move- 
ments. Linguistics likewise does not begin with meaning nor with the t 


y(u), St* 


;4) (5, 





(7) (8) 19) (10) (II) 







; t 


' I 





0,^ s/z 






Figure 12. Structural formula of the monosyllabic word in English (standard mid- 
western American). The formula can be simplified by special symbols for certain 
groups of letters, but this simplification would make it harder to explain. The 
simplest possible formula for a monosyllabic word is C + V, and some languages 
actually conform to this. Polynesian has the next most simple formula, O, C -f- V. 
Contrast this witli the intricacy of English word structure, as shown above. 

structure of logical propositions, but with the obligatory patterns made 
by the gross audible sounds of a given language and with certain s} m- 
bolic expressions of its own for these patterns. Out of these relatively 
simple terms dealing with gross sound patterning are e\ oh ed the higher 
analytical procedures of the science, just as out of the simple experi- 
ments and mathematics concerning falling and sliding blocks of wood 
is evolved all the higher mathematics of physics up into quantum theory. 
Even the facts of sound patterning are none too simple. But they illus- 
trate the unconscious, obligatory, background phenomena of talking as 
nothing else can. 

For instance, the structural formula for words of one syllable in the 
English language (Fig. 12) looks rather complicated; yet for a linguistic 
pattern it is rather simple. In the English-speaking world, every child 
between the ages of two and five is engaged in learning the pattern 
expressed by this formula, among many other formulas. By the time 
the child is six, the formula has become ingrained and automatic; even 



the little nonsense words the child makes up conform to it, exploring 
its possibilities but venturing not a jot beyond them. At an early age 
the formula becomes for the child what it is for the adult; no sequence 
of sounds that deviates from it can even be articulated without the 
greatest difficulty. New words like "blurb," nonsense words like Lewis 



y= C+sin x 






NOISES (noyz-iz) CLAWS (kiD-z) 

CATS (kat-s) 

/LEAF l/liyf*\ 
LEAVES/ Vliyv-z/' 

HORSES (hors-iz) PIGS (pig-z) 

LIPS (lip-s) 

MATCHES (.mac -iz) LAMBS (lam-z) 

NECKS (neks) 

/WIFE \ /wayf*\ 
IwiVES/ Vwayv-zj 

WEDGES Iwej-iz) ROWS (row-z) 

CHIEFS (ciyf-s) 




suf.-T^ = I(Csib)-iz ; 2(S)-z ; 

3(C-S)-s i 4(f*>v)-z. 


Figure 13. Variables and alternants: A shows by graph and by mathematical for- 
mula (equation) an interrelation of variables. B illustrates by extensible examples 
and by a pattern formula an interrelation of alternants. The formula means that 
the English suffix which is theoretically ("by root," V ) a final 's' is actualized in 
any given case by one of four alternants: after a sibilant-ending consonant, by '-iz'; 
after any sonant (vowel or consonant), by''-z,' after any voiceless (nonsonant) con- 
sonant by 's'; except that, after the special alternant 'fJf,' it is actualized by '-z,' 

the 'fS' alternating to 'v.' 

Carroll's "mome raths," combinations intended to suggest languages of 
savages or animal cries, like "glub" and "squonk"— all come out of the 
[mold of this formula. When the youth begins to learn a foreign lan- 
guage, he unconsciously tries to construct the syllables according to this 
formula. Of course it won't work; the foreign words arc built to a 
formula of their own. Usually the student has a terrible time. Not 
even knowing that a formula is back of all the trouble, he thinks his 
difficulty is his own fault. The frustrations and inhibitions thus set up 
at the start constantly block his attempts to use foreign tongues. Or 


22 = 

else he even hears by the formula, so that the English combinations 
that he makes sound to him like real French, for instance. Then he 
suffers less inhibition and may become what is called a "fluent" speaker 
of French— bad French! 

Figure 14. Flow sheet of improved process for learning French without tears. 
Guaranteed: no bottlenecks in production. 

, If, however, he is so fortunate as to have his elementan,' French taught 
by a theoretic linguist, he first has the patterns of the English formula 
explained in such a way that they become semiconscious, with the result 
that they lose the binding power o\cr him which custom has gi\en them, 
though they remain automatic as far as English is concerned. Then he 
acquires the French patterns without inner opposition, and the time for 
attaining command of the language is cut to a fraction (see Fig. 14). 



To be sure, probably no elementary French is ever taught in this way— 
at least not in public institutions. Years of time and millions of dollars' 
worth of wasted educational effort could be saved by the adoption of 
such methods, but men with the grounding in theoretic linguistics are 
as yet far too few and are chiefly in the higher institutions. 

Let us examine the formula for the English monosyllabic word (Fig. 
12). It looks mathematical, but it isn't. It is an expression of pattern 
symbolics, an analytical method that grows out of linguistics and bears 
to linguistics a relation not unlike that of higher mathematics to phys- 
ics. With such pattern formulas, various operations can be performed, 
just as mathematical expressions can be added, multiplied, and other- 
wise operated with; only the operations here are not addition, multipli- 
cation, and so on, but are meanings that apply to linguistic contexts. 
From these operations, conclusions can be drawn and experimental at- 
tacks directed intelligently at the really crucial points in the welter of 
data presented by the language under investigation. Usually the linguist 
does not need to manipulate the formulas on paper but simply performs 
the symbolic operations in his mind and then says: "The paradigm of 
class A verbs can't have been reported right by the previous investigator"; 
or "Well, well, this language must have alternating stresses, though I 
couldn't hear them at first'.'; or "Funny, but d and I must be variants of 
the same sound in this language," and so on. Then he investigates by 
experimenting on a native informant and finds that the conclusion is 
justified. Pattern-symbolic expressions are exact, as mathematics is, but 
^are not quantitative. They do not refer ultimately to number and di- 
mension, as mathematics does, but to pattern and structure. Nor are 
they to be confused with theory of groups or with symbolic logic, though 
they may be in some ways akin. 

Returning to the formula, the simplest part of it is the eighth term 
(the terms are numbered underneath), consisting of a V between plus 
signs. This means that every English word contains a vowel (not true 
of all languages). As the V is unqualified by other symbols, any one of 
the English vowels can occur in the monosyllabic word (not true of all 
syllables of the polysyllabic English word). Next we turn to the first 
term, which is a zero and which means that the vowel may be preceded 
by nothing; the word may begin with a vowel— a structure impossible in 
many languages. The commas between the terms mean "or." The 
second term is C minus a long-tailed n. This means that a word can 


begin with any single English consonant except one— the one linguists 
designate by a long-tailed n, which is the sound we commonly write ng, 
as in "hang." This ng sound is common at the ends of English words 
but never occurs at the beginnings. In many languages, such as Hopi, 
Eskimo, or Samoan, it is a common beginning for a word. Our patterns 
set up a terrific resistance to articulation of these foreign words begin- 
ning with ng, but as soon as the mechanism of producing ng has been 
explained and we learn that our inability has been due to a habitual 
pattern, we can place the ng wherever we will and can pronounce these 
words with the greatest of ease. The letters in the formula thus are not 
always equivalent to the letters by which we express our words in or- 
dinary spelling but are unequi\ocal symbols such as a linguist would 
assign to the sounds in a regular and scientific system of spelling. 

According to the third term, which consists of two columns, the word 
can begin with any consonant of the first column followed by r, or with 
g, k, f, or h followed by I. The s with a wedge over it means sh. Thus 
we have 'shred,' but not shied. The formula represents the fact that 
shied is un-English, that it will suggest a Chinese pronunciation of 
'shred' or a German's of 'sled' {si is permitted by term 7). The Greek 
theta means th; so we have 'thread' but not thled, which latter suggests 
either a Chinese saying 'thread' or a child lisping 'sled.' But why aren't 
tr, pr, and pi in this third term? Because they can be preceded by s and 
so belong in term 6. The fourth term similarly means that the word 
can begin with a consonant of the first column followed by w. Hw does 
not occur in all dialects of English; in ordinary spelling it is written back- 
wards, \vh. If the dialect does not have hw, it pronounces the spelled 
wh simply as w. Thw occurs in a few words, like 'thwack' and 'thwart,' 
and gw, oddly enough, only in proper names, like 'Gwen' or 'Gwynn.' 
Kw, ordinarily spelled qu, can have s before it and therefore belongs in 
term 6. 

The fifth term indicates that the word may begin with one of the first- 
column consonants followed by y, but only when the vowel of the word 
is u; thus we ha\e words like 'hue' {hyuw), 'cue, few, muse.' Some dia- 
lects have also tyu, dyu, and nyu (e.g., in 'tune,' 'due,' and 'new'), but I 
have set up the formula for the typical dialects of the northern United 
States, which ha\e simple tu, du, nu in these words. The sixth term 
indicates pairs that can commence a word either alone or preceded by 
s, that is, k, t, or p followed by r, also kw and pi (think of 'train, strain; 


crew, screw; quash, squash; play, splay'). The seventh term, which 
means the word can begin with s followed by any one of the consonants 
of the second column, completes the parts of the word that can precede 
its vowel. 

The terms beyond the eighth show what comes after the vowel. This 
portion is rather more complex than the beginning of the word, and it 
would take too long to explain everything in detail. The general prin- 
ciples of the symbolism will be clear from the preceding explanations. 
The ninth term, with its zero, denotes that a vowel can end the word 
if the vowel is d— which means (1) the vowel of the article 'a' and the 
exclamation 'huh?' and (2) the vowel of 'pa, ma,' and the exclamations 
'ah!' and 'bah!'— or the vowel can end the word if it is the aw sound, as 
in 'paw, thaw.' In some dialects (eastern New England, southern 
United States, South British) the vowel ending occurs in words which 
are spelled with ar, like 'car, star' (fed, sta, in these dialects), but in most 
of the United States dialects and in those of Ireland and Scotland these 
words end in an actual r. In eastern New England and South British 
dialects, but not in southern United States, these words cause a linking 
r to appear before a vowel beginning a following word. Thus for 'far 
off' your Southerner says fa of; your Bostonian and your Britisher say 
fa rof, with a liquid initial r; but most of the United States says far of, 
with a rolled-back r. For some dialects, term 9 would be different, show- 
ing another possible final vowel, namely, the peculiar sound which the 
Middle Westerner may notice in the Bostonian's pronunciation of 'fur, 
cur' (fc?, ks) and no doubt may find very queer. This funny sound is 
common in Welsh, Gaelic, Turkish, Ute, and Hopi, but I am sure 
Boston did not get it from any of these sources. 

Can one-syllable words end in e, i, o, or u? No, not in English. Tlie 
wprds so spelled end in a consonant sound, y or w. Thus, 'I,' when ex- 
pressed in formula pattern, is ay, 'we' is wiy, 'you' is yuw, 'how' is haM\ 
and so on. A comparison of the Spanish no with the English 'No!' shows 
that, whereas the Spanish word actually ends with its o sound trailing in 
the air, the English equivalent closes upon a w sound. The patterns to 
which wc are habituated compel us to close upon a consonant after most 
vowels. Hence when we learn Spanish, instead of saying como no, we 
are apt to say kowmow now; instead of si, we say our own word 'see' 
(siy). In French, instead of si beau, we are apt to say 'see bow.' 

Term 10 means that r, w, or y may be interpolated at this point 


except when the interpolation would result in joining w and y with each 
other. Term 11 means that the word may end in any single English 
consonant except h; this exception is most unlike some languages, e.g., 
Sanskrit, Arabic, Naxaho, and Maya, in which many words end in h. 
The reader can figure out terms 12, 13, and 14 if he has stuck so. far. 
A small c means ch as in 'child'; / is as in 'joy.' Term 13, which contains 
these letters, expresses the possibility of words like 'gulch, bulge, lunch, 
lounge.' Term 14 represents the pattern of words like 'health, width, 
eighth' {eytO), 'sixth, xth' (eksO). Although we can say 'nth' power or 
'fth' power, it takes effort to say the unpermitted 'sth' power or 'hth' 
power. 'Hth' would be symbolized '^eycO, the star meaning that the 
form does not occur. Term 14, however, allows both mO and mpj, the 
latter in w^ords like 'humph' or the recent 'oomph' (umpf). The ele- 
ments of term 15 may be added after anything— the t and s forms after 
voiceless sounds, the d and z after voiced sounds. Thus, 'towns' is 
tawnz, with wnz attained by term 10 plus 11 plus 15; whereas 'bounce' 
is bawns, with wns by 10 plus 12. Some of the combinations resulting 
in this way are common; others are very rare but still are possible Eng- 
lish forms. If Charlie McCarthy should pipe up in his coy way, "Thou 
oomphst, dost thou not?"; or a Shakespearean actor should thunder out, 
"Thou triumphst!" the reason would be that the formula yields that 
weird sputter mpfst by term 14 plus term 15. Neither Mr. Bergen nor 
Mr. Shakespeare has any power to vary the formula. 

The overriding factor applicable to the whole expression is a prohibi- 
tion of doubling. Notwithstanding whatever the formula says, the same 
two consonants cannot be juxtaposed. While by term 15 we can add t 
to 'flip' and get 'flipt (flipped),' we can't add t to 'hit' and get hitt. In- 
stead, at the point in the patterns where hitt might be expected, we find 
simply 'hit (I hit it yesterday, I flipt it yesterday).' Some languages, such 
as Arabic, have words like hitt, fadd, and so on, with both paired con- 
sonants distinct. The Creek Indian language permits three, e.g. nun. 

The way the patterns summarized in this formula control the forms of ~ 
English words is really extraordinary. A new monosyllable turned out, 
say, by Walter Winchell or by a plugging adman concocting a name for 
a new breakfast mush, is struck from this mold as surely as if I pulled 
the lever and the stamp came down on his brain. Thus linguistics, like 
the physical sciences, confers the power of prediction. I can predict, 
within limits, what Winchell will or won't do. He mav coin a word 


thrub, but he will not coin a word srub, for the formula cannot produce 
a ST. A different formula indicates that, if Winchell invents any word 
beginning with th, like thell or therg, the th will have the sound it has in 
'thin,' not the sound it has in 'this' or 'there.' Winchell will not invent 
a word beginning with this latter sound. 

We can wheeze forth the harshest successions of consonants if they 
are only according to the patterns producing the formula. We easily say 
'thirds' and 'sixths,' though 'sixths' has the very rough sequence of four 
consonants, ksOs. But the simpler sisths is against the patterns and so 
is harder to say. 'Glimpst (glimpsed)' has gl by term 3, i by 8, mpst by 
12 plus 15. But dlinpfk is eliminated on several counts: Term 3 allows 
for no dl, and by no possible combination of terms can one get npfk. 
Yet the linguist can say dlinpfk as easily as he can say 'glimpsed.' The 
formula allows for no final mb; so we do not say 'lamb' as it is spelled, 
but as lam. 'Land,' quite parallel but allowed by the formula, trips off 
our tongues as spelled. It is not hard to see why the "explanation," still 
found in some serious textbooks, that a language does this or that "for 
the sake of euphony" is on a par with nature's reputed abhorrence of a 
The exactness of this formula, typical of hundreds of others, shows 

>,^^that, while linguistic formulations are not those of mathematics, they 
are nevertheless precise. We might bear in mind that this formula, com- 
pared with the formulation of some of the English (or other) grammati- 
cal patterns that deal with meaning, would appear like a simple sum in 
addition compared with a page of calculus. It is usually more con- 
venient to treat very complex patterns by successive paragraphs of pre- 
cise sentences and simpler formulas, so arranged that each additional 
paragraph presupposes the previous ones, than to try to embrace all in 
one very complex formula. 
V Linguistics is also an experimental science. Its data result from long 
-Series of observations under controlled conditions, which, as they are 
systematically altered, call out definite, different responses. The experi- 
ments arc directed by the theoretic body of knowledge, just as with 
physics or chemistry. They usually do not require mechanical appara- 
tus. In place of apparatus, linguistics uses and develops techniques. 

"^Experimental need not mean quantitative. Measuring, weighing, and 
pointer-reading devices are seldom needed in linguistics, for quantity and 
number play little part in the realm of patternj where there are no 


variables but, instead, abrupt alternations from one configuration to 
another. The mathematical sciences require exact measurement, but 
what linguistics requires is, rather, exact "patternment"— an exactness of y 
relation irrespective of dimensions. Quantity, dimension, magnitude 
are metaphors since they do not properly belong in this spaceless, rela- 
tional world. I might use this simile: Exact measurement of lines and 
angles will be needed to draw exact squares or other regular polygons, 
but measurement, however precise, will not help us to draw an exact 
circle. Yet it is necessary only to discover the principle of the compass 
to reach by a leap the ability to draw perfect circles. Similarly, lin- 
guistics has developed techniques which, like compasses, enable it with- 
out any true measurement at all to specify exactly the patterns with 
which it is concerned. Or I might perhaps liken the case to the state 
of affairs within the atom, where also entities appear to alternate from 
configuration to configuration rather than to move in terms of meas- 
urable positions. As alternants, quantum phenomena must be treated 
by a method of analysis that substitutes a point in a pattern under a 
set of conditions for a point in a pattern under another set of condi- 
tions—a method similar to that used in analysis of linguistic phenomena. 

Physics and chemistry, dealing with inanimate matter, require chiefly 
inanimate apparatus and substances for their experiments. As con- 
ducted today upon a large scale, they require highly wrought physical 
equipment at every step, immense investments in physical plant. Their 
experiments are costly to conduct, both absolutely and relati\ely to the 
number of scientists. Experimental biology uses much inanimate ap- 
paratus, too, but its fundamental apparatus is its experimental animals 
and plants and their food, housing, and growth facilities. These also are 
expensive in the quantities needed. No one grudges the expense, either 
here or in the ph}sical sciences, so long as an increase in human knowl- 
edge and welfare is promised. 

The apparatus of linguistics is much less expensive than that of these 
sciences, but it, too, costs money. The experimental linguist, like the 
biologist, uses and must ha\e experimental animals. Only, his "animals" 
are human. They are his informants and must be paid for working \\-ith 
him. Sometimes he must make trips to Indian reservations or African 
villages where his informants li\e; at other times it is more economical 
to transport them to him. They provide the field for experimental in- 
vestigation. They are apparatus, not teachers. It is as important to 


study in this way languages of Indians, Africans, and other aborigines as 
it is to study the Enghsh dialects of Brooklyn, Boston, Richmond, or 

While informants are the basic apparatus, the linguist can improve 
and speed up his work with the aid of mechanical tools, just as the 
biologist studies his animals and plants with the aid of microscopes, 
X-ray machines, and other costly instruments. The linguist is aided by 
judicious use of good phonographic reproducing devices. Much could 
also be done with the help of business machines. 

Although linguistics is a very old science, its modern experimental 
phase, which stresses the analysis of unwritten speech, could be called 
one of the newest. So far as our knowledge goes, the science of lin- 
guistics was founded, or put on its present basis, by one Panini in India 
several centuries before Christ. Its earliest form anticipated its most 
recent one. Panini was highly algebraic, i.e., pattern-symbolic, in his 
treatment; he used formulas in a very modern way for expressing the 
obligatory patterns of Sanskrit. It was the Greeks who debased the 
science. They showed how infinitely inferior they were to the Hindus 
as scientific thinkers, and the effect of their muddling lasted two thou- 
sand years. Modern scientific linguistics dates from the rediscovery of 
Panini by the Western world in the early nineteenth century. 

Yet linguistics is still in its infancy so far as concerns wherewithal for 
its needed equipment, its supply of informants, and the minimum of 
tools, books, and the like. Money for mechanical aids, such as I referred 
to above, is at present only a happy dream. Perhaps this condition 
results from lack of the publicity the other sciences receive and, after all, 
fairly earn. We all know now that the forces studied by physics, chem- 
istry, and biology are powerful and important. People generally do not 
yet know that the forces studied by linguistics are powerful and impor- 
tant, that its principles control every sort of agreement and understand'- 
ing among human beings, and that sooner or later it will have to sit as 
judge while the other sciences bring their results to its court to inquire 
into what they mean. When this time comes, there will be great and 
well-equipped laboratories of linguistics as there are of other exact 


In English, the sentences 'I pull the branch aside' and 'I have an extra 
toe on my foot' have little similarity. Leaving out the subject pro- 
noun and the sign of the present tense, which are common features from 
requirements of English syntax, we may say that no similarity exists. 
Common, and even scientific, parlance would say that the sentences are 
unlike because they are talking about things which are intrinsically un- 
like. So Mr. Ever\'man, the natural logician, would be inclined to argue. 
Formal logic of an older type would perhaps agree with him. 

If, moreover, we appeal to an impartial scientific English-speaking 
observer, asking him to make direct obser\ations upon cases of the two 
phenomena to see if they may not have some element of similarity which 
we have overlooked, he will be more than likely to confirm the dicta of 
Mr. Ever)'man and the logician. The observer whom we ha\e asked to 
make the test may not see quite eye to eye with the old-school logician 
and would not be disappointed to find him wrong. Still he is compelled 
sadly to confess failure. "I wish I could oblige you," he says, ''but tn.- 
as I may, I cannot detect any similarity between these phenomena." 

By this time our stubborn streak is aroused; we wonder if a being 
from Mars would also see no resemblance. But now a linguist points 
out that it is not neccssar)- to go as far as Mars. We have not yet scouted 
around this earth to see if its many languages all classify these phe- 
nomena as disparately as our speech does. We find that in Shawnee 

* Reprinted from Technol. Rev., 43:250-252, 266, 268, 272 (April 1941). 



these two statements are, respectively, ni-l'6awa-ko-n-a and ni4'6awa- 
'ko-9ite (the 9 here denotes th as in 'thin' and the apostrophe denotes 
a breath-catch). The sentences are closely similar; in fact, they differ 
only at the tail end.- In Shawnee, moreover, the beginning of a con- 
struction is generally the important and emphatic part. Both sentences 
start with ni- {'!'), which is a mere prefix. Then comes the really im- 






? + ? 



Figure 15. Suggested above are certain linguistic concepts which, as explained in the 
text, are not easily definable. 

portant key word, VOawa, a common Shawnee term, denoting a forked 
outline, like Fig. 15, no. 1. The next element, -'ko, we cannot be sure 
of, but it agrees in form with a variant of the suffix -a'kw or -ako, de- 
noting tree, bush, tree part, branch, or anything of that general shape. 
In the first sentence, -n- means 'by hand action' and may be either a 
causation of the basic condition (forked outline) manually, an increase 
of it, or both. The final -a means that the subject (T) does this action 
to an appropriate object. Hence the first sentence means 'I pull it 
(something like branch of tree) more open or apart where it forks.' In 
the other sentence, the suffix -9ite means 'pertaining to the toes,' and the 
absence of further suffixes means that the subject manifests the condi- 
tion in his own person. Therefore the sentence can mean only 'I have 
an extra toe forking out like a branch from a normal toe.' 



Shawnee logicians and obser\crs woul.d class the two phenomena as 
intrinsically similar. Our own observer, to whom we tell all this, focuses 
his instruments again upon the two phenomena and to his joy sees at 
once a manifest resemblance. Figure 16 illustrates a similar situation: 
'I push his head back' and 'I drop it in water and it floats,' though very 
dissimilar sentences in English, are similar in Shawnee. The point of 


kwaSkwi (or kwo S k) 


$ = sh 

a = ni kwa5kwi-tepe-n-a 


i- -ho- 



n i - kvKoS k - ho - to 


Figure 16. The English sentences 'I push his head back' and T drop it in water 

and it floats' are unlike. But in Shawnee the corresponding statements are closely 

similar, emphasizing the fact that analysis of nature and classification of events as 

like or in the same category (logic) are go\crned by grammar. 

view of linguistic relativity changes Mr. Everyman's dictum: Instead of 
saj'ing, "Sentences are unlike because they tell about unlike facts," he 
now reasons: "Facts are unlike to speakers whose language background 
provides for unlike formulation of them." 

Conversely, the English sentences, 'The boat is grounded on the 
beach' and 'The boat is manned by picked men,' seem to us to be rather 
similar. Each is about a boat; each tells the relation of the boat to other 
objects— or that's our story. The linguist would point out the paral- 
lelism in grammatical pattern thus: "Tlie boat is xed preposition y." 
The logician might turn the linguist's analysis into "A is in the state x 
in relation to y," and then perhaps into fA = xRy. Such symbolic 
methods lead to fruitful techniques of rational ordering, stimulate our 


thinking, and bring valuable insight. Yet we should realize that the 
similarities and contrasts in the original sentences, subsumed under the 
foregoing formula, are dependent on the choice of mother tongue and 
that the properties of the tongue are eventually reflected as peculiarities 
of structure in the fabric of logic or mathematics which we rear. 

In the Nootka language of Vancouver Island, the first "boat" state- 
ment is tlih-is-ma; the second, lash-tskwiq-ista-ma. The first is thus 
l-U-ma; the second, Ill-IV-V-md; and they are quite unlike, for the final 
-ma is only the sign of the third-person indicative. Neither sentence 
contains any unit of meaning akin to our word 'boat' or even 'canoe.' 
Part I, in the first sentence, means 'moving pointwise,' or moving in a 
way like the suggestion of the outline in Fig. 15, no. 2; hence 'traveling 
in or as a canoe,' or an event like one position of such motion. It is not 
a name for what we should call a "thing," but is more like a vector in 
physics. Part II means 'on the beach'; hence l-U-ma means 'it is on the 
beach pointwise as an event of canoe motion,' and would normally refer 
to a boat that has come to land. In the other sentence, part III means 
'select, pick,' and IV means 'remainder, result,' so that III-IV means 
'selected.' Part V means 'in a canoe (boat) as crew.' The whole, 
Ul-YV-V-ma, means either 'they are in the boat as a crew of picked 
men' or 'the boat has a crew of picked men.' It means that the whole 
event involving picked ones and boat's crew is in process. 

As a hang-over from my education in chemical engineering, I relish 
an occasional chemical simile. Perhaps readers will catch what I mean 
when I say that the way the constituents are put together in these sen- 
tences of Shawnee and Nootka suggests a chemical compound, whereas 
their combination in English is more like a mechanical mixture. A mix- 
ture, like the mountaineer's potlicker, can be assembled out of almost 
anything and does not make any sweeping transformation of the o\ert 
appearance of the material. A chemical compound, on the other hand, 
can be put together only out of mutually suited ingredients, and the 
result may be not merely soup but a crop of crystals or a cloud of smoke. 
Likewise the typical Shawnee or Nootka combinations appear to work 
with a vocabulary of terms chosen with a view not so much to the utility 
of their immediate references as to the ability of the terms to combine 
suggestively with each other in manifold ways that elicit novel and useful 
images. This principle of terminology and way of analyzing events 
would seem to be unknown to the tongues with which we are familiar. 


It is the analysis of nature down to a basic vocabulary capable of this 
sort of evocative recombination which is most distinctive of polysyn- 
thetic languages, like Nootka and Shawnee. Their characteristic quality 
is not, as some linguists have thought, a matter of t4ie tightness or in- 
dissolubility of the combinations. The Shawnee term VOawa could 
probably be said alone but would then mean 'it (or something) is forked,' 
a statement which gives little hint of the novel meanings that arise out 
of its combinations— at least to our minds or oar type of logic. Shawnee 
and Nootka do not use the chemical type of synthesis exclusively. They 
make large use of a more external kind of syntax, which, however, has 
no basic structural priority. Even our own Indo-European tongues are 
not wholly devoid of the chemical method, but they seldom make sen- 
tences by it, afford little inkling of its possibilities, and gi\e structural 
priority to another method. It was quite natural, then, that Aristotle 
should found our traditional logic wholly on this other method. 

Let me make another analogy, not with chemistry but with art— art 
of the pictorial sort. We look at a good still-life painting and seem to 
see a lustrous porcelain bowl and a downy peach. Yet an analysis that 
screened out the totality of the picture— as if we were to go over it care- 
fully, looking through a hole cut in a card— would re\eal only oddly 
shaped patches of paint and would not evoke the bowl and fruit. The 
synthesis presented by the painting is perhaps akin to the chemical type 
of syntax, and it may point to psychological fundamentals that enter 
into both art and language. Now the mechanical method in art and 
language might be typified by no. 3A in Fig. 15. The first element, a 
field of spots, corresponds to the adjecti\'e 'spotted,' the second corre- 
sponds to the noun 'cat.' By putting them together, we get 'spotted 
cat.' Contrast the technique in Fig. 15, no. 3B. Here the figure cor- 
responding to 'cat' has only vague meaning by itself— "che\ron-like," we 
might say— while the first element is even \'aguer. But, combined, these 
evoke a cylindrical object, like a shaft casting. 

The thing common to both techniques is a systematic synthetic use 
of pattern, and this is also common to all language techniques. I ha\e 
put question marks below the elements in Fig. 15, no. 3B, to point 
out the difficulty of a parallel in English speech and the fact that the 
method probably has no standing in traditional logic. Yet examination 
of other languages and the possibility of new types of logic that has been 
advanced by modern logicians themselves suggest that this matter may 


be significant for modern science. New types of logic may help us 
eventually to understand how it is that electrons, the velocity of light, 
and other components of the subject matter of physics appear to behave 
illogically, or that phenomena which flout the sturdy common sense of 
yesteryear can nevertheless be true. Modern thinkers have long since 
pointed out that the so-called mechanistic way of thinking has come to 
an impasse before the great frontier problems of science. To rid our- 
selves of this way of thinking is exceedingly difficult when we have no 
linguistic experience of any other and when even our most advanced 
logicians and mathematicians do not provide any other— and obviously 
they cannot without the linguistic experience. For the mechanistic way 
of thinking is perhaps just a type of syntax natural to Mr. Everyman's 
daily use of the western Indo-European languages, rigidified and intensi- 
fied by Aristotle and the latter's medieval and modern followers. 

As I said in an article, "Science and linguistics," in the Review for 
April 1940, the effortlessness of speech and the subconscious way we 
picked up that activity in early childhood lead us to regard talking and 
thinking as wholly straightforward and transparent. We naturally feel 
that they embody self-evident laws of thought, the same for all men. 
We know all the answers! But, when scrutinized, they become dusty 
answers. We use speech for reaching agreements about subject matter: 
I say, "Please shut the door," and my hearer and I agree that 'the door' 
refers to a certain part of our environment and that I want a certain 
result produced. Our explanations of how we reached this understand- 
ing, though quite satisfactory on the everyday social plane, are merely 
more agreements (statements) about the same subject matter (door, and 
so on), more and more amplified by statements about the social and per- 
sonal needs that impel us to communicate. There are here no laws of 
thought. Yet the structural regularities of our sentences enable us to 
sense that laws are somewhere in the background. Clearly, explana- 
tions of understanding such as "And so I ups and says to him, says I; 
see here, why don't you . . . !" evade the true process by which 'he' 
and T are in communication. Likewise psychological-social descrip- 
tions of the social and emotional needs that impel people to communi- 
cate with their fellows tend to be learned versions of the same method 
and, while interesting, still evade the question. In similar case is evasion 
of the question by skipping from the speech sentence, via physiology and 
"stimuli" to the social situation. 


The WHY of understanding may remain for a long time mysterious; 
but the HOW or logic of understanding— its background of laws or regu- 
larities—is discoverable. It is the grammatical background of our mother 
tongue, which includes not only our way of constructing propositions 
but the way we dissect nature and break up the flux of experience into 
objects and entities to construct propositions about. This fact is im- 
portant for science, because it means that science can have a rational 
or logical basis even though it be a relativistic one and not Mr. Every- 
man's natural logic. Although it may vary with each tongue, and a 
planetary mapping of the dimensions of such variation may be necessi- 
tated, it is, ne\ertheless, a basis of logic with discoverable laws. Science 
is not compelled to see its thinking and reasoning procedures turned 
into processes merely subservient to social adjustments and emotional 

Moreover, the tremendous importance of language cannot, in my 
opinion, be taken to mean necessarily that nothing is back of it of the 
nature of what has traditionally been called "mind." My own studies 
suggest, to me, that language, for all its kingly role, is in some sense a 
superficial embroidery upon deeper processes of consciousness, which are 
necessary before any communication, signaling, or symbolism whatso- 
ever can occur, and which also can, at a pinch, effect communication 
(though not true agreement) without language's and without sym- 
bolism's aid. I mean "superficial" in the sense that all processes of 
chemistr}', for example, can be said to be superficial upon the deeper 
layer of physical existence, which we know variously as intra-atomic, 
electronic, or subelectronic. No one would take this statement to mean 
that chemistry is unimportant— indeed the whole point is that the 
more superficial can mean the more important, in a definite operative 
sense. It may c\cn be in the cards that there is no such thing as "Lan- 
guage" (with a capital L) at all! The statement that "thinking is a 
matter of language" is an incorrect generalization of the more nearly 
correct idea that "thinking is a matter of different tongue?^' The dif- 
ferent tongues are the real phenomena and may generalize down not to 
any such universal as "Language," but to something better— called "sub- 
linguistic" or "superlinguistic"— and not altogether unlike, even if 
much unlike, what we now call "mental." This generalization would 
not diminish, but would rather increase, the importance of intertongue 
study for investigation of this realm of truth. 


Botanists and zoologists, in order to understand the world of living 
species, found it necessary to describe the species in every part of the 
globe and to add a time perspective by including the fossils. Then they 
found it necessary to compare and contrast the species, to work out 
families and classes, evolutionary descent, morphology, and taxonomy. 
In linguistic science a similar attempt is under way. The far-off event 
toward which this attempt moves is a new technology of language and 
thought. Much progress has been made in classifying the languages of 
earth into genetic families, each having descent from a single precursor, 
and in tracing such developments through time. The result is called 
"comparative linguistics." Of even greater importance for the future 
technology of thought is what might be called "contrastive linguistics." 
This plots the outstanding differences among tongues— in grammar, 
logic, and general analysis of experience. 

As I said in the April 1940 Review, segmentation of nature is an 
aspect of grammar— one as yet little studied by grammarians. We cut 
up and organize the spread and flow of events as we do, largely because, 
through our mother tongue, we are parties to an agreement to do so, 
not because nature itself is segmented in exactly that way for all to see. 
Languages differ not only in how they build their sentences but also in 
how they break down nature to secure the elements to put in those 
sentences. This breakdown gives units of the lexicon. "Word" is not 
a very good "word" for them; "lexeme" has been suggested, and "term" 
will do for the present. By these more or less distinct terms we ascribe 
a semifictitious isolation to parts of experience. English terms, like 'sky, 
hill, swamp,' persuade us to regard some elusive aspect of nature's end- 
less variety as a distinct thing, almost like a table or chair. ''Thus Eng- 
lish and similar tongues lead us to think of the universe as a collection 
of rather distinct objects and events corresponding to words. Indeed 
this is the implicit picture of classical physics and astronomy— that the 
universe is essentially a collection of detached objects of different sizes. 

The examples used by older logicians in dealing with this point are 
usually unfortunately chosen. They tend to pick out tables and chairs 
and apples on tables as test objects to demonstrate the object-like nature 
of reality and its one-to-one correspondence with logic. Man's artifacts 
and the agricultural products he severs from living plants have a unique 
degree of isolation; we may expect that languages will have fairly isolated 
terms for them. The real question is: What do different languages do, 


not with these artificially isolated objects but with the flowing face of 
nature in its motion, color, and changing form; with clouds, beaches, 
and yonder flight of birds? For, as goes our segmentation of the face 
of nature, so goes our physics of the Cosmos. 

Here we find differences in segmentation and selection of basic terms. 
We might isolate something in nature by saying 'It is a dripping spring.' 
Apache erects the statement on a verb ga: 'be white (including clear, 
uncolored, and so on).' With a prefix no- the meaning of downward 
motion enters: 'whiteness moves downward.' Then to, meaning both 
'water' and 'spring' is prefixed. The result corresponds to our 'dripping 
spring,' but synthetically it is 'as water, or springs, whiteness moves down- 
ward.' How utterly unlike our way of thinking! The same verb, ga, 
with a prefix that means 'a place manifests the condition' becomes 
gohlga: 'the place is white, clear; a clearing, a plain.' These examples 
show that some languages have means of expression— chemical combi- 
nation, as I called it— in which the separate terms are not so separate as 
in English but flow together into plastic synthetic creations. Hence 
such languages, which do not paint the separate-object picture of the 
universe to the same degree as English and its sister tongues, point 
toward possible new types of logic and possible new cosmical pictures. 

The Indo-European languages and many others give great prominence 
to a type of sentence having two parts, each part built around a class of 
word— substantives and verbs— which those languages treat differently 
in grammar. As I showed in the April 1940 Review, this distinction is 
not drawn from nature; it is just a result of the fact that every tongue 
must have some kind of structure, and those tongues have made a go 
of exploiting this kind. The Greeks, especially Aristotle, built up this 
contrast and made it a law of reason. Since then, the contrast has been 
stated in logic in many different ways: subject and predicate, actor and 
action, things and relations between things, objects and their attributes, 
quantities and operations. And, pursuant again to grammar, the notion 
became ingrained that one of these classes of entities can exist in its 
own right but that the verb class cannot exist without an entity of the 
other class, the "thing" class, as a peg to hang on. "Embodiment is 
necessary," the watchword of this ideology, is seldom strongly ques- 
tioned. Yet the whole trend of modern physics, with its emphasis on 
"the field," is an implicit questioning of the ideology. This contrast 
crops out in our ifiathematics as two kinds of symbols— the kind like 1, 


2, 3, X, y, z and the kind like +, — , -^, V~ , log — > though, in view of 
0> ¥27 %j TT, and others, perhaps no strict two-group classification holds. 
The two-group notion, however, is always present at the back of the 
thinking, although often not overtly expressed. 

Our Indian languages show that with a suitable grammar we may 
have intelligent sentences that cannot be broken into subjects and predi- 
cates. Any attempted breakup is a breakup of some English translation 
or paraphrase of the sentence, not of the Indian sentence itself. We 
might as well try to decompose a certain synthetic resin into Celluloid 
and whiting because the resin can be imitated with Celluloid and whit- 
ing. The Algonkian language family, to which Shawnee belongs, does 
use a type of sentence like our subject and predicate but also gives prom- 
inence to the type shown by our examples in the text and in Fig. 
15. To be sure, ni- is represented by a subject in the translation but 
means 'my' as well as 'I,' and the sentence could be translated thus: 
'My hand is pulling the branch aside.' Or ni- might be absent; if so, 
we should be apt to manufacture a subject, like 'he, it, somebody,' or 
we could pick out for our English subject an idea corresponding to any 
one of the Shawnee elements. 

When we come to Nootka, the sentence without subject or predicate 
is the only type. The term "predication" is used, but it means "sen- 
tence." Nootka has no parts of speech; the simplest utterance is a sen- 
tence, treating of some event or event-complex. Long sentences are 
sentences of sentences (complex sentences), not just sentences of words. 
In Fig. 17 we have a simple, not a complex, Nootka sentence. The 
translation, 'he invites people to a feast,' splits into subject and predi- 
cate. Not so the native sentence. It begins with the event of 'boiling 
or cooking,' tl'imsh; then comes -ya ('result') = 'cooked'; then -'is 'eat- 
ing' = 'eating cooked food'; then -ita ('those who do') = 'eaters of 
cooked food'; then -itl ('going for'); then -ma, sign of third-person in- 
dicative, giving tVimshyaisitaitlma, which answers to the crude para- 
phrase, 'he, or somebody, goes for (invites) eaters of cooked food.' 

The English technique of talking depends on the contrast of two 
artificial classes, substantives and verbs, and on the bipartitioned ideol- 
ogy of nature, already discussed. Our normal sentence, unless impera- 
tive, must have some substantive before its verb, a requirement that 
corresponds to the philosophical and also naive notion of an actor who 
produces an action. This last might not have been so if English had 



had thousands of verbs hke 'hold,' denoting positions. But most of our 
verbs follow a type of segmentation that isolates from nature what we 
call "actions," that is, moving outlines. 

Following majority rule, we therefore read action into every sentence, 
even into 'I hold it.' A moment's reflection will show that 'hold' is no 
action but a state of relative positions. Yet we think of it and even see 
it as an action because language formulates it in the same way as it 









ITA — 'ITL — MA 

'^' i m s y a -■ i s I t a -' i "\ m a 

Figure 17. Here are shown the different ways in which Enghsh and Nootka for 

mulate the same event. The Enghsh sentence is divisible into subject and predicate; 

the Nootka sentence is not, yet it is complete and logical. Furthermore, the Nootka 

sentence is just one word, consisting of the root tl'imsh \\ith fi-\e sufHxes. 

formulates more numerous expressions, like 'I strike it,' which deal with 
movements and changes. 

We are constantly reading into nature fictional acting entities, simply 
because our verbs must ha\e substantives in front of them. We ha\c 
to say 'It flashed' or 'A light flashed,' setting up an actor, 'it' or 'light,' 
to perform what we call an action, "to flash." Yet the flashing and 
the light are one and the same! The Hopi language reports the flash 
with a simple verb, rehpi: 'flash (occurred).' There is no division into 
subject and predicate, not even a suffix like -t of Latin tona-t 'it 
thunders.' Hopi can and does ha\e verbs without subjects, a fact 
which may give that tongue potentialities, probabl}- ne\er to be de\el- 
oped, as a logical system for understanding some aspects of the universe. 
Undoubtedly modern science, strongly reflecting western Indo-Euro- 


pean tongues, often does as we all do, sees actions and forces where it 
sometimes might be better to see states. On the other hand, 'state' is 
a noun, and as such it enjoys the superior prestige traditionally attaching 
to the subject or thing class; therefore science is exceedingly ready to 
speak of states if permitted to manipulate the concept like a noun. 
Perhaps, in place of the 'states' of an atom or a dividing cell, it would 
be better if we could manipulate as readily a more verblike concept but 
without the concealed premises of actor and action. 

I can sympathize with those who say, "Put it into plain, simple Eng- 
lish," especially when they protest against the empty formalism of load- 
ing discourse with pseudolearned words. But to restrict thinking to the 
patterns merely of English, and especially to those patterns which repre- 
sent the acme of plainness in English, is to lose a power of thought 
which, once lost, can never be regained. It is the "plainest" English 
which contains the greatest number of unconscious assumptions about 
nature. This is the trouble with schemes like Basic English, in which 
an eviscerated British English, with its concealed premises working 
harder than ever, is to be fobbed off on an unsuspecting world as the 
substance of pure Reason itself. We handle even our plain English 
with much greater effect if we direct it from the vantage point of a 
multilingual awareness. For this reason I believe that those who en- 
vision a future world speaking only one tongue, whether English, Ger- 
man, Russian, or any other, hold a misguided ideal and would do the 
evolution of the human mind the greatest disservice. Western culture 
has made, through language, a provisional analysis of reality and, with- 
out correctives, holds resolutely to that analysis as final. The only 
correctives lie in all those other tongues which by aeons of inde- 
pendent evolution have arrixed at different, but equally logical, pro- 
visional analyses. 

In a valuable paper, "Modern logic and the task of the natural 
sciences," Harold N. Lee says: "Those sciences whose data are subject 
to quantitative measurement have been most successfully developed 
because we know so little about order systems other than those exempli- 
fied in mathematics. Wc can say with certainty, however, that there 
are other kinds, for the advance of logic in the last half century has 
clearly indicated it. We may look for advances in many lines in sciences 
at present well founded if the advance of logic furnishes adequate knowl- 
edge of other order types. We may also look for many subjects of in- 


quiry whose methods are not strictly scientific at the present time to 
become so when new order systems are available." ^ To which may be 
added that an important field for the working out of new order systems, 
akin to, yet not identical with, present mathematics, lies in more pene- 
trating investigation than has yet been made of languages remote in type 
from our own. 

^ Sigma Xi Quart., 28:125 (Autumn 1940). 


Tt needs but half an eye to see in these latter days that science, the 
-*' Grand Revelator of modern Western culture, has reached, without 
having intended to, a frontier. Either it must bury its dead, close its 
ranks, and go forward into a landscape of increasing strangeness, replete 
with things shocking to a culture-trammeled understanding, or it must 
become, in Claude Houghton's expressive phrase, the plagiarist of its 
own past. The frontier was foreseen in principle very long ago, and 
given a name that has descended to our day clouded with myth. That 
name is Babel. For science's long and heroic effort to be strictly factual 
has at last brought it into entanglement with the unsuspected facts of 
the linguistic order. These facts the older classical science had never 
admitted, confronted, or understood as facts. Instead they had entered 
its house by the back door and had been taken for the substance of 
Reason itself. 

What we call "scientific thought" is a specialization of the western 
Indo-European type of language, which has developed not only a set of 
different dialectics, but actually a set of different dialects, these dia- 
lects ARE NOW becoming MUTUALLY UNINTELLIGIBLE. The term 'spaCC/ 

for instance, does not and cannot mean the same thing to a psycholo- 
-gjst as to a physi cist. Even if psychologists should firmly resolve, come 

* Reprinted by permission of the Theosophical Society from Theosophist (Madras, 
India), January and April issues, 1942. 



hell or high water, to use "space" only with the physicist's meaning, they 
could not do so, any more than Englishmen could use in English the 
word 'sentiment' in the meanings which the similarly spelled but func- 
tionally different French utterance le sentiment has in its native French. 
Now this does not simply breed confusions of mere detail that an 
expert translator could perhaps resolve. It does something much more 
perplexing. Every language and every well-knit technical sublanguage 
incorporates certain points of view and certain patterned resistances to 
widely divergent points of view. This is especially so if language is not 
surveyed as a planetary phenomenon, but is as usual taken for granted, 
and the local, parochial species of it used by the individual thinker is 
taken to be its full sum. Tliese resistances not only isolate artificially 
the particular sciences from each other; they also restrain the scientific 
spirit as a whole from taking the next great step in development— a step 
which entails viewpoints unprecedented in science and a complete sev- 
erance from traditions. For certain linguistic patterns rigidified in the 
dialectics of the sciences— often also embedded in the matrix of Euro- 
pean culture from which those sciences have sprung, and long wor- 
shipped as pure Reason per se— have been worked to death. Even 
science senses that they are somehow out of focus for observing what 
may be very significant aspects of reality, upon the due observation of 
which all further progress in understanding the universe may hinge. 

Thus one of the important coming steps for Western knowledge is a 
re-examination of the linguistic backgrounds of its thinking, and for 
that matter of all thinking. My purpose in developing this subject 
before a Theosophical audience is not to confirm or affirm any Theo- 
sophical doctrines. It is rather that, of all groups of people with whom 
I have come in contact, Theosophical people seem the most capable of 
becoming excited about ideas— new ideas. And my task is to explain 
an idea to all those who, if Western culture survives the present welter 
of barbarism, may be pushed by events to leadership in reorganizing the 
whole human future. 

This idea is one too drastic to be penned up in a catch phrase. I 
would rather leave it unnamed. It is the view that a noumenal world— 
a world of hyperspace, of higher dimensions— awaits discover}' by all 
the sciences, which it will unite and unify, awaits discovery under its 
first aspect of a realm of patterned relations, inconceivably manifold 


and yet bearing a recognizable affinity to the rich and systematic organi- 
zation of LANGUAGE, including au fond mathematics and music, which 
are ultimately of the same kindred as language. The idea is older than 
Plato, and at the same time as new as our most revolutionary thinkers. 
It is implied in Whitehead's world of prehensive aspects, and in rela- 
tivity physics with its four-dimensional continuum and its Riemann- 
Christoffel tensor that sums up the properties of the world at any 
point-moment; while one of the most thought-provoking of all modern 
presentations, and I think the most original, is the Tertium Organum of 
Ouspensky. All that I have to say on the subject that may be new is 
of the premonition in language of the unknown, vaster world— that 
world of which the physical is but a surface or skin, and yet which we 
ARE in, and belong to. For the approach to reality through mathe- 
matics, which modern knowledge is beginning to make, is merely the 
approach through one special case of this relation to language. 

This view implies that what I have called patterns are basic in a really 
cosmic sense, and that patterns form wholes, akin to the Gestalten of 
psychology, which are embraced in larger wholes in continual progres- 
sion. Thus the cosmic picture has a serial or hierarchical character, that 
of a progression of planes or levels. Lacking recognition of such serial 
order, different sciences chop segments, as it were, out of the world, 
segments which perhaps cut across the direction of the natural levels, 
or stop short when, upon reaching a major change of level, the phe- 
nomena become of quite different type, or pass out of the ken of the 
older observational methods. 

But in the science of linguistics, the facts of the linguistic domain 
compel recognition of serial planes, each explicitly given by an order of 
patterning observed. It is as if, looking at a wall covered with fine 
tracery of lacelike design, we found that this tracery served as the ground 
for a bolder pattern, yet still delicate, of tiny flowers, and that upon 
becoming aware of this floral expanse we saw that multitudes of gaps 
in it made another pattern like scrollwork, and that groups of scrolls 
made letters, the letters if followed in a proper sequence made words, 
the words were aligned in columns which listed and classified entities, 
and so on in continual cross-patterning until we found this wall to be— 
a great book of wisdom! 

First, the plane "below" the strictly linguistic phenomena is a physical, 
acoustic one, phenomena wrought of sound waves; then comes a level 



of J3atter ning_in rippling muscles and spee ch nrgan , s , the physiological- 
phonetic plane; then the ph onemic^ plag g, patterning that mak eg.^ 
systenjatic-set-oLcon sonants, vowels, accen ts, tones, efc. f or^ea c^Jan- 
gLiage; then the morpTrsphuTTCmic'^ane in which the "phonemes" of 
the previous level appear combinednnfo "mdrplTemes" {words and sub- 
words like suffixes, rteT)~tb€irthe-pkfl e of morpl iotogy; then that of the 
intricate, largely uncDTtseiQ4t&-pa ^crning th -a^-gees by the meaningless 
name of syntax; then on to further planes still, the full import of which 
may some day strike and stagger us. 

^ Speech is the best show man puts on/' It is his own "act" on the (^^^^ 
stage of evolution, in which he comes before the cosmic backdrop and 
really "does his stuff." But we suspect the watching Gods perceive that 
the order Jn which his amazing set of tricks builds up to a great climax 
has been stolen— from the Universe! 

The idea, entirely unfamiliar to the modern world, that nature and 
language are inwardl}' akin, was for ages well known to various high 
cultures whose historical continuity on the earth has been enormously 
longer than that of Western European culture. In India, one aspect 
of it has been the idea of the mantram and of a mantric art. On the 
simplest cultural le\ el, a manti3J»-4s-«ief€ly uii incantatidnorprimitive 
magic, such as the crudest cultures have. In the high culture it may 
have a different, a very intellectual meaning, dealing with the inner 
affinity of language and the cosmic order. At a still higher level, it 
becomes "Mantra Yoga." Therein the mantram becomes a manifold 
of conscious patterns, contrived to assist the consciousness into the 
noumenal pattern world— whereupon it is "in the driver's seat." It can 
then SET the human organism to transmit, control, and amplify a thou- 
sandfold forces which that organism normally transmits only at un- 
observably low intensities. 

Somewhat analogously, the mathematical formula that enables a 
physicist to adjust some coils of wire, tinfoil plates, diaphragms, and 
other quite inert and innocent gadgets into a configuration in which 
they can project music to a far countr)- puts the physicist's consciousness 
on to a level strange to the untrained man, and makes feasible an ad- 
justment of matter to a very strategic configuration, one which makes 
possible an unusual manifestation of force. Other formulas make pos- 
sible the strategic arrangement of magnets and wires in the powerhouse 
so that, when the magnets (or rather the field of subtle forces, in and 
around the magnets) are set in motion, force is manifested in the way 


we call an electric current. We do not think of the designing of a radio 
station or a power plant as a linguisti c proces s, but it is one nonethe- 
less. The necessary mathematics is a linguistic apparatus, and, without 
its correct specification of essential patterning, the assembled gadgets 
would be out of proportion and adjustment, and would remain inert. 
But the mathematics used in such a case is a speclvlized formula-lan- 
guage, contrived for making available a specialized type of force mani- 
festation through metallic bodies only, namely, electricity as we today 
define what we call by that name. The mantric formula-language is 
specialized in a different way, in order to make available a different type 
of force manifestation, by repatterning states in the nervous system and 
glands— or again rather in the subtle "electronic" or "etheric" forces in 
and around those physical bodies. Those parts of the organism, until 
such strategic patterning has been effected, are merely "innocent gadg- 
ets," as incapable of dynamic power as loose magnets and loose wires, 
but IN THE PROPER PATTERN they are something else again— not to be 
understood from the properties of the unpatterned parts, and able to 
amplify and activate latent forces. 

In this way I would link the subtle Eastern ideas of the mantric and 
yogic use of language with the configurative or pattern aspect which is 
so basic in language. But this brings me to the most important part of 
my discussion. We must find ou t more jboutj anguagel . Already we 
know enough about it to know it is not what the great majority of men, 
lay or scientific, think it is. The fact that we talk almost effortlessly, 
unaware of the exceedingly complex mechanism we are using, creates 
an illusion. We think we know how it is done, that there is no mys- 
tery; we have all the answers. Alas, what wrong answers! It is like the 
way a man's uncorrected sense impressions give him a picture of the 
universe that is simple, sensible, and satisfying, but very wide of the 

Consider how the world appears to any man, however wise and ex- 
perienced in human life, who has never heard one word of what science 
has discovered about the Cosmos. To him the earth is flat; the sun 
and moon are shining objects of small size that pop up daily above an 
eastern rim, move through the upper air, and sink below a western edge; 
obviously they spend the night somewhere underground. The sky is an 
inverted bowl made of some blue material. The stars, tiny and rather 


near objects, seem as if they might be alive, for they "come out" from 
the sky at evening hke rabbits or rattlesnakes from their burrows, and 
slip back again at dawn. "Solar system" has no meaning to him, and 
the concept of a "law of gra\'itation" is quite unintelligible— nay, even 
nonsensical. For him bodies do not fall because of a law of gravitation, 
but rather "because there is nothing to hold them up"— i.e., because he 
cannot imagine their doing anything else. He cannot concei\e space 
without an "up" and "down" or e\en without an "east" and "west" in 
it. For him the blood does not circulate; nor does the heart pump 
blood; he thinks it is a place where love, kindness, and thoughts are 
kept. Cooling is not a removal of heat but an addition of "cold"; leaves 
are green not from the chemical substance chloroph}ll in them, but 
from the "greenness" in them. It will be impossible to reason him out 
of these beliefs. He will assert them as plain, hard-headed common 
sense; which means that they satisfy him because they are completely 
adequate as a system of communication between him and his fellow 
men. That is, they are adequate linguistically to his social needs, and 
will remain so until an additional group of needs is felt and is worked 
out in language. 

But as this man is in conception of the physical universe, of whose 
scope and order he has not the faintest inkling, so all of us, from rude 
savage to learned scholar, are in conception of language. Only the 
science of linguistics has begun to penetrate a little into this realm, its 
findings still largely unknown to the other disciplines. Natural man, 
whether simpleton or scientist, knows no more of the linguistic forces 
that bear upon him than the savage knows of gravitational forces. He 
supposes that talking is an activity in which he is free and untrammeled. 
He finds it a simple, transparent activity, for which he has the necessary 
explanations. But these explanations turn out to be nothing but state- 
ments of the needs that impel him to communicate. They are not 
germane to the process by which he communicates. Thus he will say 
that he thinks something, and supplies words for the thoughts "as they 
come." But his explanation of why he should have such and such 
thoughts before he came to utter them again turns out to be merely the 
story of his social njeds at that moment. It is a dusty answer that 
throws no light. But then he supposes that there need be no light 
thrown on this talking process, since he can manipulate it anyhow quite 
well for his social needs. Thus he implies, wrongly, that thinking is an 


OBVIOUS, straightforward activity, the same for all rational beings, of 
which language is the straightforward expression. 

Actually, thinking is most mysterious, and by far the greatest light 
upon it that we have is thrown by the study of language. This study 
shows that the forms of a person's thoughts are controlled by inexorable 
laws of pattern of which he is unconscious. These patterns are the un- 
perceived intricate systematizations of his own language— shown readily 
enough by a candid comparison and contrast with other languages, 
especially those of a different linguistic family. His thinking itself is 
in a language— in English, in Sanskrit, in Chinese.^ And every language 
is a vast pattern-system, different from others, in which are culturally 
ordained the forms and categories by which the personality not only 
communicates, but also analyzes nature, notices or neglects types of 
relationship and phenomena, channels his reasoning, and builds the 
house of his consciousness. 

This doctrine is new to Western science, but it stands on unimpeach- 
able evidence. Moreover, it is known, or something like it is known, to 
the philosophies of India and to modern Theosophy. This is masked 
by the fact that the philosophical Sanskrit terms do not supply the exact 
equivalent of my term "language" in the broad sense of the linguistic 
order. The linguistic order embraces all symbolism, all symbolic proc- 
esses, all processes of reference and of logic. Terms like Ndma refer 
rather to subgrades of this order- the lexical level, the phonetic level. 
The nearest equivalent is probably Manas, to which our vague word 
'mind' hardly does justice. Manas in a broad sense is a major hierarchi- 
cal grade in the world-structure— a "manasic plane" as it is indeed ex- 
plicitly called. Here again "mental plane" is apt to be misleading to an 
English-speaking person. English "mental" is an unfortunate word, a 
word whose function in our culture is often only to stand in lieu of an 
intelligent explanation, and which connotes rather a foggy limbo than 
a cosmic structural order characterized by patterning. Sometimes Manas 

1 To anticipate the text, "thinking in a language" docs not necessarily have to use 
WORDS. An uncultivated Choctaw can as easily as the most skilled litterateur contrast 
the tenses or the genders of two experiences, though he has never heard of any words 
like "tense" or "gender" for such contrasts. Much thinking never brings in words at 
all, but manipulates whole paradigms, word-classes, and such grammatical orders 
"behind" or "above" the focus of personal consciousness. 


is used to mean, however, simply the personal psyche; this according to 
Mr. Fritz Kunz is the case in the famous saying of The Voice of the 
Silence: "The mind is the great slayer of the real." 

It is said that in the plane of Manas there are two great levels, called 
the Rupa and Arupa levels. The lower is the realm of "name and form," 
Ndma and Rupa. Here "form" means organization in space ("our" 
three-dimensional space). This is far from being coextensive with pat- 
tern in a universal sense. And Ndma, 'name,' is not language or the 
linguistic order, but only one level in it, the level of the process of 
"lexation" or of giving words (names) to parts of the whole manifold 
of experience, parts which are thereby made to stand out in a semi- 
fictitious isolation. Thus a word like 'sky,' which in English can be 
treated like 'board' (the sky, a sky, skies, some skies, piece of sky, etc.), 
leads us to think of a mere optical apparition in ways appropriate only 
to relatively isolated solid bodies. 'Hill' and 'swamp' persuade us to 
regard local variations in altitude or soil composition of the ground as 
distinct things almost like tables and chairs. Each language performs 
this artificial chopping up of the continuous spread and flow of existence 
in a different way. Words and speech are not the same thing. As we 
shall see, the patterns of sentence structure that guide words are more 
important than the words. 

Thus the level of Rupa and Ndmd— shape-segmentation and vocabu- 
lary—is part of the linguistic order, but a somewhat rudimentary and 
not self-sufEcient part. It depends upon a higher level of organization, 
the level at which its combinatory scheme appears. This is the Arupa 
level— the pattern world par excellence. Arupa, 'formless/ does not 
mean without linguistic form or organization, but without reference to 
spatial, visual shape, marking out in space, which as wc saw with 'hill' 
and 'swamp' is an important feature of reference on the lexical level. 
Arupa is a realm of patterns that can be "actualized" in space and time 
in the materials of lower planes, but are themselves indifferent to space 
and time. Such patterns are not like the meanings of words, but they 
are somewhat like the way meaning appears in sentences. Tliey are not 
like individual sentences but like schemes of sentences and designs of 
sentence structure. Our personal conscious "minds" can understand 
such patterns in a limited way by using mathematical or grammatical 
FORiMULAS into which words, values, quantities, etc., can be substituted. 
A rather simple instance will be given presently. 


It is within the possibiHties of the "culture of consciousness" that the 
Arupa level of the "mental" plane may be contacted directly in an ex- 
pansion of consciousness. In Ouspensky's book, A New Model of the 
Universe, there are arresting glimpses of extraordinary mental states 
which that philosopher attained— adumbrations only, for these com- 
pletely "nonlexical" vistas cannot be well put into words. He speaks 
of realms of "moving hieroglyphs" composed entirely of "mathematical 
relations," and of the expansion and ramification of such a "hieroglyph" 
till it covered a whole aspect of the universe. Ouspensky's mathemati- 
cal predilections and his study of such things as non-Euclidean geome- 
tries, hyperspace, and the relation between time and consciousness may 
have led him to stress mathematical analogies. Mathematics is a special 
kind of language, expanded out of special sentences containing the 
numeral words, 1, 2, 3, 4, ... x, y, z, etc. But every other type of 
sentence of every language is also the potential nucleus of a far-reaching 
system. To very few is it granted to attain such consciousness as a 
durable state; yet many mathematicians and scientific linguists must 
have had the experience of "seeing," in one fugitive flash, a whole 
system of relationships never before suspected of forming a unity. The 
harmony and scientific beauty in the whole vast system momently over- 
whelms one in a flood of aesthetic delight. To "see," for instance, how 
all the English elementary sounds ("phonemes") and their groupings 
are coordinated by an intricate yet systematic law into all possible forms 
of English monosyllabic words, meaningful or nonsensical, existent or 
still unthought of, excluding all other forms as inevitably as the chemi- 
cal formula of a solution precludes all but certain shapes of crystals from 
emerging— this might be a distinct experience. 

To show the full formula for this law or pattern— a so-called "morpho- 
phonemic structural formula"— I should need a large piece of paper. I 
can however set up a condensed form of it as ^ 

0,C - ng, CiCo, C,C4, etc. . . . 

s ± C^C, + V + (Vi) O, ± (r, w, y); 

C - h, C\C'2, C'3C'4, etc. . . . 

C'„,a„ ± (t/d, s/z, st/zd). 

2 The full formula from which this is abbreviated is printed and explained in my 
paper "Linguistics as an exact science" in Technol. Rev., December 1940, Massa- 
chusetts Institute of Technology, Cambridge, Mass. (p. 223 in this volume). 


This formula requires that the Enghsh words be symbohzed or "spelt" 
according to standard phonemic spelling of the type described by 
Leonard Bloomfield in his book Language. In this system the diph- 
thongal vowels must be represented by a pure vowel (V) followed by w 
or y from the term (r, w, y), so that 'note' is symbolized nowt (or newt, 
depending on the dialect), 'date' is deyt, 'ice' is ays. That this is correct 
analysis on the physical or acoustic level is shown by the fact that, if we 
reverse a phonographic recording of 'ice' we get a sound like sya, and, 
if we say sya properly into the phonograph and reverse it, the machine 
will say 'ice.' For English this analysis happens to be exact also on the 
structural level two stages above the acoustic one, for the ys of ays (ice) 
is seen to be on the same line of pattern as the Is or els (else), the ns of 
sins (since) the ts of hats, etc.— it is part of a general architectonic 
scheme of having two consonants together. 

Now, by reading the commas in the formula as "or," we see that the 
formula is equivalent to a large series of subsidiary formulas. One of 
the simplest of these is O + V + C — h (see how it is contained in the 
big formula) which means that the word can begin without a consonant 
and with any one vowel, followed by any one consonant except h— giv- 
ing us words like 'at, or, if.' Changing the first term to the next symbol 
in the big formula, we get C — ng + V + C — h, which means that 
the word, ending as before, can begin with any single English consonant 
except the ng sound as in 'sing' (this sound ought to be written with 
ONE symbol, but, in deference to the printer, I shall employ the usual 
digraph). This pattern gives us the long array of words like 'hat, bed, 
dog, man,' and permits us to coin new ones like 'tig, nem, zib'— but not, 
be it noted, ngib or zih. 

So far the patterns are simple. From now on they become intricate! 
The formula in this abbreviated form needs along with it a series of lists 
of assorted consonants, like so many laundry lists, each list being repre- 
sented by one of the symbols Ci, C2, etc. The formula C1C2 means 
that you can begin the word with any consonant out of list Ci and follow 
it w^ith any from list Co, which happens to contain only r and I. Since 
Ci contains p, b, f, for instance, we can have words like 'pray, play, brew, 
blew, free, flee,' and the nonsensical 'frig, blosh,' etc. But suppose we 
want a word beginning with sr, zr, tl, or dl. We go to our list Ci, but 
to our surprise there is no s, z, t, or d, on it. We appear to be stumped! 
We pick up our other lists, but are no better off. There is no way of 


combining our lists according to the formula to get these initial com- 
binations. Evidently there just aren't any such English words; and 
what is more, any budding Lewis Carrolls or Edward Lears will some- 
how mysteriously refuse to coin such words. This shows that word- 
coining is no act of unfettered imagination, even in the wildest flights 
of nonsense, but a strict use of already patterned materials. If asked to 
invent forms not already prefigured in the patternment of his language, 
the speaker is negative in the same manner as if asked to make fried 
eggs without the eggs! 

Thus the formula sums up every combination that English one-syl- 
lable words or wordlike forms have, and bars out every one they do not 
and cannot have. Contained in it is the mpst of 'glimpsed,' the ksths 
of 'sixths,' the ftht of 'he fifthed it,' the nchst of the queer but possible 
'thou munchst it greedily,' and multitudes of other "rugged sounds 
which to our mouths grow sleek," but which would have "made Quin- 
tilian stare and gasp." At the same time the formula bars out numerous 
smooth but to us difficult (because unpatterned) combinations, like litk, 
fpat, nwelng, dzogb, and a myriad more, all possible and easy to some 
languages, but not to English, 

It will be evident that implicit in our one-syllable words is an un- 
dreamed-of complexity of organization, and that the old gag, "say it in 
words of one syllable," as a metaphor of simplicity, is from the stand- 
point of a more penetrative insight the most arrant nonsense! Yet to 
such insight this old cliche bears unconscious witness to the truth that 
those who easily and fluently use the intricate systems of language are 
utterly blind and deaf to the very existence of those systems, until the 
latter have been, not without some difficulty, pointed out. 

And the adage "as above, so below" applies strongly here. As below, 
on the phonological plane of language, significant behavior is ruled by 
pattern from outside the focus of personal consciousness, so is it on the 
higher planes of language that we call expression of the thought. As 
we shall sec in Part II, thinking also follows a network of tracks laid 
down in the given language, an organization which may concentrate 
systematically upon certain phases of reality, certain aspects of intelli- 
gence, and may systematically discard others featured by other languages. 
The individual is utterly unaware of this organization and is constrained 
completely within its unbreakable bonds. 



We saw in Part I that, in linguistic and mental phenomena, signifi- 
cant behavior (or what is the same, both behavior and significance, so 
far as interlinked) are ruled by a specific system or organization, a 
"geometr}'" of form principles characteristic of each language. This 
organization is imposed from outside the narrow circle of the personal 
consciousness, making of that consciousness a mere puppet whose lin- 
guistic maneuverings are held in unsensed and unbreakable bonds of 
pattern. It is as if the personal mind, which selects words but is largely 
oblivious to pattern, were in the grip of a higher, far more intellectual 
mind which has very little notion of houses and beds and soup kettles, 
but can systematize and mathematize on a scale and scope that no 
mathematician of the schools ever remotely approached. 

And now appears a great fact of human brotherhood— that human 
beings are all alike in this respect. So far as we can judge from the sys- 
tematics of language, the higher mind or "unconscious" of a Papuan 
headhunter can mathematize quite as well as that of Einstein; and con- 
versely, scientist and yokel, scholar and tribesman, all use their personal 
consciousness in the same dim-witted sort of way, and get into similar 
kinds of logical impasse. Tliey are as unaware of the beautiful and in- 
exorable systems that control them as a cowherd is of cosmic rays. Their 
understanding of the processes involved in their talk and ratiocination is 
a purely superficial, pragmatic one, comparable to little Sue Smith's 
understanding of the radio, which she turns on in such a way as to evoke 
a bedtime story. Men even show a strong disposition to make a virtue 
of this ignorance, to condemn efforts at a better understanding of the 
mind's workings as "impractical," or as "theories" if the condemner 
happens to be a yokel, or as "metaphysics" or "mysticism" or "episte- 
mology" if he happens to be wearing the traditionally correct turnout of 
a scientist. Western culture in particular reserves for the investigators 
of language its most grudging meed of recognition and its meagerest 
rewards, even though it has to counter the natural human tendency to 
find language, mysterious as it is, the most fascinating of subjects— one 
about which men love to talk and speculate unscientifically, to discuss 
endlessly the meaning of words, or the odd speech of the man from 
Boston as it appears to the man of Oshkosh, or vice versa. 


The higher mind would seem to be able to do any kind of purely 
intellectual feat, but not to "be conscious" on the personal level. That 
is, it does not focus on practical affairs and on the personal ego in its 
personal, immediate en\ironment. Certain dreams and exceptional 
mental states may lead us to suppose it to be conscious on its own 
plane, and occasionally its consciousness may "come through" to the 
personality; but, barring techniques like Yoga, it ordinarily makes no 
nexus with the personal consciousness. We could call it a higher ego, 
bearing in mind a distinctive trait, appearing through every language, 
and its one striking resemblance to the personal self; namely, that it 
organizes its systems around a nucleus of three or more pronominal 
"person" categories, centered upon one we call the first-person singular. 
It can function in any linguistic system— a child can learn any language 
with the same readiness, from Chinese, with its separately toned and 
stressed monosyllables, to Nootka of Vancouver Island, with its fre- 
quent one-word sentences such as mamamamamahln'iqk'okmaqama— 
'they each did so because of their characteristic of resembling white 
people.' ^ 

Because of the systematic, configurative nature of higher mind, the 
"patternment" aspect of language always overrides and controls the "lexa- 
tion" {Ndma) or name-giving aspect. Hence the meanings of specific 
words are less important than we fondly fancy. Sentences, not words, 
are the essence of speech, just as equations and functions, and not bare 
numbers, are the real meat of mathematics. We are all mistaken in our 
common belief that any word has an "exact meaning." We have seen 
that the higher mind deals in symbols that have no fixed reference to 
anything, but are like blank checks, to be filled in as required, that 
stand for "any value" of a given variable, like the C's and V's in the 
formula cited in Part I, or the x, y, z of algebra. There is a queer West- 
ern notion that the ancients who invented algebra made a great dis- 
covery, though the human unconscious has been doing the same 
sort of thing for eons! For the same reason the ancient Mayas or the 
ancient Hindus, in their staggering cycles upon cycles of astronomical 

3 This word and sentence contains only one Nama or lexation, maTnahl or 'white- 
race person.' The rest is all grammatical pattern which can refer to anything. The 
Nootka stem or Ndma for 'doll' with the same operations done upon it would mean 
'they each did so because of their doll-likeness.' 


numbers, were simply being human. We should not however make 
the mistake of thinking that words, even as used by the lower personal 
mind, represent the opposite pole from these variable symbols, that a 
word DOES have an exact meaning, stands for a given thing, is only one 
value of a variable. 

Even the lower mind has caught something of the algebraic nature of 
language; so that words are in between the variable symbols of pure 
patternment [Arupa] and true fixed quantities. That part of meaning 
which is in words, and which we may call "reference," is only relatively 
fixed. Reference of words is at the mercy of the sentences and gram- 
matical patterns in which they occur. And it is surprising to what a 
minimal amount this element of reference may be reduced. The sen- 
tence "I went all the way down there just in order to see Jack" contains 
only one fixed concrete reference: namely, "Jack." Tlie rest is pattern 
attached to nothing specifically; even "see" obviously does not mean 
what one might suppose, namely, to receive a visual image. 

Or, again, in word reference we deal with size by breaking it into size 
classes— small, medium, large, immense, etc. — but size objectively is not / 
divided into classes, but is a pure continuum of relati\ity. Yet we think 
of size constantly as a set of classes because language has segmented 
and named the experience in this way. Number words may refer not 
to number as counted, but to number classes with elastic boundaries. 
Thus English 'few' adjusts its range according to the size, importance / 
or rarity of the reference. A 'few' kings, battleships, or diamonds might 
be only three or four, a 'few' peas, raindrops, or tea leaves might be 
thirty or forty. 

You may say, "Yes, of course this is true of words like large, small, and 
the like; they are obviously relatixe terms, but words like dog, tree, house, 
are different— each names a specific thing." Not so; these terms are in 
the same boat as 'large' and 'small.' The word 'Fido' said by a certain 
person at a certain time may refer to a specific thing, but the word 'dog' 
refers to a class with elastic limits. The limits of such classes are dif- 
ferent in different languages. You might think that 'tree' means the 
same thing, everywhere and to everybody. Not at all. The Polish word 
that means 'tree' also includes the meaning 'wood.' The context or 
sentence pattern determines what sort of object the Polish word (or any 
word, in any language) refers to. In Hopi, an American Indian language 
of Arizona, the word for 'dog,' pohko, includes pet animal or domestic 


animal of any kind. Thus 'pet eagle' in Hopi is literally 'eagle-dog'; and 
having thus fixed the context a Hopi might next refer to the same eagle 
as so-and-so's pohko. 

But lest this be dismissed as the vagary of a "primitive" language (no 
language is "primitive"), let us take another peep at our own beloved 
English. Take the word "hand." In 'his hand' it refers to a location on 
the human body, in 'hour hand' to a strikingly dissimilar object, in 'all 
hands on deck' to another reference, in 'a good hand at gardening' to 
another, in 'he held a good hand (at cards)' to another, w^hereas in 'he 
got the upper hand' it refers to nothing but is dissolved into a pattern 
of orientation. Or consider the word 'bar' in the phrases: 'iron bar, bar 
to progress, he should be behind bars, studied for the bar, let down all 
the bars, bar of music, sand bar, candy bar, mosquito bar, bar sinister, 
bar none, ordered drinks at the bar'! 

But, you may say, these are popular idioms, not scientific and logical 
use of language. Oh, indeed? "Electrical" is supposed to be a scien- 
tific word. Do you know what its referent is? Do you know that the 
"electrical" in "electrical apparatus " is not the same "electrical" as the 
one in "electrical expert"? In the first it refers to a current of electricity 
in the apparatus, but in the second it does not refer to a current of elec- 
tricity in the expert. When a word like "group" can refer either to a 
sequence of phases in time or a pile of articles on the floor, its element 
of reference is minor. Referents of scientific words are often con- 
veniently vague, markedly under the sway of the patterns in which they 
occur. It is very suggesti\e that this trait, so far from being a hallmark 
of Babbittry, is most marked in intellectual talk, and— mirabile dictu— 
in the language of poetry and love! And this needs must be so, for 
science, poetry, and love are alike in being "flights" above and away 
from the slave-world of literal reference and humdrum prosaic details, 
attempts to widen the petty narrowness of the personal self's outlook, 
liftings toward Arupa, toward that world of infinite harmony, sympathy 
and order, of unchanging truths and eternal things. And while all words 
are pitiful enough in their mere "letter that killeth," it is certain that 
scientific terms like 'force, average, sex, allergic, biological' are not less 
pitiful, and in their own way no more certain in reference than 'sweet, 
gorgeous, rapture, enchantment, heart and soul, star dust.' You have 
probably heard of 'star dust'— what is it? Is it a multitude of stars, a 


sparkling powder, the soil of the planet Mars, the Milky Way, a state of 
daydreaming, poetic fancy, pyrophoric iron, a spiral nebula, a suburb of 
Pittsburgh, or a popular song? You don't know, and neither does any- 
body. The word— for it is one lexation, not two— has no reference of 
its own. Some words are like that.* As we have seen, reference is the 
lesser part of meaning, pat ternment the greater. Science, the quest for 
truth, is a sort of divine madness like love. And music— is it not in the 
same category? Music is a quasilanguage based entirely on patternment, 
without having developed lexation. 

Sometimes the sway of pattern over reference produces amusing 
results, when a pattern engenders meanings utterly extraneous to the 
original lexation reference. The lower mind is thrown into bewilder- 
ment, cannot grasp that compelling formulas are at work upon it, and 
resorts wildly and with glad relief to its favorite obvious type of ex- 
planation, even "seeing things" and "hearing things" that help out such 
explanation. The word 'asparagus,' under the stress of purely phonetic 
English patterns of the type illustrated in the formula cited in Part I, 
rearranges to 'sparagras'; and then since 'sparrer' is a dialectical form of 
'sparrow,' we find 'sparrow grass' and then religiously accepted accounts 
of the relation of sparrows to this 'grass.' 'Cole slaw' came from Ger- 
man Kohlsalat, 'cabbage salad,' but the stress of the pattern tending to 
revamp it into 'cold slaw' has in some regions produced a new lexation 
'slaw,' and a new dish 'hot slaw'! Children of course are constantly re- 
patterning, but the pressure of adult example eventually brings their 
language back to the norm; they learn that Mississippi is not Mrs. Sippy, 
and the equator is not a menagerie lion but an imaginary line. Some- 
times the adult community does not possess the special knowledge 
needed for correction. In parts of New England, Persian cats of a cer- 
tain type are called Coon cats, and this name has bred the notion that 
they are a hybrid between the cat and the 'coon' (raccoon). This is 
often firmly believed by persons ignorant of biology, since the stress of 
the linguistic pattern (animal-name 1 modifying animal-name 2) causes 
them to "see" (or as the psychologists say "project") objective raccoon 
quality as located on the body of the cat— they point to its bushy tail, 
long hair, and so on. I knew of an actual case, a woman who owned 

■» Compare 'kith' and 'throe,' wliich give no meaning, and a bewildering effect, 
without the patterns 'kith and kin' and 'in throes of.' 


a fine "Coon cat," and who would protest to her friend: "Why, just 
LOOK at him— his tail, his funny eyes— can't you see it?" "Don't be 
silly!" quoth her more sophisticated friend. "Think of your natural 
history! Coons cannot breed with cats; they belong to a different 
family." But the lady was so sure that she called on an eminent zoolo- 
gist to confirm her. He is said to have remarked, with unwavering 
diplomacy, "If you like to think so, just think so." "He was even more 
cruel than you!" she snapped at her friend, and remained convinced 
that her pet was the outcome of an encounter between a philandering 
raccoon and a wayward cat! In just such ways on a vaster scale is woven 
the web of Maya, illusion begotten of intrenched selfhood. I am told 
that Coon cats recei\'ed their name from one Captain Coon, who 
brought the first of these Persian cats to the State of Maine in his ship. 

In more subtle matters we all, unknowingly, project the linguistic 
relationships of a particular language upon the universe, and see them 
there, as the good lady saw a linguistic relation (Coon = raccoon) made 
visible in her cat. We say 'see that wave'— the same pattern as 'see that 
house.' But without the projection of language no one ever saw a single 
wave. We see a surface in everchanging undulating motions. Some lan- 
guages cannot say 'a wave'; they are closer to reality in this respect. Hopi 
say walalata, 'plural waving occurs,' and can call attention to one place 
in the waving just as we can. But, since actually a wave cannot exist by 
itself, the form that corresponds to our singular, wala, is not the equiva- 
lent of English 'a wave,' but means 'a slosh occurs,' as when a vessel of 
liquid is suddenly jarred. 

English pattern treats 'I hold it' exactly like 'I strike it,' 'I tear it,' and 
myriads of other propositions that refer to actions effecting changes in 
matter. Yet 'hold' in plain fact is no action, but a state of relative posi- 
tions. But we think of it, even see it, as an action, because language 
sets up the proposition in the same way as it sets up a much more com- 
mon class of propositions dealing with movements and changes. We 
ascribe action to what we call "hold" because the formula, substan- 
tive -f verb = actor -\- his action, is fundamental in our sentences. 
Thus we are compelled in many cases to read into nature fictitious 
acting-entities simply because our sentence patterns require our verbs, 
when not imperati\e, to ha\e substantives before them. We are obliged 
to say 'it flashed' or 'a light flashed,' setting up an actor it, or a light, 
to perform what we call an action, flash. But the flashing and the hght 


are the same; there is no thing which does something, and no doing. 
Hopi says only rehpi. Hopi can have verbs without subjects, and this 
gives to that language power as a logical system for understanding cer- 
tain aspects of the cosmos. Scientific language, being founded on west- 
ern Indo-European and not on Hopi, does as we do, sees sometimes 
actions and forces where there may be only states. For do you not 
conceive it possible that scientists as well as ladies with cats all un- 
knowingly project the linguistic patterns of a particular type of language 
upon the universe, and see them there, rendered visible on the very face 
of nature? A change in language can transform our appreciation of the 

All this is typical of the way the lower personal mind, caught in a 
vaster world inscrutable to its methods, uses its strange gift of language 
to weave the web of Maya or illusion, to make a provisional analysis of 
reality and then regard it as final. Western culture has gone farthest 
here, farthest in determined thoroughness of provisional analysis, and 
farthest in determination to regard it as final. The commitment to 
illusion has been sealed in western Indo-European language, and the 
road out of illusion for the West lies through a wider understanding of 
language than western Indo-European alone can give. This is the 
"Mantra Yoga" of the Western consciousness, the next great step, which 
it is now ready to take. It is probably the most suitable way for Western 
man to begin that "culture of consciousness" which will lead him to a 
great illumination. 

Again, through this sort of understanding of language is achieved a 
great phase of human brotherhood. For the scientific understanding of 
very diverse languages— not necessarily to speak them, but to analyze 
their structure— is a lesson in brotherhood which is brotherhood in the 
universal human principle— the brotherhood of the "Sons of Manas." 
It causes us to transcend the boundaries of local cultures, nationalities, 
physical peculiarities dubbed "race," and to find that in their linguistic 
systems, though these systems differ widely, yet in the order, harmony, 
and beauty of the systems, and in their respective subtleties and pene- 
trating analysis of reality, all men are equal. This fact is independent 
of the state of evolution as regards material culture, savagery, civiliza- 
tion, moral or ethical development, etc., a thing most surprising to the 
cultured European, a thing shocking to him, indeed a bitter pill! But it 


is true; the crudest savage may unconsciously manipulate with effortless 
ease a linguistic system so intricate, manifoldly systematized, and intel- 
lectually difficult that it requires the lifetime study of our greatest 
scholars to describe its workings. The manasic plane and the "higher 
ego" have been given to all, and the evolution of human language was 
complete, and spread in its proud completeness up and down the earth, 
in a time far anterior to the oldest ruin that molders in the soil today. 
\ Linguistic knowledge entails understanding many different beautiful 
|__svstems_of_ Jogical ana lysis^ Through it, the world as seen from the 
diverse viewpoints of other social groups, that we have thought of as 
alien, becomes intelligible in new terms. Alienness turns into a new 
and often clarifying way of looking at things. Consider Japanese. The 
view of the Japanese that we get outwardly from their governmental 
policy seems anything but conducive to brotherhood. But to approach 
the Japanese through an aesthetic and scientific appreciation of their 
language transforms the picture, that is to realize kinship on the cos- 
mopolitan levels of the spirit. One lovely pattern of this language is 
that its sentence may have two differently ranked subjects. We are 
familiar with the idea of two ranks of objects for our verbs, an im- 
mediate and a more remote goal, or direct and indirect object as they 
are commonly called. We have probably never thought of the possi- 
bilities of a similar idea applied to subjects. This idea is put to work 
in Japanese. The two subjects— call them subject 1 and subject 2— are 
marked by the particles wa and ga, and a diagram might show them 
with a line drawn from each subject word, the two lines converging 
upon the same predication, whereas our English sentence could have 
only one subject with one line to the predicate. An example would be 
the way of saying "Japan is mountainous": "Japaui mountain2 (are) 
many"; ^ or: "Japan, in regard to its mountains are many." "John is 
long-legged" would be "Johni leg2 (are) long." This pattern gives great 
conciseness at the same time with great precision. Instead of the vague- 
ness of our "mountainous," the Japanese can, with equal compactness 
of formulation, distinguish "mountainous" meaning that mountains not 
always high are abundant, from "mountainous" meaning that moun- 
tains not abundant relative to the whole area are high. We see how 
the logical uses of this pattern would give to Japanese great power in 

5 "Are" is in parentheses because "be many" is expressed by a single verblike word. 
The Japanese ordinarily does not use a plural. 


concise scientific operations with ideas, could this power be properly 

The moment we begin scientific, unbiased research into language we 
find, in people and cultures with the most unprepossessing exteriors, 
beautiful, effective, and scientific devices of expression unknown to west- 
ern Indo-European tongues or mentalities. The Algonkian languages 
are spoken by very simple people, hunting and fishing Indians, but they 
are marvels of analysis and synthesis. One piece of grammatical finesse 
peculiar to them is called the obviative. This means that their pro- 
nouns have four persons instead of three, or from our standpoint two 
third persons. This aids in compact description of complicated situa- 
tions, for which we should have to resort to cumbersome phraseology. 
Let us symbolize their third and fourth persons by attaching the 
numerals 3 and 4 to our written words. The Algonkians might tell 
the story of William Tell like this: "William Tell called hisa son and 
told him4 to bring hims hiss bow and arrow, which4 he4 then brought to 
hims. Hcs had him4 stand still and placed an apple on his4 head, then 
took hiss bow and arrow and told him4 not to fear. Then hcs shot it4 
off his4 head without hurting him4." Such a device would greatly help 
in specifying our complex legal situations, getting rid of "the party of 
the first part" and "the aforesaid John Doe shall, on his part, etc." 

Chichewa, a language related to Zulu, spoken by a tribe of unlettered 
Negroes in East Africa, has two past tenses, one for past events with 
present result or influence, one for past without present influence. A 
past as recorded in external situations is distinguished from a past re- 
corded only in the psyche or memory; a new view of time opens before 
us. Let 1 represent the former and 2 the latter; then ponder these 
Chichewa nuances: I camCi here; I went2 there; he waso sick; he diedi; 
Christ died2 on the cross; God createdi the world. "I ate/' means I am 
not hungry; "I ate2" means I am hungry. If you were offered food and 
said: "No, I have eateui," it would be all right, but if you used the other 
past tense you would be uttering an insult. A Theosophical speaker of 
Chichewa might use tense 1 in speaking of the past involution of 
Monads, which has enabled the world to be in its present state, while 
he might use tense 2 for, say, long-past planetary systems now dis- 
integrated and their evolution done. If he were talking about Re- 
incarnation, he would use 2 for events of a past incarnation simply in 


their own frame of reference, but he would use 1 in referring to or 
implying their "Karma." It may be that these primitive folk are 
equipped with a language which, if they were to become philosophers 
or mathematicians, could make them our foremost thinkers upon time. 
Or take the Coeur d'Alcne language, spoken by the small Indian tribe 
of that name in Idaho. Instead of our simple concept of "cause," 
founded on our simple "makes it (him) do so," the Coeur d'Alene 
grammar requires its speakers to discriminate (which of course they do 
automatically) among three causal processes, denoted by three causal 
verb-forms: (I) growth, or maturation of an inherent cause, (2) addition 
or accretion from without, (3) secondary addition i.e., of something af- 
fected by process 2. Thus, to say "it has been made sweet" they would 
use form I for a plum sweetened by ripening, form 2 for a cup of coffee 
sweetened by dissolving sugar in it, and form 3 for griddle cakes sweet- 
ened by syrup made by dissolving sugar. If, given a more sophisticated 
culture, their thinkers erected these now unconscious discriminations 
into a theory of triadic causality, fitted to scientific observations, they 
might thereby produce a valuable intellectual tool for science, we could 
imitate artificially such a theory, perhaps, but we could not apply it, for 
WE are not habituated to making such distinctions with effortless ease 
in daily life. Concepts have a basis in daily talk before scientific workers 
will attempt to use them in the laboratory. Even relativity has such a 
basis in the western Indo-European languages (and others)— the fact 
that these languages use many space words and patterns for dealing with 

Language has further significance in other psychological factors on a 
different level from modern linguistic approach but of importance in 
music, poetry, literary style, and Eastern mantram. What I have been 
speaking of thus far concerns the plane of Manas in the more philo- 
sophical sense, the "higher unconscious" or the "soul" (in the sense as 
used by Jung). What I am about to speak of concerns the "psyche" (in 
the sense as used by Freud), the "lower" unconscious, the Manas which 
is especially the "slayer of the real," the plane of Kama, of emotion or 
rather feeling (Gefiihl). In a serial relation containing the levels of 
Nama-Rupa and Arupa, this level of the unconscious psyche is on the 
other side of Nama-Rupa from Arupa, and Nama or lexation mediates 
in a sense between these extremes. Hence the psyche is the psycho- 


logical corrclati\e of the phonemic level in language, related to it not 
s^fucFiiraIlylis"is JN'ama or lexatTon, not by using it as building blocks, as 
word-making uses the phonemes (vowels, consonants, accents, etc.), but 
related as the feeling-content of the phonemes. There is a universal, / 
Gefiihl-tyi ^e wa yj jf linking experiences, which shows up in laboratory j 
experiments and appears to be independent of language— basically alike f 
for all persons. I 

Without a serial or hierarchical order in the uni\erse it would have 
to be said that these psychological experiments and linguistic experi- 
ments contradict each other. In the psychological experiments human 
subjects seem to associate the experiences of bright, cold, sharp, hard, 
high, light (in weight), quick, high-pitched, narrow, and so on in a long 
series, with each other; and conversely the experiences of dark, warm, 
yielding, soft, blunt, low, hea\y, slow, low-pitched, wide, etc., in another 
long series. This occurs whether the words for such associated ex- 
periences resemble or not, but the ordinary person is likely to notice a 
relation to words only when it is a relation of likeness to such a series in 
the vowels or consonants of the words, and when it is a relation of con- 
trast or conflict it is passed unnoticed. The noticing of the relation of 
likeness is an element in sensitiveness to literary style or to what is often 
rather inaccurately called the "music" of words. The noticing of the 
relation of conflict is much more difficult, much more a freeing oneself 
from illusion, and though quite "unpoetical" it is really a movement 
toward Higher Manas, toward a higher symmetry than that of physical 

What is significant for our thesis is that language, t hrough le xation , 
has made the speaker more acutely conscious of certain dim psychic 
sensations; it has actually produced aw^areness on lower planes than its 
own: a power of the nature of magic. There is a yogic mastery in the 
power of language to remain independent of lower-psyche facts, to over- 
ride them, now point them up, now toss them out of the picture, to 
mold the nuances of words to its own rule, whether the psychic ring of 
the sounds fits or not. If the sounds fit, the psychic quality of the 
sounds is increased, and this can be noticed by the layman. If the 
sounds do not fit, the psychic quality changes to accord w^ith the lin- 
guistic meaning, no matter how incongruous with the sounds, and this 
is not noticed by the layman. 


Thus the vowels a (as in 'father'), o, u, are associated in the laboratory 
tests with the dark-warm-soft series, and e (English a in 'date'), i (Eng- 
lish e in 'be') with the bright-cold-sharp set. Consonants also are asso- 
ciated about as one might expect from ordinary naive feeling in the 
matter. What happens is that, when a word has an acoustic similarity 
to its own meaning, we can notice it, as in English 'soft' and German 
sanft. But, when the opposite occurs, nobody notices it. Thus German 
zart (tsart) 'tender' has such a "sharp" sound, in spite of its a, that to 
a person who does not know German it calls up the bright-sharp mean- 
ings, but to a German it "sounds" soft— and probably warm, dark, etc., 
also. An even better case is deep. Its acoustic association should be 
like that of peep or of such nonsense words as veep, treep, queep, etc., 
i.e., as bright, sharp, quick. But its linguistic meaning in the English 
language happens to refer to the wrong sort of experience for such an 
association. This fact completely overrides its objective sound, causing 
it to "sound" subjectively quite as dark, warm, heavy, soft, etc., as 
though its sounds really were of that type. It takes illusion-freeing, if 
unpoetic, linguistic analysis to discover this clash between two "musics," 
one more mental and one more psychic, in the word. Manas is able 
to disregard properties of the psychic plane, just as it can disregard 
whether an equational x refers to automobiles or sheep. It can project 
parts of its own patterns upon experience in such a way that they dis- 
tort, and promote illusion, or again in such a way that they illuminate, 
and build up scientific theories and tools of research. 

Yoga is defined by Patanjali as the complete cessation of the activity 
of the versatile psychic nature.*' We have seen that this activity con- 
sists largely of perso nal-s ocial rea ction s along unperceived tracks of pat- 
tern laid down from the Arupa level functioning above or behind the 
focus of personal consciousness. The reason why the Arupa level is 
beyond the ken of the consciousness is not because it is essentially dif- 
ferent (as if it were, e.g., a passive network) but because the personality 
does focus, from evolution and habit, upon the aforesaid versatile ac- 
tivity. The stilling of this activity and the coming to rest of this focus, 
though difficult and requiring prolonged training, is by reliable accounts 

6 Bragdon's paraphrase of the Yoga Sutras, An Introduction to Yoga, Claude Brag- 
don, New York, 1933. 


from widely diverse sources, both Eastern and Western, a tremendous 
expansion, brightening and clarifying of consciousness, in which the in- 
tellect functions with undreamed-of rapidity and sureness. The scien- 
tific study of languages and linguistic principles is at least a partial raising 
of the intellect toward this level. In the understanding of a large lin- 
guistic pattern there is involved a partial shift of focus away from the 
versatile psychic activity. Such understandings have even a therapeutic 
value. Many neuroses are simply the compulsive working over and over 
of word systems, from which the patient can be freed by showing him 
the process and pattern. 

All this leads back to the idea touched upon in part I of this essay, 
that the types of patterned relationship found in language may be but 
the wavering and distorted, pale, substanceless reflection of a causal 
WORLD. Just as language consists of discrete lexation-segmentation 
(Ndma-Rupa) and ordered patternment, of which the latter has the 
more background character, less obvious but more infrangible and uni- 
versal, so the physical world may be an aggregate of quasidiscrete en- 
tities (atoms, crystals, living organisms, planets, stars, etc.) not fully 
understandable as such, but rather emergent from a field of causes that 
is itself a manifold of pattern and order. It is upon the bars of the fence, 
beyond which it would meet these characters of the field, that science 
is now poised. As physics explores into the intra-atomic phenomena, 
the discrete physical forms and forces are more and more dissohed into 
relations of pure patternment. The place of an apparent entity, an 
electron for example, becomes indefinite, interrupted; the entity appears 
and disappears from one structural position to another structural posi- 
tion, like a phoneme or any other patterned linguistic entity, and may 
be said to be nowhere in between the positions. Its locus, first thought 
of and analyzed as a continuous variable, becomes on closer scrutiny a 
mere alternation; situations "actualize" it, structure beyond the probe of 
the measuring rod governs it; three-dimensional shape there is none, 
instead— "Arupa." 

Science cannot yet understand the transcendental logic of such a state 
of affairs, for it has not yet freed itself from the illusory necessities of 
common logic which are only at bottom necessities of grammatical pat- 
tern in Western Aryan grammar; necessities for substances which are 


only necessities for substantives in certain sentence positions, necessities 
for forces, attractions, etc. which are only necessities for verbs in certain 
other positions, and so on. Science, if it survives the impending dark- 
ness, will next take up the consideration of linguistic principles and 
divest itself of these illusory linguistic necessities, too long held to be 
the substance of Reason itself. 





1925 "Purpose vs. evolution." Letter to the editors of the New Re- 

public, issue of December 19, 1925. 

1927 ["On the connection of ideas."] Printed for the first time in 

this volume, pp. 35-39. 

1928 "Toltec history." Twenty-Third International Congress of 

Americanists, New York, 1928: Abstracts of Papers, no. 109. 

1928 "Aztec linguistics." Twenty-Third International Congress of 

Americanists, New York, 1928: Abstracts of Papers, no. 116. 

1928 "An Aztec account of the period of the Toltec decline." Pro- 

ceedings of the Twenty-Third International Congress of 
Americanists, New York, 1928, pp. 122-129. 

1929 "The reign of Huemac." American Anthropologist, l>l:667-6S-\ 


1931 "A central Mexican inscription combining Mexican and Maya 

day signs." American Anthropologist, 34:296-302 (1932). 

1933 "The Mava manuscript in Dresden." Art and Archaeology, 

34:270 '(1933). 

1933 The phonetic value of certain characters in Maya writing. Cam- 

bridge, Mass.: Harvard University Press, 1933 {Papers of the 
Peabody Museum, vol. XIII, no. 2). With an introduction 
by Alfred M. Tozzer. xii, 48 pp. 

1935 Review of A. L. Kroeber, Uto-Aztecan languages of Mexico. 

American Anthropologist, 37:343-345 (1935). 

1935 "The comparative linguistics of Uto-Aztecan." American An- 

thropologist, 37:600-608 (193=;). 

1935 "Maya writing and its decipherment." Maya Research, 2:367- 






1936 Appendix to J. Alden Mason, "The classification of the Sonoran 

languages," pp. 197-198 in Robert H. Lowie (editor). Essays 
in anthropology in honor of Alfred Louis Kroeber. Berkeley: 
University of Cahfornia Press, 1936. 

1936 Notes on the "Glossary," pp. 1198-1326 in Elsie Clews Parsons, 

Hopi Journal of Alexander M. Stephens, Part 2. New York: 
Columbia University Press, 1936 {Columbia Contributions to 
Anthropology, 23). 

1936 "The punctual and segmentative aspects of verbs in Hopi." 

Language, 12:127-131 (1936). 

1936 "Notes on the Tiibatulabal language." American Anthropolo- 

gist, 38:341-344 (1936). 

1936 "Loan-words in ancient Mexico." Philological and Documentary 

Studies {Middle American Research Institute, Tulane Univer- 
sity of Louisiana), 1:1-17 (1943). Also Studies in Linguistics, 
5:49-64 (1947). 

1936(?) "An American Indian model of the uni\'erse." International 
Journal of American Linguistics, 16:67-72 (1950). 
[Reprinted in Etc., a Review of General Semantics, 8:27-33 
(1950); also in Collected papers on metalinguistics. Foreign 
Service Institute, Department of State, Washington, D. C, 

1936(?) "A linguistic consideration of thinking in primitive communi- 
ties." Printed for the first time in this volume, pp. 65-86. 

1937 "The origin of Aztec TL." American Anthropologist, 39:265- 

274 (1937). 
1937 (with George L. Trager) "The relationship of Uto-Aztecan and 

Tanoan." American Anthropologist, 39:609-624 (1937). 
1937 "Grammatical categories." Language, 21:1-11 (1945). 

1937 ["Discussion of Hopi linguistics."] Printed for the first time in 

this volume, pp. 102-111. 

1938 "Some verbal categories of Hopi." Language, 14:275-286 

1938 Review of K. T. Preuss and Ernst Mengin, Die Mexikanische 

Bilderhandschrift Ilistoria Tolteca-Chichimeca: die Manu- 
scripte 46-S8^^^ der Nationalbibliothek in Paris, Teil I, Die 
Bilderschrift nebst Vbersetzimg (Berlin, 1937). American An- 
thropologist, 40:729-730 (1938). 

1938 "Language: plan and conception of arrangement." Printed for 

the first time in this volume, pp. 125-133. 

1939 "The relation of habitual thought and behavior to language." 

Pp. 75-93 in Leslie Spier (editor). Language, culture, and per- 




sonality (Menasha, Wis.: Sapir Memorial Publication Fund, 


[Reprinted in Collected papers on metalinguistics, Foreign 

Service Institute, Department of State, Washington, D. C, 

1939 "The Hopi Language, Toreva dialect." Pp. 158-183 in Harry 

Hoijer (editor). Linguistic structures of native America (New 

York: Viking Fund, 1946). 
1939 "The Milpa Alta dialect of Aztec, with notes on the Classical 

and the Tepoztlan dialects." Pp. 367-397 in Harry Hoijer 

(editor), Linguistic structures of native America (New York: 

Viking Fund, 1946). 

1939 "Gestalt technique of stem composition in Shawnee." Appen- 

dix, pp. 593-406, to C. F. Voegelin, Shawnee stems and the 
Jacob P. Dunn Miami Dictionary. Indianapolis: Indiana His- 
torical Society, 1940 {Prehistory Research Series, vol. I, no. 9, 
April 1940). ' 

1940 "Blazing icicles." [Article on fire pre\ention] Hartford, Conn.: 

Hartford Fire Insurance Company, n.d. [Reprinted from the 
Hartford Agent.] 

1940 "Decipherment of the linguistic portion of the Ma\-a hiero- 

glyphs," pp. 479-502 in The Smithsonian report for 1941, 
Publication 3669 (Washington: U. S. Government Printing 
Office, 1942). [Also in Spanish, "Interpretacion de la parte 
lingiiistica de los geroglificos Maya." Tzunpame, Organo de 
Publicidad del Museo Nacional y Auexos [San Salvador), 
5:50-73 (August 1945), and Suplemento (Figures 1-4).] 

1940 "Phonemic analysis of the English of eastern Massachusetts." 

Studies in Linguistics, 2:21-40 (1943). 

1940 "Linguistic factors in the terminology of Hopi architecture." 

International Journal of American Linguistics, 19:141-145 

1940 "Science and linguistics." Technology Review {M.I. T.) 42-229- 

231, 247-248 (1940). 

[Reprinted in S. I. Hayakawa, Language in action (New York: 
Harcourt-Brace, 1941), pp. 302-321; T. Newcomb and E. 
Hartley, Readings in social psychology (New York: Holt, 
1947), pp. 210-218; and Collected papers on metalinguistics 
(Foreign Service Institute, Department of State, Washington, 
D. C, 1952).] 

1940 "Linguistics as an exact science." Technology Review (M.I.T.) 

43:61-63, 80-83 (1940). 




[Reprinted in Collected papers on metalinguistics (Foreign 
Service Institute, Department of State, Washington, D. C, 
1941 "Languages and logic." Technology Review {M.I.T.), 43:250- 

252, 266, 268, 272 (1941). 

[Reprinted in Collected papers on metalinguistics (Foreign 
Serxice Institute, Department of State, Washington, D. C, 
1941 "Language, mind, and reality." The Theosophist {Madras, 

India), 63:1.281-91 (January 1942); 63:2.25-37 (April 1942). 
[Reprinted in Etc., a Review of General Semantics, 9:167-188 
1940-41 Articles in the journal Main Currents in Modern Thought: 

1:1.3-5 (1940): Review of Living light, by E. N. Harvey 

(Princeton University Press, 1940). 
1 : 1.9-10 (1940): "We may end the war that is within all wars 

that are waged to end all war." 
1:1.12-13 (1940): Digest of review in American Anthropolo- 
gist, October-December 1940, of the Work of the gods in 

Tikopia (Polynesia), by Raymond Firth, as reviewed by 

E. G. Burrows. 
1:1.14 (1940): Digest of "Notes on the demonstration of 

'wetter' water," by C. R. Caryl in Journal of Chemical 

1:1.15 (1940): [Concerning descriptive linguistics at Yale.] 
1:3.4 (1941): (with F. Kunz) [Commentary regarding logic 

and science.] 
1:3.6 (1941): "H. G. Wells." 
1:3.12-13 (1941): "Interpretations of isotopes." 
1:3.15 (1941): "The Hurrians of Old Chaldea." 
1:4.10-11 (1941): Review of The ways of things, by W. P. 

Montague (Prentice-Hall, 1940). 
1:4.13-14 (1941): "A brotherhood of thought." 
1:5.12-14 (1941): "Dr. Reiser's humanism." (Review of The 

promise of scieritific Jiumanism, by Oliver L. Reiser, New 

York, 1940). 
1:6.16 (1941): [Note on shrinking glass.] 
1:7.14-15 (1941): (with F. Kunz) "Toward a higher mental 

Unknown ["On psychology."] Printed for the first time in this volume, 
pages 40-42. 





c. 1928 "A contribution to the study of the Aztec language." 43 pp. A 
detailed linguistic and literary treatment of the second poem 
found in D. G. Brinton's compilation {Ancient Nahuatl 
poetry), with Appendix A, Original text of the poem, and 
Appendix B, A list of the most common roots in the Aztec 
language. Manuscript 157, Franz Boas collection (See Lan- 
guage Monograph no. 22, 1945). Another copy in family 

1928 "Inxestigations in Aztec linguistics and Toltec history. Part II. 

The phenomenon of oligosynthesis in Nahuatl or Aztec." 13 
pp. [This paper was given before the Twenty-third Inter- 
national Congress of Americanists, New York, September 
1928.] Among family papers. 

1928 "Notes on the oligosynthetic comparison of Nahuatl and Piman, 

with special reference to Tepecano." 23 pp. Among family 
papers. [This manuscript was submitted as a supporting docu- 
ment when the author applied for an SSRC Research Fellow- 
ship, December 1, 1928.] 

1930 "Stem series in Maya and certain Maya hieroglyphs." 28 pp. 

Among family papers. [This is a re\ision, dated October 30, 
1930, of a paper read before the Linguistic Society of America, 
ClcNcland meeting, December 1929. It was further revised 
and published as The phonetic value of certain characters in 
Maya writing. (1933)] 

1930 "Notes on two recent findings from central Mexico." 7 pp. 

MS in the library of Peabody Museum, Harvard University. 
[This is the text of a paper read before the American Anthro- 
pological Association, Cleveland meeting, December 1930. It 
is a summary of material later prepared (1) in the article "A 
central Mexican inscription combining Mexican and Maya 
day signs" (1931), and (2) in the unpublished manuscript 
listed below for the year 1931. "Pitch tone and the saltillo in 
modern and ancient Nahuatl."] 

1931 "The problem of American history before Columbus." 55 pp. 

Among family papers. Annotated in pencil, "Read before 
Conn. Historical Society, Hartford, Conn., Apr. 7, 193 1." 
1931 "Pitch tone and the saltillo in modern and ancient Nahuatl." 

54 pp. Manuscript 275. Franz Boas collection (See Language 
Monograph no. 22, 1945). 




1932-5(?) "First report on Hopi." 4 pp. MS 276, Franz Boas collection 
(See Language Monograph no. 22, 1945). 

1933 "Recent determinations of phonetic characters in Maya writing." 

8 pp. Among family papers. [Read before Linguistic Society 
of America, Washington meeting, December 1933. Amplifies 
and goes beyond analysis published (1933) in the Peabody 
Museum papers.] 

1935 "The Hopi language." 59 pp. Original among family papers, 

with note "corrected and corrections rechecked — BLW." 
Carbon copy is Manuscript 192, Franz Boas collection (See 
Language Monograph no. 22, 1945). 
193 5(?) "First steps in the decipherment of Maya writing." 112 pp. 
Among family papers. [Mentioned in "Maya writing and its 
decipherment" (1935) as being worked on — "it may be a year 
before publication."] 

1936 "A comparative decipherment of forty-one ancient Maya written 

words." 10 pp. MS among family papers. [Text of paper 
read before American Anthropological Association, Washing- 
ton meeting, December 1936.] 

1938 "The reading of Maya glyph C of the Supplementary Series and 

other glyphs." Abstract and pencil draft of paper. MS among 
family papers. 

1939 "Classification of the languages of North America north of 

Mexico." Typewritten MS, 3 pp., dated December 1939, 
among papers of G. L. Trager. [Apparently the basis of a 
talk, "Linguistic groupings north of Mexico," before American 
Anthropological Association, Chicago meeting, December 

1940 "The 'parts of speech' in Hopi." 15 pp., handwritten, MS 

among family papers. Annotated "finished Oct. 12, 1940." 


Andrews, E. Wyllys. "The phonetic value of Glyph C of the Maya supple- 
mentary series." American Anthropologist, 40:755-758 (1938). 

Brown, Roger W., and Lcnncbcrg, Eric H. "A study in language and cog- 
nition." Journal of Abnormal and Social Psychology, 49:454-462 (1954). 

Carroll, John B. Foreword to Whorf's "Language, mind and reality." 
Etc., a Review of General Semantics, 9:167-168 (1952). 


• Carroll, John B. The study of language. Cambridge: Harvard University 

Press, 1953. 
Chase, Stuart. "How language shapes our thoughts." Harper's Magazine, 

April 1954, pp. 76-82. 
Chase, Stuart. The power of words. New York: Harcourt, Brace, 1954. 
Doob, L. W. Social psychology. New York: Holt, 1952. 
Feuer, Lewis S. "Sociological aspects of the relation between language and 

philosophy." Philosophy of Science, 20:85-100 (1953). 
Hackett, Herbert. "Bibliography of the writings of Benjamin Lee Whorf." 

Etc., a Review of General Semantics, 9:189-191 (1952). [The present 

bibliography represents a revision and expansion of Hackett's bibliography.] 
Hackett, Herbert. "Benjamin Lee Whorf." Word Study, 29:3.1-4 (1954). 

• Hoijer, Harry. "The relation of language to culture." Pp. 554-573 in A. L. 

Kroeber (editor). Anthropology today. Chicago: University of Chicago 
Press, 1953. 

• Hoijer, Harry (editor). Language in culture; conference on the interrelations 

of language and other aspects of culture. With papers by F. Fearing, 
J. H. Greenberg, C. F. Hockett, H. Hoijer, N. A. McQuown, S. Newman, 
C. F. Voegelin, J. F. Yegerlehner, and Florence M. Robinett. Chicago: 
University of Chicago Press, 1954. (Also published as Memoir 79 of the 
American Anthropological Association.) 

Kluckhohn, Clyde, and Leighton, Dorothea. The Navaho. Cambridge: 
Har\'ard University Press, 1946. 

Kluckhohn, Clyde. "Culture and behavior." Chapter 25, pp. 921-976 in 
Gardner Lindzey (editor). Handbook of social psychology. Cambridge: 
Addison-Wesley Press, 1954. 

Kluckhohn, Clyde, and MacLeish, Kenneth. "Moencopi variations from 
Whorf's Second Mesa Hopi. International Journal of American Lin- 
guistics, 21:150-156(1955). 

Lenneberg, Eric H. "Cognition in ethnolinguistics." Language, 29:463- 
471 (1953). 

Long, Richard C. E. "Maya and Mexican writing." Maya Research, 2:24- 
32 (1935). 

Long, Richard C. E. "Maya writing and its decipherment." Maya Re- 
search, 3:309-315 (1936). 

Mason, J. Alden. "The native languages of Middle America." Pp. 52-87 
in The Maya and their neighbors. New York: D. Appleton-Century Co., 

Murdock, George P., et al. Outline of cultural materials. New Haven: 
Institute of Human Relations, Yale University, 1938. 

Osgood, Charles E., and Sebeok, Thomas A. (editors). Psycholinguistics: a 
survey of theory and research problems. Indiana University Publications 
in Anthropology and Linguistics, Memoir 10, 1954. (Also issued as a 
supplement to vol. 49, Journal of Abnormal and Social Psychology, 1954.) 


Thompson, J. Eric S. "Pitfalls and stimuli in the interpretation of history 
through loan words." Philological and Documentary Studies {Aliddle 
American Research Institute, Tulane University of Louisiana), 1:2 (1943). 

Thompson, J. Eric S. Maya hieroglyphic writing: introduction. Washing- 
ton, D. C: Carnegie Institution of Washington (Publication 589), 1950. 
Appendix III: "Whorf's attempts to decipher the Maya hieroglyphs." 

Thompson, Laura M. Culture in crisis. New York: Harper, 1950. [Chap- 
ter 8, pp. 152-172, includes excerpts from Whorf's writings.] 

Tozzer, Alfred M. (editor). Landa's Relacion de las cosas de Yucatan. 
Cambridge, Mass., 1941 {Papers of the Peabody Museum of American 
Archaeology and Ethnology, Harvard University). 

Trager, George L. "Comments on B. L. Whorf's Thonemic analysis of the 
English of eastern Massachusetts.' " Studies in Linguistics, 2:41-45 





Cc . 



O / 

1 itr^ i 


■< V 


t_j V, 

>-, ^, 1 3» 


t-i^ ^ 

o>^ 1 -< 


cp ^ 


-^ '.^,. 




Los Angdes 

This book is DUE on the last date stamped below. 



ftB^6'% 14 





JUN 2 7 1994 
DEC 23 1994 


ftMV 3 1 1995 

^y.d, OCT 03 

^loste OCT 26 1996 

JAN 2 1997 

JAN0 6 1998 


APR 2 1998 
1 6 2008 


■it™™* ''' 




5 1 l! i*"^' =^ 

<: ^ i 


33 <3 



I— i Ll_ 

cr •< 


^. §• 

■i . 

^91-9 8Z17 900 1 

:3|a6uv SOI EIUJOIIIEQ |0 AllSjaAlun 



•^ u3^ ^ 



Z2S. •<- 

— : a; 

iV-soV^^'' %a3A: 


r^ tj^ 





S: ^' 











^ u3 

^ ^ 

^ ^<.