Skip to main content

Full text of "Race and language"

See other formats 


ill 



THE 



.^W//..- 



SCIENTIFIC Series 




t 



i 



BOUGHT WITH THE INCOME 
EROM THE 

^■SAGE ENDOWMENT FUND 

THE GIFT OF 

lientg W. Sage 

1S91 



Comell University Ubrary 



arV1851 

Race and language. 




3 1924 031 167 483 
olin.anx 




Cornell University 
Library 



The original of tliis book is in 
tine Cornell University Library. 

There are no known copyright restrictions in 
the United States on the use of the text. 



http://www.archive.org/details/cu31 924031 1 67483 



THE INTERNATIONAL SCIENTIFIC SERIES 
.VOLUME LXXII 



THE 

INTERNATIONAL SCIENTIFIC SERIES. 



Each book complete in One Volume, 12mo, and bound in Oloth. 



1. FORMS OF WATER : A Familiar Exposition of the Origin and Phenomena 

of Glaciers. By J. Ttndali., LL.D., F.E.S. WithaSllluBtrations. $1.50. 

2. PHYSICS AND POLITICS ; Or, Thoughts on the Application of the Prin- 

ciples of "Natural Selection" and "Inheritance" to Political Society. 
By Walter Baobhot. $1.50. 

3. FOODS. By Edwaed Smith, M. D., LL. B., F. E. S. With numerous rius- 

trations. $1.7a. 

4. MIND AND BODY : The Theories of their Relation. By Alkxander Bain, 

LL. D. With 4 Illustrations. $1.50. 

5. THE STUDY OF SOCIOLOGY. By Herbert Spencer. $1.50. 

6. THE NEW CHEMISTRY. By Professor J. P. Cooke, of Harvard Univer- 

sity. With 31 Illnstrations. $8.00. 

7. ON THE CONSERVATION OF ENERGY. By Balfocr Stewart, M. A., 

LL.D., F.R.S. With 14 Illustrations. $1.60. 

8. ANIMAL LOCOMOTION; or, Walkinsr, Swimming, and Flying. By J. B. 

Pbttigrew, M.D., F.E. S., etc. With 130 Illnstrations. $1.75. 

9. RESPONSIBILITY IN MENTAL DISEASE. By Henrt Maudslbt, M. D. 

$1.60. 

10. THE SCIENCE OF LAW. By Professor Sheldon Amos. $1 75. 

11. ANIMAL MECHANISM: A Treatise on Terrestrial and ASrial Locomotion. 

By Professor E. J. Mahbt. With 117 Illustrations. $1.75. 

12. THE HISTORY OP THE CONFLICIT BETWEEN RELIGION AND SCI- 

ENCE. ByJ. W. Drapbb, M.D.,LL.D. $1.76. 

i3. THE DOCTRINE OP DESCENT AND DARWINISM. By Professor OecAB 
Schmidt (Sirasburg University). With 26 Illustrations. $1.50. 

14. THE CHEMICAL EFFECTS OF LIGHT AND PHOTOGRAPHY. By Dr. 

Heumann Vooel (Polytechnic Academy of Berlin 1. Translation thoroughly 
revised. With 100 Illustrations. $2.00. 

15. FUNGI : Their Nature, Influences, Uses, etc. By M. C. Cooke, M. A., LL. I/, 

Edited by the Bev. M. J. Berkeley, M. A., F. L. S. With 109 Illustrations. 
$1.60. 

!6. THE LIFE AND GROWTH OF LANGUAGE. By Professor Willlaj? 
DwiQHT WBMiiET, of Yale College. $1.60. 



New York : D. APPLETON & CO., 72 Fifth Avenue. 



The International Scientijlc Series— {CoDtinued.) 



17. MONET AND THE MECHANISM OF EXCHANGE. By W. Stahlet 

Jbtonb, M. a., F. E. S. $1.75. 

18. THE NATURE OP LIGHT, with a General Account of Physical Optics. By 

Dr. Eugene Lommel. With ISd DlustrationB and a Table of Spectra in 
Chromo-lithOKraphy. $2.00. 

19. ANIMAL PARASITES AND MESSMATES. By MonsienrVAK Bbneden. 

With 83 DlaBtrations. $1.60. 

go. PEKMENTATION. By Professor Schotzbnbeboee. With 28 Ulnstrations, 
$1.60. 

ai. THE FIVE SENSES OF MAN. By ProfesBor Beenbtein. With 91 lllus- 
tratioDB. $1.75. 

82. THE THEOEY OF SOUND IN ITS RELATION TO MDSIC. By Pro- 
fessor Pietko Blasekna. With numerous Illustrations. $1.60. 

23. STUDIES IN SPECTEUM ANALYSIS. By J. Noeman Locktee, F.H.S. 
With 6 Photographic LluBtrations of Spectra, and numerous Engravings 
on Wood. $2.50. 

24 A HISTORY OF THE GROWTH OF THE STEAM-ENGINE. By Pro- 
fessor R. H. Thuestoic With 163 lUuslrations. $2.50. 

25. EDUCATION AS A SCIENCE. By Alexandee Bjin, LL.D. $1.75. 

26. STUDENTS' TEXT-BOOK OF COLOR; Or, Modern Chromatics. With 

Applications to Art and Industry. By Professor Ogben N. Rood, Colum- 
bia College. New edition. With 130 Illustrations. $2.00. 

27. THE HUMAN SPECIES. By Professor A, de Quateefages, Membrc de 

rinstitut. $2.00. 

28. THE CR.IYPISH : An Introduction to the Study of Zoology. By T. H. 

HuxLET, F. R.8. With 82 Illustrations. $1.75. 

29. THE ATOMIC THEORY. By Professor A. Wuetz. Translated by E. 

Cleminshaw, F. C. S. $1.50. 

30. ANIMAL LIFE AS AFFECTED BY THE NATURAL CONDITIONS OP 

EXISTENCE. By Kael Semper. With 2 Maps and 106 Woodcuts. ta.CO. 

31. SIGHT : An Exposition of the Principles of Monocular and Binocular Vision 

By Joseph Le CONTE, LL. D. With 1.32 IllustrationB. $1.50. 

32. GENERAL PHYSIOLOGY OF MUSCLES AND NERVES. By Profesf^oi 

J. RosENTHAi,. With 76 Illustrations. $1.50. 

8.3. ILLUSIONS : A Psychological Study. By James Scllt. $1.50. 

84. THE SUN. By C. A. Young. Professor of Astronomy in the College of New 
Jersey. With numerous Illustrations. $2.00. 

New York : D. APPLBTON & CO., 72 Fifth Avenue. 



The International SdmtlJUs .Series— (Contiimed.) 



m. VOLCANOES: What they Are and what I hey Teach. By John W. Judd, 
F. H. S., ProfeBBor of Geology in the Uoyal Scuooi of Mines. WiLhiliil- 
luBtratious. $:iOO. 

36 SUICIDE: An Essay in Comparative Uoral Statistics. By Henet Mob- 
SELLi, M.D., Professor ot Psychoiogical Medicine, Koyal University, 
Turin. $1.75. 

m. THE FORMATION OP VEGETABLE MOULD, THROUGH THE AC- 
TION OF WORMS. With Observations on their Habits. By CaiEiiia 
Dakwin, LL. D., F. E. S. With Illustrations. $1.50. 

38. THE CONCEPTS AND THEORIES OF MODERN PHYSICS. By J. B. 
Stallo. $1.75. 

a9. THE BRAIN AND ITS FUNCTIONS. By J. Lots. $1.60. 

40. MYTH AND SCIENCE. By Tito Vignoli. $1.50. 

41. DISEASES OF MEMORY: An Essay in the Positive Psychology. By Th. 

RiBOT, author of " Heredity." $1.50. 

42. ANTS, BEES, AND WASPS. A Record of Observations of the Habits of 

the Social Hymenoptera. By Sir John Lubbock, Bart., F. R. S., D. C. L., 
LL. D., etc. $2.00. 

43. SCIENCE OP POLITICS. By Sheldon Amos. $1.75. 

44. ANIMAL INTELLIGENCE. By Qeohge J. Romanes. $1.75. 

45. MAN BEFORE M3TALS. By N. Jolt, Correspondent of the Institute. 

With 148 Dlustrations. $1.75. 

^6. THE ORGANS OF SPEErH AND THFIR APPLTOATIGN IN THE 
FORMATION OP ARTICULATE SOUNDS. By G. H. ton Metee, Pro- 
fessor in Ordinary of Anatomy at the University of Zurich. With 47 
Woodcuts. $1.75. 

47. FALLACIES : A View of Logic from the Practical Side. By Alfeed 
SiDQWiOK, B.A., Oxon. $1.76. 

4«. ORIGIN OF CULTIVATED PLANTS. By Alphonbe ub Caneolle. $8.00. 

49. JELLY-FISH, STAR-FISH, AND SEA-URCHINS. Being a Research on 

Primitive Nervous Systems. By George J. Romanes. $1.75. 

50, THE COMMON SENSE OF THE EXACT SCIENCES. By the late Will 

lAM KlNGBON ClIFFOED. $1.50. 

51 PHYSICAL EXPRESSION: Its Modes and Principles. By Feancis War. 
NEE, M. D., Assistant Physician, and Lecturer on Botany to the London 
Hospital, etc. With 51 Ulnstrations. $1.75. 

52. ANTHROPOID APES. By Eobket Hartmann, Professor in the Unlversltj 
of Berlin. With 63 Dlustrations. $1.75. 



New Yorls : D. APPLETON & CO., 72 Fifth Avenue. 



7%€ International Scientific Series. — (Continued.) 



53. THE MAMMALIA IN THEIR EELATION TO PKIMEVAL TIMES. By 

Obc AH Schmidt. $1.50. 

54. COMPARATIVE LITERATURE. By Hittcbeson Macaulat Posnett, 

M. A., .LL. B., F. L. S., Barrister-at-Law ; Professor of Classics and English 
Literature, University College, Auckland, New Zealand, etc. $1.76. 

55. EARTHQUAKES AND OTHER EARTH MOVEMENTS. By John Miljte, 

Professor of Mining and Geology in the Imperial College of Engineering, 
Tokio, Japan. With 38 Figures. $1.76. 

56. MICROBES, FERMENTS, AND MOULDS. By E. L. Tkotjessaet. With 

107 Illustrations. $1.50. 

67. THE GEOGRAPHICAL AND GEOLOGICAL DISTRIBUTION OF ANI- 

MALS. By Angelo Heilpbin. $8.00. 

68. WEATHER. A Popular Exposition of the Nature of Weather Changes from 

Day to Day. With Diagrams. By Hon. Ralph Aeeeckomet. $1.76. 

59. ANIMAL MAGNETISM. By Alfred Binet and Chables Fee]6, Assistant 

Physician at the SalpStri&'e. $1.60. 

60. INTERNATIONAL LAW, with Materials for a Code of International Law. 

By Leone Leti, Professor of Common Law, King^s College, $1.50. 

61. THE GEOLOGICAL HISTORY OP PLANTS. With Illustrations. By 

Sir J. William Dawson, LL. D., F. R. S. $1.75. 

62. ANTHROPOLOGY. An Introduction to the Study of Man and Civilization. 

By Edwakd B. Ttlor, D. C. L., F. R. S. Illustrated. $2.00. 

63. THE ORIGIN OP FLORAL STRUCTURES, THROUGH INSECT AND 

OTHER AGENCIES. By the Rev. George Henslow, M.A., P.L.S., 
P.G. S. With 88 niustrations. $1.75. 

64. ON THE SENSES, INSTINCTS, AND INTELLIGENCE OP ANIMALS, 

WITH SPECIAL REFERENCE TO INSECTS. By Sir John Lcbeock! 
With over 100 Illustrations. $1.75. 

66. THE PRIMITIVE FAMILY IN ITS ORIGIN AND DEVELOPMENT. 
By Dr. C. N. Stakcke, of the University of Copenhagen. $1.75. 

66. PHYSIOLOGY OP BODILY EXERCISE. By P. Lageanqe, M. D. $1.75. 

67. THE COLOURS OF ANIMALS : Their Meaning and Use. By Edward 

Bagnall Poulton, p. E. S. $1.75. 

68. SOCIALISM : New and Old. By William Graham, M. A., Professor of 

Political Economy and Jurisprudence, Queen's College, Belfast. $1.76. 

69. MAN AND THE 6LACL4.L PERIOD. By G. Prbdebick Wright, D.D 

LL.D., P.G.S.A. With Maps and Illustrations. $1.76. 

70. HANDBOOK OP GREEK AND LATIN PALEOGRAPHY. By Edward 

Maundb Thompson, D. C. L., LL. D., P. S. A. |2.00. 

71. A HISTORY OP CRUSTACEA. Recent Malacostraca. By the Rev. 

Thomas E. R. Stebbing, M. A. With numerous Illustrations. 



THE INTERNATIONAL SCIENTIFIC SERIES 



BACE 
AND LANGUAGE 



BY 

ANDRE LEFEVEE 

rnoFEsaoB in the anthb!5pologioal school, fakis 



NEW YOEK 

D. APPLETON AND COMPANY 

1894 



Authorized Edition. 



^ CONTENTS 



PART I.— TUB EVOLUTION OF LANGUAGE. 

OHAP. piOj. 

I. GENERAL CONSIDERATIONS I 

II. EMBRYOLOGY OF LANQUAGB 20 

III. FORMATION OF WORDS AND THE STRUCTURE OF LAN- 
GUAGES 44 



PART IL— GEOGRAPHICAL DISTRIBUTION OF 
LANGUAGES AND RACES. 



I. THE SPREAD OF INFLECTED LANGUAGES . 
II. THE AGGLUTINATIVE IDIOMS OF CENTRAL ASIA . 

III. THE AGGLUTINATIVE IDIOMS OF SOUTHERN ASIA 

IV. THE MALAYO-POLYNESIAN LANGUAGES 
V. AI'RICAN RACES AND LANGUAGES 

VI. POLYSYNTHETIC LANGUAGES 
VII. THE SEMITIC WORLD .... 
VIIL THE INDO-EUROPEANS . 



64 
86 

no 
133 
iSS 
178 
201 
225 



vi Contents. 

PART III.— THE INDO-EUROPEAN ORGANISM. 

CHAP. PAGE 

L INDO-EUEOPBAN ROOTS ...>.. 262 

II. PABTS OP SPEECH^ — THE NODN 284 

III. THE INDO-EUROPEAN VERB. ... . 303 

IV. THE COMPOUNDS — THE INDECLINABLE WORDS . . 325 
V. INDO-EUROPEAN PHONETICS — THE CONTINUOUS LETTERS 347 

VL INDO-EUROPEAN PHONETICS— THE EXPLOSIVES . . 383 

VIL TWO ANALYTICAL LANGUAGES 405 



RACE AND LANGUAGE 

PART I. 
THE EVOLUTION OF LANGUAGE. 

CHAPTER I. 

GENERAL CONSIDERATIONS. 

Method of evolution — Ancient and modern theories of the origin nf 
articulate speech — Eleim-nts (if sound — Vowels and semi-vowels 
— Explosives or consonants proper — Long uncertainty between 
gutturals, dentals, and labials — The four classes of language — 
Isolating or syllabic, agglutinative, inflected, analytic ; correspond- 
ing originally to different degrees of intellectual capacity. 

During many centuries mankind was anxious to be 
alone in the universe, to establish between the "fallen 
god " and all other living creatures a line of demarca- 
tion, the more inviolable that his vanity overlooked of 
set purpose all intermediate degrees. 

Montaigne and La Fontaine, Georges Leroy and La 
Mettrie, sceptical philosophers or observers of nature, 
who, without any very profound study, had yet remarked 
in animals memory, reason, affections, social relations, 
the rudiments of all arts and industries, seem all to be 
infected to a certain degree by what is still termed the 
spirit of evil. They are all branded with the orthodoxy 
oF religion and prejudice. Even Linnaeus repented of 



2 The Evolution of Language. 

having classed man at tke head of apes in the order of 
Primatii. Yet, having seen the successive removal of 
the barriers raised by official geology, anatomy, and 
even by psychology, man resigns himself at last to be 
no more than tbe first of mortals. 

A very ancient doctrine, which, however, has only 
within the last forty years been put to the proof of 
experience and experiment, the theory of evolution and 
development, has led to a complete change of method. 
In the light of this doctrine students note divergences 
but seek resemblances. Leaving the barren comparison 
of extremes, they give up the easy task of contrasting 
modern civilised man, hoTno sapiens, with the ant, the 
dog, the elephant, or the gorilla. The enormous pro- 
gress made by the least imperfect of mammals is all 
the more clearly established by science, now that she 
grasps, not indeed its cause, but its point of departure. 
Science cannot, it is true, answer the insoluble ques- 
tion — Why is there progress at all ? but its existence 
postulated, science can trace its " low begianings." 
Step by step the scattered links, buried in the depths 
of the past, are recovered and joined anew, the slow 
transitions which have by degrees removed man so far 
from the animals are made manifest in their probable 
or certain succession ; and so the vain regrets for a 
lost paradise give place to the legitimate pride in an 
acquired dignity. 

We take this doctrine, this method, for lamp and 
guide in the boundless field of the science of language. 
Some parts of this domain have in our own day been 
explored with admirable wisdom ; and where we touch 
upon these discoveries of modern talent, we shall only 
need bo set forth facts which are admitted, if too little 
kaown. But here, in dealing with preliminary matters. 



General Considerations. 3 

more initiative is necessary, because of the confusion of 
doctrines. 

On one point only, and even in this we must not 
press too closely the meaning of the words, there 
seems to be complete accord between the simplest of 
men and the most subtle of thinkers, from antiquity 
down to our own time. " Articulate speech is, together 
with the use of fire, the most characteristic attribute of 
man." If we add anything to this formula, we at once 
lay ourselves open to contradiction, specious or valid. 

Does man think because he speaks ? or does he 
speak because he thinks ? The discussion of this 
dilemma is not worth the ink that has been wasted 
on it. If by thought is meant the more or less dur- 
able impression produced in the brain by sensation, 
and the more or less conscious reasoning which gives 
rise to the action consequent on the impression, it is 
evident that thought precedes the vocal act which 
renders it. If thought becomes a labour of the brain, 
independent of the immediate impression, working 
on sound - symbols, retained by memory, elaborated 
by writing, expressed or understood, substituted for 
sensations stored in recollection and analysed by the 
mind, it is no less evident that language is not only 
the instrument, but also the form and condition of 
thought. We shall see, moreover, that there exist 
intermediary stages between crude thought and elabo- 
rated thought, between certain languages and articulate 
speech. The second question is even worse formulated 
than the first. Man does not speak because he thinks. 
He speaks because the mouth and larynx communicate 
with the third frontal convolution of the brain. This 
material connection is the immediate cause of articulate 



4 The Evobition of Language. 

Many eminent men have thought, and still think, 
that man has always possessed articulate speech, nay, 
even grammatical forms. The legend of Adam giving 
names to all cattle has contributed to keep alive this 
convenient but undemonstrable theory. It may be 
said at once, that other: and not less weighty authori- 
ties have always regarded language as a conquest won 
by man, and that the studies of modern philologists 
have definitely established these ancient guesses. 

The wide-spread , faith in the divine origin of lan- 
guage is one of the arguments alleged or accepted by 
the partisans of the former theory. To begin with, it 
is worth neither more nor less than the pretended 
universal consent, so dear to hard-pressed deists and 
devout persons. We should leave it on one side with- 
out insisting further, but that it springs from a 
confusion of ideas on which it is well to throw some 
light. 

The ancients commonly attributed the invention and 
the gift of language to some special or national god, 
Thoth or Jahveh — that is to say, to superhuman men 
framed in their own image, and possessed, like them- 
selves, of mouth and throat — a childish theory whicH 
corresponded to their mental condition. It offered no 
explanation ; rather it solved the question in the sense 
of a chance acquisition of the powers of speech. An.d 
in truth the ancients were convinced that many cen- 
turies had elapsed before language was known. Some 
one had invented it, as Vulcan the art of working in 
metal, or Triptolemus agriculture; and the inventor 
was, like his fellows, a god. That was the point of 
view of the polytheist ; that of the modern metaphysi- 
cian is different. As by the slow process of elimination 
and absorption, the crowd of supernatural beings was 



General Considerations. 5 

gradually reduced to unity, the particular virtues of the 
different gods were concentrated in the God of Mono- 
theism, and finally in the vague God of the deists, 
without parts or organs. This inconsistent being has 
inherited the attributes of Tlaloc, of the Cabiri, and 
the rest. He alone sways, as best he may, the unnum- 
bered thunderbolts whose caprices taxed in earlier 
days the powers of all the gods of Olympus. Amon^ 
other functions, he has retained the distribution of all 
the evil and all the good, of which the source remains 
unknown. Briefly, it is he whom men invoke or evoke 
when all other explanations fail. The writers of 1 830, 
poets, novelists, and historians, vie with each other in 
summoning this dmis ex machina. Even men of science 
intrench themselves behind his inscrutable designs. Is 
it necessary to state that we merely confess our igno- 
rance every time that we have recourse to the deity to 
explain any fact ? The dictum that God gave to man 
breath, memory, speech, is a meaningless phrase. 

There are those who, while they do not admit it, yet 
see the inanity of such an assertion ; but they return 
to it by another road. " No need," say they, " to in- 
voke the supernatural ; nature is all-sufficient ; nature 
gave language to man." If nature be, in this connec- 
tion, simply an equivalent for the deity, the question is 
no nearer a solution. Fortunately a term so indefinite 
is open to various interpretations ; we need not quarrel 
with it if it means the natural origin of language. 
From the moment that the word iiature connotes the 
sum of things and their relations to each other, it no 
longer brings us up against the dead wall of creation 
ex nihilo. Nature in this sense lends herself to the 
research and the inductions of science. We may there- 
fore admit the harmless truism that nature in man 



6 The Evolution of Language. 

implies fclie expression of thought by means of speech. 
Such was the opinion of Epicurus, brilliantly expounded 
by Lucretius. But neither these philosophers nor 
any ancient writer, except the compiler of the Book of 
Genesis, supposed that language was a sudden revela- 
tion ; that man was at once endowed by nature with 
the noun substantive, or even with the separate syllables 
which enter into the composition of words. Their 
gradual evolution is dependent on the slow develop- 
ment of the cerebral and vocal instruments of social 
habits. 

Diodorus Siculus, a compiler of mediocre intelli- 
gence, and Vitruvius, another author of the second 
rank, have stated this simple conclusion in terms which 
Schleicher and Whitney would not disavow. 

"The voice of man," said Diodorus, " being at first 
confused and meaningless, he succeeded at length in 
framing a general system of designations common to 
all, by the constant endeavour to pronounce words 
articulately, and by agreeing together on vocal signs 
applied to each object. But as similar centres of 
organisation arose in all parts of the earth, the result 
was that absence of uniformity which gave rise to the 
diversity of tongues " (Hist. i. 8). This passage, of 
which every word should be remembered, is supple- 
mented by these words of Schleicher : " Language, 
which even during the short period of history has been 
subject to perpetual flux and change, is the product 
of a slow evolution. . , . Moreover, from the moment 
that we recognise in the physical constitution of man 
the principle of his speech, we are bound to admit that 
the development of language has accompanied, step by 
step, the development of the brain and of the organs 
of speech. But if it be language which makes man, 



General Considerations. 7 

our first ancestors were not what we understand by 
man. . . . Thus the study of language conducts us 
unmistakably to the hypothesis of the gradual evolu- 
tion of man from lower forms." 

The ancients, intuitive adherents of the theory of 
evolution, here meet modern men of science. More 
than once they allude to the day when man had not 
the gift of speech, — mutuvi et turpepeciis, " a dumb and 
servile flock," says Horace, " until the day when words 
noted sounds and impi'essions." " Utility," as Lucre- 
tius clearly understood, " called forth the names of 
things," expressit nomina rerum. What need of words 
had the anthropoid of Neanderthal or of La Naulette, 
when, alone and naked, in the thick atmosphere or on 
marshy soil, flint in hand, he wandered from thicket 
to thicket, seeking some edible plant or berry, or fol- 
lowing the traces of some female as savage as himself. 
Act followed impulse as though mechanically, and was 
accompanied by cry or gesture, joyous or plaintive. 
Constant fear, wonder, desire, hunger, and thirst ; every- 
thing that is most crude, most instinctive, least the 
result of reflection ; fleeting curiosity ; the vague and 
fugitive impression of some unexpected sensation ; 
memory at times tenacious, but extremely limited ; 
senses young and unpractised ; brain smooth and with 
few divisions, incapable of analysis — nothing here 
suggests the use of fixed and numerous symbols. 
Before man could give names to things, he must have 
observed them, distinguished them ; nay, more, there 
must have been the need and the opportunity of com- 
municating his observations and discoveries ; the germ, 
however rudimentary, of the family, of a society, of a 
public whose interest it was to understand the utter- 
ances of its members, and to join together in a common 



8 The Evolution of Language. 

undertaking. Afterwards long habit and constant 
effort were needed to retain and apply, to co-ordinate 
and multiply the vocal utterances which, to begin with, 
were uncertain and variable. Diodorus clearly per- 
ceived this ; and Vitruvius, who connects the origin of 
language with the discovery of fire, with the social 
influences of the hearth, shows us a company of men 
endeavouring by means of cries and gestures to com- 
municate to each other their admiration. " They 
uttered," he says, " various sounds %pd shaped words 
by chance ; then, using frequently the same sounds to 
indicate certain things, they began to speak to each 
other." Such are those African savages who fail to 
understand each other at night, and whose imperfect 
speech requires the aid of gesture. 

One of the strongest arguments in favour of this 
probability is the universal admiration which hailed 
the invention, or rather the acquisition of language ; 
the faith in litanies and formulas ; the magic power 
attributed to the spoken word, the revealer, almost the 
creator of the world ; the divine honours rendered to 
the personified Hymn, to poetry, to the Logos, to 
Brahma and the Word ; the inevitable confusion 1)6- 
tween light and language, between speech and reason. 

The study of the elements of speech lends its support 
to arguments drawn from sociology and such general 
considerations. We are not yet concerned with syllables, 
so variously combined in the thousands of idioms spoken 
all over the face of the earth. In their earliest form 
they do but take us back to the beginnings of those 
tongues, dead or living, which we now know of, which 
were built up from the fragments of other dialects now 
for ever vanished. The study of sounds goes yet far- 
ther back towards the source of language ; it deals 



General Considerations. 9 

with the letters, with the sounds of which syllables 
are composed. 

Some of this material is common to us and the 
brutes. It is hardly necessary to observe that the 
vowels, pure or mixed, short or long, nasal or com- 
bined into diphthongs, may be recognised in the utter- 
ances of the dog, the cat, the horse, the ox, the sheep, 
the frog, the toad and the crow. Out of the sounds 
peculiar to each species it is easy to construct, without 
omitting a single note or quality of sound, the entire 
vowel scale : a a, an (nasal) ; e, e, eu, en ; i, I, in ; 
0, 0, on ; ou, ou ; eu ; ii, u, un ; oa, oe, oi, oua, oue, ouon 
(nasal), owi, ui, &c. Note how e and are related to 
a ; to ou ; ou to u ; i to ^ and u ; while in diph- 
thongs the final vowel only is continuous, the first 
ceasing to be heard as soon as it is uttered. 

Another class of sounds give rise to similar obser- 
vations : not only the vowels are susceptible of pro- 
longation. Certain hissings and trills, which can 
give a continuous sound, and are very common 
among animals, have played so important a part in 
the formation of human speech that they cannot be 
too carefully studied. Nor can these be separated 
from other undefined utterances, midway between 
continuity and articulation, ingeniously called semi- 
vowels ; to these we may add hard and soft breathings, 
which precede or follow vowels, semi-vowels, sibilants, 
liquids, and true consonants. All these appear to 
be variants or degenerate forms of the consonants to 
which they are really related ; but the fact that they 
all, except the true consonants, may be found in the 
animal kingdom, may be urged in favour of their 
priority. They form the link between vocalism and 
consonantal language. 



lO The Evolution of Language. 

For while recognisiDg the part played by the teeth, 
the throat, the palate, and the lips in the liquids, 
r, I, Ih ; in the palatals, /, ch, sh ; in the sibilants, 
s, z ; in the semi-nasals, m, n ; and semi-labials, 
w, V, f ; it is also impossible to separate them from 
certain vowels. y, j, ch, Ik, derive from *, conse- 
quently also y, ch, sh, the liquids and the nasal n, which 
often changes with these last. Ou is the origin of 
V, w, m, f ; s and r, which are more independent, 
are not without vowel affinities through j, ch, sh and 
the liquids ; r is reckoned a vowel by the grammarians 
of India; s has something of the character of an 
aspirate, which often takes the place of this letter, 
particularly in Greek and Zend. Now the aspirate, 
considered apart from the consonant, which it streng- 
thens, is only a sort of toneless vowel ; it may be 
compared to the prefatory murmur of an old clock 
before the hour strikes. It results from the effort 
of the breath made in giving the vowel distinctly or in 
articulating the true consonant. 

In the present state of language the various semi- 
vowels often take tlie place of consonants. They have 
acquired this character by that which marks the decisive 
step towards articulation — i.e., the momentary arrest 
of the vowel-breathing by contact with the glottis, the 
tongue, the palate, teeth, and lips. From the moment 
this stoppage is produced, continuity is broken, and 
the issuing sound can only be heard together with a 
vowel or semi-vowel (however slightly audible), which 
precedes or follows it. Such is the phenomenon of 
articulation ; the word consonant, that which sounds 
with something else, expresses its essential character. 

The consonant is the substructure and the founda- 
tion of language. Man alone possesses it, and it is 



General Considerations. 1 1 

the greatest and most fruitful of his conquests. This 
treasure is composed of but six letters: h, g ; t, d; 
p, b ; the gutturals, dentals, and labials. These can- 
not give a continuous sound, however we attempt to 
prolong them ; they can only be the beginning, middle, 
or end of a syllable ; it is impossible to separate them 
from a vowel, a sibilant, a liquid, or an aspirate, with 
which they form a sort of consonantal diphthong, 
Ics, sk, kv, kh, and so forth. It seems probable in- 
deed that these double sounds were the origin of the 
pure consonants. 

Here I foresee an objection : gutturals, it will be 
said, are not unknown to animals ; a number of birds 
and mammals pronounce k, t, p, h. But this is a 
vulgar error ; it is we who attribute these articulations 
to the utterances of animals. The cock does 7wt say 
cock-a-doodle-doo, nor the rook caw, nor the sheep laa. 
They utter the breathings akin to these consonants, 
which, so to speak, lead up to them ; they come near 
to articulate utterance, but man alone has achieved it ; 
not without effort, and with varying success, according 
to the vocal and hearing power of each human race or 
group. 

This is not an assertion deduced from the logic of 
the theory of evolution. The most perfect languages, 
like the crudest, have retained the traces of a Jong 
hesitation, of a remarkable confusion, not only as in 
German, between the weak and the strong consonants, 
but between the three types of true consonants and 
the corresponding aspirate, and even between the true 
consonants and semi-vowels. We hear, as it were, 
across the ages the stammerings and hesitations of 
speech in its infancy. 

Not only have the races unequal power in the use 



12 The Evolution of Language. 

of the gutturals, dentals, and labials, but, in certain 
dialects of Africa and Polynesia, the pronunciation is 
still so uncertain that the most delicate ear can hardly 
distinguish between h and t ; the sound is doubtful, 
and approaches now the one, now the other. In like 
manner many children say tat for cat, many men fail 
to distinguish between cintiime and cinquidme. That 
which is obvious within the limits of the same lan- 
guage is seen on a wider scale in two dialects of the 
same origin, which have grown up at the same time 
and side by side. 

I give a few examples taken from the Indo-European 
languages. In all these the names of numbers np to 
ten, except the number one, are identical ; but it is 
not easy at first sight to recognise as sisters these casts 
from one mould. To be fully persuaded that eight 
and octo, zehn and deka, are the same words, we must 
have heard it stated more than once. The fact is 
certain, however, and I insist no further. Let us take 
the words/oMr and five, the only ones we need consider 
here. The Latin form quattwr, quadru, which has 
given quattro and quatre, corresponds to the Sanscrit 
tc/iatvaras ; Zend tchathwar, tchatru ; Pali tchattaro ; 
Hindustani tchar ; Lithuanian heturi ; Slav tchetvero, 
tcheturi ; Armenian tchorq, tchors ; Greek rerrapes and 
Tea-aapes; also the Umbrian and Celtic patour, pewar ; 
the ^olian irlarvpes ; - A.nglo Saxon and English fidvor, 
four ; thus we pass from guttural and dental diph- 
thongs, hv, tv, tch, to various dentals and labials, t, p, f, 
not to mention the double t alternating with the single 
or double s; or the transformation of the semi-vowel 
V into u and ou in qioatuor, patour, and into o in 
fidvor. It may be noticed that there has been a 
struggle between the guttural and the labial, and now 



General Considerations. 1 3 

the It survives, as in quatre, and now the % strength- 
ened into f ov p (in vier, fidvor, four, patour) ; it is 
less easy to understand the presence of the mixed 
dental tchatvaras and the pure dental ; but it seems 
that at the time when Latin, Umbrian, Celtic, Greek, 
German, &c., were in process of formation from the 
parent language whence they all derive, there was still 
hesitation not only between h, t, or d, and p or /, but 
_ even between these consonants and the forms tch, kv, 
tv. This conclusion becomes yet more obvious on a 
comparison of the various forms of the word five. 
Sanscrit pcmtchan ; Lithuanian penlci ; Armenian 
hing ; Umbrian pump ; Gothic fiinf ; English five ; 
German fiinf ; Greek Tre/ixTre and irevTe ; Slav panti ; 
Latin quinque ; Italian cinque ; French cinq ; Irish 
coic. From these come derivatives as various as Quinc- 
tius, Pompeiiis, Pentecost, fifty. To these examples we 
may add a few well-known permutations: Latin coquere, 
coquus (cook, V. and subs.) ; Greek weTTToo (whence pep- 
sine) ; Sanscrit patch and pah ; Low-Latin sequere, to 
follow ; Sanscrit satch and saJc ; Greek e-woixai ; hoeuf, 
bos, jSoyy, gaus, kuh, cow; ceil (eye), Latin oc-ulus, 
Greek OTTTo/xai, 6<p6a\fji.6s, Sanscrit aJceh (akchan). Tlie 
Latin forms are here simpler than the corresponding 
Greek and Sanscrit forms, the Sanscrit kch, and the 
Greek kt, ttt, era. 

Since no one of the forms adopted respectively 
by one or other of the seven or eight families which 
constitute the entire group can claim the priority, 
and since none have been borrowed by any language 
from any foreign source, we are led to believe, as we 
indicated above, that these are varying pronuncia- 
tions of a primitive form which contained the germ 
of all of them. 



14 The Evolution of Language. 

But this uncertain, confused, primitive. form must 
have existed for thousands of years in a yet vaguer 
shape before it took root in that parent tongue, which, 
ancient as it seems, was yet perhaps unborn at the 
time of the pyramids of Giseh, sixty centuries ago. 

It may be noted in passing, that this uncertainty 
of utterance does not hold good of the pure consonants 
alone ; it is true also of all the continuous consonants 
and of the vowels and semi-vowels. This variability 
occurs in each branch of a given stem, and in every 
stem of this forest which numbers three thousand 
trees ; herein lies the cause of the original or acquired 
diversity of related languages. As for the far wider 
divergence between linguistic groups of families, Indo- 
European, Semitic, Chinese, Uralo-Altaic, Dravidian, 
Basque, Algonquin, Malay, or Bantu, we must look 
deeper for the point of departure; not only in the 
possession of some special aptitude for a given sound, 
but in the variety and inequality of the intellectual 
development. 

If we disregard those Biblical and national preju- 
dices wliieh would trace the descent of the whole 
human race from a single couple, even though science, 
whether disinterested or complaisant, yet certainly 
influenced by current beliefs, has, under the name of 
monogenesis, given her sanction to the theory of ori- 
ginal unity, it has always seemed to me diSScult to 
contravene the opposite hypothesis, viz., that the genus 
homo appeared simultaneously in various places and 
under different skies. It is very true that the farther 
we go back towards prehistoric time the closer the 
analogy and the nearer the I'esemblance between rudi- 
mentary industries and ideas. But because the first 
efforts of man to raise himself above the brutes have 



General Considerations. 1 5 

everywhere led him to employ the same methods of 
constmction, to display the same social tendencies, 
and the same errors in his conception of the universe, 
it does not follow that the structure of the skeleton, 
that the form of the cranium, that the faculty of the 
brain has everywhere been identical. Just as there is 
no individual who does not differ in some respects 
from those who most resemble him, so the earliest 
ancestors of races now distinguished by colour and by 
the quality of the hair, by social polity and intellectual 
capacity, must have possessed, in order to transmit 
them to their descendants, the earliest germ of that 
divergence which now shows itself in such various and 
such manifest differences. However this may be, these 
distinguishing characteristics, primordial or acquired, 
peculiar to each group of races, pure or mixed, are to 
be found in the very structure of language. 

A cursory glance at a chart showing the distribution 
of languages reveals the fact that a third of the human 
race, not the least advanced in civilisation, the sub- 
jects of the Chinese Empire, with its dependencies, 
Burmah and Indo-Cliina, are ignorant of that which 
we call grammar. From the beginning of the world 
these people have made use of isolated syllables to 
which a strict syntax assigns in turn the value of 
verb, substantive, adjective, adverb, or preposition. 

Everywhere else, as far as we can ascertain, the 
other human groups, whether low in the scale or well 
endowed, homogeneous or of mixed race, whether num- 
bered by hundreds or by millions, have associated with 
a root-syllable other syllables which determine or 
modify its sense. This agglomeration, which does not 
indicate any degree of kindred among the languages 
to which it is common, is of two kinds. In the fami- 



1 6 The Evolution of Language. 

lies — some of them very rich and very varied — called 
Mongolian, Uralo- Altaic, Dravidian, Malay, in the in- 
numerable dialects of Oceania, Africa, and America, 
the root or central syllable remains as a rule unchanged; 
the accessory syllables, whether prefixed or suffixed, are 
more or less obliterated according to the laws proper 
to each family and dialect. These are called the agglu- 
tinative languages. Here again the classification is 
purely formal, since it ranges in the same category 
types as diverse as Japanese and Basque, Mandchu 
and Tamil, Polynesian and Turkish, Algonquin and 
Kaffir or Bashman. We insist the more on this in- 
coherence because a false air of kinship has been given 
to this class of languages by bestowing on them the 
fantastic name of the Turanian family. 

The Semitic and the Indo-European languages make 
the fusion of the agglutinated syllables more complete, 
alter and inflect the root itself, and often reduce 
suffixes and prefixes to unrecognisable fragments. 
The Semitic group, however, respect the root conso- 
nants ; so much so, that if we disregard the termina- 
tions, since the Semitic character usually omits the 
vowels, the vocabulary, really identical at bottom, 
appears not to vary at all in Assyrian, Phoenician, 
Hebrew, or Arab. In the Indo-European languages, 
on the other hand, whether in passing from one lan- 
guage to another, or within the limits of the same 
tongue, the variation affects both vowels and conso- 
nants — we only note these distinctive characters in 
passing. The two marked types which we have just 
defined constitute inflected languages ; they differ from 
the preceding as vertebrate from articulated animals. 
Inflection makes of each word an organism, a solid 
and robust individual, distinguished in form, as in 



General Considerations. 17 

meaning, from other words in which analysis discovers 
the same original root ; it gives to thought incompar- 
able freedom and precision, and furnishes it with an 
abundant vocabulary of exact terms and of derivatives 
independent of the parent root and of other cognate 
words. 

At length, and as the result of long use and close 
contraction, the words of inflected languages lose all 
trace of the syllables which have atrophied in the 
terminations of nouns and verbs, and return by an- 
other road to the syntax of monosyllabic tongues ; 
while the part formerly played by the termination in 
declension and conjugation is taken up by particles 
and auxiliary verbs. The type of these languages is 
modern English, and in a hardly less degree the group 
of languages derived from Latin ; these are called 
analytic languages, as distinguished from the ancient 
tongues, called synthetic, from which they descend. 
The evolution of language may be divided into four 
stages : monosyllabic, agglutinative, inflected, analytic. 
Without at present staying to inquire whether the 
monosyllabic tongues may not themselves be the result 
of contraction applied to confused and prolonged modu- 
lation, we may be permitted to regard as successive 
the four degrees or planes presented by the table of 
languages. No gulf separates one from another ; it 
is easy to find transitions from one to another. Just 
as our analytic are derived from inflected languages, 
so. these latter are but a variety of the agglutinative, 
which are composed of monosyllables placed in contact 
with one another, but capable of independent existence. 
The problem is to determine why certain tongues 
should have stopped at the first or second stage, while 
others reach or pass the third. Whatever weight we 



i8 The Evolution of Language. 

allow to historical circumstances, it is hardly possible 
to ignore some original inequality. 

It is clear that, however this may be, the mono- 
syllable is the embryo of every language and of every 
vocabulary which is known at the present day, and 
that, in the last analysis, it can be traced through all 
subsequent modifications. This search for the original 
elements is one of the great attractions of the study 
of language. For, if we have correctly defined the 
intimate relations between speech and thought, there 
is the chance of discovering in these syllables or roots, 
as they are termed, the trace of the first beginnings 
of intellect, the very dawn of thought. And these 
delicate researches have nowhere a greater charm than 
when prosecuted in those groups of tongues which 
are most advanced, in which the mind finds its most 
powerful instrument, its richest treasure-house. 

While indicating the attractions and the importance 
of the search, we would not conceal the difficulties 
which attend the seeker, nor the pitfalls to be shunned. 
In the first place, thre formal, morphological differences 
which separate the four great categories render all 
comparison useless. Chinese is of no avail to the 
student of Turkish ; the American, Kaffir, or Malay 
dialects throw no light on the Semitic, Berber, or 
Indo-European languages. Not only does the differ- 
ence of grammatical construction forbid useful com- 
parison, but the fundamental diversity of vocabulary 
erects another barrier. In all families of languages 
the same association of consonants and vowels may be 
met vyith, but the same sound does not correspond to 
the same sense. This is yet another proof of the 
original differences in the faculty of speech. We may 
go farther. If, in a given family of languages, in 



General Considerations. 1 9 

which the same roots form the common basis of all 
the ancient and modern dialects which compose the 
group, we analyse these roots, we shall find the sense 
evade us. Take, for instance, the Indo - European 
roots, which hardly exceed five hundred in number. 
Each one bears several very different meanings, which 
are referred somewhat arbitrarily to a vague and 
general signification. Many among them differ from 
the others only by the probable fusion of another root, 
since atrophied, which seems to have been used to 
define the sense. Take away this adventitious aid, 
and the number is reduced to a few dozen, whose 
meaning is as doubtful as the form is simple and 
crude. 

It remains to be discovered, if possible, why this 
or that instinctive utterance, vowel or consonantal, 
has characterised or evoked such and such an image, 
sensation, or relation between thought and action. 
May we hope to discover the germ of articulate and 
definite speech in interjection, in emotional language, 
in onomatopceia, or the imitation of certain sounds ? 
How should given sounds, to which a definite meaning 
was attached, have become fixed in the memory of a 
horde or clan, and communicated by degrees to the 
tribes which neighbourhood or interest brought into 
relations with it ? 



CHAPTER II. 

EMBRYOLOGY OF LANGUAGE. 

The cry, the first element of language — The cry of animals : expressive 
of emotion, the forerunner of the verb, and the name of a state or 
an action ; the call, germ of the demonstrative roots — The human 
cry : variety of intonation, stress, reduplication — Onomatopceia — 
Traces of direct onomatopoeia — Approximative, symbolical, or 
generic onomatopoeia— Onomatopceic theories of Plato, Leibnitz, 
De Brosses, and of Court de Gobelin — Onomatopoeia defended 
by Whitney, rejected by Paul Regnaud — Metaphor, founded on 
mistaken analogies, has vitiated language from its very birth. 

In order to establish the descent of man, the naturalist 
seeks to discover the living or extinct forms which 
may have served as transitions betvceen the classes, 
orders, and species. In spite of numerous breaks in 
the chain, he can nevertheless follow out the processes 
of selection and heredity, can trace the gradually in- 
creasing complication of organs and functions, the 
slow co-ordination of the members round a spinal 
cord protected by an envelope, which becomes ossified 
into vertebrae, until finally the nerve substance and its 
diverse energies are concentrated in a cerebral ganglion, 
where impressions derived from without are translated 
into ideas and movements. 

It is by similar methods that the philologist follows 
up language from the modern analytical phase to the 
inflected state, and so to the agglutinative and mono- 
syllabic ; and in all probability we have here a com- 
plete cycle, a series of which the various stages are 
each occupied by one or more groups of idioms. 



Embryology of Language. 21 

retarded or hastened in their onward progress, either 
by known facts of history, isolation, migration, inter- 
mixture, or by some ethnical and national aptitude or 
inferiority, which may be enduring or ephemeral. 

Yet this classification does not carry us far back 
into the past. Moreover, the four great categories 
which, taken together, comprehend all languages do 
not imply kindred among those idioms ranged in any 
one class. Vocabulary, and not grammar or syntax, 
marks original relationship; and the vocabularies — 
the inheritance of entirely distinct families — cannot 
be referred to a single common origin, because they 
are the result of different vocal and mental aptitudes. 
No dramatic incident like that of the legend of Babel 
dispersed the nations and languages. Each linguistic 
stem germinated apart, and grew up alone in its own 
sphere ; each family must be studied separately ; and 
in each we are completely ignorant of the changes 
which took place prior to the historic period. 

Must we then admit with M. Michel Br^al that 
the origin, not of language, but of the meaning of 
words, lies beyond our ken ? But even he, who, with 
a,ll his discretion, is so bold, has not abandoned the 
quest which he declares to be vain ; and from Plato 
down to Schleicher, Whitney, Steinthal, Noire, Paul 
Regnaud, and a hundred others, whose opinions we 
shall have to discuss, this central problem has exercised 
the minds of men. 

When experience and induction by their mutual 
aid had at length succeeded in tracing the genea- 
logical tree of mankind, a fortunate discovery was 
made by anthropology. In embryology the student 
found an abridgment, a summary, of the transforma- 
tions discovered or assumed from age to age. By the 
8 



2 2 The Evolution of Language. 

aid of the microscope, foetal life reveals to the eye 
all the phases in the development of the cell, of the 
&a^, of the very simple material aggregate which is 
destined to be clothed with the dignity of humanity — 
that is to say, to realise within a few months the 
work of a thousand centuries. Now it seems that 
language also has in some sort its embryology. Not 
that we can ever be the spectators of the formation 
of a language ; but we possess the germ nevertheless, 
the undoubted embryo of speech — the Cry, which in 
most of the higher animals, even in man himself, 
exists as an independent utterance, and suffices for 
the expression of certain sentiments, and even of a 
few ideas, and is consequently the first element of the 
crudest forms of speech. From the moment that we 
reject supernatural intervention and regard language 
as the work of time, it is not possible to seek its 
point of departure and its germ elsewhere than in the 
resonance of the air against the vocal chords, and 
in the emission of this resonance by the mouth and 
nostril. The production of the voice is at first as 
unconscious, as reflex as any other bodily movement. 
The cry, in the lower species and in the infancy of 
the higher, is invariable, like the wail of a new-born 
infant. The toad, for instance, has but one note ; that 
of the cuckoo and of a number of wild animals is 
hardly richer. Yet, since it responds to some -impres- 
sion or to some need, the sound already has a mean- 
ing, since it attracts or drives away creatures whose 
interest it is either to flee from or to approach the 
utterer of the sound. This meaning, very vague, or 
rather very elastic, grows more exact with the sensation 
of which the sound is the result ; the single note of 
the toad implies already an affirmative or imperative ■ 



Embryology of Language, 23 

proposition, embodying the sexual desire, or something 
similar. Repetition, continuance, the raising or lower- 
ing of the tone, mark the earliest efforts to attain 
to the expression of more varied and more distinctly 
realised sensation. Modulations, more or less uncer- 
tain, more or less fixed by practice, as consciousness 
dawns, come to increase the vocal resources. A given 
vocabulary will include five, six, or even ten variations 
of the specific cry, each one doubled by a stronger or 
weaker form, and susceptible of expressive combina- 
tions, comparable to our derivatives and compound 
words ; the language thus reflects^ so to speak, the 
shades of joy or pain, fear or desire, sickness or 
health, hunger or thirst, changes of temperature, the 
approach of day or night. Lucretius in his fourth 
book translates with a rare felicity all these utterances 
of birds, cows, dogs, and horses, in which are clearly 
represented those sensations and aSections which are 
common to us and to most animals. 

In animals the specific utterance is but the imme- 
diate expression of a present emotion. This obser- 
vation is fairly just, and insisted on by those who 
would accentuate the line of demarcation between men 
and brutes. It is more to our purpose to find some 
modification of a formula which is too absolute as it 
stands. For does not the language of the brutes at 
times pass the limits assigned to it ? Evoked by an 
immediate sensation, does it not sometimes correspond 
to some enduring recollection, even to a prevision 
which may be realised ? 

We are not sufficiently acquainted with the vocabu- 
lary of the anthropoids to interpret the nocturnal 
choruses of certain apes ; bijt we cannot doubt that 
the dog, which so readily distinguishes its friends 



24 The Evolution of Language. 

from enemies or strangers, summons, welcomes, or 
threatens each of these in a perfectly comprehensible 
manner. He warns, thanks, questions, asks to be taken 
for a walk or out hunting; in dreams, when memory 
excites certain chambers in his brain, he growls at the 
passer-by whom he suspects, he gives tongue when he 
imagines he sees a covey put up or a hare start from 
her form. Asleep, he does in a measure that which 
man does in his waking hours — -he specifies by sounds 
which are bat symbols certain past impressions which 
have no present objective existence. 

Nor is memory only called into action in this 
embryonic language, but foresight also, and therefore 
reflection and will. From the beginning the emotional 
cry is a summons, understood by those who hear it, if 
not by those who utter it ; it soon adapts itself to less 
elementary needs than the original instinct; it is by 
turns a warning, a command, a convocation against 
danger, for common defence, for hunting, or for 
warfare. 

These remarks on the character and use of the cries 
of animals, which may be observed by every one, apply 
without a shadow of doubt to the language of the 
anthropoid which was slowly developing into man. 
We may also add with certainty that this utterance, 
which tended to become human, was richer in modu- 
lations, more expressive, and necessarily conveyed a 
more direct intention than that of any other creature ; 
and that to the artifices, already numerous, of dnpiica- 
cation, continuance, raising or lowering the note, &c., 
were added the thousand efforts of the voice to articu- 
late the consonant, as yet beyond its mastery. 

It is said, and with reason, that the interjection is 
barren and unchanging ; that it has no place in Ian- 



Embryology of Language. 25 

guage properly so called ; that speech begins where 
interjection ends. The interjections common to all 
peoples, Ah, Oh, Eh, &c., still indeed suffice to express 
the simple emotions which provoke them — ^joy, pain, 
fear, desire, doubt, — and carry us back to the early 
period of emotional language. But it must be remem- 
bered in the first place, that many interjections may 
have disappeared, may have been absorbed into the 
words of which they were the original roots ; and in 
the second place, that the utterances even of animals 
imply memory and reasoning power, and that in man 
they have shaped themselves to the growing com- 
plexity of a progressive and social being. 

Simple sounds like A and I have been, and still 
may be, merely exclamations, but they have played a 
great part in the Indo-European languages ; long or 
short, they have formed pronouns and verbs; they 
indicate movement, place, even privation and negation; 
reinforced by a nasal, an aspirate, a liquid, one or 
more consonants, their power increases a thousandfold. 
The forty thousand monosyllables which compose the 
Chinese tongue are formed in the same way. This is 
but a single example, but it is a weighty one, of the 
ductility, of the almost infinite variability, of human 
utterance. Other languages have employed other 
methods, and have not sought to increase the number 
of monosyllables, but rather to associate and combine 
them. 

This tendency, which results in the rich develop- 
ment of grammatical forms, began by the prolongation 
or the reduplication of the syllable which is so common 
in animals and children — a custom so inveterate that it 
still obtains among us without suggesting its ancient 
influence upon the development of language. When 



26 The Evolution of Language. 

■we lay stress upon a syllable or upon a word to give 
it prominence, we use an instinctive artifice, natural 
to children and to savages, and to all people of limited 
vocabulary. Taylor, in the second volume of his 
"Primitive Culture," has collected many examples, 
taken from the languages of America and Oceania, of 
this prolongation of a sound to mark distance, import- 
ance, degrees of comparison. Vowels and liquids are 
repeated as many as five and six times. Stress, stereo- 
typed by custom, has produced accent, so diverse, so 
difficult to reduce to a common principle or law ; it 
has furnished grammar with a valuable means of dis- 
tinguishing gender, tense, and person. 

Reduplication suggests the same remarks. When 
we say no, no ; yes, yes; hip, hip; well, well; come, come; 
when children say papa, mama, dada, far far away, we 
and they obey the instinct which led our ancestors to 
enforce attention by the repetition of the same sound. 

There is no tongue in which this primitive expedient 
has not left obvious traces. Many Polynesian, Ameri- 
can, African, and other tribes call themselves, or are 
called by others, by names composed of a reduplicated 
syllable^Shoshones, Chichimecs, Niam-niam, Leleges, 
Tatar.', Berbers, without reckoning proper or common 
names, such as Unkulukulu among the KaflSrs, Tame- 
hameha among the Sandwich Islanders. 

It is incontestable that the method to which we 
owe such words as murmur, Marmar (the ancient form 
of Mars, contracted into Mamers, Ma-ors, Mavors), 
barbarus (stammerer, he who cannot speak aXaXo?), 
purpura, turtur, pipio, titio, is very wide-spread ; these 
are only surviving fragments of a method of formation 
which is still in full vigour in many contemporary 
dialects and jargons. 



Embryology of Language. 27 

Analysis discovers, moreover, in a number of roots, 
slightly differentiated in form and meaning, the origi- 
nal identity of repeated sounds which time has obli- 
terated and fused together ; sometimes one of the twin 
syllables has lost or modified its vowel or consonant ; 
sometimes they are contracted and agglutinated. Hence- 
the search is hazardous, but the fact admits of no doubt; 
it is enough to compare such forms as genvs, genui, 
genitor, with gigno, jLjvofi.ai, yiyova; or again mens, 
moneo, mania, with memini, memoria, memnon, to re- 
cognise, in the second term of the comparison, the 
reduplication of the roots gen and men, which have 
thus given rise to hundreds of derivatives. A whole 
class of verbs in Sanscrit, Greek, and Latin has been 
formed in this way : daddmi, SlSoa/xi, dedi ; cf. datum, 
Swa-w, dare. The system of the Greek conjugation 
rests in great part on the ingenious use of these varia- 
tions, where the weakened reduplication of the root 
syllable characterises certain moods and tenses. Thus 
the most rudimentary expedients of language in its in- 
fancy reappear in the elaborate devices of an advanced 
state of civilisation. 

We have said enough to show that the specific cry, 
somewhat modified by the vocal resources of man, may 
have been amply sufficient for the humble vocabulary 
of the earliest ages, and that there exists no gulf, no 
impassable barrier between the language of birds, dogs, 
anthropoid apes, and human speech. The summoning 
cry, so largely used by animals, has been developed 
and defined into command, into indication of distance, 
number, person, sex ; or demonstrative terms retained 
and exchanged by members of fixed or ephemeral 
societies, — horde, family, or tribe, — have grown beyond 
it, and been accepted, modified, or increased by neigh- 



28 The Evolution of Language. 

bouring groups. With regard to the emotional cry, it 
is perhaps, in spite of its reflex and involuntary char- 
acter, a yet more important factor. Being associated 
with all sensations, and with all the movements called 
forth by these, it affirms a state, the passing from one 
state to another, and, subsequently, an action and tlie 
result of an action. Now all this is characteristic of 
the verb ; so that, placed between two demonstratives, 
it gives to these respectively the value of what we call 
subject and object ; it makes the pivot of a proposi- 
tion, of a very elementary nature indeed, but in which 
is epitomised the fundamental mechanism of speech. 
To make ray meaning clearer, I will use English or 
Latin words, but it must be understood that each word 
should be considered as a simple emission of the voice, 
entirely uninflected. Let us take the most neuter of 
demonstratives, this, that, hoc, id, and place between 
them an utterance which indicates pain, joy, anger, 
desire, an utterance known and understood of those 
who hear it : this pain that ; that joy or ange.r or 
desire this ; add appropriate gesture, and the right 
translation of the meaning is easy in each case. He, 
you, or I am hurt or pleased by this arrow or claw, by 
this food or drink ; that he or you strikes, pursues, 
devours, or fears this, him or me. Replace the vague 
demonstrative . by the names of persons or things, and 
you have in its essentials the speech of negroes, or 
even of the civilised Chinese. • 

But the name, the substantive, or at any rate a 
large class of these, differs from the verb only by ter- 
minations or inflexions of very much later date than 
the primitive form, and non-existent in the mono- 
syllabic group. The possible verbs, of which we see 
the germ in the emotional cry, include also potential 



Embryology of Language. 29 

nouns, names of sensations, states, movements, actions. 
This is so true, that in our efforts to convey the thought, 
yet vague and undefined, of our remote ancestors, we 
have been obliged to use indifferently the words pain 
and suffer, blow and strike, fright and fear, in order 
not to use the naked root, which would have required 
a long explanation. Let us take, however, a single 
example from the Latin tongue : in dolor and dolere 
(pain and to suffer), if you suppress the substantive 
termination or and the verbal form ere, there remains 
the significant or root syllable, which is neither verb 
nor noun, but may be used for the one or the other. 
Some people may be surprised to find that so early an 
origin is attributed to nouns which are often called 
abstract ; for, on this hypothesis, what becomes of the 
received opinion that the first substantives were the 
names of concrete objects ? This distinction does not 
seem to me as valid as it is often considered. The 
faculty of abstraction is inseparable from the intelli- 
gence, which signifies, etymologically, the choice among 
several facts or qualities. A sensation realised is 
already an abstraction ; and the vocal utterance which 
corresponds to it distinguishes or abstracts it from 
other sensations ; language has no other oflace. And 
since the impression upon the subject, or subjective 
impression, precedes of necessity the knowledge of the 
object, or objective impression, it is the former which 
is first expressed in speech. An increased power of 
abstraction led to the need of designating objects 
external to man. 

Animals, which see, and even recognise, local peculi- 
arities, seem very rarely to analyse into its details the 
whole which has impressed them. Their attention is 
sluggish or ephemeral. It was the same, in his degree, 



30 The Evolution of Language. 

with man when as yet hardly raised above the bmte. 
Very slowly, after having in some degree succeeded in 
expressing his own emotions and intentions, he began 
to try to fix in his memory, by a vocal symbol, the 
fluctuating image of those objects whose contact or 
approach excited his sensations or provoked his actions. 
At first he encountered difiiculties apparently insur- 
mountable ; how describe by means of sound, colour, 
odour, taste ? how paint with the voice ? It had to 
be done, however, and man attained to this art by 
degrees and unconsciously, rendering at first, like an 
echo, sound for sound ; then referring the sound to 
the thing which gave the sound ; then to the things 
and phenomena which a given sound accompanies or 
heralds ; and finally to the thousand ideas excited in 
a brain progressively more rich and active by the 
mere mention of the symbol which connoted ah'eady 
several series of metaphors. 

The imitation of the utterances of animals and of 
the sounds of inanimate nature has been almost uni- 
versally considered as the principal source of the roots 
called attributive, with which are connected the greater 
number of substantives and verbs ; whence the name 
onomatopceia, that which creates names. It is so 
plausible an hypothesis that it has conquered most 
philosophers {pv6fx.aTa jA^kefkaTa, says Aristotle), and 
also philologists such as Renan, Whitney, Farrer, 
Wedgwood. 

Max Mliller and Paul Regnaud, speaking on behalf 
of Indo-Euiopean philology, reject this theory, the 
latter especially with vehement conviction ; but their 
criticism and their reservations fail to eliminate the 
well-known tendency of children, and even of grown 
men, to more or less accurate onomatopoeia ; and if 



Embryology of L anguage. 3 1 

such phrases as ding-dong, tic-toe, hang, &c., constitute 
only an infinitesimal and barren portion of our voca- 
bularies, it cannot be denied that they abound in a 
number of languages which have not attained to the 
inflected state. Moreover, not to mention words like 
kukuta (cock) in Sanscrit, ululare, balare, mugire, 
hinnire, &o., in Latin, we find throughout the Indo- 
European tongues roots with many derivatives of 
every kind, in which can be traced, through all 
changes .of sense and sound, the primitive onomatopoeia, 
but an onomatopoeia which is, so to speak, generic, 
and applicable to a whole class of allied sounds. It 
will be readily understood that this symbolical onoma- 
topceia is vague and doubtful, and has led the acutest 
intellects into grave errors. 

The earliest theory of this onomatopoeia, resulting 
from the adaptation of the sound to the idea, occurs 
in the " Cratylus " of Plato. " In the first place," says 
Socrates, " the letter r appears to me the general in- 
strument expressing motion. . . , Now the imposer of 
names frequently used it for this purpose ; for example, 
in the actual words peiv, pori, he represents motion by 
r ... by the letter i he represents the subtle elements 
which pass through all things ; by the sibilants fh, ps, 
s, z, all which shakes, agitates, swells ; by d and t, 
that which binds or rests in a place ; by I, all things 
smooth and gliding ; by gl, things "of a glutinous, 
clammy nature," &c. I omit the examples taken from 
the tongue which Plato spoke, and of the earlier 
phases of which he had no conception. The Stoics, ac- 
cording to St. Augustine, had entirely accepted these 
fanciful theories; they held, with Court de Gobelin, 
that the " voice indicates agreeable objects by agree- 
able sounds, and unpleasing objects by harsh and 



3 2 The Evolution, of Language. 

strident sounds." Thus in lana, lenis, mel, the liquid 
I expresses softness ; in asper, vepres, acre, sp, pr, cr 
mark roughness ; crura (thighs) renders at once the 
ideas of length and hardness. 

Leibnitz, one of the promoters of the study of com- 
parative philology, is not more fortunate in the sug- 
gestions which he brings to the support of the doctrine 
of "Cratylus." M. Paul Regnaud, in his interesting 
book " On the Origin of Language," has quoted the 
most curious of these, and we quote them from him in 
order to show that neither genius, nor serious inten- 
tion, nor real knowledge, avail to protect us from the 
wildest errors. But it matters not, since the errors of 
yesterday beget the truths of to-day ; philology has 
had its alchemists. 

" With Socrates, or rather with Plato, Leibnitz be- 
lieved that the letter r has been employed by the natural 
instinct of different peoples, such as the ancient Ger- 
mans, the Celts, &c., to signify violent movement and 
a sound like that of the letter. This appears, he says, 
in pew, to flow ; rinnen, rilren (Jlv£re), rutir, flowing ; 
Bhenus, Bhodanus, Eridanus, Bura (Rhine, Rhone, 
Po, Roer) ; rauben, rapere, to rob, ravish ; radt, rota 
(wheel) ; rauschen, to rustle ; raclchen, to stretch vio- 
lently, to rack, whence reichen, to reach ; der Bick (in 
Platt-Deutsch), a long stick or pole ; whence also Bige, 
Beihe, regula, regere, words implying length or straight- 
ness ; Berk, a long or tall person, a giant, and then 
a powerful or rich man, as is seen in Beich (German), 
and in riche and ricco (French and Italian). In Spanish 
ricos homhres means nobles or chief personages, which 
shows how metaphors have caused words to acquire a 
new meaning of which the connection is far from obvi- 
ous." This observation recoils upon its author. Of 



Embryology of Language. 33 

all the words hitherto cited, there are hardly two which 
are not out of place in each other's company. 

But to continue. The letter r is not exhausted. 
" It indicates also violent movement and noise in Riss, 
rupture, with which the Latin riimpere, Greek priyvvixi, 
French arracher, and Italian straccio are connected. 
Now, just as r indicates naturally a violent movement, 
so I signifies a more gentle movement. So we find 
children and others for whom the r is too harsh and 
too difficult to pronounce substitute the I, and say, for 
example, velly for very. This gentle movement ap- 
pears in leben, to live ; laben, to comfort ; Helen, to love 
(lubere, libido) ; lind, lenis, lentus, gentle, soft ; laufen, 
to pass rapidly like flowing water, Idbi {labitur uncta 
vadis abies) ; legen, to lay, whence liegen, to lie, and 
Lage and Laye, layer, as in Laystein, stratified stone, 
slate ; Laub, leaf, a thing easily moved ; Lap, labra, 
lip ; lenken, luo, loosen, dissolve ; lien (Piatt- Deutsch), 
to melt, whence the Leine, a river in Holland, which, 
rising among mountains, is swollen in spring by melt- 
ing snows. Not to mention a number of other similar 
words which prove that there is something natural in 
their origin, something which indicates a relation be- 
tween things and sounds, the movements and organs 
of the voice. And for this reason also the letter I 
indicates the diminutive in Latin, and in languages 
derived from Latin, as also in High German. Yet it 
niust not be supposed (a happy reservation !) that this 
tendency can be everywhere observed, for the lion, the 
lynx, and wolf cannot be styled gentle. But it may 
be that the attention has been directed to another cir- 
cumstance, and that is their swiftness, which causes 
them to be feared, and puts to flight : as though he who 
saw such an animal cried to his companion, 'Zazt/' 



34 The Evolution of Language. 

(flee) ; besides, through many accidents and changes, 
most words have become very different from what they 
were in their primitive pronunciation and meaning." 
Here good sense appears through this collection of 
childish subtleties. (New Essay on the Understand- 
ing.) 

Nothing can be more far-fetched than the genealogy 
of the word Auge, eye, as Leibnitz gives it. " A" he 
says, " the first letter, followed by a slight breathing, 
becomes Ah, and as this emission of the breath makes 
a sound distinct enough at its beginning, and then 
grows faint, this sound signifies naturally a breath, 
spiritum lenem, when a and h are but faintly heard. 
Hence the origin of "Aw, aura, haugh, halare, haleine, 
oLTfio^, athem, odem. But as water is also a fluid, it 
would seem that Ah, rendered stronger by reduplica- 
tion. Aha or Ahha, has come to mean water. The 
Teutons and Celts, for the better indication of move- 
ment, have prefixed their w to the one and the other ; 
thus welien, wind, vent, mark the movement of the air ; 
and waten, vadum, water, the movement of water or 
in water. But to return to Aha, it seems to be, as I 
have said, a root which indicates water." (Observe 
that there is no reason for the supposition.) " The 
Icelanders, whose language is akin to the ancient Scan- 
dinavian, omit the aspirate and say Aa ; others, who 
say Aken (meaning Aix, Aquas Gfranni, the waters of 
the Gallic god Grannus), have strengthened it, as have 
the Latins in their aqvM, and the German ach in com- 
pounds ; Schwartzach, black water ; Biberach, the water 
of beavers : and in ancient titles, Wiseraha (in Latin 
Viswrgis), and Ilerach (Latin llargus). From aqua, 
aigues, auue, we come at last to the French eaxt,, in 
which no trace of the origin remains. Auiue, Aiige, 



Embryology of Language. 35 

in German, means now a place which is often flooded, 
a water-meadow, and more particularly an island. . . . 
And this must be the case among a great many Teu- 
tonic and Celtic peoples, for hence it is that all which 
is, as it were, isolated in a species of plain has been 
called Avge, or Oit^e (oc-ulus). Spots of oil on water 
are so called in German ; and in Spanish ojo is a hole. 
But Auge, ooge, oculus, occhio, &c., have been applied 
moi-e particularly to the eye, that brilliant hole in the 
countenance, and doubtless the French word ceil comes 
from the same source, though its origin is not recog- 
nisable without tracing the successive links of the chain, 
as I have just done : it would seem also that the Greek 
ofifia, ovp-o-jy, come from this root. Oe or Oeland means 
island in the far North, and there are even some traces 
of it in Hebrew, where Ai means island. M. Bochart 
believed that the Phoenicians took the name given 
to the ^gean Sea from this root (signifying full of 
islands). Augere, to increase, comes likewise from 
Auge, that is to say, the rising of waters ; as ooken, 
auken, in Old Saxon meant to increase, and the imperial 
title Augustus was translated Ooker. A river of Bruns- 
wick rising in the Hartz Mountains, and consequently 
subject to spates, is called Ooker, anciently Ouacra." 

The ravings of a lunatic are not more incoherent ; 
but the philosophers of the eighteenth century did not 
think so. De Brosses, a man of the clearest intellect, 
in his treatise on the Formation Micanique des Langues, 
goes even farther than Leibnitz, and Court de Gobelin 
(in his Monde Primitif) goes yet farther than De 
Brosses. But it should be observed that, if their 
demonstration is valueless, their general theory, the 
principles which they lay down, are so full of good 
sense and so plausible, that true philologists who are 



o 



6 The Evolution of Language. 



upholders of the onomatopoeic theory — Renan, Chavee, 
Bargraft, Egger, Whitney, Henry, Hermann Paul 
— have hardly modified the expression of them. In 
their application, though the progress of science has 
removed many causes of error, the difficulties remain 
great, and often insurmountable. It is true that the 
derived languages are no longer confounded with the 
parent language, nor the modern and the past form of 
families of idioms, nor is any one likely to quote a 
Hebrew word in illustration of Teutonic or Latin usage. 
" But if we consider the untold ages which separate us 
from the earliest imil"ative cries," says Sir John Lub- 
bock, " we shall not be astonished that the derivation 
of the root words, thousands of years old, are either 
entirely lost, or at best can no longer be determined 
with certainty." We may add a very just remark of 
M. Michel Brdal : " If at times we think we recognise 
in certain sounds in our idioms an imitation of natural 
sounds, we should remember that the same sounds are 
represented in other languages by quite different utter- 
ances, in which these peoples think they recognise 
onomatopoeia ; so that we ought rather to say that we 
hear the sounds of nature through the words to which 
our ear has been accustomed from infancy.'' It is 
necessary, therefore, to mistrust resemblances which 
are too precise, too detailed, but to recognise at the 
same time that the fact of onomatopoeia, direct or sym- 
bolical, is hardly contestable, since it has left traces 
even in languages which have derived their construc- 
tion from earlier tongues, and since it may be observed 
every day in the fluctuating vocabulary of savages, the 
intellectual contemporaries of our ancestors. Does it 
not occur to every one of us at times to strive to render 
a noise, a memory, an idea, by a sound ? M. Hermann 



Embryology of Language. 2)7 

Paul (in his Frincipes de LinguisHque Historique) re- 
marks that words are created every day, in every lan- 
guage, which are the result of a vague onomatopoeia, 
and according to the rule that we must judge of those 
ages which have left no documents by what we can 
observe in later times, he concludes that since this 
method must have been in use from all time, " we may 
attribute to it the origin and the general development 
of language." 

We do not go so far. The cry is the origin ; ono- 
matopoeia is the second stage, in which language finds 
the materials which the association of ideas and meta- 
phor subsequently elaborate. We cannot do better, to 
close" the debate, than quote the opinion of Whitney, 
" The Life of Language," 4th edition, 1 892, p. 242 : — 
" If we thus accept the impulse to communicate as the 
governing principle of speech-development, and the 
voice as the agent whose action we have especially to 
trace, it will not be difficult to establish other points 
in the earliest history. Whatever offered itself as the 
most feasible means of arriving at mutual understand- 
ing would be soonest turned to account. We have 
regarded the reproduction, with intent to signify some- 
thing, of the natural tones and cries as the positively 
earliest speech ; but this would so immediately and 
certainly come to be combined with imitative or ono- 
matopoetic utterances, that the distinction in time 
between the two is rather theoretical than actual. 
Indeed, the reproduction itself is in a certain way 
onomatopoetic. It imitates, so to speak, the cries of 
the human animal, in order to imitate secondarily what 
those cries in their primary use signified directly. 
Just as soon, at any rate, as an inkling of the value of 
communication was gained, and the process began to 
4 



38 The Evolution of Language. 

be performed, a little more consciously, the range of 
imitation would be extended. This is a direct corol- 
lary to the principles laid down above. Mutual in- 
telligence being aimed at, and audible utterance the 
means employed, audible sounds will be the matter 
most readily represented and conveyed ; just as some- 
thing else would come easiest to one who used a 
different means. To repeat once more the old and 
well-worn but telling illustration : if we had the con- 
ception of a dog to signify, and the insirumentality 
were pictorial, we should draw the outline figure of a 
dog ; if the means were gesture, we should imitate 
some characteristic visible act of the animal — for ex- 
ample, its bite or the wagging of its tail ; if it were 
voice, we should say 'bow-wow.' This is the simple 
explanation of the importance which is and must be 
attributed to the onomatopoetic principle in the early 
stages of language-making.'' " The scope of the imi- 
tative principle," adds Whitney, " is by no means re- 
stricted to the sounds which occur in nature, although 
these are the most obvious and easiest subjects of sig- 
nificative reproduction. What it is may be seen in 
part fi'om the range of onomatopoetic words in known 
languages. There is a figurative use of imitation, 
whereby rapid, slow, abrupt, repetitive, motions are 
capable of being signified by combinations of sounds 
which make something such an impression on the mind 
through the ear as the motions in question do through 
the eye. And we can well conceive that, wiile this was 
the chief efficient suggestion of expression, men's minds 
may have been sharpened to catch and incorporate 
analogies which now escape our notice, because, having 
a plentiful provision of expression from other sources, we 
no longer have our attention keenly directed to them." 



.Embryology of Language. 39 

Admit, for example — and it is precisely what has 
happened — that a given sound a, ou, e, i, a given diph- 
thong, vowel or consonantal, jj, ss, ch, br, fr, tr, ps, pv, 
w, may have seemed to render a sound of wings, a 
modulation of the wind or of water; nothing was easier 
than to derive from it a thousand families of different 
words, corresponding to hundreds of objects, pheno- 
mena, sensations, and ideas ; bird, bree2re, river, stream, 
rain, foliage, tree ; flight, breath, soul, phantom ; rust- 
ling, rolling, shuddering, trembling, shivering, frost, 
winter; fever, flame, heat, vibration, light. Multiply 
these diverse productions by the infinite variety of 
vocal utterance, and from the poverty of the primitive 
language you are thrust upon the difficulty of choice, 
finding (as the result of analysis, of the association of 
ideas, fugitive or endurin;,') twenty names for one 
thing, and a hundred things to which a single name 
might be applied ; thus words undergo constant change 
of quality and character, passing from the representa- 
tion of sonorous objects to that of coloured or odorous 
objects, from movement to form, to mental image or 
concept. For the mind grows clearer and richer in 
proportion to its means of expression. But it is not 
yet able to administer its wealth ; it abandons itself to 
the flood of metaphor, to those summary and super- 
ficial comparisons which observation and experience, yet 
in their infancy, cannot correct or control. Metaphor 
connects the most discordant notions and objects ; it 
sweeps everything into its net, and creates from the 
casual juxtaposition thoughts and images. It blinds 
and it. confounds. It is a fertile source of confusion 
and error, fashioning language for the use of reason, 
and making it a supple instrument, but a misleading 
one from the beginning and for all time. 



40 The Evolution of Language. 

Throngliout this work what becomes of onoma- 
topoeia ? It has almost disappeared, its part is played ; 
man needs it no longer to give meaning to sounds. 
This is why M. Paul Regnaud seeks it in vain in our 
languages which are three or four thousand years old 
at most. The reasoning powers came into play in 
their turn, classifying the confused materials furnished 
by metaphor ; a few dozen significant syllables are 
selected, and these, by agglutination, by inflection, by 
derivation, give rise to a new vocabulary, of which the 
words may be ranged into the categories of grammar 
and syntax. 

We have been much struck with the views of M. 
Regnaud, the learned professor of Lyons, on the origin 
of language. He follows, like myself, the theory of 
evolution and development. We are the more anxious 
to show, if possible, that his criticism of the hazardous 
or impossible relationships suggested by Plato, Leib- 
nitz, De Brosses, Charles Nodier, even by Tylor, Lub- 
bock, Wedgwood, and Farrer, do not affect onomatopoeia 
considered as a persistent factor in the inferior lan- 
guages and as a prehistoric and necessary factor in 
inflected languages. 

'' It is very certain," says M. Paul Regnaud, " that 
we can hardly find the trace of these methods (onoma- 
topoeic) in Sanscrit, Greek, and Latin — that is, in 
those Indo-European languages which, though not 
spoken, have left a considerable literature. If it be 
added, as M. Pick observes, that the farther we re- 
trace towards its origin the vocabulary of the tongues 
in question, the rarer onomatopoeia becomes, we shall 
know what degree of importance we ought to attribute 
to the primitive effect of onomatopoeia. The imitation 
of the sounds of nature in all its forms, direct or sym- 



Embryology of Language. 41 

boHcal, can be regarded only as a recent and sporadic 
factor in language, and the birth and growth of its 
forms are the result, in great part, of another cause." 

It will be observed that this judgment is far from 
absolute, but rather measured and prudent. No philo- 
logist whia has rejected the supernatural denies this 
other cause of the origin and growth of language ; this 
cause is the cry. And, to minimise in passing the ob- 
jection which doubtless appears to M. Regnaud the 
most conclusive, if onomatopceia has left few traces in 
the classical languages, so has the cry ; it is limited 
to the instinctive and reflex exclamation. The reason 
urged against onomatopoeia would therefore hold good 
against the cry, of which we recognise with M. Paul 
Regnaud the ancient importance ; against the cry, of 
which the reproduction by those who heard it was, as 
Whitney remarks, an imitation, an onomatopoeia. 

But are the traces of onomatopceia so rare as M. 
Regnaud thinks in the Indo-European languages ? 
What are then these attributive roots, tu, tchid, stan, 
brh, sJcrp, kvan, dak, smr, srp, &c., to which he himself, 
and with much boldness, attaches numerous families 
of words, of which it is possible to trace the descent 
and relationships through all the dialects of our 
languages, the families frapper (strike), couper (to 
cut), etendre (to stretch), crier (to cry), creuser (to 
hollow out), chanter (to sing), mordre (to bite), glisser 
(to slip), &c. ? 

Let us take the very numerous group represented 
in Sanskrit by the root hrh (h, r, vowel, h, guttural), 
of which the pronunciation varies from hahr to hrah, 
Ireck, h'iich, and even blach, for the liquids r and I 
permute constantly. " This root," says M. Regnaud, 
" signifies generally to cry, to speak, to pray " (as in 



42 The Evolution of Language. 

JraAman, prayer, &ra7iman, priest, he who prays), and 
also " to trumpet like the elephant " (whence the Lat. 
harritus). In Greek it has given several variants: 
^pa-)(^ in the Homeric 'i^payov, to speak, to cry out ; 
jSjO^x in fip^X'^' *° ^o^'" ; ^^'^X ^" /3A>7xao/ia£, to bleat, 
in English to hark. "We can hardly omit, as the 
resemblance is so close, the French words brailler, 
lrair6,hredouiller, to brawl, to bray, to murmur. "If," 
continues our author, " as seems certain, these several 
forms are derived from a single original form, we 
must conclude that, far from distinguishing the cry 
of each animal by a special onomatopcfiia directly in 
relation with this special sound, our Aryan aneest,^is 
employed a generic term common to all, and probably 
without relation to any given cry, and that this was 
used to signify indifferently the voice of man, and for 
the utterances of the elephant, the lion, the sheep, the 
dog, &c.'' We would not assert so much ; we would 
not say " we must," but " we may " conclude from 
these facts that the single early form from ■ which 
these variants derive was either one of those generic 
onomatopceias, very vague and hardly distinguishable 
from the cry of emotion or of astonishment, or an 
onomatopoeia, at first special, chosen among twenty 
others which might have taken the same place and 
generalised for the needs of analogy and derivation. 

But it is time to sum up. Animals possess two of 
the important elements of language — the spontaneous 
reflex cry of emotion or need ; the voluntary cry of 
waruing, threat, or summons. From these two sort's 
of utterance, man, endowed already with a richer 
vocal apparatus and a more developed brain, evolved 
numerous varieties by means of stress, reduplication, 
intonation. The warning or summoning cry, the germ 



Embryology of Language. 43 

of the demonstrative roots, is the parent of the names 
of numbers, sex, and distance ; the emotional cry, of 
which oar simple interjections are but the relics, in 
combination with the demonstratives, pi'epares the 
outlines of the sentence, and already represents the 
verb and the names of states or actions. Imitation, 
direct or symbolical, and necessarily only approxima- 
tive of the sounds of external nature, i.e., onomatopoeia, 
furnished the elements of the attributive roots, from 
which arise the names of objects, special verbs, and 
their derivatives. Analogy and metaphor complete 
the vocabulary, applying to the objects, discerned by 
touch, sight, smell, and taste, qualifying adjectives 
derived from onomatopoeia. Reason then coming into 
play, rejects the greater part of this unmanageable 
wealth, and adopts a certain number of sounds which 
have already been reduced to a vague and generic 
sense ; and by derivation, composition, and affixes, 
the root sounds produce those endless families of 
words, related to each other in every degree of kin- 
dred, from the closest to the most doubtful, which 
grammar finally ranges in the categories known as 
the parts of speech. 



CHAPTER III. 

FORMATION OF WORDS AND THE STRUCTURE 
OF LANGUAGES. 

The expedients of monosyllabism : examples from Chinese — Full roots 
and empty roots — Method of agglutinative languages ; subordina- 
tion of affixed roots, which modify the sense, to a central root 
which remains unaltered — Schlegel's error with regard to the 
nature of case and verbal endings — Examples from Turkish, 
Esquimaux, and Mexican — Inflexion : intimate fusion of the full 
and empty roots ; variation of the radical vowel in the Semitic 
languages ; complete change of the root in the Indo-European 
group— Analysis of the words apercevoir, respectable, rapproche- 
ment, recueillanent — Apposition, suffixation, composition — Par- 
allel advance of the intelligence and of language. 

Having, not solved, bat thrown some light upon the 
problem of the origin of language, we now leave a 
region where induction can only attain to a general 
certainty for those which lie open to direct observa- 
tion. From the genesis of speech we pass to the 
formation and structure of languages. 

The Chinese group has been content to form from 
the raw material, with demonstrative sounds on the 
one hand, and attributive on the other, by merely 
grouping the roots, but without composition, and 
without altering the syllables, more than 40,000 
words, most of them fortunately unnecessary to the 
majority of the inhabitants of the Celestial Empire. 
Fifteen thousand are enough for the average educated 
man. Since the fundamental roots of the Chinese 
tongue amount only to 450, it follows that the same 



Formation of Words. 45 

sound is susceptible of numerous different meanings. 
Thus the form tao means indifferently to tear away, 
to reach, to cover, flag, corn, to lead, road, &c. And 
the form 1%, jewel, dew, to forge, vehicle, to turn 
aside, road. How then discover the sense ? Usually, 
by a method which is a trifle childish but very accurate, 
the Chinese determine the sense by placing two syn- 
onyms in juxtaposition ; the one certifies the other. 
T(M> and In have each one several significations, but 
taX) followed by Zm can only mean " road." The gram- 
matical value of these syllables in the proposition is 
determined by their respective positions. Ta, involving 
the notion of height, will be adverb or adjective before 
a word ; after it, a verb or an abstract noun : ta jin, 
a tall man ; jin ta, the man grows, or the man is tall, 
or the height of the man. In the same way chen will 
mefin by turns virtue, virtuous, to approve, well. 

The subject precedes the verb : Tigb ta ni, I beat 
thee ; -ni tcl ngb, thou beatest me. The relations of 
case, that which we call possessive, accusative, dative, 
&c., are expressed either by the position of the words, 
or more commonly by subordinate roots, pronominal 
or attributive, of which the proper sense is lost or 
obliterated. Y, to use, placed before tchang, y-tchanrj, 
means " with : " with a stick. Zi means interior : uo-li, 
in the house, l^chi (right, possession, the pronoun 
he) : jin-tcM-Jdun, the prince of men. Yu (to give) : 
sse yen yu jin, to give money to a man. Pa and tsiang 
(to seize, to take), i, iu, hou (to employ), often indicate 
the accusative. Fa tchoung jin teou kan, he looked 
furtively at the crowd of men ; pao hou min, to pro- 
tect the people ; i jin tsun sin, he keeps humanity in 
his heart. Thsong, yeou, tseu, hou, show origin, the 
point of departure, the ablative : thsong thien lai, to 



46 The Evolution of Language. 

come from heaven ; U hou tlden, to obtain from heaven. 
Gender is determined, as it should be (we still do it), 
by the term male and female, nan and niu : nan-tse, 
son; niun-tse, daughter ; niu-Jin, .woman. Numerous 
words signifying summit, multitude, totality, may in- 
dicate the plural, though in most cases the number 
must be divined from the context ; for instance, to jin 
.(many men), people ; jin-kiai (man all), men ; i-pei 
(stranger class), foreigners. 

In spite of tendencies towards agglutination and 
grammatical organisation, Chinese, except in certain 
southern sub-dialects, has remained obstinately faithful 
to monosyllabism ; its associations of words do not 
form true compounds, and the neutralised syllables, 
which precede or follow its. substantives, keep their 
form intact, and never become terminations of case, 
number, or gender, but they play the part of these. 
" The Chinese," says M. Hpvelacque, " have clearly 
grasped this fact, since they class their roots in two 
distinct groups — -full words and empiy words. By 
the first they understand those roots of which the 
meaning keeps all its fulness and independence, the 
roots which we in our translations render by nouns 
or verbs ; they call empty the roots of which the true 
value has by degrees become obscured, and which are 
used to determine and define the sense, and to indicate 
the grammatical relations of the full words." "What 
is grammar ? " asks the Chinese teacher of his pupil. 
" A useful art, which enables us to distinguish full 
words from empty words." 

Now, in all languages, agglutinative or inflected, 
the constituent elements of the words are likewise 
full syllables, called root syllables and empty syllables, 
which we call prefixes, affixes, or generally suffixes and 



Formation of Words. 47 

terminations. But these suffixes, altered in form as 
in sense, make a part of the word ; they are joined to 
the central root and amalgamated with each other. 
They do not differ in. kind from the roots to which 
they are attached ; when it is possible to separate 
them from it by analysis, we find them also to be 
roots, attributive or pronominal, quite capable of being 
the centre of a group of suffixes, amd, moreover, of exist- 
ing in a free state. Case-endings alone often escape 
analysis ; and this is easily understood : in their re- 
ciprocal contacts and friction words have become worn 
away at the edges, so to speak. Terminal suffixes, 
gradually obliterated and disfigured, have sometimes 
at last completely disappeared, even in ancient lan- 
guages ; sometimes they are still written, but are 
subject to elision, and- are no longer pronounced; 
sometimes the prolongation, slight or marked, of the 
syllable which preceded them alone reveals their former 
place ; then this syllable which they protected, now 
exposed, wears away and disappears in its turn. The 
word grows shorter, becomes contracted, but that 
which remains retains the accessory meanings which 
the vanished syllables added to the complete form, 
and the grammatical value which they had assigned 
to it in declension and conjugation. Thus the Sanscrit 
word a?anti is represented in Latin by sunt; the 
Latin amaverunt is sometimes altered into amavere or 
amarunt ; the primitive form paters has become the 
Greek iraTrip ; dominum has gradually contracted into 
dominu, domino, domno ; whence the modern dom, don 
(Dom Brial, Don Juan) ; the low Latin word domini- 
ariiom (suzerainty), after having dropped the termina- 
tion um, has gradually become doniffier, our word 
danger, embodying the philosophy of La Fontaine's 



48 The Evolution of Language. 

proverb, " Notre enTiemi c'est rwtre maitre " (our master 
is our enemy). There are innumerable similar cases 
which characterise sufficiently what is called dialectal 
change. They belong to a series which has been 
summed up in the convenient formula : the law of 
least resistance. In science, as we know, laws deter- 
mine nothing; they are the resultant of a certain 
number of observations which confirm each other, and 
allow of classification, and of the prevision of similar 
phenomena. This is the case here. The intelligence, 
as it gained strength, by degrees reduced and rejected 
the means which were at first necessary to guide the 
thought and assure its expression ; it has abandoned 
all useless effort, for this is the sense and value which 
we should attach to the " law " of least resistance. 

Before an almost irresistible argument from analogy 
had revealed the origin of the suflSxes, the effacement 
of the verbal and case endings had misled one of the 
precursors of comparative philology, Frederic Schlegel. 
Schlegel believed that the terminations grew from the 
body of the word through some mysterious evolution, 
as the branches grow from the trunk of a tree, or else 
as elements which had no proper meaning, but were 
employed arbitrarily and conventionally to modify the 
sense of words. This mystical conception of the life 
of language has been ably criticised and set aside by 
Max Miiller. I give the passage : — 

" Certain thinkers have considered language as an 
organic whole, gifted in some sense with life, and they 
have explained its formal elements as being produced 
by an inner natural vegetation. Languages, say they, 
should be compared not to a crystal, formed by agglo- 
meration round a fixed point, but to a germ developed 
by its internal force ; all the essential parts of language 



Formation of Words. 49 

existed in the primitive germ as truly, though only in 
the embryonic state, as the petals of the flower exist 
in the bud before it opens out to the air and sunshine. 
.... The science of language does not adopt these 
hypotheses. As for the one which represents to us a 
group of men discussing together about the manner in 
which it were best to express the relations indicated 
by the nominative, the genitive, singular and plural, 
active and passive, common sense might tell us that, 
if such abstract questions could have been discussed 
in a language destitute of inflexion, there would have 
been no reason for inventing a more perfect method 
of communication. (So thought the Chinese.) With 
regard to the supposition that there could exist in 
language — that is, in the nouns and the verbs — an inner 
principle of growth, all that we can say is that such 
a theory vanishes as soon as we look closely at it. 
Science gathers facts. Instead of regarding inflexions 
in general as conventional signs or natural excres- 
cences, she takes each termination separately, and when, 
by means of comparison, she has determined the earliest 
form of it, she treats this primitive syllable as she 
would treat any other part of language, that is to say, 
as a word which had originally its proper significa- 
tion." 

Two facts appear to us to be certain : First, that at 
a given moment of their existence, long before the dawn 
of history, the thousands of human groups scattered 
upon the surface of the earth found themselves in pos- 
session — we have seen by what probable genesis — of 
two articulate and significant vocal elements, the demon- 
strative or pronominal roots, and the attributive roots, 
substantive or verbal. Secondly, that these two classes 
of roots are the only elements of language ; that there 



50 The Evolution of Language. 

are no others, and that all tongues are the result of 
their different combinations, varying according to the 
vocal and brain power of the distinct races and sub- 
races ; either by simple juxtaposition of unaltered, 
syllables, as ia Chinese, by agglutination of several 
syllables round a central syllable, as is the case in 
all the languages called agglutinative, or, finally, by 
fusing and contracting into a single whole the central 
and the subordinate syllables, as in all the inflected 
languages. 

We have quoted some examples of th.e Chinese 
method ; we will now analyse a few forms borrowed 
from the thousands of idioms which belong to the 
agglutinative class. To love, in the most general sense 
of the word, is, in Turkish, sev ; the subordinate root 
er forming adjectives or participles, sev-er will signify 
loving; join to it the pronoun sen, thou, or siz, you, 
sever sen, sever sis, loving thou, loving yon, i.e., thou 
lovest, you love ; other syllables, of which the original 
meaning is lost, gu, i, make sevgu, sevi, love ; di, placed 
between the theme sever and the personal termination 
abridged or altered, forms the imperfect : sever-di-n, 
sever-di-niz, thou lovedst, you loved. MeJc, the sign 
of the infinitive, gives us 'sev-mek. This is not all ; 
between the two parts of the word we can insert the 
ideas of reciprocity, of causality, of passivity, of nega- 
tion, and so sum up a whole phrase in a single word. 
This is a vice which has overtaken many idioms of 
this class ; incorporation and polysynthetism, under 
pretext of seizing fine shades and the succession of 
ideas, produce words which are difficult to handle, and 
far more difficult to interpret. The verb sev-mek may 
present itself under thirty-six forms, such as sev-in-il- 
rnekj to be glad, sev-isJi-dir-il-meh, to be drawn to love 



Formation of Words. 5 1 

one another ; sev-ish-dir-il-Ke-me-mek, to be unable to 
be drawn to love one another. We are reminded of 
the Turk of the Bourgeois Gentilhomme, the Turk who 
said so much in a few words. 

Many of the above-mentioned forms are rarely used, 
and Turkish, in reality a correct and beautiful lan- 
guage, does not abuse its resources. The drawbacks 
to unlimited agglutination are especially marked in the 
Basque tongue, which incorporates with its verb, not 
only the possessive pronouns (as do also the Semitic 
languages), but even the indirect object ; and in the 
American Indian dialects, where words already com- 
pletely formed are capriciously deprived of their 
beginning or end, and, so amputated, swell with their 
unrecognisable fragments some interminable compound. 

The Greenland aulisar-iartor-asuarpok (he hastened 
to go and fish) includes aulisar, to fish, peartor, to do 
something, and pinnesiiarpoJc, he made haste. The 
Mexican no-tlazo-mahuiz-teopixcatdtzin (0 my father, 
divine and revered protectoi-) contains no, my, tlazontli, 
esteemed, mahuiztic, revered, teotl, goA, pixqui, protector, 
tatzi, father. There is an extreme variety in the 
agglutinative class of languages. There are simple 
languages, like the Japanese, like all those which 
belong to the great Malayo-Polynesian family, or again 
Finnish and Magyar ; 'others are strangely complicated, 
such as Basque and the American idioms ; some are 
extremely poor and barren, like the dialects of the 
Guinea Coast or of the Bosjesmans ; others are rich 
and regular, such as Turkish and Suomi (the language 
of the Kalevala) ; some prefer to suffix the added 
words ; others, such as Kaffir and the whole Bantu 
group, prefix them ; in some gender is wanting, in 
others number. Some are so variable that their whole 



5 2 The Evolution of Language. 

vocabulary and physiognomy changes in fifty years. 
These languages, which form the great majority of 
known idioms, proceed from a great number of inde- 
pendent sources and have but one common character — 
the subordination' of one or more roots susceptible of 
alteration in form, and deprived of their proper sense, 
to a full unaltered root which conveys the principal 
or fundamental idea of the word. 

Only two systems, two families of languages, the 
Semitic and the Indo-European, which are very rich 
and very varied, but which at least can each be reduced 
to organic unity, to one vocabulary and one grammar, 
these two groups only have passed the agglutinative 
stage. To agglutination, which they possess, and the 
methods of which they use (including incoi-poration 
and polysynthetism), they have added inflexion. 

What is inflexion ? Not much, at the first glance : 
the possible alteration of all the elements of the root 
syllable, as well as of the afiixed syllables. The root 
is not necessarily modified ; it sometimes remains un- 
changed, as in the agglutinative stage ; but it can be 
modified. This small privilege allows inflected lan- 
guages to express the relation of one word to another, 
not only by the addition of suffixes and prefixes, but 
also by means of numerous variations in the elements 
of the root itself. Hence, an extraordinary richness 
and an extreme clearness in derivation, a peculiar 
delicacy in the gTammatical notation of case, gender, 
number, person, tense, and mood (the paradigm of a 
Greek verb may include i 300 forms), and at the same 
time a great simplification in the word ; and finally, 
perspicuity and order in the sentence, insuring the 
logical connection of ideas, and due proportion in the 
expression of the thought. 



Formation of Words. 53 

Before analysing a few familiar words (for examples 
carry more weight than the most authoritative asser- 
tion), I must insist upon one point of capital im- 
portance : the irreducibility of the two families of the 
inflected class to a common origin. " Not only," says 
M. Hovelacque, " are the roots totally distinct in the 
Semitic and Indo-European languages ; they differ yet 
more in their structure. Inflexion is not the same in 
the one as the other." 

In the Indo-European idioms, viz., Sanscrit, Iranian, 
Greek, Latin, Celtic, German, Slav, Lithuanian, and their 
very numerous ancient and modern dialects, inflexion 
affects the consonants as well as the vowels. In the 
Semitic group, Assyrian, Pbofenidan, Hebrew, Chaldean, 
Syriac, Arabic, Himyarite, and Ghez of Abyssinia, the 
root consonants are unchangeable. 

Indo-European roots may be composed of a single 
vowel, long- or short, nasalised or coupled with another 
to form a diphthong, or of a vowel and one, two, or 
more consonants, and vice versd, provided they are 
pronounced by a single emission of the voice. The 
Semitic roots are composed of three consonants (at 
least analysis has not yet reduced them further). 

The characteristic of Semitism is the triliteral form 
of its roots : they are composed of three consonants; 
to which different vowels are joined as formative 
elements, indicating the various relations of the root. 
In Arabic, for instance, Mb includes the sense of writ- 
ing, dbr of speaking, Ml oi killing; gatl is murderer, 
gitl, enemy ; katala means he kills, hvMla, he was 
killed, &c. Besides this inflexion due to the use of 
different vowels, Semitism also forms words by using 
suffixes and prefixes, sometimes also inserted particles. 
But the heaping of suflBx on suffix, the formation of 
6 



54 The Evolution of Language. 

derivatives from derived words is unknown. Hence 
the close resemblance of all Semitic languages, which 
are all as nearly allied as Italian and Spanish. The 
Semitic noun can only have three cases, and these are 
wanting to most of the languages of the class. The 
Semitic verb, in the second and third person, dis- 
tinguishes the gender of the subject: gatala, he killed; 
gatalat, she killed. The antithesis of present, past, 
and future, which is essential, fundamental, in the Indo- 
European languages, does not exist in the Semitic ; it 
has only two tenses, answering the one to the idea of 
the accomplished action, the other to that of the in- 
complete action. 

These few characteristics will suffice to dispose of 
the temptation to assimilate Hebrew with Latin or 
Greek. These languages have not issued from the 
same earthly Paradise ; they have borrowed words 
from each other within the historic period, but they 
have not created a single one in common. 

We come now to the examples ; as we analyse them 
we shall learn the methods of the Indo-Europeans. I 
take, almost at hazard, a verb, an adjective, and a 
couple of nouns. 

Apercevoir. In Latin the corresponding form would 
be ad-per-cip-ere. Where is the central root ? and 
what changes has it suffered ? The root is dp, ren- 
dered in the French form by cev. The Latin labials 
are commonly softened into v in French : habere, avoir ; 
sapere, savoir ; rapere, ravir. The first form of the 
root is cap, with the sense of to seize (here to take 
with the eyes) ; in composition it becomes cip : accipit, 
incipit ; cep : inceptwtn, accepi ; cwp : aucipiwm, the art 
of snaring birds, beside auceps, bird-catcher. As for the 
c, it was hard in classical Latin ; the pronunciation 



Formation of Words. 55 

was hap, kip, hep; the Eomance tongues have con- 
verted it into a sibilant : c, soft ; and even into a 
palatal, ch6tif (Ital. cattivo, Lat. cap-ti-v-us). Note 
the relationship of regu (receipt) and recetie (recipe). 

The idea of seizing may easily be extended to the 
hearing, the sight, and the mind, capis ne, capin' ? do 
you grasp, do you understand ? Compare the Ital. 
capisco, I seize, I understand. But this idea of seizing 
may be reinforced, and it is the case here, by suffixes 
indicating direction and movement. A, Lat. ad, which 
may be traced in the old spelling appercevoir, and in 
appeler, appartenir (where the first p shows the effect 
of the labial on the dental), implies movement towards. 
Per is one of those particles of a very indefinite mean- 
ing (in Gr. irapa, beside, near, against; in Latin, in 
composition, some, about ; paulisper, parumper, a few) ; 
but in Latin, as in French, the most usual sense is by, 
through, across, by means of; thus we find, for the 
first three elements of the word we are analysing : to 
seize^through — as far as, that is, to seize from a 
distance, from afar. There remains the termination, 
which is very obscure. We note, in the first place, 
that oir is very often the French form for the Lat. ere, 
not only long, as in habere, avoir, apparere, apparoir, 
but short, as in capere, sapere, recevoir, percevoir, savoir. 
Nothing is more frequent than the substitution of the 
diphthong oi for the Lat. e, as also for and i (moisson, 
messi-o ; moi, toi, soi, me, te, se ; mois, mensis ; loi, 
lex ; roi, rex ; poison, potio ; poisson, piscio, &c.). The 
science of language is thus full of minor problems of 
which the solution is impossible, inasmuch as it would 
have been necessary to study in the throat of a living 
Gallo-Roman the point of contact, of meeting between 
the two sound movements which result in oi and e. 



56 The Evolution of Language. 

We must be content with the statement of the fact, 
which is certain. The Latin infinitive termination 
is almost equally embarrassing. We know that the 
short e, represents the i of leg-i-mus, cap-i-mus, common 
to all the verbs of the third conjugation, and that this 
i or e, which takes- the place of an ancient short a 
(Sana, hhar-a-ti, he carries), is joined to a number of 
roots, as a copulative letter, to receive the case or 
verbal ending, or a new suflSx. But it is by no 
means always present. Does it belong to an ancient 
state of language, in which no consonant could do 
without a supporting vowel, a stage which persists in 
many African idioms, and in almost all. the Malay 
languages ? Or is it the simplest addition which the 
need of some sign to indicate the noun or the adjec- 
tive could suggest : root, hhar ; hhar-a, bearer or bear- 
ing, he who bears ? A would thus be the readiest, 
the most instinctive of pronouns. And although the 
naked root retained the power to annex directly suf- 
fixes and case-endings (Lat. fer-s, fer-i, legs, lec-tus, 
reg-s, nec-s, &c.), the copulative vowel persisted be- 
tween roots terminated by consonants and suffixes 
beginning with consonants, preserving both from harsh 
contacts and from the more; diflacult assimilations. 
The final syllable re remains to be copsidered; it 
marks the infinitive in Latin and recurs everywhere in 
French : aimer, ravir, lire, fondre, risoudre, avoir. It 
has been compared to the past infinitive meminis-se, 
cessis-se, h,abuis-se ; and as the Lat. r between two 
vowels takes the place of a primitive s (floris, Jwnorii, 
generis, &c., fov flosis, honosis, genesis), it has been sup- 
posed that the two forms were originally one, se, a 
species of indefinite neuter person, joined to an inde- 
■ clinable verbal noun. So that we may translate the 



Formation of Words. 57 

word wpercewir, "the act of grasping through as far 
as," and this is the sense of the five amalgamated- 
elements. 

The adjectiV'e respectahle, re-spec-ta-b-le, is worthy of 
attention, because it includes, in the first place, one of 
the most fertile of our attributive roots, and secondly, 
sufiixes which are used throughout the Latin family. 

Spec, spic, to look, to see, which may be recognised 
in the Sanscrit of the Vedas, spac, guardian, in the 
Teut. speh-on, to see, to spy, speh-a, Eng. spy, Fr. 
espion, took in Greek the forms o-ktctt and o-kott. In 
our first chapter we pointed out this confusion between 
the labials and the strong gutturals. 

This ancient syllable shows or hides itself in suspic- 
io, Fr. soupgon ; haruspex, augiir, he who looks at and 
consults the lightning or the entrails of the victims ; 
in au-(avi)-spicium, the observation of the flight of 
birds ; in speculum, mirror (whence speculari, specula- 
tion, &c.), Ital. specchio, Ger. Spiegel ; and in the Fr. 
espi&gle, a corruption of the Ger. Vhlen Spiegel (mirror 
of owls), the fictitious hero of a collection of drolleries. 
The Greek form of the root gave e7r/cr/co7roy, overseer 
or bishop, Fr. Mque, Ital. vescovo, Span, obispo. 

Species, that which is looked at, beauty, form, charac- 
teristics, hence special, specific, specify, has given specie, 
coined money with a device, espice (in English the 
Latin word is used with the restricted sense) and ipices 
(speciar substances; Eng. spices, Ger. Spezerien), to the 
Italian words speziale and spezeria, chemist and drug 
shop, to the Yrenchipicier, dpicerie, words which do not 
look as if they were allied with hishop, but which are 
so undeniably. Ad, circum,pro, per, su, in, de, re-spic-ere, 
form as many verbs and derivatives, corresponding to 
different ways of looking. 



58 The Evolution of Language. 

The demonstrative sufSx t, followed by a vowel, 
which exists in an independent state in Sans, ta, tad, 
in Gr. to, ra, in Lat. tani, turn,, forms, by joining 
itself directly to consonantal roots, a number of deriva- 
tives, nouns, participles, and adjectives (often taken 
unchanged into French and English, as in respect, sus- 
pect, aspect, inspector, prospectus, perspective). Latin, 
adding to this new theme the termination of its first 
conjugation, has formed a whole series of verbs called 
frequentatives, aspectare, respectare, whence a new theme 
specta, the source of other adjectives and nouns, e.g., 
spectator, spectatus, spectabilis. 

We leave the s, a case- ending which adds nothing 
to the sense. 

The double suffix bili, out bh, which has formed in 
Latin, and which still forms innumerable adjectives in 
French and in Italian, should be reduced to li, a 
demonstrative which may be recognised in il-le, and in 
the diminutives in l-us, annul-us, iellus, hel, nouvel, &c. 
Bi is due to the analogy of forms like habi-lis, where it 
makes a part of the root. 

I think that there is now nothing obscure left in 
the word respectable. Respect is the look thrown back- 
wards, twice repeated, on some remarkable person or 
thing. The old form, respite, was the time necessary 
to consider, to examine anew the case of an accused 
person ; despite or spite (Fr. dipit), despectus, the look 
from above on some disagreeable object. 

We shall consider more particularly the sense of 
the particle re in another example, a substantive, 
recueillement, where it represents, however, the same 
idea of a return upon the object, or upon oneself. 

But we must first discover the central root, of which 
but a single letter subsists in our example : the letter 



Formation of Words. 59 

I. In its complete form, if such existed, recueillement 
would correspond to re-cum-leg-i-me-n-t-um. The root 
leg, lig, in Gr. Xey, Xo-y, is no less important for us 
than the roots cap and spec, since the derivatives from 
it include such words as ^lire (elect), Mite, election, 
selection, college, collection, perhaps Her (to bind, to tie), 
lien, odligation, certq.inly loi (law), religion, lire (to 
read), le<^on, lecteur ; legere, Xoyos ; and those valuable 
suffixes logg, logist, which we have borrowed directly 
from the Greek ; in brief, that whole world of ideas 
which range from choosing to repeating, reading, from 
speaking to reasoning and thinking. 

The origin is humble. Leg (which perhaps might 
be yet further analysed) had, and always has, the 
sense of taking, choosing (with the prefix e, ex, or dis, 
choosing among, selecting from a crowd, e-ligere, dili- 
gere, dilection, love). 

Gum (preposition, conjunction, and adverb) is the 
declined form of a demonstrative and relative root, ka, 
pa, ta, kv (in Gr. ttw?, kw?, ti, tis), which has given 
to Umbr. po-ei, and to Lat. qui, gum, quod, and quum. 
Since it connects propositions and ideas, it easily 
acquired the sense of " with." Here the final m has 
changed to I, attracted by the first letter of leg ; colleg 
conveys therefore the idea of taking with something, 
of assembling ; add the copulative letter i, e, of which 
we have traced the history, and the termination re ; 
you hfive the Lat. colligerc, the Fr. ctieillir, whence by 
apocope cueil, our word recueil (collection, anthology). 
L alone represents the root, but it has kept the entire 
meaning. 

In the liquid I (I mouilUe) I am tempted to recog- 
nise an intimate fusion of that letter with the guttural 
and the copulative i; so that this sound, which is 



6o The Evolution of Language. 

peculiar to the Eomance languages, and wtich should 
not be pronounced like y (a Parisian tendency), would 
represent in its entirety the theme legi. 

The suffixes which terminate the substantive are 
three in number, ma, tm, ta (we omit the case-ending 
m). These are three demonstratives which can be used 
alone: uUi-mu-s, do-nu-m, dic-tu-s, or by two at a 
time, mana-s, Gr. /xei/o?, docu-men, but already united 
in a suffix, 'mant, ment, which forms participles and 
nouns. Perhaps ma, the attributive root of fievo?, 
mens, thought, of which the ablative mente is foiind in 
most of our adverbs, is identical with this demonstra- 
tive ma. We shall study this question later. How- 
ever this may be, there is no longer any mystery in 
the word cueillement. 

Lastly re, which means return, repetition, insistence, 
permits us to translate recueillement by " reflection on 
that which has been gathered, collected by the mind." 
But a slight peculiarity allows us to take yet another 
step towards the original sense of re^ and that is the t 
which in retro unites it to the comparative suffix ro : 
retro, farther behind, deeper. The Latin ablative ended 
in d, which may be seen in inscriptions, though it had 
been dropped in classical Latin. This d, strengthened 
and preserved by the proximity of r, weakened into I 
in relligio, is thus the remains of a case-ending. He 
would be declined, red would be a case of res, thing, 
one of those words which cannot be decomposed by 
analysis, which are at once general and positive, which 
M. Breal is inclined to consider as one of the oldest 
terms of the Indo-European languages. This syllable, 
so often contracted, and reduced to the letter r, which 
we use and abuse, would thus be an ancient witness of 
that early time when man, incapable of distinguishing 



Formation of Words. 6 1 

objects by their names, termed them by one and the 
same vocal gesture, the object, the thing. Hence would 
come that force, not yet exhausted, that implicit power 
of a return to the reality which underlies all appear- 
ances, dangerous only when it produces reaction, the 
obstinate return upon things acquired for good or evil. 

One more example, which will not detain us long; 
we are already acquainted with almost all its elements. 
It is chosen to show one or two pronominal roots play- 
ing the same part as verbal or substantive roots! It 
is the word rapprochement, which may be analysed as 
follows : re-ad-pro-pe-timo-mentum. Proptimo, altered 
into proximo, is the superlative of the adverbial preposi- 
tion prope; it has been contracted in French va.\o proelie 
(whence rapprochement'). Prope, composed, it would 
seem, of pro and per, whence properare, to hasten for- 
ward, would mean " by there forward." Pro is, like 
p7'ce (prceire) and pri {primus), a case of the root pra, 
which we find in our words premier, prince, profound, 
and which implies priority, jiiroximity, advance, progress. 

These object-lessons have their philosophy. They 
show among the races which are assuredly the best 
endowed and the most capable of indefinite progress 
the parallel development of thought and language, the 
combinations of sounds responding to the association 
of ideas, the vague sense of the primitive roots being 
gradually rendered more precise by the use of meta- 
phor, which also furnished the greater variety of 
expression required by the notions and concepts of 
experience and reason. 

We have just been, as it were, spectators of some of 
the most ordinary episodes of the life of language : the 
passage of the significant sound into the word, of the 
proper sense into the figurative, the formation of the. 



62 The Evohttion of Language. 

verb, the noun, the adjective, by means of suffixes, and 
by the unlimited power of derivation. I should mention 
also two other methods, secondary in the sense that they 
really belong to the agglutinative and monosyllabic 
stages of language : these are apposition {towel-horse, 
hoot-jack, esside-main, tire-botte, ronge-lard, so largely 
used in French and English, and composition (ve(f)e\ij- 
yepera), so common in Sanscrit, Greek, and German. 
The latter method consists in treating as suffixes two 
or more words stripped of their terminations and 
made into a single whole. The polysynthetism of the 
American Indian dialects is the same thing at bottom. 
Apposition closely resembles the Chinese method of 
construction, except that it operates on ready-made 
words already shaped by grammar, instead of juxtapos- 
ing monosyllables. Apposition, composition, suffixation, 
these three methods have ruled by turns in one or 
other of the phases of language ; but the second does 
not exclude the first, the third does not eliminate the 
second, and knows how to use the first. Vocabularies 
are maintained by heredity, difierentiated by selection 
and adaptation — that is to say, by phonetic change and 
dialectic variation. Language begins by a vague pro- 
position, without apparent cohesion, continues by syntax 
(the order of words), attains to grammar by the use of 
inflexions, and when construction and wear and tear 
have altered the word and destroyed the verbal and 
case-endings, tends to return to the purely syntactic 
order, and even to the rudimentary proposition, to the 
telegraphic style which is the stenography of thought. 
But a wide interval separates the starting-point and 
the goal. To traverse this interval language has in- 
vented all the combinations, all the copulatives which 
aid thought, all the artifices of declension and conjuga- 



Formation of Words. 63 

tion, which it discards as the growing intelligence no 
longer needs them. Man casts aside his worn-out 
tools, but he keeps all that he has won by means of 
them. 

This evolution, or rather these unequal and special 
evolutions, of innumerable tongues, have been accom- 
plished in virtue of very diverse cerebral and vocal 
aptitudes, under the influence of the thousand circum- 
stances, natural and historical, which determine the 
progress of societies, apart from the human will. Hence 
some people have been led to consider speech and lan- 
guage as organisms which grow, improve, flourish, or 
degenerate in virtue of their proper qualities or their 
proper vices. It is a fascinating comparison ; but we 
must not forget that languages are also and before 
everything the product of human faculties. The share 
of intelligence and reason cannot be gainsaid. Col- 
lective intelligence, it will be urged, and impersonal 
reason. It may be so ; but collective reason is nothing 
but the aggregate reason of individuals. The life of 
languages is an unconscious life ; but from the animal 
or specific cry, there has been no modification of sound 
and of the corresponding sense which was not initiated 
by an utterance of some individual, accepted, imitated, 
and understood by two or three others, and afterwards 
by hundreds and thousands of others. There can, I 
think, be no doubt of this fact, though it has not been, 
cannot be, proved. The science of language is there- 
fore not only a natural science, but, more particularly, 
an anthropological and ethnographical science. 



PART IX 

GEOaRAPHICAL DISTRIBUTION OF LANGUAGES 
AND RACES. 



CHAPTER I. 
THE SPREAD OF INFLECTED LANGUAGES. 

Chronology and philology — The coincidences of geography and history 
with the evolution of language — Diffusion of the inflected languages, 
and especially of the Indo-European family — Retreat of the agglu- 
tinative to the borders of the civilised world — The monosyllabic 
group of the extreme East — Chinese and its written character — 
Annamite — Siamese — Burmese — Tibetan — Identity of method — - 
Difference of vocabulary. 

In the study which we are about to undertake we shall 
naturally follow the order of development — monosyl- 
labism, agglutination, inflexion, analysis. But it is 
necessai'y to point out that this division cannot be 
founded upon chronology alone. A true succession 
of the four stages of language is most undoubtedly 
proved by the analysis of grammars, vocabularies, and 
of vocal elements. But various circumstances, the 
unequal development of nations, the precocity of 
some, the tardiness of others, migrations, conquests, 
have thrown great disorder into the distribution of 
languages and into their history. Some have become 
extinct in their place of origin, and their fragments 
are buried beneath the deposits of succeeding ages ; 

others, borne to a distance, are scattered in isolated 

64 



. The spread of Inpected Languages 65 

Bpots or spread in great streams like the moraines of 
the glacial epoch over the face of the earth; Some 
vegetate in an eternal childhood, and will die without 
having grown up ; others, having made the whole cycle, 
live again in a numerous posterity. If we consult the 
oldest documents, written or preserved by oral ti-adi-- 
tion, we shall not be surprised that they do not gene- 
rally belong to the most ancient forms of language ; 
and on the other hand, a few important examples will 
force us to admit that inferiority in the means of ex- 
pression is not incompatible with true intellectual and 
literary culture. 

We will disregard for the moment those numerous 
idioms whose existence was unsuspected before the 
discovery of America and Oceania, and confine ourselves 
to the annals of the Old World, as far, at least, as 
science has been able to reconstruct them, and we shall 
see how little help they are able to afford towards a 
methodical enumeration. By far the oldest documents 
which we possess to-day are those of Egypt ; they are 
forty centui-ies old ; they witness to a long use of lan- 
guage prior to themselves, since the earliest of them 
reveal a state of transition between agglutination and 
inflexion. The hieroglyphs of , a King^ Snefrou have 
been recently deciphered on Mount Sinai, at the en- 
trance of a turquoise mine, in which the king boasts 
of his victory over the Bedouins of the mountains. 
Now this king is anterior to the great pyramids ; he 
belonged to the third Memphian .dynasty, which dates 
from four thousand two or three hundred years before 
our era. M. Bdnedite, the fortunate interpreter of 
this text, believes it to be the earliest line of writing 
which has come down to us. The most ancient cunei- 
form inscriptions cannot pretend to so great an age.; 



66 Distribution of Languages and Races. 

they are later by a thousand years, but they testify 
also to a previous long elaboration of language. It is 
known that the inscriptions on the back of statues, on 
cylindrical or conical seals, and on innumerable bricks 
dug up from the sands of Chaldea and Syria, may be 
referred to two systems, to two organisms, often given 
together in bilingual texts. The tongue of the Acca- 
dians or Sumers, the ancient inhabitants and the first 
civil isers of the valleys of the Tigris and Euphrates, 
is of the agglutinative order ; the other, Chaldean or 
Assyrian, is inflected; it is the earliest known form 
of the Semitic tongues. Now the Semitic conquerors 
seem to have owed everything, arts, beliefs, ideas, to 
their industrious subjects, their only superiority con- 
sisting in the possession of more warlike qualities and 
of a more advanced type of language. But the fierce 
and boastful proclamations of the exterminating kings, 
which properly belong to the Semitic race, have less 
interest for us than the magical divagations, the frag- 
ments of cosmogonies, even of epics, which, translated 
from Accadian into Chaldean, furnished to a certain 
degree the somewhat barren and narrow mind of the 
Semites, and left their traces in the religions of Syria, 
Phoenicia, and Judea. The Indo - European parent 
language must be attributed to the same epoch, if not 
the written documents, at least the development of the 
spoken language ; it had already left the agglutinative 
stage and was fully inflected belbre its separation into 
the different dialects. For if the Vedas, as they have 
come down to us, are relatively modern and adapted 
to the Brahminical liturgy, the idiom remains more 
archaic than the oldest Sanscrit, which was already 
extinct in the days of Alexander, older therefore than 
the Greek and Latin idioms. It is certain that the 



The Spread of Inflected Languages. 67 

parent Indo-European tongue, the common type to 
which seven families of languages are more or less 
faithful, was constituted with its grammar, its basis of 
words and ideas, more than twenty centuries before 
the Christian era, and that the tribes which spoke 
this language had traversed three stages, and mas- 
tered the most delicate shades of inflexion, while their 
nearest neighbours had stopped short, some at Semi- 
tic inflexion, some at agglutination, some at monosyl- 
labism. 

The official history of China does not go farther 
back. The Ghouhing, which purports to date from 
2356 years before Christ, was edited, with the other 
sacred books, by Confucius, towards the end of the 
sixth century. Thus historical data, except in what 
concerns the tongues of the north-east of Africa, can 
only establish the synchronism, in times of remote 
antiquity, of those great phases which the philologist 
considers as the successive stages of evolution. 

Yet the general course of civilisation may furnish 
those indications which historical lore refuses to us. 
It is at least curious to note that the part played in 
history by the different classes of language assigns 
to them precisely that rank which the science of lan- 
guage attributes to them. Suppose you have before 
your eyes a map of the globe, three facts strike you 
on the most cursory glance : the central situation and 
the growing expansion of the inflected languages ; the 
isolation and the immobility of the monosyllabic group, 
confined to its vast empire, between the mountains of 
Thibet, the Mongolian desert, the steppes of Man- 
churia, and the seas of China and Indo-China ; lastly, 
the retreat of the agglutinative tongues towards the 
confines of the world ; they are driven to the borders of 



68 Distribution of Languages and Races. 

civilisation, into the frozen regions of Siberia, into the 
heart of Africa, into the Malay Archipelago and the 
islands of the Pacific, into those parts of America 
where the remnant of the indigenous races is yet to 
be found. It is easy to see that their area diminishes 
from day to day ; once they occupied the whole of 
India, only a fifth remains to them, a part of the Dekr- 
kan ; once they covered the whole of Asia from the 
-Arctic Ocean to the Gul£ of Ormuz ; the Semites and 
the Iranians drove them thence ; perhaps tkey for- 
merly dominated in Western Europe, if it be true that 
Basque was the speech of the ancient inhabitants of 
Gaul — if it be true that the race whose remains were 
■found at the Madeleine and at the Eyzies, retreating 
with the reindeer towards the frozen North, is repre- 
sented in modern days by the Esquimaux. In America 
they give way before the Teutonic and Neo - Latin 
tongues ; in another century they will have disappeared 
from Oceania. 

With Attila, Zengliis Khan, and Timour these 
idioms made a vigorous attempt to recover the ground 
they had lost ; but they failed sooner or later, conquered 
by the Semitic, by the Indo-European tongues, even by 
Chinese monosyllabism. The ancient Bulgarians of 
Belisarius were exterminated by the Greeks and Slavs. 
iTwo exceptions seem only to accentuate this universal 
movement of retreat. In the tenth century, Magyar, 
an Uralo-Altaic dialect, succeeded in taking root 
between the Danube and the Theiss ; but it remains 
there without spreading, and though the Hungarians 
have preserved their national language, and applied it 
,t0 poetry and history, yet they only use it to express 
ideas acquired in their perpetual contact with European 
-peoples and idioms. Towards the same epoch, the 



The spread of Inflected Languages. 69 

fanatical outburst of Islam, urging forward the robber 
hordes of Turkestan, let loose upon Persia, Asia Minor, 
Syria, Egypt, and Greece the fierce invasion of the 
Turks. 

We know into what confusion this fatal conquest 
threw the political life of the East ; what disasters 
have ruined and stained with blood the richest countries 
of the world, the basin of the Mediterranean ; at the 
price of what defeats, of what efforts, Europe has con- 
trolled and kept within bounds this blind force ; with 
what infinite trouble European thought, arts, and in- 
dustries re-enter the regions which were their cradle. 
By slow degrees the health of the West has accustomed 
itself to the presence of this foreign body, now almost 
inoffensive. But the race, though no longer unmixed 
with the blood of other peoples, and in spite of its 
native virtues, and the language, in spite of its accuracy 
and harmony of sound, will always remain anomalies 
in the midst, of the civilised world. And the reason 
is that there is an anachronism, an original incompati- 
bility between the mental state of a former age and a 
more advanced intellectual condition — between a lan- 
guage which has stopped short at the agglutinative 
stage and languages which have arrived at the extreme 
limit of the inflected phase. 

The destiny of the inflected languages has been 
very different. The sub-group, which is represented 
to-day by Coptic and the Berber tongues, has not, it is 
true, progressed since the time of the Pharaohs ; it has 
remained suspended between agglutination and in- 
flexion ; it has lacked room ; wedged in between the 
dull mass of the Negroes of Central and Eastern Africa, 
and the adventurous boldness of the Arabs and of 
the Mediterranean peoples, it vegetates and dwindles. 
6 



70 Distribution of Languages and Races. 

But the Semites, whose conquering force is not yet 
spent, at any rate in Africa, have a long and important 
history. Their frontiers have receded on the North ; 
but they retain Syria, Arabia, Egypt and all the 
African coast. The nations which have spoken Semitic 
languages, Chaldeans, Assyrians, Phoenicians, Cartha- 
giaians, Hebrews, Syrians and Arabs, have all con- 
tributed something, some of them very much, to 
civilisation ; they created arts, religions, sciences — arts 
which have been superseded, religions which are dan- 
gerous, sciences which are false or incomplete, but which 
in their day have moved the world. Their influence has 
spread not only over two hundred millions of Moslems 
and Jews ; it. may be recognised in the temperament and 
the accent of the Spaniard and in the intellectual mould 
of Christendom. Even the obstacles, the material and 
moral obstacles, which Semitic influence has raised to 
the development of the Indo-Europeans, are not the 
least of the proofs of its strength. Finally, the lan- 
guages which belong to it have contributed before and 
with the Indo- European idioms to the expulsion of 
the agglutinative group ; they vanquished Accadian in 
Chaldea and Assyria ; they have made inroads into 
Turkish, and are attacking the dialects of Central 
Africa. 

The fortunes of the European idioms and of the 
peoples which created or adopted them are yet more 
significant. They have not ceased to flourish from the 
day when five or six migratory columns left the neigh- 
bourhood of the Caspian to accomplish the education 
of Europe. These bands, which increased on the road, 
spread out like the sticks of a great fan, established 
their different languages from the Gulf of Finland 
to the most outlying rocks of the Greek Archipelago. 



The Spread of Inflected Languages. 7 1 

The most fortunate, those which were the first to reach 
the temperate shores of the Mediterranean, became the 
ancestors of the Gr^co-Latin peoples and civilisation ; 
the Hellenic branch, finding a soil already cultivated, 
and precursors already somewhat polished by the 
lessons, often indirect and sometimes intermitted, which 
they had received from Egypt and Assyria, developed 
earliest with a rapid and magnificent expansion ; and 
never, it would seem, has a more supple, more rich, 
and more beautiful language been associated with a 
national genius so keen, so artistic, or so profound. 
The Latins, slower, more tenacious, hampered moreover 
by the strange mixture of races which they found 
wedged together in the narrow Italian peninsula, the 
Latins, partly by their own energies, partly aided by 
the gradual infiltrations of Hellenic culture, accom- 
plished the same work of conquest and progress ; then 
they embraced within their empire the Greeks, their 
brothers and their teachers, and the Celts and Gauls, 
also their near kinsmen, who, scattered over the forests 
of the West, had failed to turn to account, as yet, 
their courage, their intelligence, and their gift of elo- 
quence. Thus the most precocious of our Indo-Euro- 
pean ancestors had divided between them the East and 
West, when the distant pressure of the Huns and the 
Mongols, driving the belated Slavs upon the yet half- 
savage Germans, forced the frontiers of the Rhine and 
the Danube. When this terrible invasion was stayed, 
when the strong arm of Charlemagne had erected a 
barrier against that part of the horde which had not 
felt in some degree the influence of the Roman civili- 
sation, it could be seen that a large half of the new- 
comers had become assimilated to their predecessors, 
had become incorporated into the ancient Roman 



72 Distribution of Languages and Races. 

world, which had renewed its youth by the infusion of 
the new blood of the barbarians. Those who remained 
outside the Latinised West, but who, endowed with 
language and intelligence from the same common 
stock, were destined to develop in the same manner, 
the Teutons and Slavs, claimed and acquired in their 
turn their legitimate share in the direction of modern 
thought. It is to a Teutonic race, to a Teutonic tongue, 
the Anglo-Saxon race and the English tongue, that the 
honour is due of completing the work begun and con- 
tinued by the Neo-Latins, the Portuguese, the Spaniards, 
and the French. The extraordinary diffusion of the Eng- 
lish race and speech is doubtless due to very numerous 
causes ; but its coincidence with a phenomenon of the 
evolution of language is curious. English, as ancient 
in origin as Greek, Latin, or Gothic, bnt considerably 
modified in the eleventh century by an invasion of 
Trench words (with the result that two-thirds of its 
vocabulary is composed of Latin or Neo-Latin forms), 
English is the first of the group to arrive at the last 
simplification, at the analytic stage ; and it spreads far 
and wide because it is the easiest of languages, not, 
alas ! to pronounce, nor to write well, but to learn by 
ear and to speak after a fashion. 

It were ungrateful to forget the Eastern groups, 
the Aryans of Hindustan, the Iranians of Bactriana, 
of Afghanistan, of Persia, and Armenia. The first 
named, in small migratory bands, which sometimes 
became stationary for a time, after having tarried long 
in the network of affluents which form the Indus, 
descended the left bank of that great river. They 
"appear to have reached its mouth about the tenth 
century. Their manners, their social polity, their 
creed, and their tongue spread thence, and became 



The spread of Infitcted Languages. 73 

established not only in the valley of the Ganges, but 
throughout Hindustan, and even in Indo-Cbina. The 
career of the Sanscrit literature and of the nume- 
rous languages, dead and living, which are derived 
from it, have been most brilliant Hymns, inter- 
minable epics, religious and grammatical treatises, 
codes, philosophical systems, legends and love poems, 
melodrama and the comedy of manners, no style is 
lacking to this abundant literature. Between the 
first and the fifth centuries of our era, India, mistress 
of herself, and overflowing into Ceylon, Java, and 
Gambodge, listened to the learned discussions of the 
Brahmans, of the Bonzes, of the philosophers, or took 
delight in the ingenious fancies of -i^aer KlMidasa and 
of the SudraJca, and held in Asia the same rank and 
displayed the same civilising force as Greece and 
Rome in the West. Unfortunately, the numerical 
strength of the invading white race had always been 
small. Vigorous and intelligent enough to conquer a 
vast territory and to subdue inferior races, Negritos, 
Dravidians, Malays, they could not modify them, could 
not mould them to a common physiognomy and similar 
aptitudes. They could with diflSculty maintain by 
means of rigorous caste rules the purity of their own 
blood, and attacks from without brought about the fall 
of this edifice, inhabited by too many slaves and too 
few masters. In spite of a blind and deeply rooted fana- 
ticism the immense population^ of the great peninsula 
failed to ofier an efiectual resistance to the Afghans of 
Mahmoud, or j;o the Mongols of mixed race of Timour 
and Baber, or to the tenacious grasp of the English ; 
they retained only their languages, which was well, but 
also their faulty social polity, and the gross and inept 
superstitions which darken the mind of the people. 



74 Distribution of Languages and Races. 

The Iranians, some of whose idioms, Zend and 
ancient Persian, are nearly akin to Sanscrit, occupied 
the basins of the Jaxartes and the Oxus. Disturbed, 
most probably, as the Aryans had been, by the tur- 
bulent Turcoman hordes, perhaps also urged forward 
by the barrenness of the land between the Caspian 
and the Sea of Aral (a barrenness due to the ex- 
haustion of the waters of the two rivers), the future 
Afghans and Persians descended the right bank of 
the Indus, or passed to the west of the great desert 
of Khorassan, while another stream, the Armenians, 
and perhaps .the Parthians, passing between the Cas- 
pian and the mountains, gained a region which lies 
about the sources of the Euphrates, whose ill-defined 
frontiers condemn it to instability and perpetual 
subjugation. The Iranians of antiquity must not be 
judged by the modern Persians, an amiable and artis- 
tic race, the possessors of a beautiful language and of 
a fascinating literature, but weary of their past great- 
ness, and destitute of any real power to withstand 
their ancient enemy the Turk, and th_eir distant cousins 
and powerful rivals the Slavs and Anglo-Saxons. But 
the ancient Iranians were resolute and formidable ; 
they practised an austere religion which honoured 
labour, agriculture, and the family. At the beginning 
of the fifth century they dominated Asia, from the 
Punjab to Ionia, and from the Oxus to the Indian 
Ocean ; they held Syria and Egypt, and in Europe 
Thrace and the mouths of the Danube. In the course 
of sixty years they had incorporated the Medes, over- 
thrown the Assyrian empire and Chaldea, and estab- 
lished themselves solidly on the banks of the Tigris, 
between the Euphrates and the Choaspes ; they had 
vanquished the Lydian power, subdued Phoenicia, 



The Spread of Inflected Languages. 75 

ruined Ionia, dethroned the Pharaohs. They had 
raised, on the ruins and out of the ruins of the ancient 
empires, of which the languages were agglutinative 
or Semitic, a vast edifice, as frail as it was vast, a 
rapid construction, destined to a yet more rapid de- 
struction, and a terrible fall. Eendered effeminate by 
the intoxication of conquest, and yet more by the 
incorporation of exhausted races, dragging in their 
rear the hordes of savages described by Herodotus, 
they did not, when the impulse which affected all the 
Aryans drove them upon the West, bring to Europe 
the untired strength of young blood. Yet the rich de- 
velopment of the Hellenic youth was all but submerged 
beneath a deluge whose floods rolled onward the ruined 
fragments of an ancient world and the undisciplined 
energies of infant peoples. The first shock took place 
in 490 B.C. A few thousands of Athenians arrested 
at Marathon the hosts of Darius. A second and more 
terrible invasion laid waste Thessaly, Boeotia, Attica ; 
Athens perished in the flames. Greece, united to face 
the common danger, crushed the forces of Asia by sea 
and land at Salamis, Platsea, and Mycale. These 
were the critical moments of history. Montesquieu 
says : " Who would be a Persian ? " We should have 
had to submit to that fate but for the courage and 
the fortune of the Hellenes. Bat the Persians had 
nevertheless their hour of grandeur, and if they never 
recovered the power of which Alexander deprived 
them, their bravery in war and the influence of their 
religious doctrines more than once had their effect on 
their Graeco- Roman rivals. 

In this rapid sketch of the Indo-European group, 
we have sought to draw attention to two important 
truths : first, that to this group (whatever may be its 



76 Distribution of Languages and Races. 

ethnical elements), the only one which has hitherto 
shown itself capable of indefinite progress, has be- 
longed for more than a thousand years the leadership 
of the human race ; secondly, that the superiority of 
the Indo-European tongues is inseparable from the 
pre-eminence of the peoples whifch speak them ; that 
before them the agglutinative languages and the 
already inflected idioms of the Semites gave way. 
This succession is thus in perfect accord with modern 
theories of the evolution of language. 

Here China intervenes. To the seven or eight 
hundred millions of men (six hundred in the Indo- 
European group alone) who speak inflected languages, 
China opposes the motionless and compact mass of 
monosyllabism, a form of language which has sufficed 
from time immemorial to about flve hundred millions 
of human beings, who have attained by themselves to 
a certain degree, sometimes a high degree, of civilisa- 
tion. This is a fact which must be recognised and 
explained, and its causes and consequences set forth. 

The traditions of China authorise us to seek the 
cradle of the Chinese race on the eastern boundary of 
the great table-land which is connected on the north 
with the Celestial and Altai mountains, and on the 
south by the Karakoram range with the formidable 
chain of the Himalayas. .Separa,ted from the Western 
world by this vast barrier, ignorant of and unknown, 
to the races on whom they turned their backs, they 
multiplied and extended towards the east, some cross- 
ing the great desert of Gobi and the " thick forests 
of Chan-si, the others passing down the twin valleys 
of the Hoang-ho and tlie Yang-tse-kiang. There 
can "be no doubt that they found and drove out 
earlier populations, whose survivors bear the name of 



The Spread of Inflected Languages. "jj 

Miao-tse, "raw or cooked," according as they retained 
or lost their independence. Moreover, the race is by 
no means unmixed ; the types of Fou-kien and of 
Canton, in spite of the uniformity of costume and of 
the universal pigtail (adopted only 250 years ago), 
differ markedly from the inhabitants of the centre and 
south. Numerous kingdoms, often rivals, afterwards 
united in a sort of feudal hierarchy, became fused at 
last into an immense empire, the second in extent 
after that of modern Russia, under the paternal 
govei-nment of a semi-god, a son of heaven, the father 
and the mother of his subjects. This whole organisa- 
tion had taken shape and become stereotyped before 
the Christian era, without any external influence, with- 
out any communication with the rest of Asia. The 
Chinese had invented for themselves alone and to their 
own taste, all the arts and industries, all the methods 
of agriculture, of working in metals, of making pottery. 
No civilisation was ever more original, more isolated 
or more precocious. After having passed through the 
Stone Age, to which the words cJd-fao, chi-tsien, chi- 
kien, chi-jin, chi-fou, " knife, point, sword, tool, axe, of 
stone," still testify, the Age of Bronze, then of Iron, 
they have become for ever fixed in the same morality, 
the same devotion to ancestors and genii. Centuries 
have passed, and neither the intrusion of Buddhism 
and Islam, nor the Mongol devastations, neither the 
Manchu revolution nor the violent and successful in- 
cursions of modern days, have appreciably modified the 
manners and the genius of China. To a passably 
educated Chinaman, the Barbarian world is veiled- in 
a mist, in which a few and soon effaced outlines can 
barely be distinguished ; and when, at rare intervals, 
a baud of priests or of soldiers comes to convince him 



78 Distribution of Languages and Races. 

of the existence of that non-ego which is called France, 
England, or Germany, he attaches no more importance 
to it than the field-labourer to a passing hail-shower. 
Even those who have visited our lands retain as the 
impression of our civilisation only a vague wonder and 
a more definite mistrust. 

China herself, in the same way, in spite of the 
commercial relations between the Romans and the 
Seri, in spite of the narratives of the Arabs and of 
Marco Polo, notwithstanding wars, journeys, and scien- 
tific expeditions, China is for Western Asia and Europe 
merely a confused mass iu whjch can be distinguished 
for a moment, only to be lost again iu the oblivion of 
indifference, the names of a few Emperors and philoso- 
phers, scattered over an ocean of four thousand years. 
Silk, tea, porcelain, and enamels : with these we have 
from China all that the West cares about. Printing, 
gunpowder, decimal notation, an independent morality, 
and the art of government, these things we had dis- 
covered for ourselves. China has nothing to teach us 
now ; she has done nothing for us in the past. Had 
she never existed, the web of history, the long record 
of human existence, would never have shown at any 
point of its course the traces of any gap or failure. This 
isolation will doubtless cease, but is it not in itself an 
all-sufficient reason for Chinese conservatism ? And 
as concerns the language, a special obstacle may be 
mentioned — a written character almost as ancient as 
monosyllabism itself, which, adapted to the monosyl- 
labic form, has preserved it from all alteration. 

" According to tradition," says M. Vinson, " the 
first characters were rude drawings of material objects : 
a circle with a dot in the centre signified the sun ; a 
vertical stroke with two lines on either side at an 



The Spread of Injlected Languages. 79 

angle indicated a tree ; and so on. To express com- 
plex ideas, several of tkese symbols were takeu together ; 
the signs for sun and moon together represented light ; 
those for woman, hand, and broom meant a married 
woman ; to hear was rendered by the signs for ear 
and door ; to follow, by three symbols for man, placed 
one after the other. Then certain symbols were taken 
of which the pronunciation only without the meaning 
survived.' Pk,, white, together with the sign tne, took 
the sense of cypress ; fan, with earth or mountain, 
meant dyke. There are about 169 of these signs 
which have become phonetic, and of which many are 
no longer employed singly. The Chinese characters 
often have variants borrowed from an earlier period, a 
system now out of date ; the written character has, in 
effect, varied since the date, 2950 B.C. according to 
the legend, when Fou-hi invented the pictorial char- 
acter. These variations have been classed into six 
different styles. In the ordinary style the characters 
retain little of the ideograms as originally drawn ; 
they are composed of strokes of which the number 
allows of an artificial classification of the vocabulary. 
There are 214 type words, called keys: six formed 
with one stroke, twentj'-three with two strokes, and so 
on up to seventeen strokes. But this classification 
varies according to the grammarians. It is said that 
these artifices allow as many as 43,496 words to be 
written, all monosyllabic or compounds of monosyllables, 
of which about one-third compose the ordinary current 
vocabulary." I have heard it said that an educated 
Chinaman could not boast of knowing how to refid 
before he reached the age of forty years or more. It 
may be imagined how great a part caligraphy plays 
in Chinese education. It is astounding that such a 



8o Distribution of Languages and Races. 

system of writing should have been adopted also by 
the Japanese and the Annamites. The Chinese char- 
acter is written in vertical columns, or, if necessary, 
horizontally from right to left. 

The language is not nearly so alarming as the writ- 
ing, but use alone can teach it, since memory is not 
aided by grammar or by derivation. To complete the 
indications given in the chapter on vocabulary, we 
refer the student to the Linguistique of Hoyelacque 
and to the Dictionnaire des Sciences Anthrcypologiques ; 
and pass on to give some supplementary information. 

The language is by no means homogeneous and uni- 
form. Not only does the Chinese of the educated class 
differ from that of the peasant, of the sailor, of the arti- 
san or the trader, but each region has its dialect. There 
is more difference between the speech of the different 
provinces than between the various patois of France ; 
so much so that, according to M. Hovelacque, the 
Government officials sent to serve in the provinces of 
Fou-kien or of Canton, cannot, unless natives of the 
district, get on without interpreters. The speech of 
Canton is that of the south, the dialect of Fou-kien 
extends a little farther to the north along the coast 
and to the neighbouring islands ; in the central pro- 
vinces of the empire, at Pekin and Nankin, the Man- 
darin dialect prevails, which is the language of a 
wide tract of country, and also the official and literary 
language of the whole empire. 

The three principal dialects are distinguished chiefly 
by the sound. The letters 6, d, and g exist only in the 
language of Fou-kien. (All the g's so largely used in 
transliteration, tsong, tsieng, chang, represent merely 
a nasal reinforcement of the vowel.) The Mandarin 
language omits in pronunciation the initial compound 



The Spread of Inflected Languages. 8 1 

ng ; nga, ngo, ngd, ngan, are prononnced a, o, 6, an ; h, 
followed by i, kia, Jcio, kiu, which remains hard in 
the south, has become ts in the north, tsia, tsio, tsiu. 
Nothing is more common than such variations ; they 
exist in every group of languages. 

We have just seen that the g is unknown to literary 
■Chinese ; in fact, every word in it is composed of an 
initial . consonant or spirant and a vowel, simple or 
nasalised : ta, great ; fu, father ; mu, mother ; yuan, 
distant ; jin, man ; hiung, elder, &c. A single and 
very doubtful exception to this rule is the word which 
signifies "two'' and "ears'' : eul, ulh, urh, rh; the vowel 
seems here to precede the consonant, but the sound 
is confused and difficult to transcribe ; it is an efibrt 
towards the pure liquid r, which the Chinese do not 
possess. They write and pronounce France, for ex- 
ample, Folan-tsi. 

In the dialects of Canton and Fou-kien, the short 
words may be terminated by a strong explosive con- 
sonant, k, t, or f. There are in all the dialects short 
words and long, of which the quantity depends on the 
accent or tone. . These tones, invented to distinguish 
between syllables of the same sound but of very diverse 
meaning, number eight in Fou-kien, five in the Man- 
-darin dialect ; at Pelcin there are but four, three long 
and one short. These accents increase the number of 
roots from 450 to 1250. It will be seen that in this 
system, deprived of the aid of suffixes, the raw material 
of language is of the poorest. It is only by a marvel- 
lous ingenuity that the Chinese have been able, with- 
out other resource than apposition, to acquire 40,000 
signs, that is to say, 40,000 ideas, and to apply 
their imperfect instrument to every style — philosophy, 
morals, history, poetry, and the drama. But to us 



82 Distribution of Languages and Races. 

Europeans it seems that this ingemiity is displayed at 
the expense of perspicuity, logical composition, and 
also of inspiration. The thought of the Chinese, like 
their art, lacks perspective ; either it is stifled beneath 
a mass of detail, which is not properly subordinate to 
the whole, or, considering the whole as through a sort 
of fog, it loses its sense of reality ; it is either trite, 
diffuse, and prosaic, or incoherent and unreal. The 
chronicles are* interminable ; the enumerations are in- 
congruous; poetical imagination weak, and its form 
disfigured by mannerisms. Science is inaccessible, 
not indeed to the mind of the Chinese, but to their 
language and writing, and they cannot renounce these 
without losing all their past history. 

The dominion of the Chinese language is limited on 
the north by Corean, Mandchu, and Mongolian, agglu- 
tinative languages ; in the south it is found along the. 
Indo-Chinese coast in the commercial centres, where it 
contends with Malay. 

Though the type of the monosyllabic languages, 
Chinese is not the only one ; Annamite, Siamese, Bur- 
mese, and Thibetan are other examples. The countries 
where these languages are spoken are, however, but 
outlying parts of the great Chinese empire, to which 
they have been at times united by vassalage or alliance. 
Tiiey are a part of that great slice of Asia which in- 
clines towards the east, and turns away from the rest 
of humanity, the only portion of the globe where this 
fossil language could be protected in its growth. The 
hereditary tendency to monosyllabism must have been 
very strong to resist influences and conquests which 
left China intact, but which were not waTiting to Thibet 
and Indo-China. Buddhist missionaries established 
their principal sect in Thibet, and have reigned there for 



The Spread of Inflected Languages. 83 

nearly twenty centuries. In the Middle Ages the 
Aryans of India founded at Gambodge a flourishing 
kingdom and a brilliant civilisation, of which we 
admire the ruins at Ang-kor-wat, where learned men, 
Bergaigne among others, have deciphered inscriptions 
in Sanskrit and Pali ; nevertheless the old type pre- 
vailed, and with it the ancient form of language. The 
inhabitants of Thibet and Indo-China have inherited 
only the worst of legacies from their fugitive civilisa- 
tion, a narrowing and childish religion. 

Annamite, the language of the eastern portion of 
Indo-China and of Tonquin, has borrowed considerably 
from the vocabulary of Southern China, and its written 
character, which is figurative and ideographic, is of 
Chinese origin, although much modified and developed. 
Its syntax corresponds to that of Chinese. The addi- 
tion of such terms as male and female, all and many, 
to a I'oot syllable, indicates gender and number ; the ad- 
jective follows the noun (which it precedes in Chinese) ; 
various terms which signify distance, proximity, doubt, 
give to the roots the value of verbs, and determine 
mood and tense. Six tones, acute, interrogative, as- 
cending, descending, grave, equal, serve, as in Chinese, 
to differentiate words of which the sound would be 
absolutely the same, although the sense'is different. 
But, in spite of this indebtedness and the similarity of 
form, Annamite is a distinct language. The vocabulary, 
the monosyllables, which are proper to this language, 
are purely Annamite, and in no sense Chinese. 

Siamese or Thai (spoken on the north and west 
coasts of the Gulf of Siam) is separated from the 
Annamite by the language of Gambodge, which is yet 
unclassed. Siamese is rich in aspirates and sibilants, 
and has a written character of Indian origin, but has 



84 Distribution of Languages and Races. 

all the marks of monosyllabism — the use of tones (of 
which there are four), the absence of grammar, apposi- 
tion, and order of words determined by a rigid syntax. 
The language of Burmah offers the same characteris- 
tics, though it is poorer in sounds and has less variety 
of tones. The Burmese empire, which the English 
have diminished by four provinces — Arakan, Martaban, 
Tenasserim and Pegu — includes a number of tribes of 
which the origin is most uncertain, half-breeds of 
Hindus and Black Dravidians, of Malays and Negritos, 
of Mongoloids and Mois, &c., &c. These groups march 
on the north with the peoples of Yunnan, and on the 
south with the Siamese or Thai, and on the east with 
all the semi-savages of the Me-kong (Mother of Seas), 
Laotians, Stiengs, Kouis, Girais, Kharais, which Mouhot, 
De Lagrfe, and Garnier have visited and described. 
These are the relics of the ancient Gambodge^ the 
country of the Kams, Kammers, Kmers, who were so 
amenable at first, so indifferent afterwards, to the civi- 
lisation of the Hindus. These distant and interesting 
countries are now become fields of exploration open to 
our anthropologists and philologists. But the study 
of the methods and organisms of language will derive 
little profit from them. 

Thibetan, of which we have little to say, owes to 
India its rich and precious literature, consisting en- 
tirely of translations of Buddhist books (of which the 
original is sometimes lost) as well as its alphabet. Its 
method of determining case, and mood, or tense are 
once again the respective places of the words, and the 
association of full roots and empty roots. The inflec- 
tions which some have thought to discover in Thibetan 
are not more joined to the word than any other root 
.deprived in part of its primitive sense, and converted 



The spread of Inflected Languages. 85 

into a particle. He who would write a comparative 
syntax of the isolating or monosyllabic languages must 
forget all such terms as number and gender, mood and 
tense, case and person. 

After having shown that chronology nowhere goes 
far enough back into the past to furnish a basis for 
the history of language, we have nevertheless made it 
clear that the gradual elimination of the agglutina- 
tive to the advantage of the inflected idioms, and espe- 
cially the ever-growing expansion of the Indo-European 
tongues, which always tend to become more analytic, 
coincides with the discoveries of philological analysis. 
Yet one great fossil block stands apart, outside, so to 
speak, of the current which has deposited the succes- 
sive strata of language ;. the monosyllabism of Chinese, 
Annamite, Siamese, Thibetan, emerges from the depths 
of the past. We have pointed out the purely geo- 
graphical causes of its survival, and displayed the con- 
sequences of the isolation of this group — useless effort, 
complication of the written character, atrophy of the 
higher functions of the brain, incoherence and petti- 
ness of thought. We have noted the fact that these 
peoples, who have undoubtedly great gifts, have yet 
played next to no part in the history of the world and 
of civilisation. 



CHAPTER II. 
THE AGGLUTINATIVE IDIOMS OF CENTRAL ASIA. 

Languages of Corea and of Japan — ^Ethnical elements of the Corean 
and Japanese peoples — Hyperborean group : Ainus, Ghiliaks, 
Kamschatkans, Tchonktches, Youkaghirs — Uralo-Altaio family : 
I. Samoyed group ; 2. Tongouse-Mandchu group ; 3. Mongol- 
Kalmuck group; 4. Turkish group; 5. Tinno- Hungarian group ; 
the characters common to the five groups — Vowel harmony. 

In passing from monosyllabiam to agglutination, we 
have no great distance to traverse. I am not speaking 
merely of territorial distance; I mean that between 
these two phases, these two linguistic organisms, there 
are insensible transitions, the one beginning where the 
otlier ends. 

The line of demarcation is so fine that certain 
eminent philologists, Max Miiller among the number, 
hesitate to class Siamese and Thibetan among the 
monosyllabic languages. It may even be said that 
absolute monosyllabism exists no longer. The majority 
of Chinese words consist of two or three syllables, and 
we find agglutinative dialects, more especially in the 
Tongouse group, of whicli the grammar is yet so un- 
developed that it has no case or verbal endings. We 
need to fix our attention on a positive and certain 
distinction, which I have already indicated, but on 
which I must insist fui-ther, because, though appar- 
ently slight, it is yet the point of departure and the 

common characteristic of all the agglutinative lan- 
es 



Agglutinative Idioms of Central Asia. 87 

guages ; it is the change in and gradual atrophy of 
the subordinate roots. 

The syllables which the Chinese call empty, as 
opposed to the full syllables, lose in part their signi- 
ficant force, but they retain their form ; the sense is 
effaced, the sound remains invariable. The result is 
that they can neither form terminations nor serve as a 
connecting link between a root and suffixes denoting 
case or person. The words, therefore, even when poly- 
syllabic, remain sterile, and cannot produce others by 
derivation ; no Chinese, Annamite, or Burmese word 
gives birth to a series of verbs, nouns, and adjectives 
derived from a common root. 

In the agglutinative order, the root, full, or principal 
syllable, alone remains invariable ; the subordinate 
roots, those which amplify or modify the meaning of 
the full syllable, are susceptible of change in form, in 
sound, as well as in their primary sense. Sometimes 
atrophied by their close connection with the root (a 
name which the subordinate roots change for that of 
suffix), sometimes with their initial consonant or their 
central vowel affected by the influence of the root, they 
furnish a certain number of signs, applicable respec- 
tively to the different parts of speech, or else they 
form with the root an indivisible whole, a new root 
or theme, susceptible in its turn of acquiring other 
suffixes, and of giving birth to a greater or less number 
of derivative terms. 

Thus monosyllabism and agglutination have in 
common the inalterability of the root or full syllable, 
and the alteration in the sense of the subordinate or 
empty syllable ; to agglutination alone belongs the 
change in the form of the subordinate root. Inflected 
languages have, in addition, the power to change the 



88 Distribution of Languages and Races. 

root syllable. From one class to anothei: there is but 
one step ; the barrier is so slight that certain peoples 
have crossed without knowing it, so imperceptible that 
others have not sought to cross it, so decisive never- 
theless that it clearly divides the three stages of lan- 
guage. There is, I believe, no instance of a language 
tending to return to the stage which it has left, and 
it is rare that a language abandons that in which 
custom and literature have fixed it. 

China is the near neighbour, and even the titular 
sovereign, of the country of the Mandchus and of the 
Eastern Mongols ; she has been conquered by both at 
different times, but she has borrowed nothing from 
th^ir idioms, which are agglutinative, although poor 
specimens of the class, and her own influence is almost 
nil, in spite of the ascendency of her superior civilisa- 
tion. The Chinese language spreads to the north and 
west beyond the great wall, and is spoken in towns 
situated in the countries of the Mongols and Mandchus ; 
but the natives keep their own idiom, as do the mer- 
chants from the " Land of Flowers." 

Corea, a mountainous peninsula which juts out be- 
tween the Pe-tchili and Japan, was occupied from the 
twelfth to the first century B.C. by the Chinese, and 
has retained from the language of the conquerors a 
number of names of objects, of administrative divisions, 
and of occupations of all sorts ; its king, still a vassal 
of the Chinese emperor, sends a respectful embassy 
every year to Pekin to fetch the calendar of the year. 
Yet the Coreans have their language, in no way akin 
to the Chinese vocabulary, and weakly but certainly 
agglutinative from time immemorial. They have also 
an alphabet, of Indo-Thibetan origin it is believed ; 
but they do not seem to have profited to any consider- 



Agglutinative Idioms of Central Asia. 89 

able extent, any more tlian the Thibetans or the Siamese, 
by the possession of this precious instrument of progress. 
The latter have, it is true, been reduced to intellectual 
childhood by Buddhism. In Corea, Buddhism is, as 
it is in China, at once official and despised, and the 
cause which has hitherto retained the country in a 
semi-barbarous condition must be sought in the in- 
fluence of fear. The country is threatened at once by 
China and Japan, by Russia and by the Western Powers. 
This varied country, as large as the half of France, 
and peopled by at least eight millions of short broad- 
shouldered men with a type of face like that of the 
Japanese, and by a bearded race with horizontal eyes 
and light skin (the Han, descendants of immigrants 
from Nan-Ohang), opens its ports to other nations 
only under constraint. The Japanese have invaded it 
several times, notably in i 591, and exacted a tribute 
in which figured thirty human skins, and have recently 
established two trading ports on the south-east coast*; 
the Chinese, who left it alone for sixteen centuries, 
deprived it in the seventeenth century, but only for a 
short time, of its north-western provinces ; the French 
and the Americans have both made vain demonstrations 
at the mouth of the Hang-Kang, the river which 
waters Seoul, the capital. The Catholic and Protestant 
missions have made little way ; their labours have 
availed at least to give us those geographical details 
from which Anville has traced the outlines of Corea, 
and some valuable information regarding the people, 
its customs, and government, A Corean-French dic- 
tionary, the work of a priest who escaped the massacre 
of 1866, a Corean grammar in French, published at 
Yokohama, the fine collection brought back by a tra- 
veller and exhibited in 1889 at the Trocadero, and, 



go Distribution of Languages and Races. 

lastly, the perseverance of the Japanese, the English, 
and the Eussians, will sooner or later dissipate the 
obscurity which hangs over this nation, but may per- 
haps diminish the interest which is born of mystery 
and curiosity. Indeed, we know of the Corean people 
all that matters for our present purpose : the nominal 
power of an absolute king, the real power of the great 
chiefs who surround him with all the forms of a servile 
respect ; the division of the nation into nobles, plebeians, 
and slaves ; the sequestration of the married women ; 
polygamy ; the belief in genii and ancestor-worship ; 
survivals of iire-worship ; the rigour of mourning, which 
obliges a son to weep for his father three times a day 
at stated hours for three years, and to abstain for the 
same period from all public functions. None of these 
are uncommon customs, but two or three peculiarities 
deserve mention. The Ooreans do not spin or weave 
wool ; in winter they wear a greater amount of hempen 
and cotton clothing, and their soldiers wear cuirasses 
lined with many folds of similar material, which were 
proof against the bullets of old time ; violet and olive 
green are the favourite colours, white and green being 
reserved for moumingf. 

■ A few more words are necessary about the language 
of the Kaokaiuli or Korai (of which we have made 
Ooreans ; it is the name of one of the northern pro- 
vinces, but they prefer to give their country a name 
which recalls its situation between the empire of the 
centre and the land of the rising sun ; Tchiaosien, the 
clearness of morning). 

In grammatical structure, says the missionary Dallet, 
Corean somewhat resembles the Uralian and Tongouse 
idioms. The terminations of the verbs vary according 
to the sex and condition of the interlocutors. The 



Agglutinative Idioms of Central Asia. 9 1 

prbnimciation is harsh and full of aspirates, drawling 
and indistinct ; each phrase ends with a peculiar 
guttural difficult to reproduce. The liquid I is "not 
clearly heard. The vowels, fourteen in number, are 
uncertain and incline to be diphthongs. The written 
character consists of rather more than 200 signs, 
some syllabic, others alphabetical, but educated people 
disdain to use them. " The introduction of a number 
of foreign words, Chinese in the north, Japanese in 
the south, has given birth," says Elisde Keclus, " to 
various jargons which are widely spoken in the centres 
of commerce. Chinese is the official language. Just 
as in Europe in the Middle Ages, Latin, the language 
of the lettered, persisted side by side with the local 
idiom, so the written Chinese is maintained in Corea 
together with the language of the people ; but it is 
pronounced in such fashion that the Chinese could not 
understand it without an interpreter. According to 
the missionary Daveluy, the language of many districts 
is composed entirely of Chinese words, but with Corean 
terminations. In brief, every place, every person, 
every thing has two names, one Corean, the other 
Coreanised Chinese, and these synonyms enter freely 
into the speech of all classes." The vocabulary is 
mixed, not the structure ; the agglutinative character 
is found even in the elements borrowed from the 
Chinese monosyllabism. 

Over against the immobility of China and the mis- 
trust of Corea we find a people eager for civilisation. 
No sooner had treaties, extorted by intimidation, 
opened five or six ports to Europeans, than this land, 
which from the sixteenth century violently opposed 
the foreigner, Japan, or rather Nippon, which mas- 
sacred missionaries and forced the Dutch to spit upon 



92 Distribution of Languages and Races. 

the cross, became suddenly enamoured of our ideas, of 
our law, of our science, and made a vigoious effort 
towards progress. A reforming government put an 
end to the feudal daimos, to the military usurpation 
of the Syogun or Taikoun, centralised the administra- 
tion, caused a code of laws to be drawn up by French 
lawyers, projected railways, established schools every- 
where, destroyed the Buddhist temples under pretext 
of restoring the ancient worship of genii, published 
newspapers, sent students to Paris, London, and Berlin, 
to learn our languages, our manners, and institutions. 
And recently we have learned that the Mikado, the 
son of the rising sun, the ancient and divine head of 
a theocracy, has summoned an elected parliament. It 
is possible to have diverse opinions as to the future 
of a change so radical and urged forward with such 
unusual haste. In any case, it commands attention 
and sympathy; it is not possible to look coldly on 
those who welcome us with open arms. But what is 
the history of this people, which seems to be ancient 
and which yet shows all the signs of a vigorous youth ? 
A complete answer to this question would take us 
altogether beyond our subject, but some attempt must 
be made. Japan was inhabited before the dawn of 
history ; instruments of stone and of bone have been 
discovered in different parts of the Archipelago, in 
tumuli and kitchen-middens, mingled with the bones 
of monkeys, bears, boars, and deer, and of other 
animals, some of which are now extinct. Human 
bones fractured and split longitudinally even seem 
to point to cannibalism. It is not known what this 
ancient race was, nor whether it is represented by 
Negritos, who transmitted their woolly hair to some 
of the southern groups of Kiu-siu, or by the Ainus 



Agglutinative Idioms of Central Asia. 93 

(Yebiss or Mao-tsin of the Chinese), a hairy race 
which certainly long occupied the great island of 
Nippon or Hondo. At a very early period an in- 
vasion from Oorea, attacking Kiu-siu from the islands 
of Tsou-siina and Iki, drove back to the north-east 
the majority of the Ainns. These Coreans, the Kmapo 
or Ion-90, appear to have been plebeians or country 
folk of the Mongoloid type : the face wide and 
lozenge-shaped, retreating forehead, eyes narrow and 
oblique, short nose, high cheek-bones, and yellow skin. 
Finally, towards the seventh century before our era, 
tradition speaks of the arrival of a legendary con- 
queror, Kaniou-Tamato-Vare-Bixo, and of a new race, 
the Yamatos, which furnished the aristocratic element 
and the type with the oval face, straight forehead, 
narrow and often aquiline nose, horizontal almond- 
shaped eyes, and olive skin, a type which recalls in 
miniature the Malayo - Polynesian. The Yamatos, 
landing on the south-east of Kiu-siu, drove back by 
degrees the Kma90 towards the north-east. The 
fusion between the two races was slow ; the strife 
was prolonged to the middle of the second century 
of the Christian era. The Ainus, driven from Nippon 
in the seventh, held their own from the ninth to the 
sixteenth century in the island of Yeso. Then they 
lost their independence and retreated towards the 
extreme north of Yeso, and into the little archipelago 
of the Kouriles. They now number less than twenty 
thousand and are gradually dwindling ; but here and 
there atavism revives some of their characteristics in 
their ancient home. 

The civilisation of Japan was tardy and entirely 
Chinese. It was not until the sixth century that 
the worship and doctrine of Confucius, Ko-si, the 



94 Distribution of Languages and Races. 

religion of Qaka (Qakyamouni), Buddhism, and tlie 
Chinese character, all penetrated to Japan by way of 
Corea. While the Mikado, the sacred emperor, re- 
mained faithful to his ancestors, the genii or Kamis, 
and to the ancient national religion, Shintoism, his 
lieutenant, or mayor of the palace, the Syognn who 
usurped the civil and military power, embraced 
Buddhism. Confucius became and remains the 
teacher of the lettered class. These three religions 
have mutually borrowed from each other and live in 
harmony, although Shintoism has again become the 
state religion. The aristocracy has passed from Con- 
fucianism to complete scepticism. But I leave the 
description of the manners and the arts of Japan, which 
the accounts of travellers, the novels of Pierre Loti, 
and the caprices of fashion have made familiar to all. 

" Japanese literature," says M. Julien Vinson, " is 
very rich ; in the last eight hundred years innumer- 
able works of poetry, of mystic philosophy, and even 
of science, have been composed in the archipelago of 
Nippon. The earliest known works are the Kosiki, 
the sacred book or bible of Shintoism, which dates 
from the year 712, and the Yamato-boumi, or ancient 
national annals." This literary and intellectual de- 
velopment is sensibly later than the indirect inter- 
vention of China in the history of Japan. This fact 
has had a marked influence on the fate of the national 
language, Yamato, the language brought by the early 
invaders from the south-east. The intrusion of mono- 
syllabism, and especially of so inconvenient a char- 
acter, paralysed Japanese, and arrested it at the first 
stage of agglutination, between declension and con- 
jugation, cut short its tendency towards the inflected 
state, and encumbered it with a quantity of Chinese 



Agglutinative Idioms of Central Asia. 95 

words to such an extent that the populace speak a 
sort of hybrid tongue, and the pure Yamato remains 
the appanage of the aristocracy, of the lettered class, 
and of the demi-monde. 

Nothing can be softer or simpler than the pro- 
nunciation. Five vowels, a, i, ou, 4, 0; four semi- 
vowels, y, V, vj, f or h ; one liquid, r ; three nasals, 
gn, n, m; four sibilants, s, ch, z, j ; four palatals, ts, 
toh, dj, dz ; finally, the six explosives and true con- 
sonants, h, g, t, d, p,i; I is wanting. The pronuncia- 
tion clearly shows the tendencies towards inflection 
mentioned above ; letters are modified by contact, ts 
and k become kJc ; ts and t become ss ; ts and p be- 
come pp ; n and v become h, &c. The final vowel is 
aJmost mute ; m and k fall by contraction between 
u and i ; uki is pronounced wi. These are Indo- 
European phenomena. KoTe, this man, and kare, that 
man, kimi, lord, and kami, genius, seem to be modi- 
fications of a same root. 

The declension is by the aid of sufiixed particles 
which have come to signify only a sense of relation : 
tsu, no, imply possession ; m, he, ye, e, towards ; to, 
for ; te, ni, in, by ; yori, ablative ; ka, ga, nga, na, 
partitive ; Yuki-ga fwru, it snows ; ama-tsu kami, 
genius of heaven ; Yedove, to Yedo ; inisive yori, from 
antiquity ; Yedoveno missi, road to Yedo ; Yamanove, 
from the mountain. The nominative and the accusa- 
tive are marked by a species of definite article, wa, wo. 

Gender is indetermined ; number is indicated in 
the Chinese way by the addition of a word signifying 
quantity, variety, or crowd, and also more frequently 
by a reduplication which is found also in Malay : 
kuniguni, lands; tokoro-dokoro, ]p\a,ces ; JUo-hito, per- 
sons ; iroirono-fana, flowers. 



g6 Distribution of Langttages and Races. 

It is curious that the personal pronouns are want- 
ing, unless indeed, mi-ga and mi-domo, I, we, formed 
by adding a suffix to the word mi, body, can be con- 
sidered such ; it may be said too that, in expressing 
the third person by the demonstrative, Yamato con- 
forms to what is almost universal usage. Still one 
cannot but be struck by the singularity of the forms 
which stand for /, thou, we, you. A Japanese does 
not say, I see thee, we see you ; but, this man see 
illustriov^, honoured, grandeur, lord; or again, slave, 
imbecile, selfishness see height, nohility. It would 
seem that the extreme politeness inherent in the 
Japanese character has prevented the formation of 
personal pronouns, or, if they existed, has caused them 
to fall into disuse. 

For lack of personal pronouns there is no con- 
jugation. The verbs have remained simple substan- 
tives, which are declined by the aid of noun suffixes 
which allow them to be compared to Indo-European 
infinitives. Yuhu, movement ; yuku-wa, the going, to 
go ; ahe, opening, sight ; fana wo akeni, flower to 
open, to see ; to open a flower, to see a flower. Mood, 
tense, and voice are all expressed by the addition of 
different suffixes, ta, mu, tara, and by the use of an 
auxiliary, are, uru, existence, to be. The dative e 
forms the termination of the passive ; and analogous 
methods produce reflective, causative, and negative 
forms. 

Japanese, in short, is a language beautiful in sound, 
very simple and easy to learn, and capable of clearly 
expressing a great number of ideas. I mean the 
spoken language ; written, it becomes an indecipher- 
able medley. The Chinese character has been destruc- 
tive of all order and reason ; now it is considered 



Agglutinative Idioms of Central Asia. 97 

merely as a sign, and corresponds to a Japanese poly- 
syllable ; now it retains its Chinese monosyllabic 
pronunciation, and answers to only one syllable in 
Japanese. In the former case it keeps its meaning 
and loses its original pronunciation, it then needs a 
translation to render it intelligible ; in the second case 
it is only a very inconvenient syllabic character. In 
one place a single monosyllabic sign represents a poly- 
syllable ; in another several Chinese signs are required 
to express one Japanese word ^ so that the same sign 
may be pronounced in several different ways, and 
several signs conveying different meanings may be 
pronounced alike. It seems that this strange use of 
Chinese chai-acters is more especially the rule when 
dealing with abstractions and scientific matters — that 
is to say, precisely where they ai-e most inappropriate. 
So persistent an adherence to an absurd custom shows 
to what a degree the subtle and brilliant Children of 
the Sun had been struck by the Chinese power and 
civilisation, and penetrated with respect for the wisdom 
of the mandarins. They possessed themselves more than 
one syllabic character, imperfect doubtless, but a thou- 
sand times superior to the Chinese system. The 
Buddhists had even contributed an alphabet, the Sinzi 
or divine, probably of Indian origin. There are seven 
syllabic characters, of which the most used are the 
lateral, the Kata kana, and the cursive, Hira hana 
or Fira Tcana. 'J'he first consists of explanatory signs 
written in small type beside the ideograms ; the 
second has no relation to the Chinese character. 
These systems consist of forty-eight characters. In- 
stead of recognising their evident superiority, the 
Japanese taught, perhaps still teach, a minimum of 
3000 ideograms in their schools, which are even then 



98 Distribution of Languages and Races. 

insufficient ; if the scholar would acquire a real culti- 
vation, he must retain not three, but eight thousand, or 
the literature of his own country will remain closed to 
him. The modern Japanese feel the necessity of sim- 
plification ; at the first Oriental Congress, held in Paris 
in 1873, the Japanese ambassador expressed in very 
correct French the desire to see the adoption by his 
country of an international alphabet. No, language 
would lend itself more easily to our character, slightly 
modified if necessary. This wish will probably be 
realised shortly ; but if they are to reject without 
regret all the Chinese lumber, the Japanese must first 
transcribe into modern letters all their ancient authors 
and their most precious documents, and resign them- 
selves to the gradual loss of comprehension of their 
rich literature. If they come to this decision, they 
will have imposed upon themselves the wholesome 
necessity of a new renaissance. 

The Yamato language, arrested at the first stage of 
agglutination, endowed with a tendency to inflexion, is 
not akin, any more than Corean, by vocabulary, either 
to Chinese or to the idioms of the Sakhalin Island 
(ceded to the Russians) and of the coast of Asia. The 
world is full of these solitary idioms, which are born in 
and for a single tribe, or wjiich have changed several 
times, perhaps after having separated from allied dia- 
lects. We shall find in Africa and in America groups 
of savages who wear out a language in fifty years ; 
but for the moment we are concerned with North- 
Bastern Asia. 

The languages known as Hyperborean — Ainu or 
Kourilien, Ghiliak, Kamtchadale, Koriak, Youkaghir, 
Tuhouktche or Kotte or Yenissein — seem to form a 
small and poor agglutinative family. They all differ, 



Agglutinative Idioms of Central Asia. 99 

more or less, like the tribes which speak them — fishers 
or hunters, living some in tents, some in rude huts or 
in holes, dens hollowed out of the frozen soil; tribes 
which worship the bear and the whale, and believe in 
charms and sorcerers ; some of them throw their dead 
to the dogs. The Ghiliaks, who now number only 
6000, inhabit the north of Sakhalin and on the main- 
land the environs of Nikholaievsk and of the Lower 
Amur ; they will soon be engulfed by the Tongouses 
and the Russians. The Kamtchadales or Itelman, a 
dirty and inoffensive race, occupy the south of the Kamt- 
chatka ; tliey are perhaps connected by the Aleoutes to 
the Esquimaux of Alaska. The north of Kamtchatka 
belongs to the Koriaks and to the Tchoutktches. 
Beside these last, on the river Kolima, the Youkaghirs, 
tall and relatively handsome, a very mixed race, but 
distinct from the Samoyeds, have been driven back 
upon the Arctic Sea by the Tongouses and the Yakoutes. 
It is probable that all these groups are the last repre- 
sentatives of nations which formerly occupied a much 
more extended area in Eastern Siberia, and have not 
been able to resist the ancient pressure of the Chinese 
and the expansion in every direction of the Mongols 
and the Turks. They seem to be all more or less 
akin, at least from the point of view of language, to 
the Samoyed branch which borders the north-west of 
Siberia and the north-east of European Russia. 

All these tribes, who are not more wretched than 
others in their almost animal ignorance and under 
their harsli climate, are far from being or from think- 
ing themselves the lowest of humanity ; they glory, for 
the most part, in the name of heroes (this is generally 
the meaning of their name) ; they are aware of social 
distinctions ; they have their nobles and their priests, 



lOO Distribution of Languages and Races. 

their code of honour and morals. I would not omit 
them from this list ; it is useful to show how ill the 
infinite variety of human types and idioms agrees with 
the long-accepted dogma of the original unity of the 
human race and language. 

We come now to a true linguistic family, not in- 
deed closely allied, like the inflected groups, by filiation 
and constant relationship, but in which, nevertheless, 
the identity of certain pronominal roots permits us to 
suppose, if nob to reconstruct, a single ancestral form, 
a common vocabulary. This is the Uralo- Altaic family, 
of which the vast extent formerly suggested to Max 
Miiller his idea of a Turanian family, in which he 
essayed to class all those idioms which are neither 
Semitic nor Indo-European. But the hypothesis fell 
to pieces before the impossibility of ranging together 
the African, American, Malay, and Dravidian groups, 
in which there is no characteristic common to all ex- 
cept the agglutinative method. The classing together 
of these fundamentally different families only tends to 
throw the science of language into hopeless confusion. 

The vague and insufficient designation of Uralo- 
Altaic merely indicates the primitive area of the family ; 
it tells us that all the branches of this immense tree 
germinated between the Altai' Mountains and the Arctic 
Ocean, between the Sea of Okotsk and the Ural Moun- 
tains. 

The first branch is Samoyed, of which the Finlander 
Castren has made a study. Ifc extends, in Europe, 
along the eastern half of the Russian coast of the 
Arctic Sea as far as the White Sea ; in Asia, along 
the western part of the Siberian coast. Its five prin- 
cipal dialects, Yourak, Tavghi, Yenissein Samoyed, 
Ostiac Samoyed, and Kamassin, are not spoken by 



Agglutinative Idioms of Central Asia, loi 

more than 20,000 individuals. The category of gender 
is unknown to Samoyed ; the noun and the verb are 
not distinguished : Lutsa, Russian ; Lutsa-me, I am a 
Russian. Like all the agglutinative languages, it ex- 
presses by means of suffixes all the relations of num- 
ber, case, person, mood and tense. There are various 
methods of derivation. 

Tongouse, a group which is a near neighbour to 
the Samoyed, is more important from the number of 
those who speak it> The little Tongouse people, active, 
cheerful and hospitable, who live on and by the rein- 
deer — a type with round face, narrow eyes, and square 
forehead — occupies that part of Siberia which lies 
between the river Tongouska and the district of the 
Lower Amur. The Mahdchus, who number 70,000, 
to the south of the great river Amur, are really a 
branch of the Tongouses, which was formerly nomadic 
and warlike, and became in the seventeenth century 
the masters of China. To this day their eight banners 
form the nucleus of the Chinese army. The reigning 
dynasty is Mandchu, and the Mandchu gen eral-in- chief 
is still, officially, the commander of all the forces of 
the immense empire, in many parts of which, doubtless, 
the very name and existence of the Mandchus are un- 
known. These chance conquerors occupy the north- 
east of China. They have retained their Ohamanist 
religion and their language. But, as in the case of 
the Japanese, the superior influence of the monosyllabic 
Chinese has hindered this language in its natural evo- 
lution ; so much so, that the independent Tongouse, 
which is not a written language, is richer in gram- 
matical forms than its more civilised brother, which is 
promoted to the rank of a literary idiom. Mandchu 
has no conjugation, whereas the Tongouse verb abounds 
8 



I02 Distribution of Languages and Races. 

in suffixed forms. Tongonse, like Saraoyed, has uo 
gender, but it expresses very completely all casual 
relations; it forms derivatives by combinations of 
suffixes, and true compounds with a common termina- 
tion. The pronunciation of Tongouse is fluid and plea- 
sant to the ear. The principal dialects are Tongouse, 
Mandchu, Lamout, Anadyr, Kondogyr, and Vilui. 

A third group has for centre the Lake Baikal, and 
for type Bouriate, spoken by 20,000 persons. M. 
Lucien Adam, who, with M. Victor Henry, must be 
the guide of those who would make a serious study of 
the Uralo- Altaic family, ranks this language very high 
in the scale, though it is without literature and has no 
future. He thinks the grammatical development of 
Bouriate important, inasmuch as it shows the inter- 
mediate forms through which pronouns have passed in 
order to become suffixes. Side by side with Bouriate 
have grown up, on the south-east, the Mongolian spoken 
in Mongolia proper, in the central part of Northern 
China, to the west of the Mandchu territory ; and in 
the west, the Western Mongolian or Kalmuck, which 
has penetrated into Russia as far as the mouths of the 
Volga. The brilliant and terrible history of the Mon- 
gols, to-day an exhausted race, has given them a pre- 
eminence which is not justified by the organisation of 
their language. Mongolian has, however, like Mandchu, 
a written character and a literature. What a contrast 
between its present obscurity and the tumult of the 
multitudes led Iby Zenghis Khan to the conquest of the 
world ! 

From the foot of the Altai' Mountains a torrent of 
disciplined hordes under Zenghis Khan spread east- 
ward over China and deposed the Yuan dynasty, which 
Marco Polo had seen in all its power. In the west it 



Agglutinative Idioms of Central Asia. 103 

overthrew the Caliphs of Bagdad and the Sultans of 
Iconium, reached Moscow, and wasted the greater part 
of Russia, which remained during two centuries under 
the domination of the Golden Horde ; invaded Poland, 
Moravia, Silesia, Hungary (1240—41), and was only 
stayed by the combined armies of the Germans and 
Slavs. In the fourteenth century the Mongols, rallied 
by Timour, reconquered Asia. Finally, from Bactriana, 
where a Mongol dynasty had established itself, Baber 
came down to the conquest of India, and founded there 
the Mogul Empire. But now the Mongols have to 
vegetate as the subjects of the nations of which they 
were once the masters — of the Mandchu rulers of China, 
of the Czars of Russia, and of the Sultans of Turkey. 
They once were free of soul ; superstitious doubtless, 
they were not bowed beneath the yoke of any religion. 
But they have long been Mussulman or Buddhist ; their 
part is played out. 

We pass to the fourth branch, which has done the 
world no less harm than the preceding one. The 
region which it still covers with its shade is of vast 
extent ; it stretches from the river Lena and the Arctic 
Ocean to the Mediterranean. The Turkish family, the 
Hiung-nu and the Tukiu of Chinese writers, the Turan- 
ians so dreaded by the ancient Persians, were already 
known and feared two centuries before our era. Their 
warlike character, and their constant attacks upon the 
Mongols, who were a nomadic and pastoral people, con- 
tributed most certainly to precipitate upon the west all 
those invasions which destroyed the ancient civilisa- 
tions and constituted at length modern Europe. Even 
a brief summary of the history of the innumerable 
tribes — Tatars, Turcomans, Seljuks, Ottomans — which 
belong to this family would take us too far from our 



I04 Distribution of Languages and Races. 

present subject. The Turkish race is divided into five 
principal branches, to which are attached a number of 
dialects. The most northerly, scattered among the 
Tongouses and near to the Bouriates, is represented by 
Yakout, which is remarkable for the purity which it 
owes to its long isolation ; it is spoken by hardly more 
than 20,000 people. Farther west the Cossack Kirghis 
stretch to the Sea of Aral and the Caspian, and to the 
south of these and on their left, in. Chinese Turkestan, 
towards Kashgar, lie the black Kirghis or Bouroutes. 
Tchouvache, spoken in Eussia in the south-west of 
Khazan and in the neighbourhood of Simbirsk, is 
classed with these dialects. The Kirghis came origin- 
ally from the district which lies between the Yenissei 
and the Obi ; their kinsmen and near neighbours were 
the Nogais, of whom the remnant (50,000 in number) 
now inhabit Astrakhan and a few districts between the 
Caspian and the Black Sea, near Azof, in the Crimea, 
and towards the Caucasus. Nogaic, with its Cauca- 
sian dialect Koumouk, is the language of the Russian 
Tatars. Better known and cultivated is Ouigour, 
with its Djataic and Turcoman varieties; it boasts a 
literature which dates from the fifth century of the 
Christian era; it has been recently studied by Pavet 
de Courteille and M. Barbier de Meynard. The Bib- 
liothfeque Nationale has a manuscript with illuminations 
in this language, which is of great value. Finally, the 
most celebrated, and from some points of view the 
most perfect, of the Turkish idioms, Osmanli or Otto- 
man, originally from Khorassan, carried by the Sel- 
jukian bands into Asia Minor, and by the heirs of 
Othman to Constatitinople, Cairo, Tripoli, and Tunis, 
is the language of about thirty millions of people, who 
inhabit ancient Bactriana, Media, Asia Minor, Thracia, 



Agglutinative Idioms of Central Asia. 105 

and some of the Greek islands. Osmanli, which has 
gathered on its route a great number of Persian and 
Arabic words, is making some effort to return to its 
native purity, which is happily preserved among the 
Oriental Turks, and, even in Europe, in the speech of 
the populace. It is, as we have already said, a very 
attractive language, from the harmony of its vowels, 
the wealth of its verbal categories, and the regularity 
of its grammar. 

But even the Turkish branch hardly equals in abun- 
dance and in interest the Finnish or Finno-Hungarian 
family, which can boast of two literatures, valuable on 
more than one count — the Suomi literature and the 
Magyar literature. Suomi is the language of Finland ; 
Magyar is the idiom of Hungary. The latter is the 
more fortunate brother of the Ostiac (20,000) and 
Vogoul (7000) dialects of Siberia ; the former, primics 
inter pares, is the type of the Finnish peoples which 
extend westward from the Obi and the Ural : Votiacs 
(200,000), Zyrienes (80,000), Permians (60,000) ; 
Finno-Lapps, Finlanders of the Volga, Mordvines 
(700,000), Tcheremisses (200,000), confused with the 
Tchouvaches and the Nogais ; Karelians, scattered 
from the White Sea to the Lake of Ladoga ; Suomia 
(to the number of 2,000,000) in the greater part of 
Finland ; li'choudes, Vepses, and Votes, round Lake 
Onega ; Cfevines in Courland ;. Estes on the southern 
coast of the Gulf of Finland (Revel, Dorpat) ; finally, 
Livonians, reduced to a few square miles by the pres- 
sure of Lithuanians, Germans, and Russians. The 
Finnish languages are spoken by about 3,300,000 
people ; the Hungarian by perhaps 6,000,000 ; but 
they have evidently covered an immense extent of 
territory. 



io6 Distribution of Languages and Races. 

The primitive union of the Pinno-Hungarian group 
admits of no doubt; here, as in each of the other 
divisions of the Uralo-Altaic family, we find words 
which testify to an original form common to all the 
subdivisions of that family ; fish is Ttala in Suomi, 
guolle in Lapp, Iml in Mordvin, hvl in Vogoul, hal in 
Magyar. Hand is Teat in Vogoul, Mfe, Ixt, or litd in 
Suomi, Lapp, Tcheremisse, and Ostiac, lidsi, cdiz, kez, 
in Vepse, Este, Livonian, and Magyar. The slight 
differences of sound which distinguish these variants 
perhaps indicate the method of procedure in inflected 
languages ; born of the encounter of several dialects, 
they may have made use of the variants which thus 
occurred. 

Another proof of the unity of the Finnish group is 
found in the collection in one poem (as was doubtless 
the case with the Iliad) of episodes gathered by Lonn- 
rot, not in Finland only, but throughout the north and 
east of Eussia. This mosaic constitutes, in truth, the 
epic of a race ; it relates the exploits accomplished by 
the heroes of Kaleva against the Magicians and the 
monsters of Pohja, that is to say, no doubt, the strife 
of invaders from the East with either the inoffensive 
Lapps who had preceded them, or against the savage 
aborigines, those Fenni, destitute of laws, of chiefs, and 
even of gods, of whom Tacitus had heard, whose name 
the conquering Suomis took together with their ter- 
ritory. 

The Magyar literature is richer, more European, and 
more ancient, but less original than the legendary 
cycle of the Suomi Cantelar and Kaleva. The two 
languages are of equal merit. Suomi loves to multiply 
its vowels ; Magyar makes a greater use of contrac- 
tions. Both are remarkable for the richness of their 



Agglutinative Idioms of Central Asia. 1 07 

conjugation ; they surpass even Turkish in this respect. 
All the Finnish dialects can incorporate the accusative 
of the third person into the verb : I see him, I touch 
him, is said in a single word. Magyar and Vogoul 
incorporate the pronoun of the second person : I love 
thee, he loves thee. Mordvin does the same with the 
pronoun of the first person. Basque goes yet farther, 
and engulfs even the dative with the verb : I give it 
thee. These expedients are not to be envied, and may 
cause inconvenient pleonasms ; but they imply a cer- 
tain ingenuity in the peoples which have not got 
beyond agglutination. 

Can the relationship which is traceable between the 
dialects of each branch of the Altaic family be shown 
to exist between the five branches ? Not in the pre- 
sent state of our knowledge. Nevertheless, it is pro- 
bable that races which are as near neighbours and as 
mixed as the Tongouses, the Bouriates, the Yakoutes, 
the Samoyeds, and the Vogouls have spoken kindred 
dialects. But the similarities which it is as yet pre- 
mature to seek in their vocabularies appear numerous 
and unmistakable in their syntax and their methods of 
suffixing. It is especially curious to note in almost all 
the members of the family (except the Samoyeds) a 
tendency which has become more and more marked as 
the development of the intelligence demanded greater 
order and precision. It is difiBcult not to suppose 
that when the same phenomenon, vowel harmony, 
manifests itself at once, separately, in thirty different 
languages, all originating in the same region, but since 
scattered in various quarters — it is, I say, difficult not 
to suppose that these languages , have received from a 
common original this latent disposition, which only 
becomes manifest at a certain stage of growth, like 



1 08 Distribution of Languages and Races. 

those resemblances to some ancestor which may be 
nnperceived in the children of a family, and become 
evident as they grow up to manhood. 

Vowel harmony is a means of marking the subordi- 
nation of the suffix to the root ; its principle is that 
the vowel of the suffix should reflect the vowel of the 
root ; that the root se,v (Turkish), love, should have fop 
the infinitive suffix mek, and the root ha, look, for 
infinitive suffix mak ; at, horse, makes atlar in the 
plural ; ev, house, is evler in the plural. The Uralo- 
Altaic vowels, being divided into two classes, open and 
shut vowels, it follows that to an open root - vowel 
corresponds an open vowel in the suffix, and vice versd. 
Certain languages have a third order of vowels, neuters, 
which can also harmonise with the open vowel of the 
root. There are differences in the application of this 
law, which is strict or lax in proportion to the degree 
of cultivation to which the language has attained ; but, 
broadly speaking, the law has obtained for six or seven 
centuries in Mandchu, Bouriate, Mongolian, Turkish, 
Zyriene, Mordvin, Magyar, and Suomi. 

After having defined the narrow but capital distinction 
which separates agglutination from monosyllabism — - 
that is, the change in the suffix — or empty root attached 
to the unalterable root syllable, we have considered 
three sorts of agglutinative idioms : I. the isolated 
languages, of which the vocabulary is without relation 
to any other language, Corean, Japanese, or Yamato, 
arrested in its development by the Chinese civilisation 
and written character ; 2. the poor and remote dialects 
of North-Eastern Asia, Kourelien or Ainu, Ghilialc, 
Kamtchadale, Koriak, Toukaghir ; 3. a vast family 
connected at least by a grammatical relationship, 
the Uralo- Altaic family, of which the five branches, 



Agglutinative Idioms of Central Asia. 1 09 

Samoyed, Tongouse, Mandchu, Boriate-Mongol, Turkish, 
Finno - Hungarian, are all subdivided into numerous 
varieties, which may be respectively referred to a com- 
mon type, living or extinct. A few of these languages, 
Mandchu, Mongolian, Ouigour, Turkish, Magyar, and 
Suomi, have been the expression of literatures more 
or less rich, which are often interesting, and worthy of 
the part played in the world by tine peoples which 
speak them. 



CHAPTER III. 

THE AGGLUTINATIVE IDIOMS OF 
SOUTHERN ASIA. 

The Caucasian languages : Toherkesse group ; Kartvelien or Georgian 
group — The language of the Shumirs or Accadians — Brahui dialect 
• — Non-Aryan India : Kol-Aryan group (Djuangs, Birhors, Korvas, 
Moundas, Hos, Kharrias, Sonthals) ; Dravidian group — Dravi- 
dians of the North : Oraons, Paharyas, Gonds, Khonds — Dravi- 
dians of the Dekkan : dialects spoken by fifty millions of people : 
Tulu, Eanara, Tamil, Malayala, Telinga — Dravidian phonetics 
and literature. 

The violent and tardy incursions of the Uralo-Altaic 
peoples have led us far into Europe, and we must now 
return upon our steps to complete the chart of the 
agglutinative languages of Asia. Let us press along 
the northern coast of the Black Sea, where we have 
found more than one Tatar or Mongolian group, and 
re-enter Asia by the gorges of the Caucasus. It is a 
strange region, both from the place which it occupies 
in ancient tradition and from the inextricable mixture 
of the tribes which inhabit it. This region has had 
the honour of bestowing its name, of unknown origin, 
upon the whole white race. It contains the mountain 
on which, according to Jerome, the Ark of the Deluge 
was stayed, Ararat, and the summit on which the ven- 
geance of Zeus bound Prometheus, the ravisher of fire ; 
and finally, the highest northern summit of the great 
chain. Mount Elbruz, as well as the Persian Elbourz 
to the south of the Caspian, still bears the name of the 



Agglutinative Idioms of Southern Asia. 1 1 1 

legendary holy mountain known to the Persians under 
the name of Hara - Barazaiti, and to the Greeks as 
Berecynth. 

The ancient traditions collected in the Bible have 
retained for us the former names of the Tuplai or 
Tibarenians, of the Muskai or Moschians (inhabitants 
of Colchis, Georgia, during the Assyrian and Persian 
period), Tubal and Meshech, sons of Japhet, whom the 
Jewish sometimes associate with Gomer (the Cimme- 
rians), and Togarma, " who comes from the north wind 
with all his troops." From the information furnished 
by Herodotus, by Hecateus, and by the cuneiform 
inscriptions, we gather that the ancients had a very 
clear idea of the inhabitants of Armenia, who were 
gradually driven back towards the southern slopes of 
the Caucasus, and a slighter acquaintance with the 
peoples of the other side, Scythians and Cimmerians, 
who had, however, more than once invaded and dis- 
turbed Asia. 

After having been a refuge for more or less com- 
pact groups of ancient peoples, driven out and broken 
up by better armed races, the Caucasus became a pas- 
sage, at least on its eastern and western borders, not 
only to the Scythians, those multitudes of unknown 
race, doubtless of very mixed blood, who overthrew the 
first Chaldean empire, and drove the Hyksos or Shep- 
herds on to the Egypt of the Pharaohs, but also for 
the vanguard of the Hellenes, the lonians of Lydia 
and Phrygia, who transmitted to their descendants, the 
fabled Argonauts, a vivid recollection of Colchis ; and 
also probably for the future Armenians, who came and 
settled precisely within the borders of the ancient 
Alarodian or Georgian race at Van, near the great 
lake, near the tri- lingual inscriptions, of which a 



1 1 2 Distribution of Languages and Races. 

column may perhaps enlighten us as to the early forms 
of Georgian. Strabo counted in Caucasia seventy 
peoples and seventy dialects. The Eomans maintained 
as many as one hundred and thirty interpreters on the 
frontier at Sebastopol. Abonlfeda called the Cau- 
casus the mountain of languages. Many of these 
languages are in process of extinction, and the com- 
parative study of them becomes every day more diffi- 
cult. Yet, if we are guided by the information collected 
by Klaproth, by Baron TJslar, and by the Kussian 
Academician Schifner, the classification will not be 
very complicated. But we must first carefully exclude 
the Armenian of the banks of the Araxes, Ossetan, an 
Iranian idiom which has taken refuge in a central 
district to the south-east of Elbruz, modern Persian, 
and the Tatar or Turkish of Aderbaidjan, which are 
all spoken on the sonth-weEtern coasts of the Caspian, 
and also Nogai and Koumouk, which are found at 
different parts of the northern basin. We thus isolate 
the Caucasian group properly so called, represented to 
tlie north of the Caucasus by the Abazes, the Tcher- 
kesses, the Kistes, the Tchetchenes, and the Lesghians 
from the Black Sea to Daghestan ; to the south of the 
Caucasus, by the Imerethians, Mingrelians, and Lazes, 
by the Georgians and Suanians, between the Black 
Sea and the middle basin of the Cyrus and of the 
A rax (now Koura and Aras). On the maps and in 
the geography of Eeclus will be found the names of 
numerous tribes often very interesting from some 
characteristic custom, some ancient belief, from the 
beauty of their type, or from their courageous resistance 
to the Russian dominion. But from the linguistic 
point of view they probably belong to one or other of 
the two divisions which we mentioned above. 



AgglMtmative Idioms of Southern Asia, 1 1 3 

It is doubtful whether the whole of the northern 
or Circassian group has a common origin; it has been 
so disorganised, so nearly obliterated, by the Russian 
conquest, that I doubt if it now comprises a million 
individuals. The Tcherkesse nation, which was Mussul- 
man, has almost all dispersed, and has been replaced 
by Slavs and Germans. A few Tcherkesse legends 
have been collected ; the language is hard, remarkable 
for certain sounds which are peculiar to it, and for 
the incorporation of the suffixes of number. 

The southern or Kartvelian group, early converted 
to Christianity, remains intact though not independent, 
to the number of one or two millions in the neighbour- 
hood of Koutais and Tiflis. It corresponds geographi- 
cally to the Colchis and Iberia of the ancients. Its 
principal dialect, Georgian, has an alphabet. Cultivated 
in the Middle Ages, it belongs, like Circassian, to the 
agglutinative class. Itwas probablyakin to the language 
of the Aghovanik or Albanians, which disappeared 
completely in the fifteenth century, leaving no traces 
in writing of its existence. The Georgian chronicles 
have been translated into French by M. Brosset. The 
names Iberians and Albanians, Georgians, Suanians, and 
Kartvelians require some explanation. The two first, 
which must not be confounded with the Albanians of 
Bpirus and the Iberians of Spain, are somewhat ancient. 
Albanian — Alwank in Armenian — is mentioned in the 
time of Alexander. Iberian, through the forms Wirg in 
Armenian, Amr in Pehlevi, A^etpes in Greek, goes back 
to a form ^a^etpoi, ^aa-iripei, given by Herodotus. The 
Saspires made part of the army of Xerxes. Georgian 
comes from the name of the saint chosen for patron 
by the Iberians. Kartvelian, Kartouli, is really a 
national name ; Karthlos, the eponymous hero of the 



1 1 4 DistriBution of Languages and Races. 

race, was the son of Thargamos (the Togarma of the 
Bible), son of Japhet. 

Whence came these languages, which it is rash to 
class together in one Caucasian family, and of which 
the vocabulary forbids any attempt to bring them 
into relations with the other agglutinative idioms ? 
Whence came these peoples, this handsome race, similar 
in feature to the Iranian type, who were established 
in the neighbourhood of the Caucasus long before 
the development of the Assyrian Semites, before 
the arrival of the first Indo-European migrations ? 
These questions, like many others, must remain un- 
answered. I have sometimes thought that they were 
pre-Aryans — that is to say, a white race akin to those 
who wandered on the other side of the Caspian, on 
the banks of. the Jaxartes and the Oxus, and separated 
from these before the appearance, at first quite local, 
of inflexion and the Indo-European mother-tongue ; 
they would have remained at the agglutinative stage, 
protected by their mountains from the influence of a 
more advanced linguistic system. 

M. Lenormant connected them rather with the 
ancient inha,bitants of Mesopotamia and Chaldea, not 
by race however, but by language. In his view, the 
inscriptions deciphered with great difficulty at Van, 
which belong undoubtedly to the pre- Armenian tongue 
of the peoples of Mount Ararat, Urarti, or Alarodians, 
might serve as connecting link between the Georgian 
dialects, the Caucasian, and the more ancient idioms 
of Babylonia^ 

At the present day the obscurity which hung over 
the origins of Chaldea has been, if not dissipated, at 
least considerably diminished, thanks to the discoveries 
of those great cuneiform scholars Rawlinson and Opperfc 



Agglutinative Idioms of Southern Asia. 115 

An insight which is truly marvellous has been able 
to reconstruct, not altogether without gaps, but from 
authentic documents, the military, social, and intel- 
lectual history of the valleys of the Tigris and the 
Euphrates, from at least twenty centuries before our 
era. Names and dates have thus been recovered 
which had been much altered by the Hebrew writers, 
who were, nevertheless, so nearly akin to the Baby- 
lonians both in race and language. It has been found 
possible to separate the personal observations re- 
corded by Herodotus from the fables which the 
credulous historian set down on the faith of ignorant 
or duped interpreters — the stories, for instance, of Belus, 
Ninus, and Semiramis. Finally, side by side with 
a Semitic dialect which belongs to the central branch, 
midway between Aramaean and Arab, a language 
which abounds under the chisel of the scribes of 
Sennacherib and of Assourbanipal, MM. Oppert, Lenor- 
mant, and Schmidt think they have discovered, and 
established beyond a doubt, in spite of the strenuous 
opposition of M. J. Hal^vy, the presence of another 
language, anterior to the idiom of the Semitic con- 
querors, and so vigorous that it was long the official 
language of the kings of Babylon and Nineveh, aad 
that it still may be found on many inscriptions, over 
against the Assyrian text, in the manner of transla- 
tion or commentary. Some scholars have denied the 
existence of this language, which is markedly agglu- 
tinative, and of which several philologists have written 
the grammar ; it has been represented as an error in 
the deciphering, as a form, either archaic or symbolical, 
hieratic, so to speak, of ordinary Assyrian. I am not 
qualified to take a side in this debate, but whatever 
may be the truth about this second language found on 



1 1 6 Distribution of Languages and Races. 

the Chaldean inscriptions, whether we should consider 
it, with M. HaMvy, to be cryptographic, Shumirian with 
M. Oppert, or Accadian with Lenormant, there is one 
point which admits of no doubt whatever, and that is 
the existence of the peoples to whom it is attributed. 
The Shumirs and the Accadians mentioned in the 
Bible are invariably mentioned in the official formulary: 
king of the Shumir and of the Accadians is a constant 
title of the Assyrian monarchs. In Blum, in Chaldea, 
in Babylonia, they form the bulk of the population ; 
we must, therefore, recognise in them the predecessors 
and the educators of the Kaldi (the Kasdim of the 
Bible), of the Kissi or Kossei or Kushites (whom 
M. Maspero identifies with the Oriental Ethiopians 
of Homer) ; finally, the Canaanites, Aramaeans, and 
Assyrians, all Semites, and speaking Semitic dialects. 
The antiquity of the Accado- Shumirian settlements 
is clearly demonstrated by the flint implements, arrow- 
heads, axes, and hammers found in their burying-places, 
together with utensils of bronze and ornaments of gold 
and iron. To them may be attributed a considerable 
share in the invention of the cosmogonies, the obscene 
forms of worship, and the talismanic superstitions which 
are so widely spread in the Bast ; and it is also from 
them that the Semites received the deplorable cunei- 
form character, afterwards adopted by the Hittites of 
Syria, by the Cypriotes, by the Armenians, and by 
the Persians. 

The cuneiform character, which seems to be com- 
posed of wedges, nails, and arrow-heads, results from 
the alteration and abbreviation of imitative figures. 
Its use, wonderful to relate, was prolonged as late as 
the first century of our era. " Some of these signs," 
says M. Maspero, " are true ideograms, which are not 



Agglutinative Idioms of Southern Asia. 1 1 7 

always pronounced, and merely indicate the general 
sense; the greater number represent syllables, some- 
times simple, composed of a vowel and a consonant, or 
vice, versd; sometimes complex, formed of several con- 
sonants." The complex syllables may be written in 
two ways '. (i .) By decomposing them so as to form two 
simple syllables, of which the second always begins, in 
pronunciation, with the vowel of the first : thus the 
word napsat, soul, may be written na-ap-sa-at. (2.) By 
means of a special cliaracter answering to each syllable : 
nap-sat. Nabu-Kudur-Ussur may be spelt as written 
here, or thus : Jffa-bi-uv-ku-du-ur-ri-u-ts'u-ur. More- 
over, most of the signs may express several different 
sounds. Chinese and Egyptian have not imagined a 
more clumsy method. The decipherers have needed 
a hundred years to overcome the difficulties presented 
by the riddle of these inscriptions, to recognise the five 
or six extinct languages which have used the cunei- 
form character. It is but justice to recall here the 
names of the principal savants who have devoted them- 
selves to this task, and have carried it successfully 
through: Niebuhr, 1765; Tychsen, 1798; Munter, 
1800; Gri'otefend, 1802; Eugene Bournouf and Las- 
sen, 1836; and more recently Rawlinson, Hi neks, 
Fox, Talbot, Lenormant, and Oppert. Thanks to 
these last, who are the creators of Assyrian science, 
thirty centuries of history have arisen in less than 
thirty years from out of the ruined tombs. After 
having deciphered the Babylonian, Ninevite, and 
Median texts, they have discovered the remains of 
the ancient Ohaldeo-Shumirian literature. 

The Shumirs and the Accadians have mingled with 
their Iranian and Semitic successors. Their name 
was hardly known, yet now we are led to regard thera 
9 



1 1 8 Distribution of Languages and Races. 

as holding a high rank among the industrial and 
religious teachers of the world. To what race did 
they belong ? Doubtless to an autochthonous people, 
dark of skin, intermediary between the Mongol and 
the Malay, between the blacks' of the east and the 
different varieties of Caucasian whites. Their lan- 
guage, recently discovered (if it be discovered), has been, 
somewhat prematurely, connected on the one hand 
with the extinct idioms of Mount Ararat and the yet 
living dialects of Georgia, and oh the other hand with 
the Dravidian family. We will be content with noting 
the agglutinative character, which implies no original 
relationship. 

On the route to India, where we shall find isolated or 
compact groups- of languages belonging to this immense 
class, we must stay a moment to consider some islets 
of peoples lost in the obscure chaos of Beluchistan, 
from Kej, Panjgur, and Palk, as far as the limits of 
Seistan. These are the Brahui or Birrhui, whose 
patois, though much mixed with Hindi and Persian, 
shows some Dravidian affinities. Its grammar, in any 
case, is agglutinative and very simple : no gender ; 
two numbers, singular and plural ; no relative pro- 
noun ; the adjective," which is invariable, precedes the 
substantive ; the verb, neuter or active, can take causa- 
tive and negative forms, admits but a single mood 
and three tenses, past, present, and future past. All 
derivation is by the aid of suffixes. 

India, which we now enter for the first time, is a 
world in itself ; it measures twenty-six by twenty-three 
geographical degrees, and contains a population equal 
to about two-thirds of the population of Europe, more 
than two hundred and fifty million inhabitants, of 
every colour and every race. By the western frontier, 



Aggbitinative Idioms of Southern Asia. 119 

the basin of the Indus, the Aryan groups, relatively- 
few in number, but possessed of a very superior lan- 
guage and cultivation, descended slowly towards the 
affluents of the Ganges between the fifteenth and the 
tenth centuries before our era, and thence spread in 
every direction, north and eastward towards the Hima- 
layas and Indo-Oliina, to the south along the coasts, 
as far as Cape Comorin and the vast island of Ceylon. 
So great was their preponderance, that they have left 
an ineifaceable impress over this immense land ; neither 
internal wars nor invasion, nor durable conquest, has 
seriously affected the social organisation or the fanati- 
cal and scrupulous devotion of the Hindus educated 
by the Brahmans. But the number of the Aryans, 
of the white race, was too small to have any material 
influence on the blood of the multitude, or rather on 
the chaos of indigenous races. The meshes of the 
political, social, and religious net were never close 
enough to prevent all escape for the refractory groups, 
customs, and beliefs ; and even in regions which felt 
the Aryan influence most strongly, the expansive power 
of Sanscrit and its derivatives proved of no avail 
against the passive resistance of great masses of the 
population, who kept tbeir ancient languages, while 
using them to express the ideas which they learned 
from their conquerors. 

Thus, without counting the Europeans, the Jews, 
Parsees, and foreign Mussulmans, there are in India 
numerous barbarous or savage tribes which are un- 
touched by Brahmanism, tribes all the more precious 
to science that their manners and their langnages are 
a survival from pre-historic times. They have been 
called Kol-Aryans. Besides these tribes, which are 
chiefly found on the Ooromandel Coast and in the cen- 



1 20 Distribution of Languages and Races. 

tral plateau or Ghondvana, millions of civilised men, who 
occupy the lower end of the Peninsula of Hindustan, 
between the Vindhyas and Cape Comorin, retain and 
cultivate their national dialects. These are the Dra- 
vidas or Dravidians, a race of very tnixed blood, whom 
certain ethnographers consider to be the resultant of 
two invasions, Thibetan and Uralo- Altaic, operating on 
an indigenous or, at any rate, a more ancient race. 

Some separate and some class together the Kol- 
Aryans and the Dravidians in regard to race and lan- 
guage; it would seen that the latest opinion of science 
inclines to separation, and forbids us to call the first 
Dravidians of the north, or even Proto- Dravidians. It 
is true that the name invented for them by Mr. George 
Campbell in 1866 is hardly more suitable. Kol might 
be allowed, since it is the name of one of the tribes in 
question ; but Aryan is misleading, because one of the 
characteristics of these peoples is that they have not 
been Aryanised. 

The Djuangs are the most savage ; their solemn oath 
is made upon an ant-heap or upon a tiger-skin. Little, 
naked, tattooed, red-brown in colour, bowmen or slingers, 
these poor wretches can neither spin nor weave, are 
ignorant of the potter's art and of the use of metals. The 
Birhors of the district of Hazaribagh and the Korvas 
of Chota Nagpur dispute the lowest place with them ; 
little, dark, tattooed, they live in the forest and build 
huts on steep rocks. The neighbours of these last 
Kols or Mundas, Hos and Bhumidjs (the name of the 
Bhumidjs seems to be Aryan ; it comes from bhumi, 
the earth, and is perhaps akin to the Latin homo for 
Jiumo), all these Kols form a total of about a million, 
some strong, thick-set, and chocolate-coloured, others 
tall, copper- coloured, with long coarse hair. They 



Agglutinative Idioms of Southern Asia. 121 

neither spin nor weave, but can work in metals. They 
vow themselves to the tiger should they come short of 
their oath (which they nevertheless forget very readily). 
Some of their superstitions are curious ; if the shadow 
of a passer-by cross their food, they will not eat it, 
but throw it away; an evil spirit is in it.. The 
Kharrias, the Kurs, and especially the twelve tribes 
of the Sonthals, appear to be less unapproachable. 
They have houses ; some of them are cultivators, 
others readily leave their homes and enter service. 
The Sonthals number about a million, are smaller than 
the Aryan Hindus, have a round face, straight eyes, 
hair black and thick, snub nose, and large mouth. 
They are fond of music and dancing, bamboo flutes, 
rings, necklets, bracelets, and fine clothes. Their 
manners are not austere. 

The only written documents in the Kol dialects are 
some partial translations of the Bible into Mounda 
and Sonthal, and a few legends or songs collected by 
the curious. These present some interesting pecu- 
liarities, from the phonetic and grammatical point of 
view. They are very rich in vowels, and in addition 
to the spirants, palatals, and explosive consonants, 
they have other sounds difficult to define and imitate, 
which seem to be introduced into the Sanscrit alphabet 
under the name of cerebral or lingual letters. In the 
body of a word the consonants are separated from each 
other by supporting vowels, long, short, or even neuter, 
like the French e mute. Derivation is by sufBxes and 
infixes : dal, the action of beating ; da-pa-l, cushion ; 
da-na-pal, covering. The genders are not distin- 
guished, but the number has four or five forms — 
singular, plural, dual, the plural particular, and the 
plural .general : ain, I; abon, we all; cda, we others; 



122 Distribution of Languages and Races. 

alin and aldn, we two, &c. The declension is very- 
full — genitive, dative, ablative, instrumental, and 
locative. 

The pronouns are personal, demonstrative, interro- 
gative. Placed after the noun, the personal pronouns 
give a possessive sense : apu, father ; apu-ling, our 
father ; apu-pe,, your father ; hopon, son ; hopon-in, 
my son. The relative pronoun is wanting ; the ad- 
jective, which is invariable, precedes the substantive. 
The verb, properly speaking, does not exist, since 
suffixes of person, place, and time can convert every 
noun and every adjective into verbal expressions. 
This is the case with all agglutinative languages, but 
the number of possible combinations and the use of 
auxiliaries place the Kol dialects on the same level as 
Turkish and Finnish, or even higher ; for they can not 
only make of their pseudo-verb a preposition which 
incorporates the direct and indirect objects ; not only 
can they make it active, passive, middle, causative, 
intensive, &c., but they have six or seven moods — 
indicative, imperative, potential, conditional, infinitive, 
gerundive, and participle ; and as many tenses — three 
presents, a preterite, an imperfect, and a pluperfect. 
Many of these methods are found also in Basque and 
the American dialects, and are wanting in Dravidian. 
It is not rare to find among uncivilised peoples a 
linguistic faculty superior to that of their neighbours 
or of their civilised kindred, but it is often difficult 
to explain this apparent anomaly. Here a probable 
solution has been found. The Kol-Aryans are the 
remnant of a fallen people who were at the time of 
the Brahmanic invasion at the head of the races of 
India. Though crushed and destroyed by the Aryans, 
they were yet powerful enough to modify and enrich 



Agglutinative Idioms of Southern Asia. 123 

tlie Sanscrit pronunciation ; their influence is still to 
be discerned in the use, confined to India, of the 
so-called cerebral consonants, and perhaps in the 
complexity of the Sanscrit conjugation. 

The races to which the Aryans give the name of 
Drayidas stop short, on the contrary, at the first stages 
of the evolution of agglutination. Their northern and 
central groups, Oraons of Bengal, Ghonds of Ghond- 
vana, Khonds of Orissa, &c., have remained in their 
primitive condition, and the great masses of the 
Dekkan, fifty millions, while they accept with docility 
the education, ideas, and beliefs of their conquerors, 
have yet kept, and very cleverly utilised, the poor 
organisation of their rudimentary languages. 

It is probable that the Dravidians of the north and 
centre are nearer to the primitive type than their Aryan- 
ised kindred of the Dekkan. The Oraons of Bengal, num- 
bering about 600,000, say themselves that they come 
from the west, but they have nothing of the Turanian 
or Mongol ; they have low and narrow foreheads, curly 
hair, eyes large and well opened, long eyebrows, pro- 
minent teeth and jaws; their colour is dark-brown 
and their body well proportioned. They are fond of 
copper ornaments, and load their heads, necks, and 
arras with them. They often intoxicate themselves 
with a spirit distilled from rice. They have their 
dances, their banners, their feasts, their tribal gods, 
and their thousand superstitions and rites common all 
over the earth to all races whose creed is a vague 
animism. The Oraons live with their animals in miser- 
able huts. In the villages where the ancient customs 
are preserved, two exactly contrary to each other may 
be noted ; in one tribe the unmarried of both sexes 
sleep under the same roof, in another the young men 



124 Distribution of Languages and Races. 

pass the night in a special cabin under the guard of 
an old man, the girls being under the charge of the 
elder widows. The primitive Oraons hesitated between 
promiscuity and decency, their descendants have not 
yet made their choice. 

From the banks of the Ganges to those of the 
Brahmaputra, the Eajmali^ls, Meiers, or Pah§,ryas 
(mountaineers), 400,000, build bouses of bamboo 
surrounded with gardens and orchards ; they also like 
strong waters made from rice and sorghum. They 
practise divination and have an animistic creed. The 
Aryans took from them the doctrine of metempsy- 
chosis. They are polygarnists ; they bury their dead. 
They differ from the Oraons by their oval face, their 
thick lips, and their long hair knotted up on the 
head. There are Ghonds (50,000) in Bengal, but 
their principal habitat is Ghondvana, a dangerous cen- 
tral district. Their twenty tribes or castes are among 
the most savage of India ; they are half naked, they 
shave their heads, their weapons are an axe and a 
pike ; they set fire to the forests to sow their crops, 
aud poison the waters to obtain fish. These are the 
ogres or Ratchas of the Brahmanic legends. There is 
a natural confusion between these Ghonds with flat 
face, with thick black hair, smooth or slightly waved, 
very da.rk skin, and fragile lower limbs, and the Khonds, 
who are smaller, but equally dark ; they live in the 
south of Bengal, on the coast of Orissa, are full of 
sanguinary superstitions, and practise human sacrifice. 

These diSerent tribes speak dialects which are akin 
to the Dravidian languages. But the Dravidians 
proper, whom we shall find in Mysore and in the 
Dekkan, who are, moreover, very much crossed witli 
Aryan blood, have retained nothing of the savagery 



Agglutinative Idioms of Southern Asia. 125 

of their congeners. They are a civilised people, who 
have their cities, their monuments, their industries, 
and their literature ; they were the first to enter into 
relation with the Europeans, with the Portuguese and 
Dutch ; and the little which remains to the French of 
the empire which Dupleix and Labourdonnais souglit 
to found is all, with the exception of Chandernagor, 
situated on their coasts. Sixty thousand Dravido- 
French electors send a deputy and a senator to the 
French Assembly. 

One of our most eminent philologists, M. Julien 
Vinson, was brought up among them, and writes and 
speaks their language as he does Basque and French. 
We can have no surer guide, and we will take from 
him our account of the history, domain, languages, and 
literature of the Dravidians. 

" The existence of the Dravidian tongues is proved 
by history from very early times. Tamil words, geogra- 
phical names, Sangara, Pandion, Madoura, occur in 
Ptolemy, Strabo, Pliny, and in Arrian in his ' Circum- 
navigation of the Red Sea.' A Sanscrit writer of the 
seventh century, KumS,rilabhatta, quotes a few common 
Tamil words, nader, step ; pdmb, snake. The name 
given to the peacock, mentioned in the Book of Kings 
as among the birds brought from Ophir to Solomon, 
thuJd, togei, is believed to be Tamil. From the time 
of the arrival of the Portuguese at Goa, the Jesuits, for 
the purposes of their propaganda, studied the native 
idioms. Towards 1550 they were teaching Malaysia 
and Tamil in their seminary at Ambalakkadu, near 
Cochin. In 1577 they published a Dodrina Chris- 
tiana in Malaysia by means of characters engraved on 
wood by a lay brother of their order ; in 1578 they 
printed in Tamil a book of devotions." 



126 Distribution of Languages and Races. 

It was through the Dravidians that these missionaries 
studied the manners and the religions of India and 
the Sanscrit language. It was natural that, later, 
when the French Jesuit Cceurdoux and the first Eng- 
lish rulers had pointed out the relations of Sanscrit 
with Latin and Greek, the southern languages and 
traditions should be neglected. Interest centred on 
the Ganges and the Indus. Following the general 
tendency of their day, and misled besides by the con- 
siderable number of words borrowed from Sanscrit by 
the Dravidian vocabulary, Indian scholars (Carey in 
1 8 1 4) treated Tamil and its congeners as derived, 
as descended from Sanscrit. This error is disproved 
by the Indians themselves. More accurate ideas pre- 
vailed, and in 1 8 1 6 Ellis first aflBrmed the original 
independence of Tamil, Kanara, and Telinga, which is 
now universally admitted. 

There is no doubt that the Dravidas, whom Vinson 
thinks identical with the ancient Parias, about whom 
so many fables are told, once occupied a much more 
extended area, beside the Kol- Aryans. Their present 
domain is more extensive than Italy, Prance, or Spain. 
It stretches from the tropic of Cancer to Cape Comorin, 
and into the northern half of Ceylon. There are five 
principal dialects : in the north-west, in the upper 
valley of the river Krishna, Kanara, Kanada, Kama- 
taka, is spoken by nine millions of people ; to the 
north-east Temougou, Telougou, Telinga, by fifteen 
millions, of whom 5000 inhabit the French settlement 
of Yamaon ; Telinga is spoken in the lower and middle 
basin of the Godavery and of the Krishna, and on the 
Coromandel coast. It is a language which has been 
tnuch modified, and is very soft and agreeable; it has 
been called the Italian of the Dekkan, and it is near 



Agglutinative Idioms of Southern Asia. 127 

neighbour to the most archaic dialect, Kanara. On the 
east coast, and in the interior of the country, between 
Lake Pulicat, Bangalore, and Trivanderam, in the pro- 
vinces of Madras, Tanjore, and Travancore, in the 
French towns of Pondicherry and Karikal, fifteen mil- 
lions of men speak Tamil ; to the west, about Cochin 
and Cananore, and in the settlement of Mah4 Malaysia 
or Maleolum is the speech of three. and a half millions, 
separated from Kanara on the east by the Nilghiris, 
where the dialect Toda shelters itself, and towards the 
north by Tulu and Kudangu. 

Two or three slight indications seem to point to an 
original unity of these languages, and even to show 
that this unity was prolonged to a comparatively recent 
date. The name Kanara or Karanata has been given 
to the Tamil side, the Carnatic, and Tamil is often 
called Malabar by the earliest European visitors ; now 
at the present day it only occupies the extreme south 
of the Malabar coast, the rest belonging to Malay§,la. 
The Indians of Malacca and Singapore are called Kling 
— that is, Telinga ; they are, however, Tamils. The fact 
is, that the separation, now very marked, of the Dra- 
vidian idioms disappears as we approach the ancient 
forms in Kanara and Tamil, and as we recognise in 
Malaysia a derivative, a corruption of Tamil, and in 
Tulu, Kudagu, and Toda intermediaries between Tamil 
and Kanara. Telinga, which is the most altered of 
all, is also a descendant of Tamil. Tamil, in short, 
from the richness of its vocabulary, and from the 
priority of its culture, holds in the Dravidian group 
the same rank as Sanscrit among Indo-European lan- 
guages. It is also, like the people which speaks it, 
the only idiom of the group which retains any vitality, 
a certain power of expansion. Tamil has almost taken 



128 Distribution of Languages and Races. 

the place in the north of Ceylon of Pali, an Aryan 
dialect. 

The Dravidian pronunciation is soft; it has no 
aspirates ; moreover, it grows weaker and more unde- 
cided every day. Many sounds which exist in the 
spoken Tamil language do not appear in its alphaljet, 
and where we read a pure vowel, such as a, e, % u, or 
a distinct diphthong like ai, we hear something unde- 
cided and muffled. The true consonants seem also to 
have been originally few in number, since the Tamil 
written character does not distinguish between &, g, 
and d, between t, p, and h. On the contrary, the 
trills, nasals, palatals, and sibilants have always existed 
in abundance, and have communicated to the explo- 
sives a species of hesitation and uncertainty which it is 
difficult to reproduce; the dentals especially are affected 
by it. Finally, the lingual consonants are found in the 
Dravidian as in the Mounda and Sonthal languages ; 
they result from what we should call a defect of pro- 
nunciation, the inopportune contact of the tongue 
either with the teeth or with the palate; the con- 
sonant is not clearly given, it is strangled, unfinished, 
like I and r in pickle and lord, and in the Provenpal 
chival. We have said above that Sanscrit adopted 
these incomplete sounds. Two marked peculiarities 
distinguish the words borrowed from Sanscrit ; no 
word can begin with a soft explosive ; no hard explo- 
sive can stand isolated in the body of a word. Thus 
the Tamil equivalent of the Sanscrit word gati is kadi 
(the German method is similar, brocket for projet). 
The consonant r cannot begin a word ; it requires too 
much effort ; it must be introduced by a vowel. The 
Sanscrit word rajah becomes in Tamil irayan, iragan. 
The dialectic variations lie generally between explo- 



Agglutinative Idioms of SoiUher7i Asia. 129 

sives and palatals of the same order (this is a 
general rule). In Tamil and Malayslla the dentals 
have an increasing tendency towards the English ih, 
hard or soft. In Telinga tch and dj often pass into 
tz and 2, a phenomenon very common in Italian giorno, 
Venetian zorno, Neapolitan yorno. Vinson gives Icevi, 
ear, in Kanara, tchem in Telinga, cevi in Tamil. 

The derivation is clearly agglutinative, and need 
not delay us except to note a few new facts. Every 
declension, and that which the grammarians wrongly 
term conjugation and voice, is effected by suffixes 
accumulated and interlaced. There are not, properly 
speaking, any verbs, but derivatives indicating state, 
action, frequency, causation, negation, &c., actuality, 
distance in the past or ftiture. One peculiarity I think 
we have not yet encountered — the declension of forms 
already furnished with verbal suffixes. In old Tamil 
poems, says Vinson, we find forms such as garndayak 
ku: fa?',, to reach ; n, euphonic; d, sign of the past 
tense ; ay, thou ; kku, a sign of the dative : to thee 
who hast drawn near. It is the absence of the rela- 
tive pronoun which entails such constructions. One 
more example : tevar-ir signifies god-you, you are god, 
but also, you who are god, and is thus susceptible of 
all suffixes of declension, possessives, locatives, &c. 
The radical tevar is already declined (plural of majesty), 
and in such compounds remains invariable. 

The distinction of the genders is not common in 
the agglutinative class, and it seems to have been 
originally unknown to the Dravidian languages ; even 
now it only applies to adult human beings. Women 
have a right to the feminine gender only in the plural ; 
in the singular their name is neuter, like that of 
children. In Tamil there are really only two genders, 



1 30 Distribution of Languages and Races. 

the noble gender and the inferior gender. For the 
rest, the intellectual evolution of the primitive Dra- 
vidians does not appear to have been very advanced, 
for their proper vocabulary does not contain the words 
which may be translated by to he, to have, soul, will, 
God, priest, hook, writing, grammar ; but by borrow- 
ing the conceptions, the ideas, and the terms which 
they lacked, they have acquired a very rich idiom, 
capable of lending itself to the subtleties of religious 
philosophy and to the fantasies of a brilliant poetry. 

The Kanara, Tulu, and Telinga alphabets are de- 
rived from the Sanscrit character employed, in the third 
century before our era, in the inscriptions of the Bud- 
dhist King Afoka. Malaysia has similarly adapted 
to its own use an old Sanscrit alphabet called Grantha. 
Tamil seems to have received its alphabet from the 
Phoenician and Arab merchants. The most ancient 
inscriptions (ninth century of our era) exhibit these 
different types. 

Dravidian literature is later than the Aryan influ- 
ence. The principal dialects have been cultivated, but 
the palm belongs to Tamil both for age and merit. 
Literary Tamil, which differs considerably from the 
spoken language, and is much purer, possesses mystic 
poems composed by Jaina, Sivaist, and Buddhist sec- 
taries, and epic poems encumbered with metaphor, 
among which is a long history of Joseph, written by 
the Jesuit Beschi in the last century. There are also 
collections of maxims, modern lyrics, solemn and very 
monotonous hymns, and licentious tales ; treatises on 
astrology, divination, and medicine belong to modern 
times. 

Vinson believes that all the Dravidian dialects of 
the south will become absorbed in Tamil, and those of 



Agglutinative Idioms of Southern Asia. 131 

the north in Telinga, the one the best preserved, the 
other the most changed of this interesting and vigo- 
rous family. 

Tamil, as we have said, thanks to the energy and 
initiative of the people of the south of the Dekkan, 
is spoken in the northern half of Ceylon. The south 
of that great island is the home of another aggluti- 
native language, Cingalese or Elou, which contains a 
great number of Tamil and P§,li words more modern 
than the rest of the vocabulary. It is not yet known 
whether Cingalese should be considered as a branch 
very early separated from the Dravidian stem. 

Before quitting the Asiatic continent, let us cast a 
glance over the road we have travelled. Prom the 
Caucasus to the southern extremity of the Penin- 
sula of Hindustan we have found (omitting the Sem- 
ites and the Aryans) four groups or types of the 
agglutinative class : the very various dialects of the . 
Caucasus, which have been classed, with more or less 
certainty or probability, in two families which are 
related to each other, the northern or Tcherkess family, 
and the southern or Karfcvelian family ; they belong to 
races driven into the mountains, on the one side by 
the Altaics and the Slavs, on the other by the ancient 
Assyrians and Iranians. It seems that, under the 
name of Urarti, people of Ararat, the Kartvelians 
formerly occupied Armenia, and were the near neigh- 
bours of the ancient inhabitants of Lower Mesopotamia 
and Ohaldea, the Accads and the Shumirs. We have 
seen how skilfully modern research has reconstructed 
the civilisation and the language of these peoples, the 
inventors of the cuneiform character. Crossing the 
Indus, we have found in India, early conquered and 
organised by Aryans, two strata of agglutinative idioms. 



132 Distribution of Languages and Races. 

the one destined to disappear, in spite of a relatively 
advanced development, the Kol- Aryan group; the other, 
the Dravidian group, vigorous and capable of hold- 
ing its own among the numerous dialects of Sanscrit 
origin. 

Separated by vocabulary, by the physical and in- 
tellectual diversity of the races which have used them 
or who still speak them, these four expressions of 
human thought are united by two features only which 
are common to both ; they belong to the same lin- 
guistic class, and to nations which occupied the soU of. 
Asia before the arrival of the Semites and Aryans^ 
that is to say, of inflected languages. 



CHAPTER IV. 

THE MALAYO-POLYNESIAN LANGUAGES. 

Ethnographic theories of the peoples between Madagascar and the 
Paschal Islands : Negritos, Papuans, Australians, Indonesians, 
Polynesians, Malays — The spread of the Malays on the Indo- 
Chinese coasts and in the Indian Archipelago — Softness and 
simplicity of the Malay dialects (Eastern group : 'l^gala, with 
which is connected Hova, Bisaya, Formosan ; Western group, 
Malayo-Javanese) — Character, manners, and literature of the 
Malays — The Polynesians : physical indolence ; effacement of 
consonants ; poetical and mythical tendencies. 

In the whole of the vast Malayo-Polynesian domain, 
extending from Madagascar to the Sandwich Islands 
in one direction, and in another to New Zealand, 
passing by the Sunda Islands, a common speech 
reigns, of which the groups and sub-groups not 
only belong to the same class, but possess the 
elements of the same vocabulary. Only three 
languages or families of languages are foreign to it, 
and these, moreover, are too little known for philo- 
logists to pronounce upon their origins and aflSnities. 
How did the dominant idiom come to extend over so 
vast a space ? Did it appear first at some central 
point ? Was it imported from Asia or Polynesia ? 
from the north or from the east? Is it the language 
of a conquered race which has absorbed that of the 
conquerors, as Anglo-Saxon imposed" itself upon the 
Normans ? Or the language of invaders, of ^migrat- 
ing tribes, like the languages of the Indo-Europeans ? 
10 '3' 



134 Distribution of Languages and Races. 

A cursory review of the races of Malaysia and 
Oceania will throw light upon these questions, if it 
does not solve them. 

Disregarding secondary distinctions, we find a 
central black mass, Australia, New Guinea, and the 
adjacent islands, Melanesia, between two wings of a 
lighter tint, olive and coffee-coloured to the west, 
copper and reddish - bronze towards the east. The 
black mass, which we must regard as autochthonous, 
is yet very far from being homogeneous. There are 
three types : the true Papuan, of middle height and 
robust frame, bearded, with long head and frizzled 
hair ; the Negrito, little and frail, with round head 
and wavy or smooth hair ; the Australian, of mixed 
race, of varying height, hair sometimes frizzled, 
sometimes stiff and straight, more or less dolicoce- 
phalous. 

The eastern lighter wing appears to be nearly 
homogeneous, more or less tinged towards the left 
by contact with the Papuans or Australians. It 
presents from the Tonga Islands eastward a fine 
race, tall, well made, and well endowed. These are 
the long-headed Polynesians, who people, in small 
scattered groups, those islands of the Pacific which 
are perhaps the relics of a submerged continent. The 
western wing, on the other side of New Guinea, 
includes the Philippines, Celebes, Borneo, Ceram, Bali, 
Sumbava, Java, and Sumatra ; here we find, sub- 
stituted almost everywhere for the ' Negritos, a faii'ly 
tall race with slightly lengthened cranium, corre- 
sponding to the Polynesians, and who can hardly be 
separated from these; they have been calted Indo- 
nesians. Their principal subdivisions are : the Battaks 
of Sumatra, who practise agriculture and keep flocks ; 



The Malayo- Polynesian Languages. 135 

they are still cannibals as regards the bodies of 
criminals ; the Redjangs and the Lampoungs of the 
same island ; the Macassars and Boughis of Celebes ; 
the Dayaks of Borneo, obstinate head-hanters ; the 
Bisayas and Tagals of the Philippine Islands, of more 
or less mixed blood. 

Lastly, around, beside, amidst the Negritos, who 
are reduced to a savage condition, and the stronger 
and better armed Indonesians, the Malays, little, 
round-headed, with yellowish skin, active and coura- 
geous in spite of their slight frame and small ex- 
tremities, traffickers and pirates, occupy the coasts of 
the Malay Peninsula, of Borneo, of the Philippines ; 
haunt the ports of Indo-China and of Southern China ; 
and people the greater part of Sumatra and of Java, 
either pure or crossed in varying proportion with the 
Indonesians and Melanesians. Thus the Javanese 
proper, who fill the centre of the island to the number 
of thirteen millions, are of very mixed blood, while 
the Madurais of the east and the Sundeans of the 
west appear to belong to the true Malay type. We 
have seen that the Tamatos or Japanese aristocracy 
are supposed to be of Malay origin. 

It seems to me that from the preceding notes, 
incomplete and summary as they are, we may conclude 
that the Malays are the latest comers in all the places 
in which we find them established, and that they 
nowhere found the land uninhabited on their arrival. 
Even in Sumatra, of which they occupy the centre, 
they are wedged between the Atchinese and the 
Battaks on the west and the Lampoungs on the south- 
east. Even in Java, of which they occupy the two 
ends, they have only been able to modify the central 
group. Elsewhere, except in small islands, they only 



136 DistribtUion of Languages and Races. 

occupy the coasts. Checked in Timor and Ceram, they 
have completely failed to establish themselves in New 
Guinea. Driven northwards, they left important 
groups in the Philippines, and thence perhaps gained 
Japan. We seem almost to see their invasion, their 
wanderings ; and probability here approaches certainty, 
inasmuch as everywhere they have driven out or pene- 
trated among Negritos or Indonesians, or else fallen 
back before immovable and dense populations, and 
that no invasion has followed theirs, or rather only 
well-known contingents of Klings or Tamils and Arab 
merchants, who have not sensibly modified either the 
distribution of races or the geography of the Malay 
world. 

What was the cradle of this race ? Was it the 
Philippine Islands, where the Tagal dialect preserves 
the purest and most developed forms of the Malay 
language? Or Sumatra, which the Malays themselves 
regard as their country, and whence, if we are to trust 
their chronicles, they set forth in the twelfth century 
to conquer the Indo - Siamese Peninsula, and to 
found Singapore and Malacca ? The first hypothesis, 
which finds few supporters, is hard to reconcile with 
the fact that the idiom of "the Hovas of Madagascar 
belongs to the Tagal branch of the Malay stem. Now 
the Tagal people appears never to have left the 
Philippines, unless perhaps to visit the Marianne 
and Pelew Islands ; moreover, they are not of Malay 
blood ; and if they have kept the language pure, it is 
because they received it before it had been mixed with 
Indian and Arabic. This applies also to the Hovas, 
Indonesians crossed with Papuan and Malay blood, who 
were probably driven out from one of the Sunda 
Islands in prehistoric times. With regard to Sumatra, 



The Malayo- Polynesian Languages. 1 3 7 

everything tends to prove that this was an early, but 
not the original, centre of expansion. 

It is now generally agreed that the Malays are 
of Asiatic origin, and bear a general resemblance' in 
shape of skull, &c., to the Mongols or Mongoloids, 
Burmese, Laotians, Miao-tse ; and a probable cause 
for their emigration may be found in the great dis- 
turbance occasioned in the far east by the Chinese 
conquest and expansion. 

However this may be, the arrival of the Malays 
in the Sunda Islands must have taken place in very 
remote times. It must have taken long centuries for 
them to assimilate numerous Indonesian groups and 
teach their language, or at least the elements of their 
language, which was not completely developed, to 
those wandering tribes who carried it with them and 
scattered it in more or less altered form throughout 
the islands of the Pacific. For the inhabitants of 
Polynesia are for the most part Indonesians driven out 
by the pressure of the Malays and mixed in varying 
proportions with the blood of Papuans and Negritos, 
who are themselves of mixed race, and also with 
Australians (in New Zealand) ; perhaps also with 
indigenous races and with Americans of Peru. 
Polynesian tradition points to the Island of Bolotu as 
" the land of the souls " — that is to say, of their 
ancestors — an island of the west which is identified 
with Bouro near Ceram, one of the Moluccas. Thence 
rounding New Guinea, touching at the Solomon 
Islands, at Fiji, at Samoa, scattering themselves from 
island to island, they came to the central position of 
Tonga, the sacred island, Tongatabu, the land of 
Tangaloa or Taroa. From Tonga, from Savaiki, they 
went southward as far as New Zealand, northward as 



138 Distribution of Languages and Races. 

far as the Sandwich Islands and Hawaii; then gain- 
ing Tahiti, the Marquesas, the Pomotu Islands, they 
made their way to the distant Paschal Islands. A 
later movement to the left carried some of their long 
canoes towards the lesser islands of the Carolines and 
of Micronesia. 

It is after the double exodus of the Hovas towards 
the Kaffir-land of Madagascar and of the Polynesians 
of the future towards the Pacific that the Malay 
world properly so called enters into history and 
begins to be sensible of external influences. The 
fame of the two-horned Alexander (Alexander the son 
of Ammon), the king of Rome, born in Macedonia, has 
penetrated as far as Sumatra. This fabled conqueror 
had visited the Malays, and the kings of P.alernbang 
trace their origin to him. It is necessary to add that 
this legend, which came, no doubt, through India, has 
acquired Persian and Arabic elements, and that the 
chronicles in which, it is embodied have a family 
resemblance to the " Thousand and One Nights " 
{Sedjarat Malayan, translated by Marcel Devic ; 
Leroux, 1878). From the second to the sixth cen- 
tury of our era the Malays were influenced by the 
Tlings, the Tlihgas of the Coromandel Coast ; the 
impression they made is most sensible in Java. They 
were also from this date in perpetual contact with the 
Chinese. In the tenth century their flourishing realm 
attracted the merchants of Arabia and Persia, as we 
see from the "Marvels of India," a curious little Arabic 
compilation, translated by Marcel Devic, and from the 
history of " Sinbad the Sailor." The names of Singa- 
pore,. the lion's town, and Malacca, which is the name of 
au Indian fruit, towns founded in the twelfth century, 
■show that Hindu influence was still predominant ; but 



The Malayo- Polynesian Languages. 139 

the work of the Moslem began in the following 
century. Yet religions were not wanting in Malaysia, 
Buddhism, Sivaism, not to mention animism, occult 
but never extinct. Mohammedanism, introduced at 
Atchin in 1206, at Malacca in 1276, was established 
in the Moluccas and in Java towards the middle of the 
fifteenth century. Celebes embraced Islamism just at 
the time when Vasco di Gama threw open the rich 
" Spice Islands " to European commerce and to the 
somewhat tardy and superfluous Christian propaganda. 
The rest belongs to modern history ; all that we need 
retain is that neither the Portuguese conquest, nor 
the Dutch, Spanish, and English occupations, nor the 
commercial rivalry of the prolific and swarming 
Chinese, have diminished the domain of the Malay 
idiom, which remains the international language of a 
very important part of the far east. 

Malay is in the Indian Archipelago, to borrow the 
phrase of the learned John Crawf urd, what the French 
language has been in Western Europe. All the 
nations who transact business there understand it ; all 
new-comers make haste to learn it ; and among the 
immense number of idioms spoken in the two con- 
tinents there is not perhaps one so well fitted to 
serve as the means of communication between the 
various peoples who meet in that part of the world. 

A language destined to play such a part must before 
all things be sonorous, easy to hear and pronounce, 
devoid of those aspirates of various kinds among 
which only a practised ear can distinguish at once, 
and of those guttural and clucking sounds which seem 
so natural to the aborigines and are the despair of the 
stranger whose throat and lips cannot fashion sounds 
so unfamiliar. In these respects Malay is a perfect 



140 Distribution of Languages and Races. 

language. Its vowels, a, e, i, 0, u, are sounded as in 
Italian, and give rise to no confusion. Its consonants, 
which are in sufficient number to give richness to the 
vocabulary, include none which are difficult of arti- 
culation, even by the Chinaman, whose organ is so 
imperfect. At Singapore and in the other commercial 
centres of the Archipelago, Chinese, who have come 
from different parts of China, use the Malay tongue 
among themselves, because of the mai'ked divergences 
of their own dialects. An Englishman, Dutchman, 
Frenchman, Arab, Spaniard, Siamese, Hindu, hearing 
a Malay word pronounced, can repeat it at once 
without the smallest difficulty. 

To this quality, invaluable in an international idiom, 
is added another which is hardly less necessary — 
simplicity of structure. The great majority of the 
radicals are composed of words of two syllables, which 
are absolutely invariable : rati, bread ; padi, rice ; 
kayou, tree ; or any, man ; makan, to eat ; minom, to 
drink ; hetoul, true. These words cannot be distri- 
buted into grammatical categories ; the same term 
may be noun, adjective, or verb ; no gender, no 
number ; no declension or conjugation. The feminine 
and masculine are indicated when necessary by the 
addition of such words as man, woman, male, female. 
The plural is indicated either by the word many, 
much, or by reduplication ; orang-orang, men ; radja- 
radja, kings. Monosyllabic particles placed at the 
beginning, end, or in the body of a word may define 
the substantive or verbal sense. For example : in 
Dayak, lauk, fish, forms palauk, fishermen ; in 
Boughi, nasu, to cook, gives panasii, cook ; ^a is a 
prefix signifying action, condition. In Tagal, paligo 
means to bathe, paligo- an, the bathing-place; niog. 



The Malayo- Polynesian Languages. 141 

palm-tree, niog-an, grove of palm-trees. The enclitic 
ni, na, makes of sipii, to seize ; tajMy, to knead ; 
sinipit, anchor ; tinapay, bread. Tense and mood are 
similarly rendered by the words already, still, to 
wish, &c. 

Translated word for word, a Malay phrase resembles 
what is familiarly called pidgin English. Here is 
the beginning of a collection of fables : " Live once a 
man merchant, Pouti his name ; very much rich, but 
no child him, therefore much wish for child," &c. 

It will be easily understood that so simple a 
language, which was suflBcient for the needs of peoples 
such as the Javanese, Boughis, Atchinese, hardly 
inferior to other Moslem races, suited also the Indone- 
sian and Melanesian tribes, scattered and hidden in the 
forests and mountains of the larger islands. These 
tribes, yellow, brown, or black, do not speak exactly 
the same language ; on the contrary, the idioms 
change from one to the other with an extreme variety ; 
but there exists a visible connection between them, 
which, if it does not at once show community of 
origin, points at least to a remarkable analogy of 
method. And in truth a brief study of the eighty 
dialects enumerated by Robert Gust will prove that 
the differences existing between them are far less than 
those which we find between the Romance languages. 

The Malay languages, properly so called, are 
divided into two branches, the eastern or Tagala, the 
western or Malayo-Javanese. To the first belongs 
Tagala, spoken in Luzon and Bisaya, in the islands 
immediately to the south of Luzon ; Formosan in the 
east and centre of Formosa (the western portion is 
Chinese) ; here there is no mixture of Hindu words, 
which proves the great antiquity of the arrival of the 



142 Distribution of Languages and Races. 

Malays at Formosa. In tlie Marianne Islands the- 
language is still Tagala-Malay. Malagasy or Hova,. 
spoken in the extreme west of the Malay domain, is 
connected with this branch of the family. 

The second branch includes Malay proper, spoken 
in the Malay Peninsula, in the greater part of Sumatra, 
in the little neighbouring islands, and on the coasts 
of Borneo; Battak, Atchinese, and Larapoung in 
Sumatra ; Dayak in the centre and north of Borneo ; 
Boughi and Macassar in Celebes ; then the important 
Javanese sub-group : Javanese, spoken in the centre 
of Java by thirteen millions of people ; Sundean, 
spoken round Batavia by four millions, Madurese, 
Bali, &c. ; employed by almost equal numbers in the 
east of Java, Madura, Bali, and other smaller islands. 
Javanese is the most cultivated of the group, and its 
religious and poetic literature, inspired by Indian 
ideas, is not without value. But since we cannot 
treat of everything, we will concentrate our attention 
upon Malay proper. It gives me an opportunity of 
doing homage to the memory of my friend, Marcel 
Devic, linguist and philologist, who took pleasure in 
translating and making known this simple and liquid 
tongue. 

One word in the first place on the character used 
by the various dialects. When they accepted the 
lessons of the Persian and Arabian missionaries, the 
Malays adopted their alphabet. This detestable in- 
strument, which certainly makes half the difficulty of 
the Turkish, Persian, Hindustani, and Arab languages, 
this alphabet, destitute so to speak of vowels, and full 
of aspirates, of guttural sounds, of emphatic articula- 
tions, is ill suited to an idiom as sonorous as Spanish, 
as soft as Italian and Portuguese. The other peoples 



The Malayo-Polynesian Languages. 143 

of the family have made use of an old Hindu character. 
The alphabet of the Javanese is agreeable to the eye 
and points to a true sense of art among them ; but 
their representation of the vowels is very singular ; to 
give only one example : the sound is expressed by 
two characters, of which one precedes and the other 
follows the consonant. The Tagala and Bisaya alpha- 
bet, which is incomplete in the vowels, has a certain 
strange grace. That of the Battaks is very ugly, as 
becomes a nation among whom literary cultivation has 
not put an end to cannibalism, which has become, on 
the contrary, a legal institution. One would not wish 
the Malays to adopt such a character, but their language 
"would profit much from the use of the Latin alphabet, 
which is so obviously suited to it. 

The introduction of Islamism has not flooded Malay 
with Arabic to the same extent as other Mussulman 
idioms. About a hundred and sixty Arabic words 
are reckoned, and there are about thirty Persian ones. 
We have more than that number in French, as any 
one may convince himself by consulting the Oriental 
part of Littr^'s Dictionary, which was revised by Marcel 
Devic. Hindu idioms have had a more sensible in- 
fluence; but the proportion of Hindu words is not more 
than five per cent. Finally, Portuguese and Dutch 
commerce and colonisation have introduced a few terms 
which it is hardly worth while to notice. Perhaps 
also China, Cochin-China, Burma, Siam, and Annam 
have furnished a small contingent. It is none the less 
true that Malay, as employed for several centuries, has 
not undergone that alteration which might have been 
expected in an idiom spoken as a lingua franca by 
peoples come from the four quarters of the horizon. 
There is no comparison, from this point of view, 



1 44 Distribution of Languages and Races. 

between Malay and the lingua franca of the Levant, 
in which Arabic, Turkish, Greek, and Italian are 
mingled in a species of formless patois, suitable for the 
verbal exchange of a few ideas which are always the 
same, but insufficient for a business letter. 

The Malay vocabulary includes about five thousand 
radicals, of which half are pure Malay, a fourth part 
common to Malay and Javanese, and the other quarter 
is made up from the foreign sources mentioned above. 
The language is rather poor than rich ; it does not 
possess, like our tongues, a great variety of terms to 
express the different shades of the same idea ; so that 
where one word suffices us, a long periphrasis is neces- 
sary to them. On the other hand, it is not encumbered, 
like the Tagala of the island of Luzon, with dissyllabic 
terms which bear no resemblance to each other in form, 
and express the same action accomplished under some- 
what different circumstances. For example: to eat 
meat, to eat fruit, to eat in company, to eat in the 
morning, to eat in the evening, to eat a little, quickly, 
by mouthf uls, with appetite ; each of these phrases is 
expressed by a single word ; there are forty such words, 
and no two are derived from the same root. 

The lack of synonyms and of synthetic expressions 
is the cause of a certain slowness and repetition ; the 
language is diffuse and full of circumlocutions, and this 
defect, which is not very sensible in conversation, is 
very marked in the most careful writing. 

The literature is abundant, but very little of it is 
original ; it abounds in translations and imitations of 
Hindu, Persian, Arabic, and especially of Javanese 
originals ; thus it has borrowed from India the MaTia- 
bharaia, from the Arabic the famous collection " Kalila 
and Diinna," and from a species of current opinion, 



The Malayo-Polynesian Languages. 145 

fragments of legendary history mixed with a few tradi- 
tions which have some probability. But every people 
assimilates more or less what it borrows, adding some 
traits of manners, national character, and domestic life. 
A fable of foreign origin, a fragment of a chronicle, will 
reveal to us the real character of these Malays, who 
find means to conciliate gentleness and barbarism, 
gaiety and the wildest delirium, piracy and commerce, 
an extremely lax morality and the precepts of the 
Prophet, and persistence in animism with a devout 
monotheism. 

There are Malay tribes who are very inoffensive 
and comparatively industrious — in Java, for instance ; 
but at bottom (and indeed not far below the surface) 
you would find, I think, the classic Malay, the brigand 
without faith or mercy, who sweeps the seas in search 
of his booty, attacks every solitary ship, massacres the 
crew, and makes off with the cargo to his lair on the 
shore. Such are those pirates who astound even the 
Arabs, who, armed with their kriss, seize a rich mer- 
chant in the open market at Timor or Taneh, and 
ransom or murder him then and there, run amuck, 
massacring all who resist, and if taken kill themselves. 
It is hardly likely that a people of this nature should 
understand justice and virtue as we do. 

The popular, and more or less justified, belief is 
well known, that a child who is precociously intelligent 
will not live to grow up. If Taylor undertook to show 
us that this is a survival from the time when too clever 
children were destroyed, he would find an argument in 
a passage in a Malay chronicle. A king of Sumatra 
was attacked by monstrous animals, which at certain 
hours came and bit the legs of his army. A child 
advised that a palisade should be raised breast-high. 



1 46 Distribution of Languages and Races, 

This counsel, which filled the wisest ministers of the 
prince with wonder, discouraged the evil beasts and 
saved the camp. On this a song was composed, with 
the refrain, " No harm came of it, thanks to the wisdom 
of a child." Now as the blessed Maharajah Padouka 
was on his way home, the nobles said to him, " Sire, 
this child is already very clever though so young ; 
what will he be when full grown ? It were better to 
get rid of him." So all thought it right of the king 
to have him put to death. Nothing could be more 
simple, and neither the king nor the historian feel 
more remorse than would a soldier of Radetski, the 
judge or executioner of his vanquished and rebel fellow- 
countrymen. For there are lacunes also in our code 
of morality. 

So much for justice ; I pass to virtue. It is not 
absent, but it manifests itself with a brutality which 
recalls the most flourishing epoch of the Middle Ages. 
A moralist thus recounts the fable of two friends. 
Two men, brothers in heajt, travel together. At a 
tournament one of them, as is the custom, gained 
the hand of the princess by his skill at the games. 
His friend, who is in love with the lady, goes away, 
and after a long time comes back afflicted with a 
horrible disease. The wise men declare that there 
is no other remedy than to rub him with the blood 
of the little son of his ' friend, who has become king. 
The friend takes the knife, kills the child, and him- 
self rubs the sick man with the blood of the victim. 
" Such were the friends of old time," sighs the Malay 
writer ; " there are none like them now." 

Again, a young man wishing to put to the proof 
the devotion of his friends, pretends to have killed 
his mistress, and goes from door to door with the 



The Malay o-Polynesian Languages. 147 

supposed corpse on his back asking who will help to 
dispose of it. In a fairly long romance, the heroine, 
wife of a king and mother of three children, is pursued 
by the attentions of a minister. The queen, who 
is devout, objects in vain that infidelity is forbidden 
by the Prophet, and that such a crime would lead 
them both straight to hell. The minister insists, 
and threatens to kill a child if the mother does not 
yield. He cuts the boy's throat in very deed, bat 
without gaining his end. The same threat in the 
case of the second son, with the same result. In 
brief, when all her children lie on the ground with 
their throats cut open, the mother asks permission 
to wash off the blood which has splashed her, and 
to bathe and perfume herself; "after which, she says, 
the valiant minister shall have his desire." Irresistible 
minister ! but he should have at least killed the hus- 
band while he was about it. 

The people which takes pleasure in such tales does 
not lack pedantic doctors, nor professors of social and 
political philosophy. Among them was Bokhari of 
Djohor, who, in the seventeenth century, wrote a 
treatise celebrated in Malaysia, entitled MaJcota Radja, 
or the Crown of Kings. This Bokhari possessed, it 
would seem, a learning which is rare among Malays. 
He knew Arabic and Persian ; and his book is full of 
anecdotes borrowed from writers in the two languages. 
It is a species of manual, filled with puerile and 
minute details about all that concerns the administra- 
tion of a monarchical state ; duties of subjects towards 
the sovereign, whether Moslem or infidel ; the etiquette 
and hierarchy of the court ; the office of the mini- 
sters, ambassadors, and functionaries ; the education of 
children; the qualities of a believer, justice, bene- 



1 48 Distribution of Languages and Races. 

volence, true dignity ; even anatomy and physiognomy 
find a place in the work. Poetry is mingled with 
prose, and the author is so proud of his work that 
he puns upon his own name, Djohori, which means 
native of Djohor, and also jeweller. " Bokhari," he 
says, "is a jeweller, this may be seen from the orna- 
ments of the crown." In spite of these ornaments, 
the crown is wearisome, and the genius of the Malay 
is less suited to moral treatises than to tales and 
legendary histories, in which are set forth the force 
and cunning of heroes of adventure who resemble 
Ulysses or Hercules. 

The popular poetry of the peoples of the Sunda 
Islands is their most original contribution to literature. 
At their feasts two singers are pitted against each 
other, like the shepherds of Theocritus and Virgil. 
The one presents in an impromptu distich an image 
or an allusion ; the other replies in the same rhyme 
and metre, giving a similar or contrary idea. " What 
use is the lamp without a wick," says the one. "Why 
make play with the eyes if nothing is meant ? " retorts 
the other. And so on, piling up witty or fanciful 
notions, of which, it is true, the meaning and the charm 
often escape us. These quatrains are called pctintoums. 
Victor Hugo has quoted a few in the notes to Les 
OrientaUs ; and it would be easy to collect whole 
series at Sumatra and Borneo. 

We have said that New Guinea and the continent 
of Australia are a world apart, between the Malay 
and Polynesian worlds. The languages of Australia, 
which are imperfectly known, will soon have disap- 
peared with the natives who speak them. Tasmanian 
has lately become extinct with the death of an old 
woman, a queen, the sole survivor of her race. It 



M^r^.^ 



The Malay 0- Polynesian Languages. 149 

is doubtful wlietlier these dialects belonged to a single 
family. In any case, the only thing which they have 
in common with the Malay group is their aggluti- 
native character. We are not able to say more of 
the Papuan languages ; but we shall at least be able 
to study them at leisure, when Europeans have taken 
possession of New Guinea. These eastern Negroes, 
much stronger and more vigorous than the wandering 
Australian tribes, much less savage and more in- 
dustrious, have there their centre and their home, 
in an island which is at least as large as Borneo ; 
they occupy also the islands of New Britain, the 
Solomon Islands, and the New Hebi-ides ; and it is 
believed tltat their language maintains itself in New 
Caledonia. In several places, however, a mixture of 
Malay or Indonesian blood is apparent. 

Passing to the north of this group, the emigrating 
tribes have just touched its borders and settled in 
the smaller islands. A first zone, called Melanesia, 
though inhabited by a race which is akin to the 
Papuans, a dark - skinned, hairy people of middle 
height, has yet been penetrated by Malay customs 
and language. Samoa and the Fiji Islands form the 
transition between Melanesia and Polynesia. From 
the Tonga Islands the whole of the Pacific belongs 
to one of the finest races in the world, tall, slender, 
deep-chested, often with regular features, and of noble 
or pleasing outline. Unfortunately, these well-made 
men and these attractive women lose their vitality 
in proportion to their distance from the Sunda Islands. 
The facility of gaining a livelihood, infanticide, tribal 
warfare, cannibalism, and lastly, traders and mission- 
aries, bringing in their train clothes and spirituous 
liquors, phthisis, small-pox, and other diseases, have 
11 



1 50 Distribution of Languages and Races. 

rapidly decimated these peoples, which are so worthy 
to survive. The most civilised, even those of the 
Sandwich Islands, dwindle without apparent cause. 
The most energetic, those of New Zealand, fought 
valiantly for their independence, and are now dying 
out, though unmolested, before the English and Scotch 
settlers. It is difficult to refrain from useless regrets 
over the gradual disappearance of this fine race ; it 
would not be safe to affirm that more than a million 
of Polynesians are scattered throughout these thou- 
sands of islands, which ai'e by no means all of small 
size : the New Zealand group is hardly smaller than 
Great Britain. 

Oook was the first to recognise the linguistic unity 
of Polynesia ; the native of Tahiti who went with him 
to New Zealand conversed without difficulty with the 
natives of that country. La Perouse noted the affinities 
of Polynesian with Tagal and Bisaya. Finally, Porter, 
Mariner, Dumont d'Urville, Ellis, Sir George Grey, and 
Eienzi have collected data which are precise and con- 
vincing. But the original Malay type tends to become 
effaced with distance. The Malay words are fewer in 
number in proportion to the distance from the point 
of departure. The methods of formation remain the 
same, but a free use is made of the common elements, 
of the monosyllabic roots which are hidden in the 
Malay dissyllables. This phenomenon is supposed to 
be caused by the detachment of the Polynesian branch 
of the family before the complete development of the 
race and of its language. But a yet deeper cause of 
the change may be found in the physical indolence of 
the race ; the law of least resistance may be seen in 
all its force in the modifications introduced. 

At the outset of these studies, when we sought to 



The Malay o-Polynesian Languages. 1 5 1 

explain the origin of the explosive consonants, the 
bases of articulation, we pointed out the primitive in- 
decision of language between the sounds k, t, and p, 
between the gutturals, dentals, and labials, then between 
the liquids and the nasals, between the sibilants and 
the aspirates. Now nowhere is this confusion more 
apparent than in Polynesian between the different 
dialects, and sometimes in the same. Man is in- 
differently tanata or kanaka; shade or spirit is akoua, 
atoua, apoua ; kalo is the same as taro, Samoa as 
Hamoa, Sawaiki as Havaiki. Moreover, the consonants 
are perpetually dropped : tanata becomes tane ; Sawa- 
'iki, Hawa/i ; Ariki, lord, arii, areoi ; tiki, demon, tii ; 
potcarka (the Spanish pnsrco, pig) becomes houaga or 
pouaa. The supreme god is Tangaroa, Tangaloa, 
Taaroa. Constantly there are double vowels and diph- 
thongs : aa, ee, ii, ea, oa, oahou, which suppose inter- 
mediary consonants which have been dropped. The 
repetition of syllables and dissyllables, mea-mea, oro- 
oro, Sac., tends to replace everywhere particles and 
affixes. The examples are sufficient to establish our 
point. The facts are constant. There are twenty 
consonants in the Malay alphabet ; there are only 
fifteen at Tonga, ten at Tahiti, and even less in some 
other islands. The nasals and liquids, few in number 
at Tonga, are almost unused at the Marquesas Islands. 
Lastly, the dialects which are relatively the roughest 
belong to the nations which are the least weakened, 
the Maoris and the Hawaians. 

The Polynesians have no written literature ; but 
their oral traditions and cosmogonies are numerous. 
From their traditious, transmitted from age to age by 
the harepos or historians, committed to memory by 
Hawaian princesses, handed down from father to son, 



152 Distribution of Language^ and Races. 

from mother to daughter, we gather the little that is 
known of their history. The deeds of the gods — that 
is to say, of the chiefs and ancestors — were sung with 
every solemn accessory. These chants constituted the 
title of kings and chiefs ; an attempt has been made, 
in their interpretation, to separate truth from fable. 
In New Zealand the English administrators have ad- 
mitted, as constituting a valid claim in suits relating 
to the possession of land, genealogies and evidence 
contained in these traditional songs. 

The Polynesians share with their kinsmen of the 
Surida Islands the gift of poetical improvisation, which 
is rendered easier for them, by the sonorous fluidity 
of their language.. The arrival of a friend or of 
the convoy of a chief was saluted by stanzas or 
elegies, monotonous and diffuse, but not wanting in 
sentiment and grace. .Dumont d'Urville gives one, 
a funeral chant improvised by a woman of the Sand- 
wich Islands : — 

" Alaa ! alas ! my chief is dead. 
Dead is my lord and my friend, 
My friend in the time of famine, 
My friend in the drought, 
My friend in the rain and the wind, • 
In the sun and the heat, in the cold of the mountain, 
In the calm and in the storm. 
My friend in the eight seas. 
Alas ! alas ! gone is my friend, 
Gone never to return." 

The last line is the cry of nature, so simple and so 
natural, that it is the utterance of all peoples, savage 
or civilised, without regard to the contradiction it 
offers to all the fictions of animism and of religion. 
And it may be noted in passing, there is no people 



The Malayo-Polynesian Languages. 153 

which has a firmer belief in. ghosts — that is, in another 
life, in the immortality of the soul — than the Poly- 
nesians. Here is another fragment, the description of 
a volcano in eruption : — " The summit has long been 
on fire. The land of Touha-Ehou was deserted. The 
bird perched on the rocks of Ohara-hara. During 
eight nights, during eight days, those who till the soil 
held their breadth, looking round them anxiously. 
By the wind, by the storm laden with rain, the dust 
has been carried to Hoina. The eyeballs were red- 
dened with this dust. Tavai, Tavai, blessed be 
thou, land in the midst of the sea, who sleepest peace- 
fully on the bosom of the waters, and turnest thy face 
to the pleasant breezes. The wind had reddened the 
eyeballs of the men with the tattooed skin; the sand 
of Taou is at Poha-Touhoa ; the lava at Ohia-Ota- 
Lani. The path is over the sea to the shores of 
Taimou. Inland the path to the mountains was hidden ; 
Kirau-Ea was hidden by the tempest ; Pele dwells in 
Kirau-Ea, in the gulf, and feeds for ever upon flames." 
The Malay family of languages is one of the simplest 
and most convenient of the agglutinative idioms, as 
it is the most extensive and the most clearly defined. 
It constitutes a perfectly independent group, or at 
least its relationship to any other has not been dis- 
covered. It is doubtful that such can be established, 
whether the Malays, who came originally from regions 
occupied since by the Chinese, brought it with them, 
or whether they found it in Indonesia, in the Sunda 
Islands, the Moluccas, and the Philippines. The first 
hypothesis appears the more probable ; it agrees with 
Polynesian traditions, all of which seem to take their 
rise in the west. The diffusion of the Malay race 
would seem to have determined the successive migra- 



154 Distribution of Languages and Races. 

tions which have visited and occupied one after the 
other the different archipelagoes of the Pacific. Amidst 
the Malayo- Polynesian world subsist the remains of the 
Negritos and of the Australians, and the compact mass 
of the Papuans or eastern Negroes. The little known 
languages of these peoples hardly lend themselves to 
comparative study. The details we possess upon the 
literary culture of the Malays and Polynesians, un- 
happily arrested in their original development by Islam 
and by Christianity, show us specimens of humanity 
which range from the acceptable to the charming. 



CHAPTER V. 

AFRICAN RACES AND LANGUAGES. 

The past of Africa — Distribution of races— Linguistic map of Africa — 
General characteristics of the African languages — Idioms of the 
BoBJesmans and Hottentots — The clacking sounds or klicks — KafBr 
or Bantu family — Prefixing of the syllables which denote case or 
person — Phonetic peculiarities of the dialects — UkuJiZonipa—TTan- 
sitiim from the proper to the figurative sense — Bornu group : 
Haonssa — Senegambian group : Mandingue, Eiwe, Wolof — Peul 
or Poul — Upper Nile group : Dinka, Nouba, or Kensi — Gallaic 
group— The Berber languages — Coptic and Ancient Egyptian. 

We found a certain pleasure in reducing to a single 
family the various Malay and Polynesian idioms, and 
in showing by examples, drawn from the annals and 
imaginative poetry of these peoples, what good use had 
been made of these simple and sweet-sounding lan- 
guages by the different races, often well gifted, which 
speak them. Our study of Africa will not, I fear, 
oifer us the same kind of interest; for, with the ex- 
ception of the Mediterranean region. Lower Egypt and 
Barbary, nowhere in Africa has man risen to the 
intellectual level attained by the Malay or the Poly- 
nesian ; and in this great mass, which covers more 
than twenty-two millions of square miles, numerous 
ifroups of savage languages form what at the first 
glance is a hopeless chaos. We need to throw the 
light of history upon this confusion of races and 
tongues ; but history is arrested at the desert of 
Sahara. Two-thirds of the immense continent remain 



156 Distribution of Languages and Races. 

buried in an obscurity wbicli is scarcely diminished by 
the vague reports of rare Carthaginian merchants trans- 
mitted to us by Herodotus and Diodorus, nor by the 
somewhat legendary accounts of Arab merchants about 
the country of the Zendj, doubtless Zanzibar and the 
Somali coast. For generations the boldest and most 
fortunate travellers hardly explored anything but the 
coasts; and the great discoveries which have been 
made in the last thirty years have brought to light 
no documents, no monuments of an historic character. 
Thus, the past of Africa (which is at least as ancient 
as the rest of the world) is hidden not only from the 
civilised nations of Europe, but also from the millions 
of savages who swarm and vegetate on a soil which is 
nevertheless fall of wealth and resource. These tribes, 
even the most advanced among them, were arrested so 
early in their development that they have not yet 
arrived at that point at which a people, either by 
writing or by some material sign, fixes the memory of 
its vicissitudes, and finds in the consciousness of its 
previous inferiority the desire and capacity for progress. 
The African negro is certainly capable of improve- 
ment, but not of his own initiative or in his own 
country. His memory is short, his foresight almost 
rdl. Present enjoyment suffices him, poverty or death 
affect him little.. His morality results from his im- 
mediate interest, especially from fear of his master. 
He is animal in the spontaneity of his instincts. 
Many of them are by nature gentle and indolent, but 
deceitful, and too often ferocious at times. Many of 
them laugh easily, but they bite as readily when 
hunger urges them. It is impossible to rely upon 
their promises and their most solemn oaths. There 
are among them brave soldiers, clever hunters, here 



African Races and Languages. 157 

and there intermittent cultivators of the soil (who,, 
however, leave all the work to the women), travelling 
smiths ; lastly, and especially in the valley of the Niger, 
artisans, potters, weavers, tanners, enough to satisfy 
the limited wants of half-naked populations; but, 
generally speaking, from the Guinea Coast to the great 
lakes in which the Nile takes its rise, and from Bornu 
to the Orange River, these tribes stagnate in imme- 
morial savagery. The most ingenious of the Zulu 
tales are very poor stuff, and so is the mythology of 
the Dahomeyans. One is inclined to prefer to these 
laborious and futile inventions the hungry naivete of 
the Hottentot, who considers the sun as a piece of 
lard which has unfortunately been hung out of reach. 

The first impression caused by the narrow foreheads 
above the protruding jaw and thick lips, by the sooty 
bodies anointed with every species of evil-smelling 
grease, is uniformly unpleasant, but it is easy to per- 
ceive marked differences in conformation, stature, colour, 
physiognomy, among these tribes who live side by side, 
in a singular confusion, and generally without other 
frontier than the palisade or mound of earth which 
surrounds their village. It is easy to understand that 
there are intruders, conquerors, who are either absolute, 
or the suzerains of subjects whose condition varies 
between servitude and vassalage. We try to follow up 
the path taken by the invaders, and the geographical 
distribution of the victors and vanquished, and especi- 
ally the amount of mixture between them (which is a 
measure of the length of time during which forced 
relations have existed between the indigenous race 
and the later comers), will supplement to some extent 
the missing historical data. 

We thus are enabled to see, in a very remote age, 



158 Distribution of Languages and Races. 

the north, of Africa, perhaps then joined to the Canaries 
and to Spain, inhabited on its borders by a white race, 
the Lybians or Berbers, whose domain extends to the 
delta of the Nile on the one hand, where they march 
with tribes, also white probably, of Asiatic origin, 
Khamites and Semites, and higher np the great river 
they encounter a black, smooth-haired race, the Noubas 
and Barabras. WhUe the mixture of these three races 
forms the Egyptian nation, whose colour varies from 
reddish - brown to a yellowish - white, the western 
Lybians, rounding and crossing the Sahara, find them- 
selves in presence of true Negroes with woolly hair, the 
Yolofa and others, who occupy the basins of the Senegal, 
the Niger, and the Ogooue, and whose dense masses are 
not much tinged with Berber blood. 

On the extreme east another white stream, issuing 
from the point of Arabia, takes the Noubas, already 
driven back by the progress of the Egyptian people, in 
flank, and leaves upon the coast, and in the highlands 
of Abyssinia, the Gallas, the Somalis, the Ethiopians, 
all mixed with native blood in varying proportions. 
This east central invasion -has had two important con- 
sequences : it drove towards the west, by slow degrees, 
to the south of the Sahara, towards the basin of Lake 
Tchad, and towards Guinea, a part of the Nouba 
population, already somewhat tinged with Lybian and 
Asiatic blood, and already somewhat awakened by 
contact with superior races. These Noubas are the 
Peuls or Pouls, studied by Faidherbe, who, scattered 
over Senegambia and Guinea, and further mixed with 
the Lybians of the south and with the Berbers of the 
desert, constitute the dominant class or caste of the 
Niger country. 

On the other hand, the pressure of the Nubian 



African Races and Languages. 159 

Arabs, of the Gallas and the Somalis, determines a 
movement towards the south. A tall black race, the 
finest of all, with woolly hair but Caucasian features, 
the Bantus or Abantus, since named by the Mussul- 
mans Cafr, Kaffirs, or infidels, descend along the 
Zanzibar and Mozambique coasts, people the west of 
Madagascar, and cover the shores of the Indian Ocean 
from the Zambesi River to the river of the Great Fish ; 
important fractions of Bantu people ascend the great 
Zambesi River, and even gain the Atlantic coast : these 
are the Bechuanas in the centre, the Damaras in the 
west. These Kaffirs, destined later to fall under the 
yoke of England or of the free republics of the Dutch 
Boers (Orange Free State and Transvaal), took the 
place of the earlier occupants, the Hottentots or Khoin, 
and the Bosjesinans or Bushmen, doubtless of some- 
what mixed blood, forming the bulk of the population 
in the west of Cape Colony ; the latter were driven 
back into the desert of Kalahari, and confined -on the 
north and east by the Bechuanas, on the west by the 
Damaras and by the Namaquois Hottentots, and on 
the south by the Griqua Hottentots and the white 
inhabitants of Cape Colony. 

Neither the Hottentots nor the Bushmen are 
Negroes. It has been conjectured that the first are 
half-castes of Bushmen and Kaffirs, and also an attempt 
has been made to connect them with the mixed races 
of the north of Africa; but differences of language 
and feature render both opinions doubtful ; it is only 
proved that the names of places in Kaffraria are still 
Hottentot. It is probable that before they were driven 
out by the Kaffirs the Hottentots themselves had dis- 
possessed the Bushmen, whom they call 8ah and San, 
or natives. These last are interesting by their very 



1 60 Distribution of Languages and Races. 

depth in the scale of human beings ; they are among 
the poorest specimens of the genus homo. Fritsch 
affirms that those who gave them their name, " men of the 
bush," wished to intimate that they are creatures inter- 
mediate between the man and the ape. Without shelter, 
even the most rudimentary hut, without chiefs, laws, 
or worship, neither tillers of the soil nor shepherds, 
wandering in small clans or isolated families, they live 
solely by hunting and pillage, on roots, fruit, honey, 
ostrich eggs, the larvae of ants, locusts, reptiles, &c., 
gathered by the women. Always hungry, they eat 
all that they can find, and their flattened bellies 
become enormous in a short time, to return to their 
original condition in a few hours. These alternations 
of repletion and inanition furrow their skin into pro- 
found wrinkles, in which collects the grease with 
which tbey anoint their bodies as a protection against 
mosquitoes. They weave a few mats, and manufacture 
their weapons, which, however, they do not forge ; they 
work the cold iron with flints. The dog is their only 
domestic animal. The Bushman is little, pot-bellied, 
his skin of a dirty yellowish-brown. His forehead is 
straight, but his brain very small ; his thin hair is 
rolled up into little balls like pepper-corns ; his nose 
flat, his mouth protruding ; his chin retreating under 
his thick lips, which do not meet. The women are 
frightful ; tlie famous Hottentot which may be seen 
at the Museum is a faithful copy of the form of a 
Bosjesman woman who died in one of our hospitals. 
This unhappy race, which is unfortunately of pure 
blood, has nevertheless certain qualities. The mother 
loves her children ; the man is lively, gay, obstinate ; 
hunted by hunger, killed without mercy by his stronger 
neighbours, whose territory he is constantly invading, 



African Races and Languages. i6i 

he is accused of being fierce and revengeful Wlio 
would not be so under like circumstances ? 

We have made the tour of the coasts of Africa ; 
the centre remains to be considered, the great plateau 
bounded on the east by a series of great lakes, Nyassa, 
Bengueolo, Tanganyika, Victoria and Albert Nyanza, 
traversed at the equator by the vast curve of the Congo, 
watered on the north by the marshy affluents of the 
Nile. This immense region has been traversed and 
partly made known by explorers, Speke, Livingstone, 
Baker, Stanley, Cameron, Brazza, and others, whose 
narratives are familiar to us. This region abounds in 
inhabitants of every height and build, of every shade of 
colour between ebony and light chocolate ; dwarfs like 
the Akkas, who were figured on the Egyptian monu- 
ments, and who appear to be only less savage than the 
Bushmen ; cannibals like the Niam-Niam of Schwein- 
furth, a race whose peculiarity of costume caused them 
to be taken for men with tails; courageous tribes, 
such as the Monboutous; finally, some attempts at 
absolute monarchy, notably Uganda. But nowhere is 
there any trace of what we call civilisation, of artistic 
or intellectual culture. The future of all this inferior 
humanity, vigorous and perhaps susceptible of improve- 
ment, if Islam and Christianity would abstain from 
fighting for its unconscious soul, if drunkenness, theft, 
and murder were not encouraged by the Arab traders 
in their greed for ivory and slaves, is one of the great 
problems which the Northern nations have to solve. 
Are there enough men in Europe to rule and educate 
these inert multitudes, and would it not have been 
wiser to leave them to themselves ? These are prob- 
lems of which the solution will not be seen by any 
one now living. 



1 62 Dislributien of Languages and Races. 

The distribution of the four or five hundred dialects 
spoken in Africa corresponds fairly well to the above 
rough sketch. They may be classed in six or seven 
groups, according to the scheme laid down by Earth, 
Appleyard, Bleek, Fr. Mtiller, Hovelacque. In the 
north the Semitic and Khamitic languages prevail ; to 
the first belong ancient Ghez and Amharic, or modern 
Abyssinian ; to the second, the Berber idioms, the 
Egyptian of the Pharaohs, Coptic, and finally the 
Ethiopian branch : Somali, Galla, Bed] a, Saho, Dan- 
kali, Agaou. Immediately to the west the Nubian lan- 
guages are spoken by the inhabitants of the basin of 
the Upper Nile and of a part of Khordofan ; Nubian 
or Kensi, Dongolavi, Toumali, Koldadje. From Lake 
Tchad to the middle basin of- the Senegal, a distance 
of 2250 miles, extends Peul or Poul, entirely distinct 
from the families which it traverses or borders. Be- 
tween the equator and the Sahara, from the lakes of 
the Upper Nile to the Atlantic, the Negro dialects, 
properly so called, prevail: (i.) The Dinka group 
(Bari, Bongo Ohillouk, Nouer, &c.), the poorest of all, 
hardly issued from the monosyllabic stage ; (2.) The 
Bornou of Lake Tchad ; (3.) The Haoussa of the 
Soudan, a more advanced language, rich in dialects ; 
(4.) Sonrai, towards the great elbow of the Niger; 
(5.) Wolof on the Senegal, Mandingue or Malinke on 
the Gambia, Feloup in Guinea ; (6.) Krou, Egbe, and 
Ibo along the Gulf of Benin and the ocean. 

The most important and clearly limited family is 
the Kaffir or Bantu, which extends over all Eastern 
Africa, and south of Zanzibar penetrates as far as the 
Atlantic, between the Zambesi and the Congo, and 
even, crossing the equator, comes in contact with the 
Guinea languages. Its eastern branch comprehends 



African Races and Languages. i6 



o 



the dialects of Zanzibar and Mozambique, of the Zam- 
besi and Kafifraria, Swali, Zulu, and Kaffir; a second 
central branch is represented by Tekesa and Setchuana. 
To the third branch belong, beginning in the north, 
the language of Fernando-Po, Mpongue (spoken in 
Gaboun), Dikele, Isubu, Congo, Angolian, and Herero 
or Damara. South of Herero the Nama, Kora, and 
Griqua dialects form the Hottentot group, the neigh- 
bours of the Bushmen. 

Before giving the characters of a few of these lan- 
guages, of which I have wished to give at least the 
principal names, let us see first if there are any features 
common to all. Here is one, very general in the 
agglutinative class, a dislike to the accumulation of 
consonants ; the African prefers syllables terminated 
by vowels, and in the groups of the north or of the 
extreme south, where final consonants exist, it is easy 
to trace the language back to an earlier period when 
the final vowel had not been dropped ; just as, by 
poetical license or rapid pronunciation, most Italian 
words may lose their final vowel. From this rhythmic 
and euphonic point of view the African languages are 
called alliteral. But it would be a mistake to regard 
the multiplicity of vowels as a certain guarantee of 
softness and harmony. Most of the African dialects 
possess gutturals and very hard aspirates, and espe- 
cially a number of confused nasal consonants, which 
our alphabets are obliged to render by two letters, ng, 
nk, nd, nt, mb, mp, &c. ; these commonly occur at the 
beginning of words. 

In so far as they are agglutinative and alliteral, the 
African tongues resemble the Dravidian, Malay, Fin- 
nish, and Turkish groups. It is a moral resemblance, 
the sign of the same intellectual level, manifested 



1 64 Distribution of Languages and Races. 

at the moment when the languages were fixed by a 
similarity of method. 

Africa was a centre of production; its human and 
linguistic types, wide as are the differences among 
them, are manifestly autochthonous. The Mediter-' 
raneans of the north and the Semites who have pene- 
trated the eastern frontier have exercised on the black 
and yellowish masses a certain influence, but rather 
physical than intellectual. 

The avoidance of an accumulation of consonants is a 
trait common to all these languages ; there is anotherj 
this time a matter of grammar. They have a strange 
conception of number and gender. The African lan- 
guages generally divide objects into two categories, 
animate and inanimate. They again divide the animate 
into two classes, not according to sex, but according 
to intelligence, that is to say, into men and brutes ; 
thus they have a neuter, and two degrees correspond- 
ing to a rude classification of the living world, but 
they have no masculine and feminine properly so called. 
With regard to number, some have two plurals, apply- 
ing the one to things of the same nature, the other to 
a collection of iniscellaneous objects. 

From likenesses we pass to differences. One is 
sufficiently marked to claim consideration at the 
outset, since it separates into two irreducible groups 
the Guinea system and the Kaffir system. The latter 
places before the root or theme the syllables which 
modify or define its sense, the other employs suffixes, 
or rather places after the radical the particles which 
correspond to our verbal terminations and case-end- 
ings. Prefixation, which is not rare in the languages 
of Europe and Asia, but which here is exclusively em- 
ployed, constitutes the originality of the Bantu group. 



African Races and Languages. 165 

Bosjesraan and Hottentot, whose relationship, al- 
though not proved, is nevertheless probable, are 
distinguished by a very remarkable peculiarity of 
pronunciation, the clucking of the tongue against the 
palate, cheeks, or the teeth ; these sounds are called 
Milts, and are very varied and difBcult to reproduce. 
There are six or seven in Bosjesman ; Hottentot has 
only four left, of which some traces are found in 
certain Kaffir dialects. Livingstone reports that be 
recognised the Bosjesman patois in the neighbour- 
hood of the great lakes, far to the north of their 
present home. Other authors think that they trace 
analogies between Hottentot and some of the Nile 
dialects. It is on such data, somewhat uncertain, 
that is based the probable opinion of the slow retreat 
of the Bushmen before the Bantu invasion. The ex- 
treme antiquity of these tribes is moreover attested 
by the kWks, in which we trace a resemblance to the 
sounds produced by angry or excited monkeys. 

The language of the Bushmen proper is very little 
known ; that of tlie Hottentots, Bushmen with an ad- 
mixture of other blood, and somewhat more civilised, 
has been a good deal studied. It is rich and varied in 
sound. Although complex in appearance, the formation 
of the words does not exceed the ordinary methods 
of agglutination. The root is always placed first, fol- 
lowed by the derivative elements. Tims, since the 
suffix differs with the subject, object, or vocative, and 
since each suffix has three forms corresponding to 
the singular, dual, and plural, it follows that a single 
word can have nine different forms ; but the root 
remains and gives the sense. The function of the 
various suffixes is easily recognised, and, compared to 
the simplest Indo-European declension, the Hottentot 
12 



1 66 Distribution of Languages and Races. 

xiiachinery is simple to childishness. Like Chinese 
and Annamite, Hottentots have numerous homophones, 
that is to say, words which have the same sound and 
correspond to several meanings; these are distinguished 
by intonation. Thus the word hoQ) signifies, according 
to the intonation, obscurity, place, or linen. Accent 
also helps to the comprehension of the language ; it is 
always placed on the first or root syllable ; in com- 
pound words, that is to say, when two or more roots 
precede the suffix, the accent remains on the princi- 
pal word, on the first syllable. Hottentot is, like the 
language of the Bushmen, in process of extinction ; 
its principal dialect, Nama, is spoken by not more 
than twenty thousand individuals. 

We have enumerated above the principal divisions 
of the Kaffir family. Its dialects may be traced back, 
both by grammar and by vocabulary, to a common 
origin, a mother - tongue of which these are the 
varieties. It is remar liable not only for the use of 
prefixes, but also for the almost inflected character of 
its vowel system. This group is in this particular 
much in advance of most agglutinative languages. 
Here are a few examples of both characteristics. In 
Kaffir the prefixes of the singular are v/m and Hi) of 
the plural, aha, ama. Ntu, man, gives um-ntu, the 
jnan, aha-ntu^ the men ; zvi, word, ili-zvi, the word, 
amazvi, the words. Hence the name of the Zulus is 
Ama-Zulu, and the prefix ama constantly recurs in the 
narratives of travellers, when they give the names of 
the tribes of East Africa. The suffixes of case are 
also prefixed. 

Various forms of the word man in the singular and 
plural will give a sufficient idea of the phonetic 
variation. The word is tu, often nasalised into ntu. 



African Races and Languages. 167 

The suffixes are, as we have just said, um and aha or 
ama. Now we find in the singular in Zulu umu-ntu ; 
in Congo, omu-ntu; in Tete, mu-nttu ; in Kisambala, 
mu-ntu ; in Isubu, mo-tu ; and in the plural respec- 
tively aba-ntu, wa-ntu, ia-tu. Herero, which is softer 
in sound, has ova-ndu, va-ndu. The Va-Herero have 
the unfortunate custom of filing the front teeth of 
the upper jaw, and of extracting the four correspond- 
ing teeth of the lower jaw. Hence their lisping pro- 
nunciation, which resembles the imperfect speech of a 
child, has no liquids and no true sibilants. L, r, s,f, are 
wanting, and their z halts between the hard and soft th. 
These Bantu languages, Max Mtiller tells us, from 
the data furnished by Bleek, are generally alike in the 
simplicity of their syllables, which begin by a single 
consonant, preceded by a half-articulated vowel, 
perhaps the remains of an atrophied suffix, or by a 
double consonant (pt, ht, ks), or by a nasalised con- 
sonant, or accompanied by a clucking of the tongue, 
or followed by the semi-vowel w. All these groups are 
considered very simple. Lastly, the syllable cannot 
end in a consonant. Baptize becomes hapitizesha ; 
gold, igolide ; camel, nlcamela; bear, ibere; priest, 
mperesite; kirk, ikerike; apostle, mposile; sugar, 
isugile; English, ama-nge-si. These examples are 
given by Appleyard. The differences between Kafiir 
and its dialects consist almost entirely in changes 
of consonant, often very unexpected changes. Thus 
Sechnana is wanting in the hard g and the soft s, both 
found in Kaffir ; on the other hand, it possesses the r 
where Kaffir has only an I. Kaffir prefers the sounds 
b, d, g, V, z; Sechuana the stronger consonants p, t, k, 
f, s. The consonantal diphthongs of Kaffir and the 
Mpongue group, such as mb, ts, are hardened in 



1 68 Distribution of Languages and Races. 

SecBuana into 'p. The dentals permute with the 
linguals. 

We must note a peculiarity which has contributed 
not a little to the confusion of the Bantu dialects. 
"The Kaffir women," says Appleyard, "have many 
words peculiar to themselves. This is the result of a 
custom called ukuUonipa, which forbids them to pro- 
nounce those words in which is found a sound which 
also occurs in the names of their nearest male re- 
latives." An analogous custom, tepi, which banishes 
from the language of Tahiti the syllables which com- 
pose the names of the kings and queens, has also 
existed among the ancient Kaffirs. Thus the Ama- 
mbalu, out of respect for their chief U-la-iiya, replace 
the word ilanga, sun, by the word isota. For a similar 
reason the Amagqunu-kwebi use the word 'mmela, 
immela, instead of 'si-she-tshe, which is the general 
term for knife. It is easy to imagine the confusion 
which such quaint customs have produced, repeated 
throughout long generations. It is curious to find 
these puerilities among two such different races as the 
Kaffirs and the Polynesians. Max Miiller, in pursuit 
of his chimeras, Turanianism and Monogenesis, sees in 
the phenomenon the result of I know not what ethnic 
relationship. I am rather inclined to regard this coin- 
cidence as the effect of a similar social and mental 
condition, the servile and superstitious respect of 
chiefs and ancestors. We may also add, a naive ten- 
dency to create new words, to vary and maltreat old 
words; a tendency which is visible in the various slang 
dialects, and noted also among the Indians of America. 

We will not leave the Kaffirs, who represent the 
highest elements of the negro race, without some 
further study of their intelligence, without giving some 



African Races a?id Languages. 1 69 

examples, for instance, of the way in whicli they pass 
from the concrete to the figurative sense. It is a 
well-known phenomenon, and may be seen in all 
languages. But it is impossible to insist too much 
upon the metaphorical origin of language. Beta, to 
beat, to strike, becomes, to punish, to judge; dhle-la- 
7ia, to eat in company, is to have friendly relations ; 
fa, to die, to be ill, to languish ; Mala, to be seated, 
to live, to dwell, to remain; ihladi, bush, shelter (a 
reminiscence of the Bushmen); ingcala, winged ant, 
skill, rapidity ; inncwadi, a reed, book, vessel ; inja, 
dog, an inferior; Icolwa, to be satisfied, to believe; lila, 
to weep, to deplore; mnandi, soft to the touch, content, 
agreeable ; gauka, to be broken in two, to be dead or 
stupefied; umsila, tail, courtier, or court messenger; 
akasiboni, he does not see us, he despises us ; nikela 
indhlehe, to give ear, to listen ; ukudhla uhomi, to eat 
, life, to live ; ukudhla umntu, to eat a man, to confiscate 
his goods ; ukumgekeza inholoh, to break the head, to 
weary, to bore ; tikunuka umntu, to smell some one, 
to accuse him of witchcraft. 

The Bantu peoples are not related to the Negroes 
of Guinea, of Senegambia, of the Soudan, at any rate 
in language. There is no similarity between their 
vocabulary and that of the numerous Negro groups 
of which I have given the most important. Their 
grammar also separates them by the exclusive use 
of prefixation. They have the privilege of forming 
a linguistic family, and of lending themselves to the 
study of comparative philology. 

Elsewhere in Africa we find nothing but dispersed 
and isolated groups ; more than twenty little groups 
divided up into tribes and dialects succeed and inter- 
mingle with each other Irom Gaboun to Morocco, 



170 Distribution of Languages and Races. 

from Lake Tchad to the Atlantic. The most central, 
towards Bornou and Kanam, has been scientifically- 
studied by Dr. Barfch (1862). The most important 
dialects are Kanuri, with five cases and numerous 
verbal forms. Earth thinks it has some relations with 
Egyptian, Coptic, and even Finnish, but especially with 
the languages of the coast, Odji, Fanti, and Ashanti. 
Teda, in spite of the difference of the pronouns, is 
closely allied to Kanuri. Haoussa, a harmonious 
language, belongs to a mixed race, industrious and 
evidently superior to the neighbouring tribes ; it is 
understood in the markets from Timbuctoo and even 
into Senegambia. Barth has translated into Haoussa 
the second chapter of St. Matthew. Shall I name 
also Fufulde or Fulfude, Songui, Logona, Wandala, 
Bagrimma, and Maba, Tibbu, Goura, Legbe^ Roama, 
Kasm, Gbali, and the nine barbarous patois which are 
spoken round Lake Tchad ? 

Among the innumerable dialects spoken in Guinea 
and Senegambia we will mention only those of which 
the knowledge is important to the French military occu- 
pation : Mandingue, Malinke, Dialonke, with thirteen 
dialects, on the Gambia and the Niger ; Eiwe or Egbe, 
studied by the missionary Steinmann, with which is 
connected the idiom of Dahomey ; finally, Wolof, Serere, 
Bidchoro, &c., spoken in Cayor and in the French 
colony of Senegal. Eiwe is alliteral ; the English word 
school becomes su-ku, and the German Fenster, fesre. 
Wolof presents the same character ; it is, moreover, 
very nasal, without being therefore less harmonious 
and rhythmical. Markedly agglutinative, it obtains by 
means of divers suffixes seventeen voices to the verbs, 
and several shades in the meaning of the nouns, 
according as the object is near or far. 



African Races and Languages. 171 

Throughout the Malinke territory we find Peul or 
Poul, which is supposed to be connected with the 
Nubian group, and through that perhaps with ancient 
Egyptian. But too many changes have taken place 
during three or four thousand years, over a distance of 
more than 2000 miles, to make it possible that any of 
these hypotheses should ever be confirmed. However 
this may be, the eastern origin of the race admits of 
no doubt. The language is totally foreign to the 
peoples which this race has conquered or dominated ; 
that which it has in common with Wolof and Serere 
is the result of reciprocal borrowing, and is noticeable 
especially in certain dialects of Poul : Foutatoro, Fouta- 
djallo, Bondu, Sokoto. It were as reasonable to connect 
it with Arab because Islam has introduced into it a 
number of terras relating to religion, law, and similar 
subjects. 

Poul has no guttural aspirates, and it rejects also 
ch and j. Its conception of gender is approximately 
that which we indicated above. Beings are divided 
into two categories, which Faidherbe calls the human 
and brute genders : in the one animals and inanimate 
things, in the other all which belongs to humanity. 
This capital distinction gives to the declension an 
appearance of complexity ; there are two singulars and 
two plurals. The nouns, adjectives, and participles which 
belong to the human gender, end in in the singular, 
this vowel is an agglutinated pronominal root : gorko, 
man ; in the plural these words end in le (they). In 
the brute gender the singular is marked by a vowel, 
or by the suffix am ; the termination is rare. The 
plural of the brute gender varies, and certain euphonic 
laws seem to play a great part in the agglutination of 
the terminations with the root. The initial consonants 



1 7 2 Distribution of Languages and Races. 

of tke word in the singular may permute with others 
when the word is in the plural. The verb remains 
much simpler, and the analysis of its component 
elements is easy. The syntax is not complicated ; the 
order of the words in the sentence is determined by 
the succession of ideas. Thus the name of the pos- 
sessor is preceded by the name of the thing possessed ; 
the object, direct or indirect, follows the verb. It will 
be seen that the real difficulty of Poul consists in the 
great variety of the laws of euphony., 

On the left flank of Poul, and not far from its pro- 
bable birthplace, the Dinka family vegetates ; poor 
and almost monosyllabic, it suffices for the true savages 
of the Bahr-el-Abiad and of the left bank of the Upper 
Nile, Bongos, Dinkas, Monbottos, Nouers, Niam, who 
are still cannibals. A little farther north we find the 
Nouba group, of which it would be of great value to 
have a thorough knowledge, in order to determine 
whether it has any connection with the Libyan dialects. 
The districts which it occupies were formerly the refuge 
of the Pharaohs when driven out by the Hyksos ; 
Egyptian civilisation ascended the Nile as far as above 
Meroe, and left pyramids and temples in those regions, 
now once more given over to barbarism. The Shabak 
and the Tahraka came from Napata towards the end of 
the seventh century before our era to defend Egypt, 
enervated by the theocracy, against the Assyrian inva- 
sion. A few years ago, M. Benedite, in a stay of some 
length at Philse, made a study of Kensi, but without 
much result ; the dialect is in course of extinction, and 
the whole language consists in a few soft and easy 
phrases which are used by the boatmen of the Cataracts. 

We have several times mentioned the Libyan or 
Egypto-Berber group, which has occupied the whole, 



African Races and Languages. 173 

and still occupies a part of tte Mediterranean zone, 
and the eastern point of Africa as far south as 
the Gulf of Aden. It is regarded as the transition 
between the agglutinative and the inflected languages ; 
nevertheless, it remains much nearer the first, and we 
cannot leave Africa without having given at least a 
sketch of it. We will, therefore, briefly define the 
three principal branches of the Libyan family : Berber, 
Ethiopian, Egyptian. 

Phoenician, Greek, Latin, and finally Arabic have been 
spoken in succession or simultaneously on those shores 
where the Libyan tongue was formerly spoken by the 
Numidians, the Getulse, and the Mauritanians. Never- 
theless, certain dialects have survived and are still 
extant in Algeria, in Morocco, Tunisia, and beyond 
the Sahara in the Upper Soudan. Such are Kabyl, 
Mozabi, Chaouya, Zenatya (in the environs of Con- 
stantine), Tamachek, Touareg. The Berber or Amazig 
language — for the resemblances . between its various 
vocabularies is suflSciently close to allow us to consider 
the language as a whole — still occupies a very vast 
domain, which formerly extended to the Canaries, the 
land of the Guanchos. It is a rude irregular idiom, 
modified by Semitism, but African in the power to use 
prefixes, and by the polysynthetism of its verbs ; it 
has, like Basque, a doubly reflective voice, expressing 
in a single word such phrases as je rrien doute. But 
the verb has but one tense, a sort of aorist, to which 
the ideas of present and future are attached by methods 
which are altogether accessory. The sign of the 
feminine is t : amaher, a Touareg man ; tamaher, a 
Touareg woman. This sign is often both prefixed and 
sufiixed : akli, negro, taklit, negress ; ekahi, cock, 
tekahit, hen. The Berber language has been written ; 



1 74 Distribution of Languages and Races. 

there are a few fables and poems in this idiom, and 
Touareg inscriptions graven on rocks have been dis- 
covered, with twenty-eight peculiar characters, evi- 
dently of Semitic origin, and composing perhaps the 
Numidian alphabet mentioned by Valerius andMaximus. 
Tamachek has still a written character, fairly regular 
in appearance, but destitute of signs for the vowels ; 
it is necessary to know the language in order to be 
able to read it. 

The Ethiopian branch, Galla, Bedja, Saho, Dankali, 
and Somali, must not be confounded, it seems, with ^ 
the decidedly Semitic idioms of Abyssinia, Tigre, 
Amharic, and others, which are connected through 
Ghez with Himyarite of the coast of Arabia. Although 
influenced by Semitism, the Ethiopian dialects (very 
little studied from the point of view of comparative 
philology), belong to the Libyan family by the use of 
the sign t for the feminine, which may be either a 
prefix or a suffix. We note, however, that the two 
tenses of Bedja and Saho are expressed in a manner 
which is purely Semitic. The one, the aorist, is- 
indicated by the prefixing of the personal pronouns ; 
the other, the present, by the post-position of these 
pronouns. The same methods are employed, but indif- 
ferently, by Coptic, which uses auxiliaries to distinguish 
the tenses. 

Coptic, extinct since the seventeenth century, but 
still in use as the sacred language of the Monothelite 
sect, had from the second to the eighth century a 
fairly rich literature, very precious to those who are 
attracted by the minutiae of Christian exegesis. It 
was, as the name indicates (ha-ka-ptah, aiyvirro^ ; 
Guptos, Copt), the popular form of Pharaonic Egyptian, 
and it is through Coptic that Egyptologists have found 



African Races and Languages. 1 7 5 

out how to decipher the annals of the Snefrou and the 
Rameses. 

The discovery of ancient Egypt is one of the finest 
conquests of the century, and it is a French conquest. 
The essays of Scholtz and of Barth^lemy presented 
only hypotheses which were hardly even ingenious 
(1775)- We must look in the works of Champollion 
the younger, the first reader of the hieroglyphs (1790— 
1832), and of his successors Rosellini, Salvolini, Lepsius, 
Brugech, Rouge, Maspero, for the certain and estab- 
lished principles of decipherment. The reading has 
changed more than once ; the syllabic, alphabetic, or 
ideographic character of the signs has had to be deter- 
mined, and the hieroglyphs checked by the hieratic 
writings, which are more summary in form, and by 
the cursive or demotic, the instrument of the language 
of affairs and of common speech. 

There were in Egypt, from a very early epoch, two 
languages, the one sacred, the other popular, which 
soon presented marked differences, of which the prin- 
cipal (Lepsius proves this from the Rosetta inscription) 
was the preference of the demotic form for prefixation. 
That which the priests wrote after the root or theme, 
pronominal signs, affixes of tense, number, and gender, 
the people preferred to place at the beginning of the 
word, as in the Bantu group. 

The Egyptian language is very simple : a feminine 
of which the sign is ^ ; a plural in m, vA ; no oases. 
A syntax which rules that the verb shall occupy the 
first place in the sentence, followed by the subject, the 
direct object, the indirect object, and lastly the adverb, 
write I letter to you to-morrow ; a formula which does 
not lend itself to eloquence or poetry, and which yet 
has sufficed for the emphatic proclamations of the 



176 Distribution of Languages and Races. 

kings, for the precepts of a lofty morality, and for the 
religious and philosophical lucubrations preserved by 
the priests in the " Book of the Dead." Triumphal 
odes have also been found, romances, and even medical 
treatises, which make the joy of Egyptologists. 

But the really important discovery is the probable 
relationship, now recognised by Fr. Miiller and by 
Maspero, of the Libyan and Semitic languages. In 
the two groups there are the same pronouns, the same 
method of forming the plural by the addition of a 
termination. The two families must have separated 
at an epoch when their common language was at a 
very early stage of its development. The one came to 
a standstill, the other advanced towards the inflected 
state. But where were they together, and wbere did 
they part company ? Did the Berbers come from 
Asia, or are the Semites a Mediterranean people, who, 
crossing the delta of the Nile, spread towards the 
Euphrates and into Arabia ? The first opinion has 
in its favour the prejudice which considers Asia as 
the cradle of the human race, at least of the white 
and yellow races. But the second seems to derive 
some support from the data of pre-historic anthro- 
pology. The question is unsolved, and the field open 
to conjecture. 

For the rest, everything is vague and obscure in the 
unknown past of Africa. We have endeavoured to 
present its probable phases ; in the north a flux 
and reflux of white peoples, causing the Nubians to 
retreat upon the Bantus, and the Bantus upon the 
Bushmen and Hottentots. In the east, a Semitic 
invasion, mingling the languages and the races of the 
Gallas, the Somalis, and the Abyssinians, and deter- 
mining the exodus of the Penis across Bornou and 



African Races and Languages. 177 

the Soudan as far as Senegambia. On the extreme 
west, a descent of the Moors or Berbers, adding to the 
confusion of the groups and idioms of Nigritia and 
Guinea. In the centre, a remnant of divers savage 
races, tall or dwarfish, black or chocolate, the Bongos, 
the Dinkas, Nouers, Niam-Niain, Akkas, pressed to- 
gether pell-mell between the Nubians and the Peuls, 
between the Abyssinians, the Bantus of the lakes and 
of the Congo, and the Negroes of the Lower Niger 
and of Gaboun. A great civilisation in the valley of 
the Nile ; semi-barbarism on the northern coasts, in 
the basin of the Niger, and on the south-eastern coast ; 
everywhere else every shade and degree of moral and 
intellectual poverty, of a merely animal existence. 



CHAPTER VI. 
POLYSYNTHETIC LANGUAGES. 

The Basques — Complete isolation of the Basque or Uskara tongue — 
Incorporating or abbreviating character of this agglutinative idiom 
— Persistence of Basque customs — Origin of the Basques — 
Songs of Altabiscar and of the Cantabri — The American races 
and idioms — Has America an indigenous race — Probable Asiatic 
origin of the successive strata of the population — Fanciful com- 
parisons between the American religions and the Hindu or Egyp- 
tian beliefs — Table of races, general characteristics, and variety of 
the families of languages — Examples and decomposition of poly- 
synthetic terms — Life and language of the Inuit or Esquimaux — 
Iroquois and Algonquin group — The plateau of Anahuac — Central 
America — Peru — General review of the agglutinative languages. 

Before leaving tlie Old World we must say a few- 
words about a curious idiom which bears some resem- 
blance to the American languages, and which is found 
quite isolated in a district of the Pyrenees, between 
the Pic d'Anie and Biarritz, between Pampeluna and 
Bilbao. This is Eskuara, Euskara, or Uskara, spoken 
in three French arrondissements (Bayonne, Oloron, 
MauMon), and three Spanish provinces (Alava, Gui- 
puzcoa, Biscaya). Uskara, which from the tenth cen- 
tury has been the wonder of the Gallo-Romans and 
Gallo-Pranks, is simply an agglutinative language, at 
once poor and complex, like all the dialects belonging 
to the same class ; poor by its vocabulary — omitting 
of course the words borrowed from Latin, Spanish, 
Arabic, and French — complex by the richness of its 

sounds, by the delicacy of its euphonic laws, by the 

178 



Polysynthetic Languages. 179 

double use, in conjugation, of suflBxes and auxiliaries, 
and by the abbreviation of words in compounds. The 
contrast of such a language, understood by hardly half 
a million of individuals, with the Neo-Latin dialects, 
which drove it back into the mountains, caused its im- 
portance and difficulty to be exaggerated by observers 
who were accustomed to quite different conceptions of 
the word and the sentence. It is barely thirty years 
since Uskara, more seri(jusly studied, has ceased to be 
a mystery. By degrees it has been discovered that its 
declensions consists in the post- position of numerous 
suffixes : of, in, by, for, with, without, towards, as far 
as, &c. ; that its primitive conjugation admits of the 
incorporation of the direct and indirect objects, and 
of numerous shades obtained by the accumulation of 
suffixes ; that the moods and tenses, at first consisting 
only of the- indicative, of the past and present, were 
enriched by the relatively modern use of the auxi- 
liaries naiz and dut, to be and to have, of which the 
termination varies according to the person addressed, 
whether a man, a woman, or a superior (this is known 
as the periphrastic conjugation); finally, that the famous 
compounds, called polysynthetic, -orteana, thunder, for 
ortz-azanz (cloud-noise) ; arkume, lamb, for ardi-hume 
(sheep-little) ; sagarno, cyder, for sagar-arno (apple- 
wine) ; Yainkoa, god, for Yaim-Goikoa, the lord on 
high, or rather the lord-moon, — it has been discovered, 
I say, that these famous compounds by apocope (which 
are yet more common in the American dialects) have 
parallels everywhere, and do not differ from idolatry 
for idololatry, hidalgo for hijo de algo, usted for vuestra 
merced, mamzell for mademoiselle. Thus, thanks to the 
labours of Prince Lucien Bonaparte, of MM. Van Eis 
and Julien Vinson, Uskara, in spite of its complicated 



i8o Distribution of Languages and Races. 

forms, its irregularity, and its constantly inverted 
phrase, has lost much of its strangeness. 

Its real intei-est lies, not in a few Uskara names 
of places mentioned in charters, letters-patent, Papal 
bulls', &c., from the year 980, nor in the little speech 
of Panurge in tbe second book of Eabelais, 1542, 
nor in the love-poems of a priest (1545), nor in 
the translation of the New Testament printed at 
La Eochelle in i 5 7 1 by the order of Jeanne d'Albret, 
nor in the extant sermons and pious tracts, nor even in 
the curious pastorals recently published and translated 
by Julien Vinson ; it lies in the long, the immemorial 
existence in a Gallo-Roman country of an idiom 
made for Bantus, Dravidians, or Algonquins. Suomi, 
Magyar, and Turkish have been imported into Europe 
by invasions of which the date is known ; but the 
establishment at the foot of the Pyrenees of the 
Uskara language and of those who speak it, the 
Escualdunac, is a fact anterior to history, and one of 
which neither anthropology nor ethnography render 
any account. These Vascons, Vascongados, Bascli, 
Basques — such is the name by which they have been 
known from antiquity^ — have lost since the seventeenth 
century their peculiar customs, their habit of carrying 
three little javelins, and their fierce manners; they have 
retained in Spain those fueros of ancient date which 
testify to their long independence, and in France the 
love of the game of tennis, of certain dances, and of 
open-air fetes; but they have become very peaceable and 
honest citizens, if very ignorant and very devout. As 
for their type, studied by Broca, it is mixed, passing 
from a marked dolichocephalous skull to a true sub- 
brachycephalous. At the utmost, the occipital develop- 
ment of their skull can be taken for a sign of race. 



Polysynthetic Languages. 1 8 1 

The Basques, notwithstanding the similarity of the 
name, are not Gascons or Aquitani. Are they autoch- 
thonous, the. last remains of an age when the aggluti- 
native form of language prevailed in Europe as in the 
rest of the world ? And while their kindred retreated 
towards the north with the mammoth and the reindeer, 
did they take refuge with the bear in the Pyrenees, 
together with the Navarais and the Asturians or Can- 
tabri ? Proof is wanting, and can never be obtained, 
but the hypothesis is plausible. 

The most general opinion, which has perhaps been 
too lightly adopted, connects the Basques with the 
Iberians, an ancient Mediterranean race, doubtless Afri- 
can, which occupied the south of Western Europe 
before the coming of the Ligurians and Celts. Tlie 
evident fact that the Basques are established in regions 
necessarily occupied, or at least traversed, by the 
Iberians or Celtiberians is urged in support of this 
opinion. Some ancient names of places in Spain, 
[lliberis, Ebro, have been connected with Basque roots 
or terminations, as also a few words which have been 
deciphered on coins and medals or on inscriptions, 
Iberian words in the Latin character. But the inter- 
pretation of these documents is most uncertain, and the 
arguments produced are therefore iiot convincing. It 
is not known what language "the Iberians spoke, nor 
whether the Basques did not precede them in Spain or 
in the Pyrenean district. As for the comparisons which 
have been attempted, with regard either to the race 
or to the language, with the PhcBnicians, the Finns, 
the Magyars, the Berbers, or the Indians of North 
America, they cannot be considered seriously* Among 
agglutinative dialects resemblances do not, as we have 
already said, imply relationship. With, regard to the 
13 



1 82 Distribution of Languages and Races. 

race, whatever its origin, it has become completely 
European. The Basques for a long time took pride 
in their isolation, in their language unknown to other 
men, which they considered as the original language 
of humanity, the universal mother-tongue. These 
dreams may be found in the work of their gramma- 
rians, Ohaho and Inchauspe. The Basques resisted the 
Eomans, as they had perhaps resisted the Celts and 
the Iberians ; they destroyed at Roncesvalles the rear- 
guard of Charlemagne. The song of Altabiscar, 
which was at one-time believed to be a contemporary 
poem, but which is really several centuries later, and 
seems to have been thought in French, celebrates 
worthily this savage victory. It is the finest page in 
their literature : — 

" A cry was heard in the heart of the Basque 
mountains; it is the tumult of an army. Ours replied 
from the summits, and the chief of the clan sharpened 
his javelins. Why did they come to trouble our 
peace ? The Lord who made the mountains willed 
thfit men should not cross them. The rocks fall, roll, 
crushing the troops. Blood flows in torrents. What 
broken bones ! what a sea of blood ! Flee, King 
Carloman, with thy black plumes and thy red cloak. 
Thy loved nephew, the brave Roland, lies slain below. 
Chieftain, it is over ; go and embrace your wife and 
children. The eagles will devour these remains, and 
their bones will whiten there to all eternity." 

The song of the Cantabri, even less authentic, if that 
be possible, celebrates a strife with Augustus C^sar : 
Octavius is the master of the world, Lecobidi of 
Biscay a. . , , "For five years they have blocked 
our mountains ; they are many, and we are few. They 
will flee, if they can, towards the banks of the Tiber." 



Poly synthetic Languages. 183 

This is very well ; but we need to know whether the 
Cantabri were Basques. The refrain, Ld U lelo leloa, 
appears to be a reminiscence of the Arabic formula 
Allah, il Allah. 

If we are unable to solve the problem of the origin 
of the Basques, we are yet more bafHed by the mystery 
oE the origin of the Indians of America. The Escu- 
aldunac have at any rate a long history, and the 
Europeans only landed in America four centuries ago, 
to exterminate the inhabitants and destroy all traces 
of their history. Thus the discussion as to the origin 
of the different tribes and nations, scattered over a 
continent four times as large as Europe, might be end- 
less. It is generally doubted that America had any 
indigenous race ; the absence of great apes, even of 
fossil ones, seems to exclude the possibility of a pre- 
liminary anthropoid evolution. Yet in America, as in 
Europe, and throughout the different pei'iods of the 
quaternary epoch, man has gone through the same 
industrial stages, passing from the rough stone imple- 
ment to tools of polished stone and of copper, beyond 
which he had not got at the time of the European inva- 
sion. Man in America seems to have had very humble 
beginnings, if, as appears established, his refuge in 
the Pampas consisted in holes hollowed out beneath 
the carapace of giant reptiles, now extinct ; and he 
appears also to have been endowed with a very slight 
degree of perfectibility, if we are to recognise him 
in the fierce Charrua, in the stupid B'otocudo, in the 
hateful Apache, or the hungry Californian. Asia, by 
way of the Behring Straits and the Aleutian Islands, 
seems to have sent to Polar and Northern America 
slightly more advanced specimens of humanity : first, 
the Esquimaux, the near relations of the Tchbukches 



1 84 Distribution of Languages aii'd Races. 

and Koriaks ; then various Mongoloids of dark colour, 
the Sioux, Dacotas, Pawnees, Natchez, Seminoles, fol- 
lowed by the Algonquins, Hurons, Iroquois, Delawares, 
and behind these, closing the procession, the peoples 
of Athabasca, of Mackenzie, all of whom, pressing 
forward to Florida and Mexico, drove towards the 
extreme north the unhappy Esquimaux. This hypo- 
thesis offers an ex;planation of the progress along the 
Ohio and the Mississippi of the constructors of the 
mounds or tumuli, and the successive immigrations of 
the Guatemaltecs, Yucatecs, Othomis, Quiches, Mayas, 
phichimecs, and Aztecs, gradually heaped up between 
the Isthmus and Texas. But in South America the 
marked differences of stature, colour, physiognomy, 
and cranial capacity are a puzzle to the ethnologist. 
The wisest course is merely to note the thick-set form 
of the Brazilian savages, both the conquerors of ancient 
times, Guaranis, Caribs, and Tupis, and the conquered 
races, Botocudos and Tapuyas ; the middle height, the 
aquiline nose, and retreating forehead of the Peruvians; 
the tall stature of the Patagonians, and the miserable 
character of the Fuegians, who sometimes eat an old 
woman for want of something better, and who replied 
to Darwin's question as to whether they preferred 
their wife or their dog, that they preferred the dog 
because he could catch the otter. 

It is, of course, possible that among these diverse 
types some should be of foreign origin, but whence 
could they have come ? What wind bore them to this 
immense peninsula, isolated between two vast oceans ? 
There is no answer to this question. Some of the 
beliefs, and some of the constructions of Mexico and 
Peru have been compared with the religion and the 
arts of Egypt and India. This, in my opinion, is quite 



Polysynthetic Languages. 185 

cMmerical. If there are coincidences, they are for- 
tuitous, or they result from evolution, which leads alt 
the human groups through the same stages and by the 
same steps. 

Observers have also sought to trace in North 
America some sign, physical or intellectual, of the 
Scandinavians, who towards the tenth century probably 
discovered, and to some extent colonised, the coast of 
Greenland and of North-Eastern America. Without 
disputing the fact, which appears to be established, 
and even admitted, that a slight infusion of European 
blood has been propagated in the scattered tribes of 
the Eed Indian hunters, we do not see that the 
dolichocephalous Esquimaux have been modified by it, 
or that the deep-set eyes, the heavy jaw, the narrow 
foreheads of the Assiniboines and Cherokees have been 
altered. In order to improve upon these faces, for 
which nature has not done much, the recent but 
repeated crossing with the French Canadians was 
necessary. 

No part of the world was so sparsely peopled as 
America. Tribes which live chiefly by hunting need 
a great deal of space, and the continual wars which 
they wage in order to keep their territory clear from 
the encroachments of their neighbours do not further 
the increase of the groups. Only in the region of the 
centre were there compact populations, in Mexico, the 
Antilles, the Isthmus, Columbia, and Pern, where a 
true civilisation flourished, unhappily extinguished by 
the Spanish conquest. The jealous isolation main- 
tained by most of the tribes prevented the fusion of 
the dialects, even of those which were near neighbours; 
and linguists reckon, between the Arctic Ocean and 
Cape Horn, as many as twenty-seven families of Ian- 



] 86 Distribution of Languages and Races. 

guages, not related to each other, and belonging for 
the most part to the agglutinative class. 

Omitting certain independent and little known 
idioms of Sonora and Texas, of Mexico, the Andes, 
Guatemala, and of the Antilles, which have not got 
beyond the monosyllabic phase, we find the following 
twenty-three groups : — On the shore of the Arctic 
Ocean and in the Aleoutian Islands, Inuit or Esquimaux; 
to the north-west, Kenia, in Alaska; in the west of 
the Dominion, on the shores of the Athabasca, Ata- 
pache ; to the south of Hudson's Bay, Algonquin ; on 
the St. Lawrence, around the great lakes, Iroquois ; 
Koloche, to the west of New Britain ; in the United 
States, Oregonese, Calif ornian, Yuma, Dakota or Sioux, 
Pawnee, Alapache ; Maya, in Yucatan ; Mexican, 
Nahuatli or Aztec ; Ohiboha, among the Muyscas 
of Columbia ; Carib and Arevac, in the Guianas ; 
Quichua-Aymara, in Brazil and Parana; Guaycuru and 
Abipone, in La Plata ; Araucan, in Chili ; Pnelche, in 
the Pampas to the west of Buenos Ayres ; Tehuelche, 
in Patagonia ; and the idioms of Terra del Puego. 

Before considering any of these groups, all rich in 
dialects, we must point out those characters which they 
have in common. For their general resemblance is 
as striking as the variety of their constituent parts. 
They are all incorporating and polysynthetic. " The 
American languages," says Frederick Miiller (" General 
Ethnography "), " repose as a whole upon the principle 
of polysynthetism or of incorporation : thus the sepa^ 
rate ideas, which, in our languages, the phrase unites 
together" under the form of detached words, are, on the 
contrary, in the American idioms, united together in an 
indivisible unity." This method has seemed to many 
students of these languages to be sufficiently typical to 



Poly synthetic Languages. 1 87 

justify th.e creation of a new class of languages. But 
it is easy to see that many other languages, agglutina- 
tive or inflected, have made use of the methods which 
the American dialects abuse. 

The incorporation of the objective pronouns, and 
even of nouns, exists in Basque, as we have mentioned 
above (hence by some it has been connected with the 
American languages) ; we also noted it in the Finnish 
dialects. The incorporation of the subject pronouns is 
the basis of the Indo-European conjugation, dadami, 
dadasi, dadati, I give, thou givest, he gives ; the in- 
corporation of the object prevails in some Romance 
languages : in Itfdian, portandovi, portandovelo, carrying 
to you, carrying it to you ; in Gascon, dechemdroumi, 
let me sleep. The Semitic languages incorporate the 
direct object. In short, the Hebrew sabachthani, thou 
hast forsaken me ; the Magyar latlak, I see him ; the 
Basque deniogu, we give it to hiui ; and the Iroquois 
keiavis, I give it to them, differ in reality, says M. 
Hovelacque, only in the order of the elements of which 
the word is composed. The American languages go far- 
ther, they amalgamate the noun subject with the verb ; 
Algonquin, nadholiTieen, bring us the canoe, formed of 
naten, to bring, amochol, canoe, i euphonic, neen, to us ; 
Iroquois, sogininjinitizoyan, if I do not take the hand, 
into which enter sogena, to take, and oningina, hand. 

The possessive declension of the noun (Iroquois, 
onkiasita, the foot of both of us; Algonquin, nindawema, 
my sister) is used in Hungarian, atynnlc, our father ; 
in Hebrew, eli, my God. There is nothing here 
peculiar to the American dialects ; the French m'amie, 
m'dmour, present a sporadic case of the same incor- 
poration ; in the word tante, for instance, is concealed 
the possessive ta placed before the Latin amita. 



1 88 Distribution of Languages and Races. 

The variation of the verb, when the speaker wishes 
to express changes in the manner of the action or in 
the object, points especially to tlie absence of general 
ideas; it is a mark of inferiority often observed in 
those languages which are not far advanced in their 
evolution, and of which the inflected stage has retained 
some traces. Prom this phenomenon are issued the 
many verbal variants, causative, intensive, desiderative, 
&c., which Arab and Sanscrit obtain by means of the 
insertion of atrophied suffixes, and which make the 
inconvenient and barren wealth of Turkish, and Wolof. 
Why then should we be astonished that the Chilian 
elu%, to give, ramifies into ehyuen, to give more, eluz- 
qmn, to seem to give, eluvcden, to be able to give, 
eluduamen, to wish to give ; or that Tamanacan has 
JUCU7-U, to eat bread, jemeri, to eat fruit, janeri, to eat 
■ cooked things ; and Cherokee, kutuvo, I wash myself, 
tsekusquo, I wash some one else, takiiteya, I wash dishes, 
takuTigkala, I wash my clothes ? 

Polysynthetism proper remains to be considered, a 
species of oral stenography, which amalgamates into a 
compound word four, five, or six words, capriciously 
abridged, with the beginning, middle, or end omitted. 
It is an agglutination of words instead of an aggluti- 
nation of suflSxes ; but we need not go far to find 
examples, very simple ones it is true, which we hardly 
remark. In German, heiTn, for hei dem, zur for zu der ; 
in French, du, au, ds, for de le, it, Je, en les, anjoiir d'hui 
for a le jour de hui (hodie) ; in English, lady for hlaf- 
dige, loaf-giver, are the result oP contractions in every 
way similar to the Chippeway toto-chdbo, wine, formed 
of toto, milk, and cJwminabo, bunch of grapes. Finally, 
in many of our complex derivatives, rapprochement, 
recueillemcnt, &c., the constituent parts of the woi'd 



Polysynthetic Languages. 189 

are hardly less obliterated than in the Algonquin 
pilwpe, bachelor, formed of pilsitt, chaste, and lenape, 
man. These observations, chosen among many, will 
render less singular words such as amaTujanachqioi- 
mincM, large-leaved oak, an Algonquin word composed 
of amangi, large ; nachic, hand ; quim, termination ap- 
plied to fruits with a shell, and aelvpanti, tree trunk ; 
as the Esquimaux .aulisariartora suarpok, he has 
hastened to go fishing ; as the Mexican notlazomahuiz 
teopixcatatzin, my father, divine protector, esteemed 
and venerated. 

In short, the faculty of incorporation and of poly- 
synthetism, very inconvenient when it produces these 
immensely long words, which can hardly be pronounced 
in one breath, is not unknown to the languages of 
Asia and Europe. Developed in America to an extra- 
ordinary degree, it gives to these very various idioms 
a superficial resemblance, an appearance of unity; but 
it does not place them out of, still less beyond, the 
agglutinative class ; it does not bring them nearer 
to the inflected ordei», where what remains of these 
antique methods is of the nature of an atavic 
survival. 

We have only lately been able to consider the 
American languages with anything but extreme re- 
serve and mistrust. The study of these idioms has 
been brought back to its proper place by such men 
as Lucieri Adam and Victor Henry, whose work has 
confirmed the previsions of Hovelacque and of Vinson, 
but it has been given over for three hundred years 
either to the pious fancies of missionaries, seeking to 
find in Ontario or Chili some dialect escaped from 
Babel, or to the illusions of etymologists such as 
Brasseur de Bourbourg, who would connect Nahuatl 



iQO Distribution of Languages and Races. 

with the Teutonic languages, or the jargon of Van- 
couver with both English and French, never inquiring 
whether a given word has not been borrowed, very 
naturally, from the foreign colonists; or, finally, to 
the preconceived theories of the upholders of Tura- 
nianism and Monogenesis, and others who seek 
Aryans, Oopfcs, and Buddhists in Peru, and the 
remnant of the ten tribes of Israel in the Par West. 
The essay of Duponceau on the Canadian languages 
(1836) marks the first step towards the scientific 
classification of the northern group, and it was only 
at the second American Congress at the Luxem- 
bourg that, in his studies on the sixteen groups of 
American idioms, M. Lucien Adam drew up a table 
of which the arrangement will not be disputed by 
specialists. 

We have just seen what common features give to 
the American languages a marked family resemblance. 
But this apparent likeness conceals an extreme diver- 
sity in the vocabularies, and in composition and syntax, 
which take the place of grammar in the agglutinative 
idioms. A detailed study of these differences would be 
out of place in these summaries, of which the chief 
aim, is to determine the extent of the domain of language, 
and the ancient or actual area which the principal 
varieties of language occupy in it. I shall, therefore, 
confine myself to a few groups chosen in the north, 
the centre, and the south of the New World; 

The Inuit (" men," so they call themselves), whom 
in disdain the Mohicans called Eskimanzik, eaters of 
raw meat, whence Esquimaux, seem to deserve our 
attention for more than one reason. First of all, tliey 
liave the rare if slender privilege of constituting a 
pure race ; secondly, they represent, by their extreme 



Polysynthetic Languages. 191 

dolichocephaly and their manner of life, the man of the 
quaternary age, the man of the reindeer ; thirdly, in 
spite of their dirty-yellow skin, their narrow eyes, 
their coarse hair, their flat round face, their thick-set 
ungraceful frame, they are not wanting in courage, in 
gaietyj or in moral and intellectual qualities. The 
Esquimaux of Greenland have become very acceptable 
Danish citizens, municipal councillors and merchants. 
They display a real taste for geographical science. Since 
i860 they have had at Godthaab a printing-press and 
an illustrated journal, of which the text and drawings 
are furnished entirely by the natives ; and they publish 
in their own language collections of popular traditions. 
To the Esquimaux of Greenland belong the fifteen 
hundred natives of Labrador. 

Along the shores of the Arctic Ocean, from Hudson's 
Bay to the northern point of Asia and to Cape Sche- 
lagskoi, are scattered the Great Esquimaux, the Onki- 
lones, the Aleoutes, and the Tchoukches, divided into 
numerous tribes, but numbering in all not more than 
twenty or thirty thousand individuals. They extended 
formerly as far as the valley of the St. Lawrence, per- 
haps as far as Massachusetts. But the Red Indians, 
their hereditary foes, have driven them back towards 
the desolate regions, where they will one day die out, 
in spite of their strength to resist the miseries of their 
hard life. The soil produces nothing but grass in 
the summer and a few stunted shrubs. Reduced to 
hunting and fishing, especially the latter, having for 
assistants only dogs and reindeer, which are hardly 
tamed ; for weapons, whalebone, walrus horns, imple- 
ments of stone, and the drift timber from wrecks or 
forests, brought by favourable currents to their frozen 
shores, they have been able to provide for all their 



192 Distribution of Languages and Races. 

needs. Warmly covered in impermeable clothing, and 
in boots made of sealskin, they embark in their frail 
boats and hurl against the sea-beasts their harpoons 
furnished with the bladders which permit them to 
recover these precious weapons. In the winter they 
make holes in the ice and keep watch for the walrus, 
which comes there to breathe ; they attract them by 
imitating their cry. At night they take refuge in 
huts built of earth and ice, which long galleries 
and furs protect against the cold ; lamps abundantly 
furnished with blubber give out such a heat (not 
without smell) that all clothing becomes superfluous. 
Here they dwell by twenties, naked, dirty, and happy ; 
there is no thought among them of decency, of modesty, 
or of monogamy ; yet the women are chaste, they are 
loved and they love their children, who are buried 
with them if they die before these are weaned. 
Everything is in common among these people, though 
each one has his fishing canoe, his ummiach or trans- 
port boat, his harpoons and nets, his wives and his 
little ones. The voracity of the Esquimaux is in 
direct ratio to the rarity of satisfying food. When a 
whale is thrown up on the coast, they rush upon the 
immense prey, cut it to pieces, tear away the meat, 
remove the bones, take out the intestines, run along 
under the ribs of the carcase. When this happy task 
is completed, the men lie down upon their back witji 
open mouth, and the favourite wife feeds them with 
the nicest morsels. He seems nearly to resemble our 
ancestors of the Magdeleine. Yet the Esquimaux haiS 
what these lacked, a religion, an ideal, and even a 
powerful clergy, initiated by long and hard trials to 
the mysteries of the invisible world. He worships the 
.whale, which is saluted by the hymns of tlie young 



Polysynthetic Languages. 193 

girls; he fears the spirits of the dead and their great chief 
Togarnsuk ; he aspires to a paradise situated beneath 
the ice, at the depths of the sea, where unnumbered 
seals come of their own accord to be slain. 

The Angakok, priest or sorcerer, has visited this para- 
dise more than once, swallowed like Jonah by a friendly 
whale, or conducted thither by familiar spirits. The 
credulity of the Esquimaux is greater, though it is diffi- 
cult to believe it, than that of all past or present divots. 

The language of the Esquimaux bears no resem- 
blance (except in its polysynthetism) to that of the Red 
Indians of the great lakes. It is not known whether 
it belongs through Aleoute and Tchoukche to the 
Samoyed or Uralo-Altaic group. Its dialects, from 
Greenland to the mouth of the Mackenzie, are sepa- 
rated only by insignificant differences. Its sounds are 
simple, a little. too guttural: the g, h, g nasal, the 
groups rh, rkr, tch, dj, ch (the German ch), the dentals, 
and a sort of palatal I abound, while s and the labials 
are rare. The vowel a dominates the others ; e and 
i, ou and 0, seem to alternate according to accentuation 
and number : angakok, priest ; angekut, priests. There 
are a dual and a plural, numerous suffixes denoting 
case, which are always placed after the noun, a con- 
jugation poor in moods and tenses, but abounding in 
intensive, negative, and frequentative forms, as we 
have repeatedly found in the agglutinative languages. 
Their method of counting merits a special mention. 
At first, says M. Victor Henry, they only reckoned up 
to five, the hand; to express six the Inuit says "one 
upon the second hand," &c., for eleven, "one on the 
first foot ; " for twenty-one, " one on the first hand of 
the comrade;" for forty-one, "one on the first hand 
of the second comrade." It is rather long, but this 



194 Distribution of Languages and Races. 

method of calculation leads to the decimal system, 
which is found in Aleoute. Many races, who long 
tarried at the notion of three, have wandered into 
complexities of the duodecimal system. M. Henry 
makes a very just observation about the interminable 
words of the Esquimaux, which may, I think, be ap- 
plied also to many cases of American polysynthetism. 
They are, he says, phrases in which the words are 
abbreviated by the rapidity of pronunciation, as in the 
French phrases Quest ce qu'il dit, pronounced keskidi ; 
Qu'est ce que c'est que cela, prgnounced kekckcga. Ohal- 
luktueksarsekautit, for ckaUuk/.uek-sakharsek-autit, " thou 
hast experienced many things worthy of memory." 

To the south-east of the Esquimaux lie two con- 
federations of Red Indians, the Algonquins and the 
Iroquois, whose numerous tribes furnished Cooper with 
most of his heroes, cruel, subtle, brave, magnanimous, 
and sententious. Compared with the Esquimaux these 
people are handsome, well-made, active, picturesquely 
clothed, their skin of a beautiful Florentine bronze 
under their family blazon and their variegated paint- 
ing. More favoured by climate, they are less dirty 
and less greedy, but they have not passed the moral 
and intellectual level of those whom they formerly 
drove out of their territory ; it is even doubted whether 
their most adaptable tribes can learn, like the Green- 
landers, to take part in political and municipal life. 
Among them agriculture was abandoned to the women, 
and was rare and rudimentary ; they lived only for 
and by hunting, always at war with their neighbours, 
in order to extend their domain as they killed out the 
game. Their religious ideas and practices, dances, 
songs, and exorcisms are pretty much the same, at 
bottom, as we find among civilised as among savage 



Polysynthetic Languages. 195 

peoples. It is one of tlie great benefits of ethnography 
that it has shown ns over all the earth the equivalence 
of the genuflexions, the mummeries, and the beliefs 
called consoling or sublime. So Iroquois, Algonquins, 
or Esquimaux of the five parts of the world believe alike 
in the intervention, always capricious, often malign, 
of spirits or of gods, in the power of certain clever 
seers, well-paid interpreters of the will of those above 
or below; and, finally, in a second life in happy hunting- 
grounds which more than one has visited in dreams, and 
where the creatures come to be killed by the weapons 
which have been buried in the tombs of the warriors. 

The Iroquois or Hurons of the lakes were divided 
into six nations, speaking Mohawk, Onondaga, Seneca, 
Oneida, Cayuga, and Tusiarora. The Algonquins, far 
more numerous, included about thirty tribes and as 
many dialects, related to each other by grammar and 
vocabulary; in Canada proper: Algonquin, Chippeway, 
Ottaway, Menomeni, Knistemaux ; in Acadia : Souri- 
quois, Micmac, Etchemin, Abenaki, &c. ; in the States 
of Massachusetts, Rhode Island, New York, New 
Jersey, Pennsylvania, Delaware, Maryland, Georgia : 
Narragansett, Mohican, and Lenape, Miami, Saki, &c. 
The simple phonetic system of these languages pre- 
sents no remarkable peculiarities except perhaps a sort 
of labial sibilant w, wdanis, his daughter, in Lenape, 
but oudanis in Ottaway. 

The verb and the noun, as in many agglutinative 
languages, are not distinguished ; the verb is only a 
noun accompanied by suffixes which mark possession. 
The pseudo-conjugation is none the less very rich in 
forms, in variants, though not in moods and tenses. 
An English missionary, Edwin James, fancied that he 
conld attribute seven or eight thousand to the Chippe- 



1 9^ Distribution of Languages and Races. 

Way verb, but his eyes must have deceived him at that- 
moment. The possessive pronoun and the first adjec- 
tive are prefixed to the noun ; kuligatchis, thy pretty 
little foot {hi, thou ; w%dit, pretty ; wichgat, foot ; chis, 
diminutive) ; hitanittowit, the Great Spirit (Jcitama- 
nitou) ; wit is an adjective termination. 

Algonquin has no genders ; Iroquois has two, the 
one for gods and men, the other for everything else, 
women and children, animals, plants, or mountains. 
Yet there are particles and affixes to distinguish 
animate and inanimate. The vocabulary is poor in 
abstract terms, and even with the aid of borrowed- 
words from English, Spanish, French, or German, the 
orator is condemned to have recourse to the strange 
metaphors which travellers have remarked with admira^ 
tion without always understanding them. 

Iroquois is strongest in numeration ; it has separate 
words for the first ten numbers. The Algonquin, like 
the Esquimaux of Hudson's Bay, stops short at five ; 
but as he calls ten Jive more, a hundred ten times ten, 
and a thousand the great ten of ten?, it will be seen 
that he is fairly well endowed with tlie arithmetical 
faculty. An Iroquois manuscript is said to exist ; it 
would be important to know its date. The missionaries 
were the first Iroquois and Algonquin litterateurs, and 
poor ones : at the present day the Cherokees, who 
appear to accept civilisation, publish newspapers. But 
all these tribes, in spite of the relative gentleness of 
the Canadian immigrants, will disappear, either by the 
intermixture of their race with others, or destroyed by 
drink, before they have learned to submit to regular 
work and to social servitude. 

It was, however, from their borders, perhaps even 
from one of their groups, that the mound -builders 



focysyninettc L.angiiages. 1 97 

started to go down the valleys of the Mississippi and 
the Ohio, and their tombs testily to a certain degree 
of industrial civilisation. Mingling with the Natchez, 
the Pawnees, and the Oomanches, they invaded in 
successive waves the plateau of Anahuac, the Olmecs, 
Toltecs, Chichimecs, and Aztecs, whose different features, 
monuments, and religions we are now iDeginning to dis- 
tinguish. Tlieae conquerors from the north encountered 
in Yucatan, in Chiapas, in Guatemala, and in the 
Isthmus, the Mayas and the Quiches, the builders of 
Izamal and of Pallanque, who had perhaps come from 
the north before them, or perhaps from Columbia, 
where the Chibchas already formed a theocratic state, 
with four kings and two popes. The brilliant Aztec 
empire overthrown by Cortez was not very ancient; 
but it is certain that for some centuries a considerable 
civilisation had existed in these regions, which were 
full of opulent cities and majestic monuments. For 
my part, I have not yet been able to admire in them- 
selves the hideous carvings of the Mexican temples, 
yet they are tolerable from the decorative point of 
view. These were the dwelling-places of innumerable 
gods — of the air, of the spring, of fire, of lightning, 
of the sun — whom a sincere piety worshipped with 
human sacrifice. To-day the poor peons of the pueblos, 
although they retain a vague memory of Montezuma 
and their past grandeur, genuflect before other gods, 
who have also drunk deep of the blood of the vanquished. 
The god of the Inquisitors was a worthy successor to 
the fierce Huitzilopochtli. But though the people of 
the Anahuac were terribly oppressed in the sixteenth 
century, they are gradually recovering their place in 
humanity, and through the mixture of races are again 
taking their place in civilise'd life. The Christian zeal of 
the Spaniards destroyed the cities, the arts, the inscrip- 
U 



igS Distribution of Languages and Races. 

tions, and the books of Mexico, except, if it be authentic, 
the Popid-vuh, a puerile rhapsody on cosmogony. But 
the results of scientific study of the various ethnic 
strata and of the thousand languages which are heaped 
up in these districts are full of interest. According to 
the fine work of M. V. A. Malte Brun (1877), there 
are, in au area only four times the size of France, not 
less than 2 80 dialects, classed in eleven great families, 
which comprehend thirty-five idioms and sixty-nine 
principal dialects, besides sixteen languages not classed, 
and sixty-two lost idioms. Most of these eleven families 
are still extant sporadically among the peoples of the 
centre, but overshadowed by Nahuatl or Mexican, 
which is spoken as far as the Isthmus. The oldest 
authentic Mexican text is a catechism printed at 
Antwerp in 1558. 

On the other side of the equator, on the slopes of the 
Andes, another purely American civilisation had been 
formed, completely separated from Central America, 
the civilisation of the Quechuas and of the Aymaras, 
the vast empire of the Incas, theocratic and commu- 
nistic (to the profit of the kings, children of the Sun), 
with its harems, its towns, divided into four quarters 
by walls at right angles, its grandiose festivities, its 
mystic cake and liqueur, its legends, in which a little 
history was mixed with a good deal of self-adulatory 
fiction. The great originality of ancient Peru was 
this state communism, unknown elsewhere, this paternal 
exploitation of happy serfs, lodged, fed, and married, at 
the expense and for the benefit of one man. Quichua 
or Ketchua, the principal language of the Pacific coast 
and of the valleys of the Andes, seems to have come 
originally from Quito, from the upper basin of the 
Amazon; one dialect, Tchintchaysouya, occupies the 
centre of Pern ; a second, Cuzco, is spoken in the 



Poly synthetic Languages. 199 

extreme south and towards Chili ; finally, Gochabamba 
in Bolivia, and Caltchaki on the eastern slope of the 
Andes, surround the somewhat narrow domain of 
Aymara, which appears to be entirely distinct from 
these. The mechanism of Quicha seems to be nearly 
identical with that of the other agglutinative idioms of 
America ; relatively rich in compounds of a reasonable 
length (tchimpu, cloud, rasu, block of snow, Chimhorazo), 
it abounds in forms with suffixes and in derivatives of 
great length. It has no gender ; the noun and tlie 
verb are confounded ; the particles of case, the personal 
and possessive pronouns, suffice for all the shades of 
thought. The language is guttiiral and affects doubled 
initial letters, ttanta, bread, ppatclia, dress. Spoken 
side by side with Spanish in the towns of Ecuador, 
Peru, and Bolivia, and almost exclusively in the 
mountainous districts, and in the north-west of the 
Argentine Rt^public, Quicha has furnished to Spanish 
a number of geographical and local terms, among 
others, the names of the llama, the vigogne, and the 
alpaca. It was not written before the conquest ; the 
Peruvians, in spite of their advanced civilisation, had 
not even attained to the riddles of the hieroglyph ; 
they still used knots of different coloured ribbons 
as mnemonics. These were called quipos, and have a 
family likeness to the rows of seeds or shells which 
the Red Indians of the north threw in front of them 
to mark the different stages of their discourse. 

The Peruvians, like the Mexicans, have survived in 
great numbers the terrible Catholic invasion, and though 
long overwhelmed by the blow which had fallen on 
them, long stupefied by superstitions far inferior to 
their ancient religious beliefs, they are now raising 
their heads and claiming their place among free 
peoples. The rest of the American peoples, the most 



200 Distribution of Languages atid Races. 

vigorous, the most worthy to live, have perished or are 
about to disappear before the greed of the European 
immigrants. Without hunting-grounds, without game, 
they are condemned to die out. Only the poorest 
specimens of American humanity, the Abipones, the 
Charruas, the Botocudos, and in the extreme north the 
Kienas and the Athabasks, the dwellers in thickets, in 
deserts, in the torrid or the frozen zones, may count 
upon a respite of a few centuries. 

Oar summary review of the agglutinative languages 
is terminated. We have seen that, simple or complex, 
their structure is founded solely upon the addition, to 
one or more invariable themes or radicals, of subordinate 
roots, emptied of their proper sense and reduced to 
aflBxes, suffixes, infixes, and prefixes. The immense 
majority of these languages have never been written ; 
driven out to the borders of civilisation, into countries 
not easily approachable by Europeans, they continue to 
Vegetate obscurely. But a few more favoured groups, 
preserved, and even developed, - by civilisation, have 
attained to some literary life. Japanese, Mandchu, Mon- 
golian, Finnish, Hungarian, Turkish, Dravidian, Malay, 
Georgian, Basque, Greenland, Algonquin, Mexican, and 
Quicha have all contributed, in very different measure 
indeed, to the progress of human intelligence. Their 
names are worthy to be remembered. 

We pass by an easy transition from the agglutinative 
class to the inflected class. We shall find in the latter 
all that can be done by suffixation. A single thread 
separates inflexion from agglutination — that is, the 
possible variation of the root syllable ; a very slight 
acquisition, which, while it stops the abuse of suffixa- 
tion, allows of the expression of every shade of the idea 
without lengthening the word, and also of attaching a 
general idea to all the derivatives from the same root. 



CHAPTER VII. 
THE SEMITIC WORLD. 

Noah, Hain, and Shem — Conjectures on the origin of the Semites — • 
Ethnical variety, lingmstlc unity — Exodus of the Canaanites : 
Hyksos, Fhceniciatis, Hebrews— Jews and Syrians crushed in the 
struggle with Egypt and Assyrians — Rule of the Fersiaus, Greeks, 
and Romans — Appearance of the Arabs — Christianity and Islam — 
Tardy revenge of the Semites — Character of the inflexion and 
structure of the word in the Semitic languages — Northern branch 
of the Semitic family — i. Arameo-Assyrian group : Chaldean, 
Nabatean, Syriac, Syro-Chaldean, Assyrian — 2. Canaanitigh 
group : Phoenician, Punic, Samaritan, Moabitish, &c., Hebrew — 
Southern branch : Arabic, Himyarite, Ghez or Ethiopian. 

The peoples whom we are accustomed to call Semitic 
have always ignored their relations with the Biblical 
patriarch Shem, son of Noah. Bnt if we disregard 
the lettet of the precious record, compiled and recast 
many centuries after the events which are therein 
transformed into legendary fables, if we consider in 
themselves the names of Noah, Ham, Shern, and Gush, 
we shall readily overlook the inexactitude of the name 
given by the moderns to the Chaldeans, the Arameans, 
the Oanaanites, and to the Arabs. For Noah is a 
Semitic god of great antiquity, Nouach, a genius with 
four outspread wings, god and saviour, the spouse of 
Tihavti, the fecundity of the abyss ; Ham was Khemos, 
the god of the Moabites, and perhaps identical with 
the Egyptian Khem ; we find Gush among the Cos- 
sians or Kissians of the Euphrates, and among the 
southern peoples whom the Pharaohs fought on the 



202 Distribution of Languages and Races. 

two shores of tlie Red Sea ; " the vile Gush," said the 
Egyptians*; but they none the less gave to their royal 
princes the title of Prince of Gush, which shows the 
importance which they attached to the subjugation of 
these Gush or Gushites, the Ethiopians of Herodotus, 
cut in two by the Semitic expansion ; as for Shem, it 
is difficult not to recognise in him Samas, Samson, the 
sun-god of the Assyrian pantheon. 

Gushites, Hamites, Semites are far from being 
synonyms ; but it is hardly possible to doubt their 
relationship, or at least the intimacy of their primitive 
connection. Only it is very difficult to determine the 
vicissitudes of their prehistoric life. Experts differ ; 
some, M. Eenan, for example, assigning to the Semites 
a northern origin ; others think, with Echrader, that 
the nucleus of the race was formed in the centre and 
west of the Arabian peninsula, where the language 
approaches most nearly to the supposed mother-tongue, 
where the Chaldean legends and divinities have least 
penetrated, though they form the common ground of 
thought among the other Semites. Finally, since 
philologists are agreed in recognising affinities, rudi- 
mentary but probable, between the Khamito-Berbers 
and the Semites, it is hard to conjecture where they 
both came from, or where we should place the common 
country where they possessed a common idiom. 

We must be content to know that their separation 
was accomplished at the time when Menes came down 
from Upper Egypt to the Delta to found the ancient 
empire and Memphis, about five thousand years before 
our era. At that date the languages of the Nile and of 
the Libyan desert had reached the extreme limit of the 
agglutinative stage, which they have not over-stepped ;■ 
and the Semites were doubtless progressing towards 



The Semitic World. 203 

the inflected period. Thenceforward the two races 
have no point of contact except the Isthmus of Suez. 
The one, without advancing farther than Mount Sinai,- 
develops its precocious yet enduring civilisation, 
builds towns, pyramids, and temples, and, from the 
worship of animals and of the Nile, rises to the religion 
of the sun, of fire, and to the belief in immortality. 
The other, wandering without name or route, given up 
to the worship of stones and of the heavenly bodies, 
fluctuates between Nedjed and the Euphrates ; for two 
thousand years it is lost to history. At most, we may 
attribute to some attack on the part of the nomads 
the fall of the first Egyptian empire and the retreat 
of the Pharaohs to Thebes. 

When history first takes cognisance of the Semites, 
the practically unchanging unity of their linguistic 
organism was constituted, and much more strongly than 
the Indo-European unity. The dialectic differences do 
not affect either the formation of the words or the 
vocabulary, but only a few details of grammar and 
pronunciation. But ethnical unity exists no longer. 
If the Arab with hie high, long head, his slender, 
nervous body, his profile at once strongly marked and 
refined, may be considered as the faithful guardian of 
the racial type, the thick-set build of the Chaldean, the 
tendency to fat and the massive face of the Assyrian, 
point to various mixtures with more ancient peoples. 

We have already mentioned the very probable 
existence of non-Semitic races and languages, Shumi- 
rians and Elamites, round about the Persian Gulf, in 
Babylonia and Susiana. There, in these regions of 
ancient civilisation, several Semitic groups obtained 
their industrial and religious education. At the time 
when the Shumir Likbagas (3000) reigned in Chaldea, 



204 Distribution of Languages and Races. 

Bab-ilou, Babylon, the gate of El, was already a 
tiourishing city under kings who were also priests ; 
and the shores and islands of the Gulf of Ormuz were 
occupied by the Canaanites, the Poun, Poeni, Punici, 
the future Phoenicians, whose territory on the other 
hand reached to the Himyarites of Southern Arabia. 
Mesopotamia and Armenia were also full of Semites, 
Arameans to the north and west, the future Jews to the 
north and in the centre, in Arrapachitis (Arphaxad) ; 
lastly, the Assyrians of the middle valley of the Tigris. 
It would seem from the legend of Nimrod preserved 
by the Bible, that the Assyrians were a Chaldean 
tribe, whose national god or eponymous hero, Assour, 
was the male of the Canaanitish Aschera. All these 
tribes, more or less compact, more or less powerful, 
received from Lower Chaldea their gods and their 
beliefs, the tradition of the deluge, the worship of 
winged bulls, transformed at a later date into cherubim, 
the Elohim and Baalim of every species, the goddesses 
of fertility and of the spring sunshine (Adonis, Tham- 
muz), dead and resuscitated. A cataclysm of which 
the causes are not known, an Elamite invasion under 
the pressure of the Persians, an incursion of Scythians 
from the other side of the Caucasus, destroyed about 
the year 2300 B.C. the earliest Assyrian empire, drove 
the Arameans back towards Syria, the Israelites 
towards Lower Chaldea, and decided the Canaanites 
to cross the desert ; while the torrent of the Hyksos 
(robber chiefs), bearing onward in its course Edomites, 
Ammonites, Moabites, was hurled upon the delta of 
the Nile. 

The Canaanites, driving before them the ancient 
inhabitants of Palestine, the Pelestes or Philistines, 
massed themselves upon the Syrian coast around Arvad 



The Semitic World. 205 

and Tyros ; these were the names of their ancient cities 
of the Persian Gulf. Finally, under the name of 
Hebrews, people of the other shore, the clan of a 
certain Terah and of a certain Nahor, decided to quit 
Ur in Clialdea, bearing away their gods like Anchises, 
eagerly pursued by the Elamite chieftains or kings ; 
among others by Chedorlaomer. The leaders of the 
fugitives, Abraham and his nephew Lot, underwent 
some misfortunes in the neighbourhood of the Jordan, 
of Sodom and Gomorrah ; others whose names have 
come down to us, of doubtful wisdom and uncertain 
morality, Isaac, the dishonest Laban, Esau the simple, 
and the astute Jacob, continued to live with difficulty, 
surrounded by other nomads, until famine or their 
vagabond habits drove the Hebrews to the confines of 
Egypt, into the land of Goshen, beside the Hyksos. 

Meanwhile Egypt had not abandoned the hope of 
revenge ; her national kings had not ceased for five 
centuries to harass the foreign conqueror. Ahmes, 
the founder of the eighteenth dynasty, finally expelled 
from the delta the armies and the government of the 
Hyksos ; and his successors, returning upon Asia the 
attack which they had thence received, subjugating, 
or ratheik putting to ransom, all the Canaanites of 
Judea, Phoenicia, and Syria, crossed the Euphrates 
and the Tigris. Nineveh twice fell into their power, 
and the whole Semitic world became vassal to the 
.Pharaohs. The influence of Egypt was real though 
temporary, but in the reciprocal dealings which were 
the result of the conquests of the Tutnes and the 
Araenhoteps, the share of the Semites was on the 
whole the larger. Marriages with the daughters of 
kings or vassal governors brought into Egypt and 
established Asiatic types, ideas, and customs on the 



2o6 Distribution of Languages and Races. 

'Theban throne. Amenliotep IV. was purely Semitic ; 
lie endeavoured to replace the religion of Auimon by 
the sun-worship of Syria. In 1887 were discovered 
the fragments of a correspondence exchanged be- 
tween the kings of Syria, Armenia, and Babylonia 
and the Pharaohs Amenhotep III. and IV. ; all these 
letters are written in cuneiform character and in 
Semitic or other dialects ; it is probable that the 
answers were drawn up in the same character and in 
the same languages. For the rest, the subjugated 
nations had soon recovered. Saryoukin I. had recon- 
stituted the Chaldean empire ; the Assyrians, ever at 
war on their eastern and western frontiers, had more 
than once crossed the Upper Euphrates and pene- 
trated Asia Minor as far as Troad, where the name 
Assaracus seems to be a relic of an Assyrian dynasty. 
The Hittites or Itlietas occupied the north of Syria ; 
find when liamses II., Sesbstris, desired in the 
fifteenth century to renew the exploits of his ances- 
tors, he was checked at Kadech by the Hittites and 
forced to retreat after an undecided battle. Tlie great 
expansion of Egypt was stopped, at least towards the 
north. The Semitic peoples, on the contrary, were 
everywhere in the ascendant. Phcenicia was colonising 
the European and Libyan coasts of the Mediterranean ; 
Bylos (Gebel), Tyre, and Sidon had commercial settle- 
ments in many places, where the potteries, the stuffs, 
and the jewels of the East were exchanged for the raw 
products of Gaul, Spain, and Africa. Their boats, 
navigated by oar and sail, had even passed the Columns 
of Hercules and coasted round Europe as far as Eng- 
land and Denmark. Not only did they leave with the 
Etruscans, the Sards, the Pelasgians, the Siculi, and 
the Hellenes of the /Bj'can and of Ionia the rudiments 



The Semitic World. 207 

of the arts and of philosopby, but they also brought 
them an inestimable treasure, the alphabet, sixteen or 
eighteen signs, extracted from the chaos of the hiero- 
glyphs. It is disputed at the present day whether the 
Phoenicians were really the authors of this famous in- 
vention ; but it was certainly they who spread it over 
Europe, and who unwittingly, for they, considered only 
the usefulness of a commercial writing, gave to the 
West this necessary instrument of intellectual progress. 
As for the great empires of the Euphrates and the 
Tigris, in the midst and in spite of bloody revolu- 
tions, now pretty well known and dated, they rose to 
a considerable degree of power and dignity. The 
excavations of recent years have laid bare their palaces 
and temples; their books graven on thousands of bricks, 
their seals with magical formulas, and the great 
triumphal inscriptions engraved by conquering kings 
on statues, walls, and on the living rock, have been 
deciphered. Their artists excelled in the minute and 
in the colossal ; their gigantic statues, rude and 
grandiose, sustain comparison with the finest Egyptian 
work, and their influence can easily be traced in the 
archaic monuments of Asia and Greece. 

The Hebrews bad as yet held no place in history. 
It was only towards the end of the fourteenth century, 
under one of the successo"s of Sesostris, that, urged by 
oppression and by want of room in the land of Goshen, 
they left Egypt, with difficulty avoiding disaster on 
the sandy shores of the Gulf of Suez. From the 
peninsula of Sinai they had to pass through the tribes 
of Midian, Moab, and of Edom, and then force their 
entrance into the promised land ; they returned to it 
late, the country was already occupied. Hence those 
ex^terminations, those servitudes, and all the adventures. 



2o8 Distribution of Languages and Races. 

naively exaggerated later in the Book of Judges, which 
is often legendary, but full of interest from the double 
point of view of ethnography and ethics. By courage 
and perseverance the Benou-Israel to the north and 
east, the sons of Benjamin and Judah to the south, 
ended by subduing and absorbing in part the other 
Canaanites who had preceded them, without passing 
the bounds of Syria on the one hiand, and on the bther 
the Philistine towns, without piercing the narrow 
band of the Phoenicians. Like all the Semites of the 
north, they had their sacred stones, their Baalim, male 
and female, Baal, Moloch, Aschera, Dagon, their 
winged gods with bulls' heads, their bronze lions and 
serpents, worshipped in the high places ; but they all 
rallied more or less round a coffer or ark, which con- 
tained their national god, who was of the heavens or 
solar, named El Jahve, the Phoenician Jao, and vari- 
ous symbolical objects, a seven-branched candlestick 
(representing the seven planets), a table, bread, and 
sacerdotal ornaments. Their neighbours had also 
their favourite patron, some Dagon, others Astarte, 
Marna, Derketo, ' Moloch, or Chemos. We see what 
becomes of the primitive monotheism of the Hebrews. 
It was not till the tenth century, when the brave and 
not very virtuous David, and his son, the splendid and 
not less voluptuous Solomon, had, thanks to favourable 
but ephemeral circumstances, constituted the brilliant 
and brief Jewish empire, that the ark, transported 
with great pomp to the newly conquered capital, 
became the obligatory centre of religion. Even in 
the temple of Solomon (if this wonder of the world 
ever really existed), there were quarters reserved for 
puofetitutes and eunuchs, the sacred servants of the 
goddess Aschera ; not to mention the great serpent, 



The Semitic World. 209 

the cherubim, and other animals representing the 
ancient creed of polytheism. Jahve had to himself 
only the holy of holies. The unity of creed was, 
moreover, so little established that the jealousies raised 
by the pretensions of the high priest brought about the 
division and the ruin of the empire ; and in the little 
principality of Jndah, retained by the tribe of Judah, 
together with the Levites or sacerdotal tribe, the elo- 
quence of the prophets and the effoi-ts of two pious 
kings, Hezekiah and Josiah (622), could not assure to 
Jahve a complete triumph over the strange gods. This 
triumph was only secured to him by the ruin of the 
people chosen by him from all eternity. 

The existence of the Jewish tribes had always been 
precarious and threatened by many foes. Unable to 
engage in serious strife even with the kings of Syria, 
their position was hopeless when, divided against 
themselves, they became the battlefield for the two great 
rivals, Assyria and Egypt. Towards the eighth cen- 
tury the victory declared itself in favour of the Assyrians, 
who in the following century invaded Egypt ; then 
the latter, reanimated by the Ethiopian princes, renewed 
the fatal war, of which one result was the destruction 
of Samaria by Saryoukin and the ruin of the kingdom 
of Israel. All who did not perish in the massacre 
•were transported into Mesopotamia at the end of the 
eighth century, 708—7 1 0. A few fugitives gained 
Jerusalem and Egypt. Finally, in the sixth century, 
587—581, Jerusalem, attacked by Nebuchadnezzar, the 
greatest king of a new Chaldean empire, was taken and 
burned. The fierce courage of the unhappy Zedekiah, 
the last prince of Judah, could not prevent the second 
captivity. It was in humiliation and misery that the 
relics of this much-tried race put together their tradi- 



2 lo Distribution of Languages and Races. 

tioas, not without some admixture of foreign elements, 
and rallied for ever to their god Jahve. 

Assyria had been conquered by the Medes ; Chaldea 
fell in her turn before the Persians (536 B.C.). Then it 
was that Zerubbabel, Esdras, and Nehemiah, 5*36-430, 
were able in the course of a hundred years to bring 
back two or three columns of exiles to reconstruct the 
Temple with great difficulty, and finally to compile 
those ancient fragments, completing tliem, interpolat- 
ing tbem, reconciling them as best they could with 
orthodoxy, the poems adjudged to David and Solo- 
mon, and the dithyrambs and revelations attributed to 
the different prophets. This work, which was nearly 
finished at the time of the Greek translation of the 
Septuagint, was begun under the Ptolemies for the 
AlexandrineJews,and continued through the timeof the 
Maccabees and up to the beginning of the Christian era. 
Meantime the Jewish nationality outlived that of the 
powerful Semites who thought to destroy it, and sur- 
vived alone ; alone it kept a species of independence. 
The Persian monarchy, the brilliant passage of Alex- 
ander, the Seleucidje, the dominion of the Parthians, 
had already buried the ancient glory of Assyria and 
Chaldea ; even the language of Sargon and Nebu- 
chadnezzar had ceased to be spoken above Babylon. 
Syria and Phcenicia had accepted in turn the yoke of 
Egypt, Greece, and Rome. Finally, the cruel siege 
and sack of Jerusalem by Titus, and a last convulsion 
under Hadrian, put an end to the unhappy destiny 
of the people of Israel, or rather marked the beginning 
of a long and terrible agony, borne with invincible 
energy and patience. In fact, the three northern 
groups, Aramaic, Chaldean, Canaanitish, had been 
destroyed at the time of the Persian conquest, and 



The Semitic World. 2 1 1 

the sceptre of the civilised world had passed into the 
hands of tlie Indo-Europeans. And it was lon» 
before men saw that the maladies inherited by the 
peoples of the West from these dying nations, — Baby- 
lonian corruption, orgiastic religions of Asia Minor, 
the enervating mysticism of despairing souls, assured 
but too well the vengeance of the conquered. 

Moreover, the Semites of the south, the Arabs, had 
not been touched. The Pharaohs by the Red Sea, the 
Sargonides by the desert, had attacked, pierced even 
here and there, this block of Arabia, and had annexed 
it to their empires. But this mattered little to the 
nomad Bedouin, to the hardly, formed tribes wliich 
floated about Mecca and Medina. Some carried their 
tents into another part of the desert, the others fell 
back or paid some small tribute, and continued to 
make war among themselves for women or horses. 
From the time when the Syrian dynasty came to 
hasten the decomposition of the Roman empire, and 
especially when Zenobia and Odenatus all but realised 
•their dream of an Oriental empire, the Arabs of the 
north, with whom were mixed the half Jewish Canaan- 
itish tribes, Idumeans, Moabites, &c., began to take 
some part in Western life, and to be influenced by 
Judaism and Christianity. Mahomet appeared with 
his incoherent and inoffensive book, but also with his 
terrible doctrine of the identity of the two powers, 
religious and civil ; and Islam, let loose upon a world 
still shaken by the fall of the Roman empire and by 
the struggle for its territory of the swarming new 
races, Islam gave a tremendous power to the Semitic 
races, a power far more fatal to the Mediterranean 
world than the domination of the cruel Assyrians or 
that of the superstitious Chaldeans. In less than a 



2 1 2 Distribution of Languages mid Races. 

hundred years the Arabs had conquered Syria and 
Persia, Egypt, the African coast and Spain, and France 
as far at Poitiers. This conquest of the East and 
South by races which had not attained a high degree of 
intellectual cultui-e jeopardised the future of tlie world, 
until the day when John Sobieski, in the seventeentii 
century, forced the Ottoman vizier to raise the siege of 
Vienna. It is true, indeed, that the part played by 
the Arabs was not lacking in brilliancy, and the evils 
of which they were the cause are not without compen- 
sation. There was a brief and splendid civilisation at 
P)agdad, at Cairo, at Kairouan, at Tlemcen, at Fez, at 
Cordova, and Granada ; a rich literature ; an active 
commerce which reached to China, the Malay Archi- 
pelago, and India ; finally, a shock which, sending 
back upon Europe the translations of forgotten Greek 
authors, driving from Constantinople the last custodians 
of Hellenic science, determined the Renaissance in 
Italy, France, and Germany, the revolt against the 
humiliating, stupefying yoke of the Christian theocracy. 
But, and the fact is curious, that which is commonly 
called Arab philosophy, astronomy, and architecture, 
belongs in truth to the peoples roughly awakened by 
the sword of the Arab, to the people of Bactriana, of 
Mazenderan, of Persia, of Syria, of Egypt, of Barbary, 
and Spain. The Arab, unlike the , Assyrian, is no 
artist : no Arab has ever painted or carved the human 
face. He is a musician and a poet, a witty story- 
teller, with a taste for maxims, anecdotes, apologues, 
and pithy sentences ; but his mind lacks both breadth 
and concentration. No dogma could suit him better 
than the arid and empty formulae of Islam, than the 
Koran with its medley of maxims and narratives, its 
contradictions, its. idle and endless controversies. At 



The Semitic World. 2 1 3 

the present dav, Seniitistti may be occasionally a source 
of trouble, but it is no longer a danger ; even Mussul- 
man fanaticism, its te.rible creation, though it may 
spread among the inferior races of Africa, seems only 
an anachronism, which we must know how to reckon 
with indeed, but which is powerless against civilised 
Europe, and its allies America and Australia. For 
the second nd the last time, Indo-European culture 
has conquered; first, in 537 B.C. with the Persian 
Cyrus, and in 330 B.C. with Alexander; and, secondly, 
in 732 with Charles Martel, and in 1683 with John 
Sobieski. 

Let us now endeavour to establish the general 
characteristics, and to sketch out a table of the lan- 
guages spoken by this important section of the human 
race. 

There is no stronger or more unchanging unity 
among any group of languages than that which exists 
in the Semitic group. The dead and living languages 
which compose it hardly differ from each other so 
much as the various Romance or Sclavonic dialects. 
Not only are the elements of the common vocabulary 
unchanged, but the structure of the word and of the 
phrase has remained the same. The persistence of 
the radical consonants is the most striking feature of 
the organism. The radical, as the agglutinative phase 
had left it, admits usually of three consonants, sus- 
tained by one, two, or three variable vowels, of which 
the diversity indicates tense, mood, voice, the form of 
the verb, the adjective or substantive character of the 
noun ; hence the Semitic roots are called triliteral, 
because the imperfect writing, not noting the vowels, 
puts in evidence the. three fundamental letters or con- 
sonants ; but Ihese roots may have one, two, or three 
IS 



2 1 4 Distribution of Languages and Races. 

syllables. Moreover, various particles, generally mono- 
syllabic, pronouns, case endiags, verbal prefixes, com- 
plete the grammatical organism, which, is of extreme 
simplicity. 

From the same triliteral root, qtl., to kill, Idh, to 
write, dhr, to speak, &c., a change in the vowel produces 
nouns like qatl, murderer, qitl, enemy, qitalu, blow ; 
verbal forms like qatala, he killed, qutala, he was 
killed, qotla, qotel, qtal, &c. Suffixes or prefixes indi- 
cate the tense : takhoba, thou writest or wilt write, 
hatahta, thou hast written. Tlie Semitic verb has but 
two tenses, a perfect and an aorist, according as the action 
has been accomplished or is in course of accomplish- 
ment ; it admits only two moods, the indicative and 
the imperative, whence the Arab has drawn a subjunc- 
tive and a jussive. The two voices, active and passive, 
have each fifteen, thirteen, seven, or five forms, accord- 
ing to the dialect, characterised eitlier by the doubling 
of the second consonant (qattala, quttaia, he has killed 
many, he has been entirely killed), or by lengthening 
the first or second vowel, or by the prefixation of 
various syllables : hiqtil, hithqattel, hithqotal, niqtal, &c., 
to give an intensive, causative, desiderative meaning, 
an expedient already well known to the agglutinative 
languages, which we shall find again in the Indo- 
European. The second and third persons of the verb 
express the sex of the subject. 

The possessive and personal pronouns, whether sub- 
ject or object, are suffixed to the verbs and nouns ; 
7d, me, ta, thou ; Sabachthani, thou hast forsaken me ; 
Mi, my God. The demonstrative pronouns appear to 
be formed from the vowels a, i, u, which the written 
character expresses, and which had an aspirated or 
consonantal character : h, soft, ye, oue. 



• The Semitic World. 2 1 5 

The declension has two genders ; the termination t 
is the sign of the feminine ; the neuter appears to 
have existed, but has disappeared ; the plural and 
dual masculine are indicated by m or n (v/m, un, im, 
in) ; the feminine plural keeps the final t. Three 
cases, the nominative in ou, genitive in i, accusative 
in a (um, im, am) : abd-u, abd-i, abd-a (servus, servi, 
servum), are retained in Arabic. Hebrew replaces 
them by particles, I, b, and et ; sometimes the simple 
juxtaposition of the determined and determinant, that 
which is termed the constructed state, takes the place 
of the genitive : melekh Israel, king of Israel ; biti 
iamin, son of the right ; and again with a pronominal 
suffix, ben-on-i (son, sorrow, me), son of my sorrow. 
Lastly, the definite article, which is invariable, Aa(^) 
in Hebrew, al in Arabic, is prefixed to the word, 
doubling the initial consonant, or assimilating itself 
with it : in Hebrew hammelehh, the king ; in Arabic, 
ar-rahman, the merciful. 

All the processes we have just enumerated have 
nothing new for us except one, and that indeed 
important — the change in the root vowel, together with 
the invariability of the consonant. This ingenious 
artifice both brings the Semite near to the Indo- 
European and yet separates them ^profoundly ; so 
completely that, in supposing them to have had a 
common period, monosyllabic or agglutinative, it would 
be impossible to establish any relation between the two 
systems either in the conjugation, or — and this is far 
more important — in the conception of the root and in 
the formation of the word. 

The Semitic phonetic system, very simple in the 
vowels, abounds in aspirated gutturals and emphatic 
consonants, h soft, h hard, hh. gh, kh, &c., which are 



2i6 Distribution of Languages and Races. 

diflBcult to render in our smoother tongues. We will 
content ourselves with these general indications ; the 
subject is too vast for us to engage upon it without 
transgressing our limits. It has been so long and so 
completely studied, that we refrain from even citing 
the hundreds of authors who have thrown light upon 
its smallest particulars. The Ristoire GiniraU des 
Langues SimifAqms, by M. Kenan, though unfortunately 
incomplete, is yet in French the surest and the most 
open-minded guide which we can indicate. 

Tlie Semitic languages form two great branches, 
each subdivided into two groups. The northern branch 
comprehends the Aramaic-Assyrian group and the 
Oanaanitish group ; the southern group includes the 
Arabic group, properly so called, and the Himyarite 
group. 

The name Aramaic is given to two dialects which 
are very nearly allied — Chaldean and Syriac, which 
are separated by preferences for certain vowels and by 
a difference of accentuation. As the remaining Baby- 
lonian inscriptions are deciphered, we shall attain to 
a better knowledge of ancient Chaldean. In the 
present state of science, this language is chiefly 
represented by certain parts of the Jewish Bible, espe- 
cially the Books of Esdras and of Daniel, which were 
doubtless written during the captivity, soon after the 
fall of the last Babylonian empire, from the fifth to 
the second century ; then, towards the Christian era, by 
the Targum, translations and paraphrases of the Hebrew 
books ; the Talmud, which is of somewhat earlier date, 
contains also some Aramaic elements. Nabatean aud 
Mend&ite or Sabian, southern forms of Chaldean, have 
left us a treatise on agriculture, translated into Arabic 
in the tenth century of our era, and the curious book 



The Semitic World. 2 i 7 

of Adnm, which is perhaps posterior to Islamism. M. 
Renati found in the Museum of Naples some Nabatean 
inscriptions, dating from before the Christian era, 
which bear witness to the flourishing condition of this 
Chaldean colony, which had emigrated to Petra and- 
was governed by independent kings. 

The Aramaic which was spoken at the time of 
Christ was divided into two sub-dialects : that of 
Galilee, which resembled the Syriac pronunciation, 
and that of Jerusalem, of which the pronunciation 
was more marked and nearer to Chaldean. Jesus 
and his disciples evidently spoke the dialect of their 
country, as appears from certain passages in the New 
Testament. It was in this dialect, called Syro- 
Chaldean by Jerome, that the notes of Matthew and 
Mark were written, the point of departure of the 
tirst and second Gospels. 

Syriac, in its primitive state, is unknown to us, as 
also Syro-Chaldean. We know that it was one of 
the principal dialects spoken in Judea before and 
after Christ. It may claim the inscriptions of Pal- 
myra, which date from the three first centuries, and 
the version of the Bible called Pechito (the simple), 
which is attributed to the second. It is the language 
of the Aramaic Christians, of a whole literature of 
controversy, in which the works of St. Ephrem, poet, 
controversialist, and commentator, hold, we are in- 
formed, the first rank. The great schools of Nineveh 
and Edessa, species of theological faculties, ceased not 
from the fourth to the sixth century to send forth 
Gnostic, Monophysite, and Nestorian writers, whose 
disputes fill the bibliography of the Maronite Assemani 
(three folios). Syriac poetry and medicine were 
deeply tinctured with theology. History, or rather 



2i8 Distribution of Languages and Races. 

the chronicle of contemporary events, is represented 
in this literature by the valuable works of James 
of Edessa and of Barhebraeus (tbirteentli century), 
doctor, bishop, primate of the East, known also by 
■the Arabic name of Abulfarage. Barhebrseus is the 
last name in Syriac letters ; for as early as 853 the 
Caliph Mottewakkel had forbidden them to be taught. 
The language was declining also, having long been 
eked out with Greek, Latin, Frank, and Arabic words. 
But in its golden age (fifth to the ninth centuries) it 
had served the instrument for the beginnings of Arab 
philosophy. It was through Syriac translations of 
Aristotle, Proclus, Porphyrius, by Abraham the Great, 
Ibas, Sergius, and James of Edessa, that the works of 
the Greek mind became known to the barbarous West. 
We therefore owe some gratitude to this language, 
which lingers obscurely in the neighbourhood of the 
lakes of Van and Urmia. It is still used in the 
liturgy of the Nestorians and of the Maronites. 

Assyrian is a discovery of this century ; it was 
revealed by the third column of the Persian inscrip- 
tions. The labours of Rawlinson, Oppert, and Menant 
have assured its place, for some time disputed, in the 
Semitic family. It is sufficient to transcribe it into 
the Hebrew character to make clear its relationship 
to the other kindred dialects. The cuneiform char- 
acter, so difficult to read because of the mixture of 
ideograms with phonetic signs, has at least this 
advantage, that it is syllabic, and makes it evident 
that the famous neuter consonants of the triliteral 
root were not mute, and that the Semitic word is 
the result of a slow agglutination, and not of a pre- 
conceived abstraction. 

To the Canaanitish group belong Phoenician, 



The Semitic World. 2 1 9 

Samaritan, the languages of the left bank of the 
Jordan, notably Moabite, known by the stele of 
Mesha, and lastly, Hebrew. The first and the last of 
these dialects are almost exactly alike. In Phoenician 
there are several primitive forms which exist in 
Hebrew only as archaisms ; it presents also, even in 
its colonial dialects, traces of Aramean, which the 
Jews in their Egyptian exile naturally avoided. Of 
Phoenician literature there only remain fragments trans- 
lated from the history of Sanchoniatho, and a Greek 
version of the circumnavigation of Hanno. But 
numerous inscriptions, collected from all the coasts of 
the Mediterranean, at Carthage, and in the islands of 
Cyprus, Malta, Sardinia, &c., allow the language to 
be classed with absolute certainty. Phoenicia itself 
has furnished epigraphic texts of great importance, 
the long inscription which may be seen at the Louvre 
on the totnb of Eschmounazar, king of Sidon, and 
the stele of Jehawmelek, king of Byblos, interpreted 
in 1875 by Vogue and Kenan. These little princes, 
vassals of Egypt or Persia, seem to have lived in the 
sixth or fifth century ; conquerors were commonly 
tolerant of these local sovereignties. " It is I," says 
Jehawmelek, " I, son of Jeharbaal, grandson of Adom- 
melek, whom our lady Baalath has made king of 
Byblos," &c. This text, which briefly describes the 
portico and the sanctuary dedicated to Baalath, proves 
the exactitude of the records of Lucian and Plutarch. 
We have mentioned Phoenician first, because it 
developed itself before the arrival of the Hebrews in 
Palestine ; but no one is ignorant of the fact that the 
language of the immortal Job has played, after its 
extinction, however, a far more important part in the 
world than its Canaanitish sister. We have spoken 



2 20 Distribution of Languages and Races. 

oFtlie late date of tlie compilation of the Bible, tlie 
revision from the point of view of orthodoxy of all the 
fragments, of the books of all the ages which were 
carried to Babylon, the work" of unification due to 
Esdras and Neheraiah, which betrays itself by the 
almost perfect unity of the language. If there are, 
as it is thought, some parts of the Bible anterior to 
the eleventh century before Christ, if a few psalms, 
the Books of Job and of Judges, really date from the 
age of David and Solomon, it is by induction only 
that they can be distinguished from the parts which 
were re- written, like the Pentateuch, or quite modern, 
like the Chronicles and Judith, Ruth and Tobias, 
written long after Chaldean and Syriac had taken 
the place of Hebrew as spoken languages. Aramaic 
reigned in Palestine long before the epoch of the 
Maccabees. Hebrew was, however, the written language 
up to the first century before our era, and was main- 
tained in the rabbinical schools as late as the twelfth 
century. Its principal monument of this period was 
the Michna, a collection of traditions, a sort of second 
Bible. After the decline of the Arabs, especially in 
Spain, the Jewish priests returned to their national 
language, and wrote, spoke, and taught it still. 

Ancient Hebrew is poor in abstract terms, and 
from this relative poverty arises its principal beauty, 
its metaphorical energy of language ; there are few 
works fuller of colour and power than Job. Certain 
mythical psalms, generally misunderstood, several of 
the visions of Isaiah, Ezekiel, and others, when the idea 
does not disappear in a delirium like the frenzy of 
an oracle ; lastly, the simple narratives of the Book of 
Judges, so pagan, so cruel, and so courageous, may 
be read, some certainly with admiration, others with 



The Semitic World. 22 1; 

pleasure. PTebrew has the concision and the strength, 
of Latin with the simplicity of the analytical languages. 

Arabic, the prototype of the southern branch, has, 
on the other hand, the subtlety and richness of Greek. 
It has retained almost intact all the resources of 
Semitic speech, the three cases, various verbal forms 
(lost in Hebrew), the plurals,, obtained by modifying 
the vowels of the root {abd, servant, ihd or xbad, ser- 
vants) ; and we may add that we are far from knowing 
all the vocabulary of the other Semitic languages, 
whereas the vast Arabic literature, which embraces 
every subject and every style, employs all the wealth 
of its dictionary. 

This literature, of the purest and doubtless the 
oldest branch of the family, presents no monument 
comparable to the Bible, but it is brilliant and abun- 
dant. It is not ancient, and seems to have taken its 
rise only after the exhaustion of the other branches of 
the Semitic family. The famous Gacidns, poems which 
were crowned and suspended to the vaulted roof of 
the temple of Mecca, are little earlier than the time 
of Mohammed. Several of these describe with fury, 
so to speak, the sufferings and the joys of the fierce 
Bedouin : — ■ 

" I am not, says Chanfara, one of those beings who 
are stupid and timid like the ostrich, whose heart rises 
and falls in their breast like the lark in the air. . . . 
I swallow a handful of dust, without a drop of water, 
rather than give to an arrogant man the right to say 
that he has done me a service. ... I strangle hunger 
in the coils of my bowels, twisted like the cords of 
the spinner. ... I sleep on the hard earth, my back 
supported by the projecting bones of the spine. My 
pillow is a sinewy arm of which the joints stand out 



2 2 2 Distribution of Languages and Races. 

like the little bones tossed by the gambler. Know 
that I am a man of patience, that I wear its mantle 
over the heart of a hyena or of a wolf, and that hardi- 
hood serves me for sandals. How many times have 
I plunged into the rain and the darkness, with hunger, 
cold, and ten-or for companions." So lived the hero 
Antar, the robber of horses and of women, whose 
adventures have been so well translated by Marcel 
Devic. 

■- Islam and the Koran inspired these savage souls 
first with fanaticism and afterwards with a taste for 
mystical subtleties. Later the establishment of luxu- 
rious courts changed the poets of adventure into 
polished courtiers, servile parasites, clever narrators of 
Indian and Persian traditions, collectors of historical 
anecdotes, mixed with witticisms and verbal .jesting. 

Inspiration has failed and talent lias diminished, 
biit the written language is unchanged. The same 
cannot be said of the Arabic of the people, which alone 
has dialects, those of Barbary, Arabia, Egypt, and 
Syria differing very little the one from the other, but 
all characterised by the rejection of grammatical forms. 
This spoken Arabic lias arrived by degrees at the same 
stage as ancient Hebrew. The Mosarabic of Spain, 
which became extinct in the last century, and Maltese 
had long become formless patois. 

Arabic, being the language of Islam, has deeply 
penetrated all the Mussulman nations, Turkish, Persian, 
and Hindustani, and has contributed a considerable 
number of words to European vocabularies (see the 
Supplement to Littre's Dictionary) : '^ro, ciplur, cotton, 
sirup, algebra, magazine, crimson, &c. 

Himyarite reigned to the south of Arabic ; it was 
the language of the Queen of Sheba, and is now well 



The Semitic World. 223 

known through a great number of inscriptions, and is 
perhaps still spoken under the name of Ekhili in the 
district of Marah. But Islam carried Arabic to the 
shores of the Indian Ocean. It is in Abyssinia that 
we must seek for the last vestiges of Himyarite. 
Several centuries before our era, the African coast of 
the Red Sea had received Semitic colonies, and a 
language known as Ghez or Ethiopian, which was 
very developed and still had cases and thirteen verbal 
forms. In the fourth century, when Christianity pene- 
trated into Ethiopia, the Bible was translated into 
Ghez, together with other Jewish, Christian, Greek, 
and Arabic works. For some time Ghez has only 
existed as a learned or liturgical language, but a cer- 
tain number of allied dialects, Amharic, Tigre, Harari, 
are still spoken in parts of Abyssinia. 

In this rapid summary, which is all that my general 
scheme will permit, I have tried to sketch the historical 
and intellectual destinies of the Semites, noting succes- 
sively the Chaldean education of the Aramaic and 
Canaanitish nations, their certain and probable migra- 
tions, the movement caused by the invasion of the 
Hyksos among the Phoenician or Punic tribes, among 
the Aramaic or Syrian nation, and, lastly, among the 
family of Terah, Abraham, and Jacob ; the great 
maritime and commercial expansion of the Phoe- 
nicians ; the late establishment of the Israelites 
in Judea ; the crushing of the Canaanites between 
Egypt and Assyria ; the northern Semites subjugated 
and annihilated by the Persians, by the Greeks, and 
the Romans ; Greco-Roman civilisation, undermined 
by Eastern corruption, by enervating mysticism, by 
Christianity, the vengeance of the vanquished Semites; 
finally, when invasions were dismembering the Roman 



2 24 Distribution of Languages and Races. 

colossus, Mussulman fanaticism threw upon the world 
tlie Arab hordes which had remained till that time 
intact and free in their deserts. 

In enumerating the general features of Semitic 
speech, the invariability of the consonants and the 
flexion of the vowels, we have said that this important 
particular alone renders vain any attempt to establish 
a relationship between the Semitic family and the 
Indo-European. 

Finally, by a few dates and quotations we have 
characterised their groups and sub-groups, pointed out 
their points of resemblance and of difference, noted 
their duration, their respective merits, the importance 
and the bearing of their literatures. 



CHAPTER VIII. 
THE INDO-EUROPEANS. 

The science of language leaves untouched the domain of ethnography — 
Inattention of the ancients with regard to the manners, languages, 
and origin of their neigh bciurs — Philology, long forbidden by Chris- 
tian prejudice, was thrown open by Leibnitz — Discovery of Sanscrit 
— The Indo-European family of languages constituted by F. Sohlegel 
— Summary sketch of its eight branches; Celtic, Teutonic, Slav, 
Lettic, Italic, Hellenic, Iranian, and Hindu — Original unity, 
dialectic alteration — The mother-tongue and the organic forms — 
The cradle of the language should be sought between the two 
great sub-groups, Eastern and Western — The social, moral, and 
intellectual condition revealed by the elements which are common 
to all the Indo-European idioms — The Semitic history of Bossuet 
is effaced by the history of the Indo-Europeans — The Aryans 
reign throughout the world. 

The indigenous populations of quaternary Europe have 
been replaced, or rather over-laid, several times in 
succession, by migrations coming from the south and 
from the east. Successive crossings, modifying at 
once the type of the conquerors and of the conquered, 
have resulted in an extreme diversity in the height, 
build, and physiognomy, not only of the fifteen or 
twenty peoples who have shaped themselves during 
the historical period, but also of the far more numerous 
elements which geographical necessities and the course 
of events have united into nations. There is no 
doubt that these ethnical differences have had a very 
considerable influence on the construction and aspect 
of the idioms which have prevailed in a given region 



2 26 Distribution of Languages and Races. 

of Europe. It is to them that we must attribute the 
peculiarities of pronunciation, of accent, and of syntax 
which separate and characterise the Hellenic, Latin, 
Teutonic, and Slav groups and sub-groups. 

But ethnography and the science of language do 
not coincide. It is very rarely found that a people 
speaks the original language of its ancestors. Unless 
the disproportion in numbers or cultivation be too 
great, the language of the immigrant conquering 
minority is imposed upon the conquered majority, and 
even survives the race which imported it. Hundreds 
of millions of men may employ an idiom, altering it 
more or less, but without destroying the basis of it, 
which has been created in its entirety by a race, a 
people, a tribe, which disappeared thousands of years 
ago from the distant and unknown land where it had 
its birth. And when a wonderful discovery reveals 
this fundamental identity between the languages of 
rival nations, enemies, or at least separated by manners, 
aspirations and distance, it arouses, together with a 
legitimate astonishment, erroneous confusions, protes- 
tations, controversies, which are occasionally useful, 
more often idle or exaggerated. Some infer brother- 
hood of race from kinship of tongues ; others deny 
the existence of the human group which has invented 
this uniqiie vocabulary and grammar ; others again 
claim for their country and their ancestors the honour 
of having conceived and propagated them. Philology 
is, I submit, in a position to resolve these difficulties 
and to put aside these objections ; and that without 
trespassing on the domain which properly belongs to 
ethnography ; it does not minimise the differences, 
the specii,! characters of peoples ; it does not maintain 
that at a given time, at any time, the inhabitants of 



The Indo-Europeans. 227 

Western Asia, of Italy and Germany had a common 
ancestor ; it only establishes that they owe to a single 
definite group, and not to their own initiative, their 
languages, their .institutions, the germ of their destiny. 
The ancients were not unaware that the world was 
peopled before their arrival in Asia Minor, in Greece, 
in Italy. In many regions their predecessors main- 
tained themselves beside and amongst them. From 
the texts collected by M. Arbois de Jubainville credible 
traditions showed that the Iberians, nearly related to 
the Atlantides, were established in the west as far as 
the Rhone, and even threw off a branch into Italy, the 
Sicani ; the Pelasgians, under the name of Phrygians, 
Sardinians, Lydians, Lycians, Cares, Leleges, Tursenes, 
were scattered over the coast of Asia, in the archipela- 
goes of the JEgean, throughout Greece, and in southern 
Italy ; then came the successive arrivals of the Ligu- 
rians and the Siculi, of the Illyrians, Thracians, and 
Bithynians, closely followed by the little group of 
the Hellenic tribes. These vague traditions were all 
sufficient for the most enlightened Greeks. As for 
the different languages, which they certainly knew, 
and which were not extinct in the sixth century before 
our era, in the time of Peisistratus and Solon, it does 
not appear that they ever thought of collecting them. 
Their own idiom was enough for them ; all others were 
barbarous jargons, useless and negligible. Plato 
having remarked the resemblance of the names for 
fire and dog in Greek and Phrygian, contents himself 
with supposing that the Hellenes had perhaps received 
certain words from the autochthonous races. Even 
the prolonged contact with the Persians, whose lan- 
guage was learned by a few Greeks, notably by 
Alcibiades, did not win them from their indifference. 



2 28 Distribution of Languages and Races. 

The expedition of Alexander taught the invaders 
nothing ; the Sanscrit dialect spoken by Porus re- 
mained a closed book to the learned men who sur- 
rounded the king of Macedonia ; and if we did not 
know that the Etnperor Olaudian had written sixteen 
books on the history and the language of the Etruscans, 
we might affii-m that the sense of language was as 
absolutely unknown to the Latins as to the Greeks. 

We shall not expect to find the Middle Ages more 
enlightened than antiquity. It took the ancestors of 
modern peoples centuries to learn that which intelli- 
gent humanity had already acquired before, them. 
Christianity, retaining a few scraps of Latin, the 
science of the day, preached to the new populations 
resignation, humility, obedience, and ignorance. The 
fall of Constantinople, the exile from thence of thp 
scholars, and the dispersal of the Byzantine manu- 
scripts, the discovery of printing, were necessary to 
rouse Europe from its torpor. This was the Renais- 
sance ; the veil was lifted, at least for a few, and day 
began to dawn on Europe. Man turned again, to 
things of earth, and regaining an interest in all the 
manifestations of human activity, leaving faith for 
reason, recognised in speech the necessary instrument 
of thought and analysed its organism. Nevertheless, 
despite the efforts of Bibliander, Henri Estienne, 
Roccha, and Scaliger, who attempted some comparisons 
between Greek, Latin, and French, and of Guichard, 
who in his ITarmonie Etymologiqm (1606) distin- 
guished the Teutonic and the Romance dialects^ 
and constituted a separate family, including Hebrew, 
Chaldean, and Syriac, a capital error long turned 
philology from the right path. Orthodox logic could 
not seek elsewhere than in Hebrew the origin of all 



The Indo- Europeans. 229 

languages. Was it not in Hebrew that God spoke 
to Adam, and that the serpent tempted Eve ? More- 
over, God had dictated the Decalogue in Hebrew, and 
the creature made in his image could only speak in 
Hebrew. Even the boldest dared not doubt it. It is 
true that the adventure of Babel had happened since, 
but should there not exist, in the dispersed and con- 
fused languages, at least the traces of the primitive 
tongue ? One can but admire the ingenuity displayed 
by commentators and etymologists in the endeavour 
to, extract from the Bible the names of the gods of 
the heathen, and even Latin and French words. In 
order to bring Greek nearer to Hebrew, Guichard read 
it backwards, from right to left. 

Leibnitz was the first to oppose this inveterate 
prejudice. "There is," he says, "as much reason to 
consider Hebrew the primitive language as to adopt 
the opinion of Goropius, who in 1580 published a 
work at Antwerp to prove that Dutch was the language 
spoken in the Garden of Eden." He was the first to 
propose, in his " Dissertation on the Origin of Nations," 
the application of scientific methods to the science of 
language. Surmising that, in the absence of written 
history, the analysis of words might yield authentic 
information on the ideas and manners of primitive 
peoples, he proposed to Peter the Great, in 17 13, the 
plan of a collection of vocabularies. He drew up 
himself a list of common terms and encouraged the 
work of the German Bckhardt. His hypotheses, as 
we know, were too tentative, too little methodical to 
succeed ; but by their very failure they pointed out 
the way ; they showed that the first essential of fruit- 
ful comparison is the collection and classification of a 
sufficient number of facts. 
16 



230 Distribution of Languages and Races. 

The example of Leibnitz was followed by others. 
And if guess-work played the principal part in the 
clever study of Pr^ret on the " Origin and Mixture of 
Ancient Nations," if the premature philosophy of lan- 
guage, as displayed in the " Primitive World " of Court 
de Gdbelin, could throw no light on the affinities of 
European idioms, it was because there was wanting 
a standard of comparison which should explain their 
divergence. The sacred books of India concealed this 
standaud ; it lay there unknown and unexpected, 
until this century discovered it and realised its im- 
portance. 

Sanscrit, the language of the Brahmans, known 
before our era to the Buddhists of China, had been 
studied from the eighth century by Persian, Arab, 
and Turkish translators. Some fragments of its rich 
literature had even reached us and have remained in 
our tales and apologues. But although towards the 
end of the fifteenth century Filippo Sachetti had 
noted some points of resemblance between Italian 
and Indian words, it is doubtful that even the name 
of Sanscrit was known in Europe before the middle of 
the eighteenth century. 

Vasco di Gama, meanwhile, had landed in Calicut 
in 1498; the Portuguese missions, throwing them- 
selves at once on the rich Indian prey, must have 
learned the language of the countiy, Tamil, and from 
the year 1559 the priests of Goa knew enough of the 
doctrines of India to invite the Brahmans to public 
controversy. In 1606, Roberto de Nobili, who dis- 
guised himself as a Brahman, and cleverly presented 
himself as the interpreter of a fourth Veda, read in 
the original the Laws of Manii and the Puranas. It 
was doubtless under his influence that the Ezur- 



The Indo-Europeans. 231 

Veidam was composed in India, a Christian imita- 
tion of tlie Vedas, which holds a certain place in the 
erudition of Voltaire. 

Fr. Pons in 1 740 sent to Fr. Duhalde an exact 
description of the four Vedas, of the grammatical 
treatises, and of the six great systems of philosophy. 
Lastly, in 1767, another Frenchman, Fr. Coeurdoux, 
sent to the AbW Barth^lemy, who in 1763 had 
asked him for some historical information, two papers 
on the analogies and the kinship of the Samscroutan 
language with Greek, Latin, German, and Sclavonic ; 
he gave four lists of similar words and grammatical 
forms, noted the presence of the augment in Sanscrit 
and of the a privative ; he refused to attribute to 
borrowing and commercial dealings resemblances which 
affected not only isolated terms, but the formation of 
the words themselves. If these precious documents 
had been made public, France would have had the 
honour to inaugurate the comparative study of Indo- 
European languages. Unfortunately they remained 
buried in the archives of the Academy, and only 
appeared in 1808, at the end of a memoir of Anquetil 
Duperron. In the interval science had progressed ; 
England and Germany had made the discovery which 
might have belonged to us. 

The affinities recognised by Hahled, 1778, Sir 
William Jones, Paulin de Saint-Barthelemy (Philippe 
Wesdin), 1790, were admitted by Lord Monboddo 
(1792— 1795). Dugald Stewart, it is hard to say 
why, was obstinate in denying the existence of San- 
scrit. But his incredulity was unavailing against the 
grammars published from 1790 to 1836 by Wesdin, 
Colebrooke, Carey, Wilkins, Forster, Yates, Wilson, 
Bopp, Benfey ; against the texts edited, beginning in 



232 Distribution of Languages and Races. 

1784, by the first Asiatic Society, founded at Calcutta. 
The contrary exaggeration prompted the enthusiastic 
Oriental scholars to regard Sanscrit as a universal 
mother-tongue. Sir William Jones avoided this error; 
he supposed for Sanscrit, Greek, and Latin a common 
source, which perhaps, he says, exists, no longer. 
Modern science has confirmed his hypothesis, and, 
while recognising the general priority of the Sanscrit 
forms, notes in the other idioms" of the family peculi- 
arities which cannot be traced farther back, which 
are, so to speak, collateral, and point to the necessary 
existence of an earlier language, of a type which is 
yet visible through the alterations suffered by its 
various forms. 

While the mysteries of India were being revealed 
to English investigators, two vast collections, the " Cata- 
logue " of Hervas, and the " Mithridates '' of Adelung, 
came to furnish philology with the treasure of facts 
which alone can change hypothesis into certainty. 
Hervas, a Spanish Jesuit, and a missionary in Ame- 
rica, collected three hundred vocabularies and thirty 
grammars, discovered the unity of the Malay group, 
the independence of Basque, the relationship of 
Hungarian, Lapp, Finnish, ^nd suspected the relation- 
ship of Greek and Sanscrit. His work, in six volumes, 
dates from 1 800. The " Mithridates," founded in part 
on the " Catalogue," in part on vocabularies collected by 
order of Catherine II., appeared from 1 806 to 1 8 1 7. 
Adelung died in 1 809, but his son finished the work. 
The classification of languages could thenceforward go 
on with a more rapid and assured step and in the 
right direction. The most brilliant, the richest, the 
most vigorous group, that which was the first to be 
clearly defined, claiming for itself the place till then 



The In do-Europeans. 233 

abandoned to the Semites, was the group to which 
our European languages belong. In 1808, the poet 
Frederic Schlegel, who had studied Sanscrit under 
Hamilton (180 1 — 1802), constituted clearly, in his 
book on the " Language and Wisdom of the Hindus," 
the Indo-Germanic family. Though the work is out 
of date, like the symbolism of Kreutzer and of Herder 
which inspired it, though the bold guesses of the 
author have fallen before the demonstrations of gram- 
matical analysis, yet Schlegel is to Adelung, even to 
Sir William Jones, what Copernicus is to Ptolemy. 
He conceived a new world ; he created one of the 
richest domains of the human mind, or rather he 
opened its doors. His book, which is no longer read, 
gathers dust on the threshold of the science of which 
he was the inaugurator. 

Before studying the organism of the Indo-European 
speech, such as we are ^ble to reconstruct it from the 
features common to its numerous varieties, it is indis- 
pensable to glance over the immense area which it 
covers, and to indicate, in space and time, the place 
occupied by each of the groups of languages which 
have issued from it. If we disregard its modern 
annexes, which include the two Americas and Australia, 
we shall find that it reigns from the months of the 
Ganges to Iceland, and from Sweden to Crete, compre- 
hending five-sixths of Hindustan, Afghanistan, Persia, 
Armenia, three-quarters of Eussia, of Sweden, and of 
Norway, and all the rest of Europe, except the Basque 
country, Hungary, and a portion of Turkey in Europe. 

In the extreme west, in Scotland, in Ireland, in 
Wales, and in Brittany, we find the remains of the 
Celtic group, generally subdivided into Gaelic (in- 
cluding Erse and tlie dialect of the Isle of Man) and 



234 Distribution of Languages and Races. 

Cymric (including Cornish and Breton). Save for a 
few inscriptions which are not yet completely ex- 
plained, these languages are only known to us by 
relatively recent texts. 

Some Irish glosses of the eighth century, and a few 
Breton and Cornish documents of the eleventh and 
thirteenth centuries, are the most ancient remains (all 
the rest is hypothetical) of a language formerly spoken 
in the north of Italy, throughout Gaul, and in the 
British Isles, a language which, in spite of the ■ illu- 
sions of Celtic enthusiasts, has only left to the French 
tongue a few names of places, historical names men- 
tioned by the Latin writers, and about two hundred 
and forty authentic words in addition to these. 

Eacially the Gauls and Celts can be distinguished. 
The latter were round-headed, with dark hair and 
eyes, of middle stature, and strongly built. The Gauls 
were very tall, very fair, warlike and adventurous. The 
Celts probably occupied before the dawn of history 
the whole of Central Europe, the valley of the Danube, 
Savoy, Auvergne, Brittany, Ireland ; traces of them 
are found in Roumania (or Dacia), in Austria, and in 
Bavaria. Did these bring with them the Celtic dia- 
lects, or did they receive them from the Gauls or Bel- 
gians ? This question is insoluble, for it is impossible 
to give the date of the arrival of the Gauls, who were 
doubtless the first wave of that great flood which bore 
the Teutons to the north of the- Alps, the Latins to 
the south. Towards the sixth century they certainly 
occupied a great part of Northern Germany, dominated 
Gaul from the Ehine provinces to the Pyrenees, and 
Italy as far as the Po, perhaps as far as the Tiber. 
They destroyed Rome at the beginning of the fourtk 
century, Delphi a hundred years later, and even pene- 



The Indo- Europeans. 235 

trated into Asia Minor, into Galatia. It was to put an 
end to their incursions that the Eomans, after having 
with difficulty subdued them in Cisalpine Gaul, an- 
nexed the Transalpine provinces to the republic in the 
middle of the first century before our era. It is well 
known how rapidly the Gauls and the Celts adopted 
the languages and civilisation of their conquerors. 
Gallic, the most ancient of the Celtic dialects, had 
completely disappeared by the fifth century of our era, 
and the others are but the degenerate descendants 
of an extinct language which some consider to be 
related to Latin, others to Teutonic. However this 
may be, their literature, which is fairly abundant, has 
been carefully studied by Luzel, Gaidoz, D'Arbois de 
Jubainville, and the Indo-European origin of their 
vocabulary and grammar has been established by 
Pictet {Be, I'AffiniU des Langues Celtigms avec le Sans- 
crit, 1837), by Bopp (" The Celtic Languages from the 
Point of View of Comparative Philology," 1838), and 
by Zeuss (Grammatica Celtica, 1853). 

The powerful German branch had quite another 
destiny ; its historical existence is not very ancient, 
but it has itself ramified into vigorous and cultivated 
branches which cover a great part of northern Conti- 
nental Europe, the British Isles, and the United States. 
The earliest known name of the Germans or. Teutons 
(Teotisk) seems to be Bastarnes. From the year 182 
B.C., they wandered between the Niemen and the Ehine, 
from the Alps to the Black Sea. Soon appeared the 
Teutons of Marius, the Suevi of Ariovistus, then the 
Germans of Varus, the Quadi, Alamanni, Franks, of 
Marcus Aurelius, Probus, and Julian. Owing to the 
strange lack of curiosity in the ancients, nothing of the 
earliest times of the German languages has come down 



236 Distribution of Languages and Races. 

to us. By a fortunate chance, a precious manuscript of 
the fifth century, preserved at Upsal, the Codex Argen- 
teus, has retained for us the fragments of a Gothic trans- 
lation of the Bible. The author was a Cappadocian, 
brought up among the Western Goths on the Lower 
Danube, and under the name of Ulfilas he became their 
bishop and their chief (3 1 1-38 i). The Goths, Wisi- 
goths, and Ostrogoths, who played so fatal a part in the 
sad drama of the fall of the Roman empire, were the 
rearguard of the German invasion; they barred the 
passage between the Black Sea and the Baltic. Under 
the shock of the Slav or Wendic invasion, in the year 
77 of our era, they were driven partly into Sweden, and 
in part between the Dniester and the Balkans, whence 
they hurled themselves upon Greece, Italy, and Southern 
Gaul. Gothic became extinct in the .ninth century. 
By its less mutilated forms it may be classed almost at 
the same stage as Latin and Greek ; it is not the father, 
but the elder brother of the other Teutonic dialects ; 
its relationship with the Scandinavian languages and 
the Low Dutch dialects is specially marked. 

The most anciently cultivated of the Scandinavian 
idioms, Norse or Norrois, carried to Iceland in the 
ninth century by pagans fleeing from the Christian 
propaganda, has preserved for us the most precious 
traditions on the mythology of the North. The Hliods 
and the Quidas, which were recited in the seventh and 
eighth centuries in Norway before the emigration, 
were collected in the eleventh in the poetical Edda of 
Soemund. The prose Edda of Snorri Sturleson in the 
following century, and then numerous Sagas, complete 
the cycle of national legends, which are for the most 
part common to all the Teutonic tribes. Danish and 
Swedish, which developed side by side with Norse, 



The Indo-Europeans. 237 

form independent though nearly allied branches of the 
Scandinavian family. 

In the north of Germany there are certain spoken 
dialects which are no longer written, Platt-Deutsch or 
Low Dutch, which are intermediate between Scandi- 
navian, German, and English. This was the language 
of Wittikind, and two manuscripts of the eighth cen- 
tury have transmitted to us a Christian poem written 
at this epoch for the conversion of the Saxons, the Hel- 
jand or " Saviour." Frisian, cultivated in the twelfth 
century, Flemish, the language of the Burgundian 
court in the fifteenth century, and its twin, Dutch, be- 
long to the same group, and are intimately connected 
with Anglo-Saxon. The English tongue, which has 
received from Latin and French more than half of its 
rich vocabulary, is none the less essentially Germanic 
in what remains to it of grammar and in the core of 
the language. It was introduced in the fifth and 
sixth centuries by the Jutes and the Angles. Anglo- 
Saxon, very nearly allied to Gothic, is represented 
by the epic poem of Beowulf, which is attributed to 
the seventh century ; it was spoken until the time 
of William the Conqneror (1066). Thanks to the 
simplification which is the result of time, this old 
idiom has renewed its youth ; the la,nguage of Shake- 
speare, of Bacon, of Walter Scott, and of Shelley has 
produced a magnificent literature, and has spread itself 
over the whole earth. It is the conquering idiom. 

The Teutonic tribes destined to form the German 
nation, properly so called, have gone through many 
vicissitudes, which partly account for the absence of 
ancient documents in their dialects. That which the 
Romans and the Gauls called the Germanic invasion 
was commonly merely a forced emigration under the 



238 Distribution of Languages and Races. 

double pressure of the Slavs and the Huns. Prom the 
fifth to the sixteenth century there were no Germans 
in Eastern Germany. Slavs occupied Silesia and 
bordered on Saxony ; the Avari approached the Ehine 
and harassed the frontiers of Charlemagne. Inde- 
pendent Germany in . the eighth century was reduced 
to Saxony, then conquered and annexed by the Frankish 
emperor. The Franks themselves, who had spread in 
great numbers over the Rhine provinces, were, so to 
speak, lost in the Latin empire, to which one of their 
families, which was much crossed with Belgian blood, 
had furnished the chiefs. And though the kings of 
Austrasia had kept their national dialect, although 
Charlemagne spoke it and took care to collect Ger- 
manic songs and traditions, the domain of the true 
Teuton was extremely limited. It comprehended Ala- 
man, Bavarian, Suabian, and Frankish dialects. The 
Prankish of the Merovingians and of Charlemagne 
no doubt held the first rank in Old High German. 
We may mention, as belonging to this period, the text 
of the sermon pronounced by Charles the Bald in 843 
before the battle of Fontenay, and, in the tenth cen- 
tury, a poem which celebrates the victory of Louis 
III. and of Carloman over the Normans. 

In the thirteenth century Suabian prevailed and 
constituted Middle High German ; it was the language 
of the Minnesingers, and has been rendered famous by 
the creators of the national poem of the Nibelungen. 
Finally, literary German arose with the translation of 
the Bible by Luther, as did classical Arabic with the 
Koran, and became the universal language of a far 
larger Germany. I need not praise German poetry, 
philosophy, and science. But we may be permitted to 
regret, in language as in religion, the extreme timidity 



The Indo-Europeans. 239 

of the Reform. Luther did not venture to rid the 
language of the silent or nasal terminations, of the 
clumsy construction, of the relics of declension which 
trouble the ear and weary the mind. 

The northern provinces of Prussia were long occupied 
by the Letts and Lithuanians, who had taken the 
place of the Vandals, the Heruli, and the Lombards. 
The greater part of them were attached to Germany 
by conquest, by the crusade of the knights of the 
Teutonic order. Russian Lithuania shared the fate 
of Poland. The Lettic group, interesting by its archaic 
forms, is only known to us, as so often happens, by 
modern documents. It comprehends Old Prussian, 
which became extinct in the seventeenth century, and 
is represented by the eight hundred words of a lexi- 
con of the fifteenth century, and by -a catechism 
dated 1561. On the frontier of Eastern Prussia and 
in Russian Lithuania, about 150,000 people speak 
Lithuanian, which is often better preserved than 
Sanscrit itself. Its literature consists of the works of 
a poet, Donalisius (17 14— 1780); a few prose fables 
have also been collected, together with proverbs and 
popular songs. Lettic, which is more corrupt, is 
spoken in the north of Courland and in the south of 
Livonia by about a million of people. 

These languages are akin to one of the largest" 
groups of the whole family, the Wendic or Slav group, 
which came into Europe during the first five centuries 
of our era ; it is divided into two great branches. 
Eastern and Western. The first includes Russian, 
Great Russian in West Central Russia ; Little Russian, 
Rusniac, or Ruthene in the south of Russia and even 
into Austria (spoken by fourteen millions of people ; 
there are documents of the eleventh century), Ser- 



240 Distribution of Languages and Races. 

vian, Croatian, Slovenic, and Bulgarian, of which the 
most ancient form is to the whole group what Gothic 
is to the German dialects; modern Bulgarian is, on 
the contrary, very much altered. Old Bulgarian or 
ecclesiastical Slav, which Miklosich, the author of the 
Wendic grammar, declared to be the father of all the 
Slav idioms {lingua Palceo-Slomnica) was fixed in the 
ninth century by the apostles Cyrillns and Methodus 
in their translation of the Bible. Slovenic has left 
fragments which date from the tenth century. 

The western branch covered from the seventh to 
the ninth century vast districts of Germany in which 
only German is now known : Pomerania, Mecklen- 
burg, Brandenburg, Saxony, Western Bohemia, Austria, 
Styria, and Northern Oarinthia. Though now much 
restricted, it can still boast numerous dialects ; among 
others the Wendic of Lusatia, which is dying out, 
Tzech or Bohemian, which is very vigorous (ten mil- 
lions), of which a variety, Slovac, is found in Hungary ; 
lastly, Polish (ten millions), of which the very important 
literature begins at the end of the tenth century, and 
numbers, from the twelfth onwards, many chroniclers 
and poets. Tzech has been cultivated from the eighth 
century ; its first documents are the celebrated manu- 
scripts of Kralovdor and of Zelenohora, discovered in 
1 8 17. Since they date from the transition period 
between Christianity and Paganism, they are as valuable 
to the student of mythology as to the philologist. 
The time of Huss gave great prominence to Tzech 
letters ; but conquered and given over to the Jesuits, 
Bohemia's language was proscribed ; it has, however, 
been revived from the end of the last century. The 
relatively modern cultivation of the Slav languages 
does not alter the fact that they date from the earliest 



The Indo-Europeans. 241 

period of the Indo-European speech ; their grammar 
has a very archaic character, especially in the declen- 
sions. The various branches of the group are closely 
connected. Safarick tells us that a Bohemian under- 
stands Slovac, a Slovac Polish, a Pole the Wendic of 
Lusatia. A modern Eussian can still with a little 
attention follow the Bulgarian office of the ninth cen- 
tury. Eussian and Polish, though belonging to two 
distinct classes, hardly differ from each othei- more 
than do Spanish and Italian. 

The South of Europe belongs to the Italic and 
Hellenic families. The one has given birth to our 
languages, the other by its literature has formed our 
mind. They are in the first rank, — the finest impres- 
sion of the Indo-European type. 

Latin was at first a very small central group of 
dialects, Sabine, Volscian, Latin, superposed upon 
the unknown languages of the aborigines, Ausonians, 
Auronci, and Siculi. Its history is that of Eome 
itself. From the eighth to the fourth century B.C. it 
was written only in certain Annals, in a few liturgical 
books and songs, and in the Law of the twelve tables, 
and remained confined to Latium, between two kindred 
languages, Samnite to the south, Umbrian to the north, 
surrounded by Etruscan in Tuscany and in the Cam- 
pania. Celtic reigned in the valley of the Po, Greek 
in the two Sicilies. Samnite or Oscan, spoken and 
understood in Rome as well as Latin, and Umbrian, 
still heard on the right bank of the Upper Tiber at 
the time of the Antonines, have left valuable inscrip- 
tions, deciphered by Mommsen, Aufrecht, and Kirch- 
hoff (1845- 185 l), and completely elucidated by Michel 
Br^al. The tables of Agnona and Iguvium show us 
very peculiar forms and a remarkable phonetic system. 



242 Distribution of Languages and Races. 

wticli certainly influenced the Latin pronunciation. 
Latin very early outgrew its primitive rudeness, of 
which a few inscriptions have preserved examples, and 
took that gravity, that harmonious strength, which 
command our admiration. By the first century of our 
era this tribal dialect, the language of Plautus, of 
Ennius, of Lucretius, of Cicero, of Virgil, and of 
Tacitus, had conquered not only Italy, but also Spain, 
Gaul, and Northern Africa ; and until the eighth 
century it remained the idiom of the civilised or half- 
barbarous We.st. But this language of literature and 
of the government had not suppressed the provincial 
patois, the Latin of the people, imported into the 
countries conquered by the legionaries. The Latin 
of the country and of the camp {riisticus, castrcTisis), 
modified, contracted, mutilated by Dacians, Germans, 
Gallo-Franks, Oeltiberians, gave birth, towards the 
ninth century, to seven new groups of dialects, 
called Neo-Latin or Romance : French (of the Isle of 
France, Burgundian, Picard, Walloon, Norman), Pro- 
venpal (Dauphinese, Genovese, Piedmontese, Limousin, 
Toulousain, Bearnais, Catalan), Spanish, Portuguese, 
Italian (Venetian, Lombard, Tuscan, Corsican, Sardinian, 
Neapolitan, Sicilian), Latin (Friulian, Tyrolese), and 
Roumanian (Moldo-Wallaphian), which, more or less 
mixed with foreign words, have kept the vocabulary 
and the accent of their mother-tongue, but have carried 
the Indo-European elements from the synthetic to the 
analytic stage. The history of the development of 
these languages, all daughters of provincial Latin, 
shows us the gradual transformation of an idiom into 
fcee and oi-iginaL derivatives. This phenomenon, which 
has taken place, as it were, before our eyes, will explain 
what took place when the Indo-European speech split 



The Indo- Europeans. 243 

into the various families : history throws light even 
upon pre-historic time. 

Greek, the most complex, the most subtle, and the 
most learned of the languages of antiquity, was de- 
veloped centuries before Latin ; the traditions of the 
Hellenes take us back 1800 years before our era; the 
name of the Achaians figures on an Egyptian inscrip- 
tion of the fourteenth century B.C. Asia Minor was 
colonised in the eleventh century, the epoch of the 
Homeric poems, which were collected in the sixth. 
Tradition tells first ot the legendary heroes, ^olus, 
Acheeus, Ion, and Dorus, who descended by Mount 
Haemus among the Thracians, the Pelasgians, and the 
Epirotes, whose languages are perhaps preserved for 
us in Albanian and Etruscan, and established them- 
selves in the mountainous districts of Thessaly, of 
Pieria, and of Phtiotide, round Dodona and Delphi ; 
then from the Hellad and the Peloponnesus, which they 
rapidly conquered, the four or five tribes sent out 
swarms into Asia, Africa, into Italy and Gaul, where 
they everywhere succeeded the Phoenicians. The 
various co-existing dialects, -iEolian, the link between 
Greek and Latin, the Ionian of Homer, Hesiod, and 
Herodotus, the Attic of Plato and Demosthenes, the 
Dorian of Pindar, the choruses of the tragedians and 
the idylls of Theocritus, Cretan, Laconian, Macedonian, 
&c., preserved either by an imperishable literature or 
in abundant inscriptions, allow us to study in a most 
complete manner the structure and the history of Greek. 
Towards the time of Alexander its dialects, though 
they had not completely disappeared, were confounded 
in a uniform literary language, that of Polybins^ 
of Plutarch, of Lucian, which was spoken and under- 
stood from Marseilles to the Euphrates, from Byzantium 



244 Distribution of Languages and Races. 

to Alexandria and Gyrene; it was characterised by 
the predominance of Attic. Towards the fifth century 
the Greek pronunciation became corrupt ; this was the 
Byzantine age, from which the language passed by 
degrees into Romaic or modern Greek, now spoken 
in Greece, in the Archipelago, and on the coasts of 
European and- Asiatic Turkey. The vast river at 
whose waters all thinking humanity slaked its thirst 
has now dwindled to this little stream. 

Leaving Europe, we find in Asia Minor a few extinct 
and little known languages, of which the inscriptions 
will doubtless determine the character, Phrygian, Garian, 
Lycian ; they were certainly related to Greek, and 
perhaps also to the Iranian group, of which we find 
the vanguard in Armenian, Khurd, and Ossetan or Iron 
of the Caucasus. We cannot tarry over these, which 
are, however, very interesting (especially Armenian). 
We can only enumerate also the Iranian dialects of 
the Bast, Afghan or Pushtu, and Beluchistan ; ancient 
Persian claims our, attention. 

The discovery of the Iranian group, which is re- 
presented by the charming language of Firdousi (tenth 
century), of Hafiz, and of Saadi, is one of the most 
glorious achievements of modern philology. The ad- 
venturous Anquetil Duperron, a Frenchman, at the 
price of unnumbered and unimaginable fatigues, after 
having learned Tamil, Persian, and Pehlevi at Surat 
and at Pondicherry, acquired from the Destours, or 
priests, i8o manuscripts, among others the Zend- 
Avesta, accompanied by Pehlevi, Sanscrit, and Persian 
translations, and escaping with them from the hands 
of the English, who had taken him prisoner, deposited 
them at last in the Royal Library at Paris (1754— 
1762). The translation which he published in 1772, 



The Indo-Europeans. 245 

made from the Persian, is extremely imperfect ; but it 
arrested the attention of the philosophers and philo- 
logists. Rask, a Dane, was the first after him to 
attempt a translation of the original text ; but the 
honour of founding the study of Iranian belongs to 
Eugene Burnouf. By comparing the Zend with the 
bad Sanscrit of the translator, Neriosengh, Burnouf 
discovered, with its grammar as well as its vocabu- 
lary, a language which enabled him to read the cunei- 
form inscriptions of Xerxes and Darius. His works 
("Commentary on the Yafna," 1833; "Memoir on 
the Inscriptions of Hamadan," 1836; " Studies on the 
Zend Texts," 1840— 1850) have been taken up and 
completed by Brockhaus, 1850, Westergaard, 1853, 
Haug, Kossowics, Justi, Spiegel (185 1—56— 63), and 
lastly by MM. Michel Br^al, Hovelacque, and De 
Harlez. 

The most ancient Iranian monuments of which the 
date is certainly known are the inscriptions of the 
Achemenides (Hamadan, Bisoutoun) ; they belong to 
Persian proper. The Zend texts, in the state in 
which we have them, are probably later than the 
origin of the Husvarech or Pehlevi of the Sassanides, 
and of Parsee ; their date may be fixed at the third 
centmy of our era (226). Yet their language presents 
forms of the highest antiquity, almost always twins of the 
Sanscrit forms. This is because these Gathas, liturgical 
litanies, the remains of a literature which had already 
been extinct for perhaps five centuries, belong to an 
epoch which witnessed a restoration of Magism, and 
have preserved for us an idiom spoken by Zoroaster, in 
Media and Bactriana, some three thousand years ago, 
and carried eastward by the ancestors of the Medes 
and Persians when they came to establish themselves 



246 Distribution of Languages and Races. 

to the north of the Elarnites and of the Assyrians, while 
the Ossetes and Armenians, connected with the Slavs by 
the Scythian dialects, passed along the western coast 
of the Caspian and the mountains of the Caucasus. 

Meanwhile the Aryans of India (this title will not 
be denied them any more than to the kings of Persia, 
who claim it), the future conquerors of Bengal, advanced 
slowly into the Punjab among the affluents of the 
Upper Indus, stopping here and there to build houses, 
to till the ground, pasture their fl&cks, to wage war 
among themselves, celebrating the discovery of fire 
upon the sacred hearth, associating in their sacrifices 
their ancestors, the forces of nature, and the brilliant 
gods of the storm and of light. Towards the tenth 
century they reached the Ganges and the mouths of 
the Indus ; they penetrated into the great peninsula, 
took possession of Ceylon, and overflowed into Burmah, 
Gambodge, and the Malay Islands. They acquired 
the art of writing late, only in the third century, when 
classical Sanscrit had ceased to be spoken, and was 
merely the language of literature ; they fixed the text 
of the Vedas, preserved by oral tradition, and composed 
in a language older than Zend. But though replaced 
in common use by Prakrit, by Maghadi, the language 
of Buddhism, by Pali, the sacred language of Ceylon, 
it remains the idiom of the philosophers, grammarians, 
and poetSj of the great epics, of the drama, of the 
Puranas ; it has not ceased down to modern times to 
be the sacred language of the Brahmans, who still write 
and speak it. Around it flourish the modern dialects, 
its children and grandchildren, Hindi, Hindustani, 
Bengali, Mahratta, Guzerati, the Eomany of the gypsies, 
and its powerful influence is felt even in Malay, through 
Kawi, the sacred language of Japan. 



The Indo-Europeans. 247 

In order to prove the original unity of all these 
languages it would suffice to compare a few hundreds 
of words taken at hazard from the several grammatical 
categories ; but even more convincing is the study of 
the transformations undergone, from age to age and 
from nation to nation, by the elements, roots, and 
suffixes which are common to them all. It is, in fact, 
the constancy of the formal and phonetic changes in 
each group and each idiom of the family respectively 
which has served as the basis of comparative grammar. 
This phenomenon has allowed Bopp and Schleicher to 
measure as it were the degrees of relationship between 
kindred languages, to distinguish between the elements 
common to all, and the particular use of these elements, 
from which results the original development of each 
idiom ; to bring etymology into accord with the law 
of dialectal alteration ; and lastly, to discover, for each 
root, and for a great number of words which are con- 
jugated and declined, a primitive form, or, if not primi- 
tive, at least anterior to the variants of which it is 
the point of departure and the source. So that the 
divergences of the dialects furnish the surest proofs 
of their genealogical affinity, and by bringing back 
the student to the type of which they have blurred 
the outlines, they reveal to him the features, certain 
or probable, of the ancient Indo-European organism ; 
just as the numismatist traces in certain Merovingian 
or feudal coins the features of Probus, Aurelian, or 
Philip, disfigured by the clumsy tool of the ignorant 
and barbarous copyist. 

Thus it is that comparative grammar is enabled to 
re-establish, according to all probability, the organic 
forms of the Indo-European idiom, at the moment 
when, having already attained to the inflected state, it 



248 Distribution of Languages and Races. 

was about to undergo those alterations and trans- 
formations which cleft it into eight mother-tongues. 
Suppose that Latin had disappeared ; an attentive 
comparison of the seven Eomance idioms which arose 
from it would enable us to reconstitute it. So with 
Indo-European ; bold or prudent philologists, Chav^e, 
Schleicher, analysed its mechanism ; Fick drew up its 
dictionary. 

The organic form so discovered becomes the term of 
comparison among all those which are more or less 
different from it, without, however, losing all trace of 
it. And it becomes clear at once that no idiom tends 
towards the organic state, but that all tend away from 
it ; all in varying degrees are, not sketches, but modi- 
fied effigies ; not embryos, but remnants and vestiges of 
an earlier unknown type. 

Again, it is easy, as we compare root with root, 
termination with termination, to show that the altera- 
tion, the wear of the elements common to the different 
vocabularies, increases as we go westwards, from the 
Sanscrit of the Vedas to Zend, from Zend to Slav, 
from Zend to Greek, from Slav to German, from Greek 
to Latin, from German and Latin to Celtic. Partial 
exceptions are assuredly numerous, but there is a 
general law. Avoiding absolute formulas, we may say 
that the eastern branch of the Aryan tongues, Sanscrit 
and Persian, is in a far better state of preservation, far 
nearer the organic state than the north-western and 
south-western branches. 

If, then, Indo-European has existed, with its roots 
and terminations, its declension and conjugation, with 
its typical grammar, it must have taken rise in a region 
where the ancestors, the linguistic ancestors, of the 
Hindu and the Persian, of the Greek, the Latin, and 



The Indo-Europeans. 249 

the Celt, of the Goth, the Scandinavian, the Teuton, 
the Slav, and of the Lithuanian could know and under- 
stand each other. The trunk can only be found where 
the branches spring. 

"What does tradition or documentary history tell us ? 
With regard to the past of the Hindus and the Persians, 
we have the testimony of the Vedas, of the epics, and 
of the Zend-Avesta. We can determine the march 
from the gorges of the Hindu-Kush of the wandering 
tribes who advanced by slow degrees from the Punjab 
to Bengal and the Dekkan, and conquered the great 
peninsula, without, however, destroying the conquered 
races. They came from the north, and none of their 
traditions points to a distant Western origin. Their 
primitive country, which they call Aryavarta, is the 
same as the Arya-Vsedja of the Iranians^ who, there is 
no sort of doubt, remained in it longer tha*^ they — long 
enough to forget them. Now the Persian Aryans 
came from Bactriana, where the Gathas of ti(e Avesta 
were composed ; from Bactriana, whence th^y were 
driven by the Turanians, the Turks, their legendary 
and historic enemies, who were still cursed ip. the 
tenth century of our era in the Shahnameh. The pame 
Arie, Ariana, given to a region which lies between 
Afghanistan and Media, marks the second stag^ of 
these Persian Aryans, some of whom decided to go 
round the Caspian to gain Media and Armenia, ^he 
others massing themselves by degrees in Persia proper, 
until the day when the Medes and the Assyriai^s 
yielded them the empire. 

Loner before the arrival of the Persians in Western 
Asia, and even before that of the Hindus in the valley 
of the Ganges, history shows us relations of the Greeks, 
the Phrygians, and the Lycians installed in Asia Minor, 



250 Distribution of Languages and Races. 

and the Hellenes themselves in the valleys of the 
Hsemus and of the Pindas, having already left Thrace, 
of which they knew and celebrated the mountains, 
Ehodope and Ismara. There is no doubt a vast hiatus 
between the Caspian and Thrace, and the Hellenes did 
not recollect the journey that they must have made. 
Here, however, mythology and philology come to the 
aid of history ; it is impossible to separate the language 
of Homer and that of the Vedas ; the legends crystal- 
lised by the Hellenes, the Phrygians, and the Cretans 
round Olympus and Ida come direct from the fount at 
which the Aryan rhapsodists dranlf. The march of 
the Hellenic tribes towards the Hellad and the Pelo- 
ponnesus is the . continuation of the movement which 
bore them into Phthiotide and into Pieria. Lastly, the 
fable of the Argonauts, the expansion of their colonies 
on the northern shore of the Black Sea, denotes an 
earlier acquaintance with Colchida and tlie Chersonese 
of Taurus. It was to the Caucasus that Zeus bound 
Prometheus. Ai'e not these reminiscences of the lands 
which they had travelled over, pressed by the Cimerians 
(the Cymri or Gauls) and the Scythians, who were 
doubtless Slavs, mingled with the Finno-Mongols ? 

The Latins are even more ignorant than the Greeks 
of their origin, since their history only begins in the 
eighth century B.C., and their Trojan traditions were 
borrowed from the Etruscans and from the Greeks of 
Cumse. But they could not have learned or created 
their language in Italy ; it proceeds directly from the 
Indo-European source, and is connected with the most 
ancient form of Greek, with j3Eolian. Nor did they 
invent their Jupiter, the Dyauspitar of the Aryans. 
Few in number, a small tribe, lost between the Hellenic 
nations and the Gallic mass making its way up the 



The Indo- Europeans. 251 

Danube, they must have passed unperceived along the 
Alps and the Adriatic, borne onwards by the migra- 
tion of the Umbriaus (fourteenth century B.C.), or 
urged forward in the tenth century B.C. by the exodus 
of the Pelasgians or Etruscans. Then they encamped 
between the Albi and the Curii and vegetated there, 
until the day when they took part in the foundation 
of Rome. 

The diirk-haired Celts, of whom ethnography finds 
the traces from Dacia to Artriorica and Ireland, the 
fair-haired Gauls {Volk, Bolg, whence Belg-ian and 
Welsh; Eag. folh), the Gauls, who at the time of 
Ambigat and Biturix occupied the whole of Germany, 
and soon after of Gaul, Great Britain, and the west and 
centre of Spain, the Cymri, who were probably identical 
with the Cimmerians, all these peoples, who spoke Indo- 
European dialects, certainly progressed from east to 
west ; so much so, that they were driven to the Ehine 
and the Atlantic by the Germans and the Slavs. The 
Gallic language, which has almost entirely disappeared, 
was, it is well known, related to Latin, which explained 
its rapid disappearance. As for the Neo-Celtic dialects, 
in spite of the modifications they have undergone, they 
are none the less marked with the family features. 

The Germans, Slavs, and Lettic race remain to be 
considered ; their origin cannot admit of a doubt. The 
first did not occupy their present country at the time 
of the Gallic dominion, or at most they were disputing 
the coasts of the Baltic with their Finnish predecessors. 
It was only in the middle of the first century before 
our era that the Suevi appeared in force on the Lower 
Rhine ; Caesar kept them to the right bank. By 
degrees Germany filled up between the Oder and the 
Rhine, between Jutland and the Alps ; it swarmed 



252 Distribution of Languages and Races. 

with once-famous tribes, whose names have now dis- 
appeared from our maps, Cherusci, Irminoni, Isc^voni, 
Ingsevoni, Quedi, Marcomanni, who were either de- 
stroyed in their long wars with the Roman world or 
in their intestine broils ; they existed obscurely up to 
the fourth century in the Decumatian lands, more or 
less subject to tribute, and penetrated as far as the 
Weser by the Roman legions and influence. Behind 
them stretched the land of the Goths, the most 
powerful of the Germanic races, who in the fourth 
century covered, from the Baltic to the Dniester, what 
was afterwards the Polish dominion ; they came from 
more distant lands. The proof of this is found in the 
pressure of the Slavs, themselves harassed by the Huns 
in their Scythian pastures ; this flood threw one branch 
of the Goths upon Sweden, the Visigoths and the 
Ostrogoths on to the right bank of the Danube, and 
penetrated into the heart of Germany. 

The Germans of the north, the Lombards, the Rugi, 
Heruli, Vandals, urged by the Borusses and the Lithu- 
anians, were already in movement, wandering where 
chance led them, some to Italy, some to Spain, and 
even Africa ; then the Suevi, the Burgundians, and 
the Franks arrive in turn in the valleys of the Meuse, 
of the Scheldt, of the Somme. The Teuton tribes 
who remained in Germany, Alamanni, Suevi, Pranconians, 
Saxons, were crowded between the Rhine and the Weser, 
sometimes attained to the Elbe ; everywhere the Huns 
dominated, followed by the Slavs. The ancient Teutons 
dwelt in more or less scattered or dense masses in 
Scandinavia, England, the north-west of Gaul, Spain, 
and Cisalpine Gaul. 

But I have already sketched the table of these 
complicated events, which will sufiBce to destroy the 



The Indo- Europeans, 253 

pretensions of our neighbours to the lands to which 
emigration once carried them, and where their bands, 
dominant for a time, ended by becoming absorbed into 
the earlier population, and also to show the Eastern 
origin of the Teutons. 

In the case of the Slavs, the point is not likely to 
be called in question ; but it may be necessary to 
insist upon the fact that the numerous and very rich 
languages of these peoples could not have been framed 
in the lands where they are now spoken, or have issued 
the one from the other ; they betray their close rela- 
tionship with the organic Indo-European. 

As long as no sign shall have been discovered of a 
Western origin in the case of the Slavs and Iranians, 
of the two groups which remained together longest in 
the neighbourhood of the common cradle, so long as it 
has not been demonstrated that the Celts and the 
Gauls progressed without an obvious reason from the 
west towards the east, or that the German hive, solid 
in its centre from all time, sent forth swarms to the 
right and the left, Celts, Gauls, Slavs, and Persians, even 
Latins and Hellenes, so long we shall be constrained 
to place the Indo-European country somewhere between 
the eastern and western branches. But even sup- 
posing that the smallest particle of evidence could be 
alleged in support of one of these hypotheses, com- 
parative grammar would still be there to tell us that 
no one of these idioms can render an account of its 
forms and its rules. None can explain itself, but all 
can be explained by each other; none of them is a 
sketch of a type towards which it tends ; all are the 
■ various modifications of a common stock, of an earlier 
language, which has disappeared just because all have 
carried it away with them. Languages only travel with 



254 Distribution of Languages and Races. 

those who speak them. Those which concern us now 
have therefore been imported by immigrants, who 
were probably too few in number to modify materially 
the mixture of more ancient ethnic elements, but 
powerful enough to impose their language, their intel- 
lectual discipline, and in some cases the corresponding 
civilisation. 

For speech, being the expression of thought, assuredly 
reveals the aptitudes, the facilities of the brain, the 
industrial, aesthetic, and social condition of each human 
race and group. The Indo-European unity was not 
merely a matter of grammar and lexicon, but of in- 
tellect and morals. If the majority of the sister 
languages designate by the same word a thing, a ' 
being, a relation, a sentiment, an abstract idea, is it 
not evident that these were already known to the 
primitive group ? Among all the roots which might 
characterise them, it had already chosen those which 
appeared to be the most expressive. In order to 
ascertain what some have kept or lost, what others 
have acquired, it is sufficient to gather together the 
terms which are common to all or to some of these 
idioms. This is what Adolphe has attempted to do, 
sometimes rashly, bat generally with success. 

Centuries of pastoral life had preceded the separa- 
tion of the idioms. The names of such things as 
flock, ox, sheep, pig, dog, shepherd, pasturage ; of the 
yard, the stable ; of meat, wool, milk, butter, cheese, 
present marked agreement; and from these terms are 
derived the words which imply wealth, property, the 
family, the master, the host. The bull and the cow 
are the principal actors in the myths, the prize in the 
battles between the sun and the clouds, the thunder 
and the winds. Though a shepherd people, the 



The Indo-Europeans. 255 

Aryans were no longer nomads ; they knew barley, 
the art of ploughing, mills, flour, and perhaps bread. 
They drank fermented liquors, hydromel, and perhaps 
wine. The fabrication of the cart, the axle, and of 
the yoke necessitated the use of the chisel, the wedge, 
the axe, and the knife ; there were carpenters and 
smiths. Tiie anvil was of stone. There is no evidence 
of the use of iron, but there is no doubt that silver 
was known and bronze in use. Spinning and weav- 
ing of a rude kind existed ; the dress was sewn ; 
necklaces and bracelets adorned the neck and wrists. 

The art of building, or rather of hollowing boats, 
was known ; the names of the hull and of the oar are 
as old as our languages, but there is no question of 
mast, of sail, or of keel. The sea was distant or un- 
known, and the navigation was only upon lakes or 
rivers ; boats were required for crossing the rivers, as 
bridges were not invented. 

The house, which, by its principal name (dama, 
domus), recalls perhaps the bundle of poles which 
carried the primitive tent, was the chief work of the 
carpenter. It is doubtful whether masonry had any- 
thing to do with the construction, unless it contributed 
a little plaster or mortar. The bed, the seat, a few 
utensils constituted its furniture. The house was 
surrounded by a ditch, which seems to have furnished 
most of the words which signify enclosure, yard, or 
garden. Within this boundary was the well or the 
cistern, and the hearth where the food was cooked. 
It may also be conjectured that fire was also lighted 
in the hut, since some of the words which signify 
house seem to be connected with a root which means 
to burn. The door has kept its name throughout the 
ages, dvar, door, Oupa, fores, but the key only appears 



256 Distribution of Languages and Races. 

among the Westerns. The houses belonged to families, 
to households, monogamous, at least in theory, in 
which the husband and wife were equals and master 
and mistress {jpati, iroa-is, patui, ttotvio) ; a tribe, a 
clan (djariana, yevos, gens, chunni) must therefore have 
occupied a certain number of huts, which formed a 
village (trapd, tribus, thaurp, dorf), or even a town 
(poura, ttoXk). There was no nation, no people ; the 
words which have since expressed this idea signified 
then number and multitude only. Everywhere there 
were small groups, sometimes united for their common 
defence, often separated by their quarrels among each 
other. War, for these barbarians, was already' the 
action, par eiccellence (adji, adjma, agon, agmen, the 
combat, the army), whether it were a battle in the 
open country, on horse with the javelin, on foot with 
the sword, or the assault of the enclosure, the burg, 
where the enemy had withdrawn with his troops and 
his booty. The hero {vira, vir, baro) fought upright 
on his chariot or mounted on the horse, which he 
excited with the sound of the horn {grina, cornu), the 
rude precursors of the trumpet. Owner of horses, 
tamer of horses, friend of horses, were the most coveted 
titles among the Persians, Greeks, and Gauls. 

The conquered enemy was carried into slavery, and 
was reckoned, with the flocks, in the wealth of his 
master. The powerful man and his wife were styled 
indiiferently dampati, gopati, dasapati (Secnrortii, Sear- 
TTOtva), masters of the house, of oxen, of slaves. 

The love of war, without which no robust and lonar- 
lived race has ever arisen, implies the love of glory, 
the true motive of every courageous action and of 
every great work. No human group has ever felt it 
more strongly than ours. To be known, to be sung, 



The Indo-Europeans. 257 

the desire is innate in the Hindu as in the Greek, in 
the Persian as in the German. There is a root hru, 
Idu, ql%, which in a thousand diiTerent modifications 
has furnished the names of peoples, of heroes, even of 
gods ; the Slavs are the glorious ; all the qrwoas, all 
the slaf, all the KKri% the chl%, the Mod, and the Jirua, 
Ladislas, Herakles, Cloms, Louis, and Roland, were 
famous men, or aspired to justify the paternal pride 
which had bestowed upon them these high-sounding 
names. 

The tribes had chiefs of war and peace, kings, 
except perhaps the Hellenes, who made hardly any 
use of the root rag or reg, common to all the other 
sister languages. Their social organisation was founded 
upon property, common and individual. Inheritance 
was known, but was doubtless confined to the rank, 
the house, the product of toil, and the booty taken 
in war. Exchange was the only form of econo- 
mical relations. The oxen served for money. Law, 
right, debt, crime, judgment, evidence, and fine were 
named before the Indo-European expansion.- Most 
of the roots which express these notions allow us to 
discover the entirely material origin of the highest and 
most abstract ideas ; this may be said of all the terms 
which relate to the life of the mind ; the soul {animd), 
simply the breath of life ; thought, merely the power 
of measuring and weighing objects ; will, memory, 
knowledge, the power of creating things a second time 
(roots gan, to engender, gnd, to know). Religion is 
not the least striking characteristic of our ancestors ; 
freed from the minutiae of animism and from the 
enervating practices of chthonism (of which, however, 
some traces are retained), it had already attained to 
the adoration of the forces of nature and of atmos- 



258 Distribution of Languages and Races. 

pherical phenomena. Lastly, the language, the result 
of an extreme degree of agglutination) simple in its 
roots, indefinite in its power of derivation, presents 
itself as a complete organism, cultivated and at the 
same time free. 

What name should be given to this aggregate of 
the immediate ancestors, of so many different races, 
which are nevertheless endowed with the same polity, 
the same language, the same creed, and the same 
culture ? There is one which, in spite of opposition, 
has finally prevailed, since it no longer misleads any 
one, no longer implies unity of race ; it is that of 
Darius, " Arya, son of Arya," that which the Brahmans 
claim, and which they have translated to the heavens 
in the person of the god Ahriman. It is true that 
the other peoples, except the Iron or Ossetes of the 
Caucasus, have not adopted it, but they knew it and 
used it. What are Ares, Avion, Aristos, Arete, Arte- 
mis, the strong, the best, virtue, the most honourable 
of the Greeks ; what the Ario-vist of the Germans, 
but doublets and derivatives of the name borne by so 
many great persons among the Hindus and Persians. 
The original meaning is vague ; among the true Aryas 
it is noble, famous ; but it also means traveller and 
labourer. 

This last acceptation is universal. Lat. arare, ara- 
trum, aratio ; Gr. apouv, apoTpov, ApotrK ; Litli. 
ar-ti, arhlas ; Slav, ora-ti, oradlo ; Goth, arj-an ; 
Anglo-Sax. erjan ; Eng. ear; Irish, ar ; Cornish, ara- 
dar ; Welsh, ardd. The ancient names of the earth 
are connected with it : Gr. epa ; Sans, ira ; Ger. 
ero, airtha, eortha, earth ; Gael, ire, irionn. How- 
ever this may be, Aryo-European would be better 
than Indo-European, and, for the sake of simplicity, I 



The Indo-Europeans. 259 

shall use without scruple the words Aryans, Aryan 
languages and nations. 

I must meet in advance a possible reproach : for 
Bossuet's Semitic history you would substitute an 
Aryan history. You only dethrone the chosen people 
in order to put forward another privileged, predestined 
group. Yes and no. I suppress no fact of history. 
I attribute nothing more to the Aryans than the 
science of language allows them ; they began, like all 
others, in savagery ; but since they came late in time, 
and with an already developed language and intelli- 
gence, they rapidly reaped the benefit of the inventions 
of their predecessors. The Egyptians, the Chaldeans, 
the Semites in Western Asia, the Chinese in the 
extreme East had reached a level which the Aryans 
have since passed in their institutions, in the arts, and 
in the expression of thought. For two or three thou- 
sand years the direction of the world has fallen to the 
Aryans, and, in spite of momentary failures, of Arabic, 
Mongolian, or Turkish incursions, they have kept the 
torch, they have carried it into America, into Aus- 
tralia, and returning to their cradle by sea and by 
land, they bear the light even into the heart of Africa, 
even into the dim twilight of the East. 

Compare the false and incoherent history to which 
Bossuet lent the support of his eloquence — history 
modified to suit the Jewish Bible as it was revised in 
the fifth century, and the prophecies after the event 
of Daniel and of John ; compare it with the realities 
unveiled by the discovery of the Indo-European group. 
Note how the movements of the peoples are ordered 
and illuminated by it. While from the eastern slope 
of the great Asiatic plateaus the ancestors of the 
Chinese descended their rivers, the Blue and the Yellow, 



26o Distribution of Languages and Races. 

and multiplied in tkeir immense empire, isolated, use- 
less, and unknown, two centres of civilisation arose, 
on the banks of the Nile and at the mouth of the 
Euphrates. Separated from these Egypto- Semites by 
the Himalaya and the desert, slowly increasing tribes 
of white men, part shepherds and part agriculturists, 
monogamous, worshippers of the heavenly bodies, gra- 
dually, under the pressure of the Mongols, leave their 
common country, forgetting each other as they travel, 
but retaining their idioms and their acquired culture 
exactly in the proportion of their increasing distance. 
The Celts are driven westward by the Gauls, the G-auls 
by the Germans, these by the Slavs and Lithuanians, 
themselves urged forward and finally overrun by the 
Mongols and the incursion of the Huns. The future 
Hindus are already making their way among the 
affluents of the Indus. Lastly, the Greeks and Latins, 
passing south of the Celts, Germans, and Slavs, and 
north of the Semitic world, follow the right bank of 
the Danube, and one stream of them flows towards 
Thrace and Thessaly, the other towards the Tiber. 
The Iranians alone remain, harassed by the continual 
attacks of the Turks ; they reach Media, Persia, con- 
quer and take the place of the old Semitic empires, 
and come into collision in Ionia and at Marathon with 
their old neighbours, now forgotten, with the Hellenes, 
already masters of the Mediterranean basin. 

This large and simple view gives the true meaning 
of history. It explains the successive effacements of 
the ancient civilisations, the encounters and the strifes 
of the Gauls and the Italiots, of the Hellenes and the 
Persians, of the Germans and the Grseco-Roman world, 
the Mongolian incursions into the field left clear by 
the Aryan migrations, and the equilibrium slowly 



The Indo-Europeans. 26 1 

established by the mutual resistance of the various 
races, occasionally disturbed by these passing irrup- 
tions. It explains also the movements of the Germans, 
checked by the Celtic block, and returning against 
the Slavs, who, long the victims of shock and counter- 
shock, fluctuated, without a fixed frontier, between 
Germany and the Tartar chaos. The various German 
invasions declare themselves as the consequences of 
the primeval impulse. Even the conquest of the 
Americas and of Oceania may be said to proceed from 
the impulse communicated by the pressure of the 
Mongols, four thoasand years ago, to the tribes who 
dwelt between Turkhestan and the Oxus. 

Such is the new conception of history, which rejects 
as a chimera the divine plan and the biblical genea- 
logies ; it is the creation of philology. 



18 



PART III. 

THE INDO-EUROPEAN ORaANISM. 

CHAPTER I. 

INDO-EUROPEAN ROOTS. 

Inflexion a higher degree of agglutination — The Indo - European 
material consists of full roots and empty roots, demonstrative or 
pronominal, and attributive or verbal — Pronominal roots : pronouns 
and suffixes — Attributive roots, primary, secondary, and tertiary — ■ 
Reduction of the variants to a small number of ancestral forms 
— Roots expressing an action of the mind : the ma family — The 
naked root, the tlieme or radical, often preserved by the com- 
position of words. 

The kinship of the Indo-European languages is a 
phenomenon of the same order as the close affinity 
of the Bantu, Berber, Turkish, or Semitic dialects. 
Their area is not more extensive than that of the 
Malay idioms. The ethnic differences between the 
nations which have acquired them are not greater 
than between the various Malayo-Polynesian groups. 
The pre-existence of a common speech and of a 
human aggregate where this mother-tongue was 
formed, is not less evident and less necessary here 
than in the case of the other independent families. 
All the earlier types have perished, because writing 
was unknown ; but they may yet be traced in the 
idioms derived from them. Indeed, it is not only what 

may be called the first types that have disappeared ; 

262 



Indo-European Roots. 263 

even secondary forms have often passed away without 
leaving any monument of their existence ; nothing is 
left of the Teutonic, Slav, and Italic stems, which 
have produced so many different branches. Fortu- 
nately we possess in their earliest as well as in their 
latest forms Sanscrit, Iranian, Greek, and Latin ; and 
it is enough to compare with these the Hindu, Persian, 
Romaic, and Romance groups, which have issued from 
them, to understand the formation of the eight Indo- 
European branches, and to determine that they are all 
issued from a common trunk. 

Hence there is no mystery about the origin of these 
languages, save that obscurity which belongs to vast 
distances in time. The peoples which have carried 
along with them, or received in the course of their 
wanderings, the elements of the same intellectual 
cultivation, have, more or less rapidly, risen above the 
level of other races and nations. This fact is histori- 
cal and patent ; it need not surprise us, since we find 
everywhere these inequalities in aptitude and destiny. 
Is not the Malay superior to the Papuan, the Moor 
to the Negro, the Aztec to the Abipone, the Semite 
to the Berber, the Chinese to the Mongol ? Latest in 
time, the Aryan is heir to the conquests of his pre- 
decessors, and their apogee was his point of departure. 
Egypt, Chaldea, Assyria had given their measure and 
spent their force ; with livelier energy, with greater 
breadth and subtlety of mind, he took in hand the 
abandoned task ; and this intellectual and aesthetic 
superiority is manifest to us in his speech. 

But his language is not separated from others by 
any gulf ; it only differs from them by a more inti- 
mate combination of the same original elements ; it 
has outstripped them, but on the same road. It is 



264 The Indo-European Organism. 

not possible to donbt this, since it is the comparative 
study of these great dialects and the analysis of its 
forms which have suggested the theory of language, 
and enabled philologists to determine the general 
development of speech, the successive phases of mono- 
syllabism, agglutination, and inflexion. 

Inflexion, as we have said, is but the fusion of the 
agglutinated syllables; it necessarily supposes, there- 
fore, an agglutinative period, but agglutination could 
not have taken place without a supply of roots 
which were already susceptible of juxtaposition, as in 
Chinese, that is already classed as fvll and empty, or 
principal and subordinate roots. Finally, this rudi- 
mentary use of the vocal elements bears witness to 
ages yet more remote, in which the sounds adapted 
to the designation of things and beings, relegated to 
the second place, into the class of demonstratives and 
auxiliaries, the ancient cries of pain, joy, summons 
and warning. Even the long trials and experiments 
which resulted at last in the true consonants are 
revealed to us by the numerous variants of a single 
root, either in the same language or in the passage 
from one dialect to another. Before attaining to the 
clear pronunciation of the simple guttural, dental, or 
labial, before uttering the sounds h and g, t and d, 
2? and 6, the ancestors of the Indo-European long hesi- 
tated among consonantal diphthongs, such as sk, kch, 
lev, M, kh, gv, gj, gd, and gh, such as tch, tj, tv, dj, dv, 
or pt,pv, mp, id ; so with the sibilant s, so diiBcult for 
the Greeks and Iranians to pronounce at the beginning 
of a word, and for the Latins between two vowels, that 
these peoples replaced it either by an aspirate or by 
an r ; so with the liquids r and I or n, which are con- 
stantly interchanged or replaced by a dental. The 



Indo-European Roots. 265 

acquisition of the vowels was the final result of similar 
hesitations and efforts ; thus the short e and were not 
familiar' to the Aryans of India or to the Goths, and 
these peoples only know the long e and as combina- 
tions of an a, short or long, with i and -y (vowel or 
semi-vowel). Then, according as a vowel was uttered 
with more or less force, followed or preceded by an 
aspirate, it annexed by degrees a guttural, a sibilant, 
or even a trill like the Sanscrit vowel r. Finally, if 
the soft palate rose at all towards the nasal passages, 
the vowel took a peculiar sound — it was nasalised, as 
an, in, on ; the insertion of an n was the inevitable 
result of this phonetic accident, and it is often diffi- 
cult to determine whether the added n is due to the 
nasalised vowel or to a suffix na. 

These inductions, which M. Paul Regnaud has in- 
terpreted I think somewhat too absolutely, are not 
merely logical processes — they repose on ascertained 
facts. These variants of which I spoke are not hypo- 
thetical ; they subsist and co-exist, forming groups in 
which the intimate relationship of the members is 
evident. The Indo-Europeans preserved them, and 
came by degrees to use them to express shades of 
meaning, and the new acquisitions which enriched 
thought as time went on. Sanscrit grammarians 
reckoned in their language about seventeen hundred 
roots, apparently irreducible by analysis ; the first com- 
parative study brought down to the number of five 
hundred the elements common to all the languages of 
the family. A more thorough analysis reduces them to 
one-fifth, perhaps one-tenth of this number, the narrow 
base on which reposes the most vast and the most fertile 
of linguistic organisms. 

From these preliminary but not a priori considera- 



266 The Indo-European Orgajiism. 

tions let us retain the following points : The Indo- 
European mother-tongue is not a concept of the mind, 
nor a species of miracle ; it is a reality, the- product 
of the elaboration of ages, as is proved by the wasting, 
the atrophy of its case and verbal endings. Its genesis 
and material are those of other languages, which it has 
distanced only by a more vigorous and intelligent use 
of the same methods and artifices. 

Two classes of monosyllabic roots compose it — de- 
monstrative or pronominal^ roots, attributive or verbal 
roots ; the first furnished all the pronouns, most of the 
prepositions, conjunctions, and suflSxes ; the second, 
all nouns and adjectives, all _ verbs, and most of the 
■ adverbs. 

The pronominal roots often occur in the naked 
isolated state, the primitive monosyllable : sa, ho, ta, 
to, this, that ; sya, tya, sva, snia ; ma, me, tu, te, I, 
thou ; ya, ka, ku, who, which ; dva, tri, two, three ; 
da, ga, as in the Greek and German particles ge, de ; 
often they have acquired case-endings, s, m, t, hliyam, 
hhyas, which are easily detached ; often again they are 
agglutinated and coagulated together, without chang- 
ing the indicative relation or personal meaning : for 
instance, in Sanscrit, ima, esa, ata, eta, ana, ena, eva, 
eka, into which enter the simplest sounds that man can 
utter, a, t, and which are naturally reinforced, doubled, 
with other syllables, vaguely associated with some ges- 
ture, with some idea of distance and of place ; under 
this form the pi'ominent roots have given birth in all 
the Indo-European idioms to many series of indispen- 
sable words, varied and changed to an indefinite ex- 
tent, in which it is difiicult to trace the amalgamated 
syllables. Who would think to find Jwc illud in the 
French word oui, or hie hie in iei, or hie ille hwic hie 



Indo-European Roots. 267 

hie in eelui-ci, if these fusions were not revealed to us 
by forms such as o-'il and i-ce-lui? Bopp and his 
followers, Pott, Benfey, Kuhn, guided by similar in- 
dications, have been able to find and follow up with 
certainty the clue to these labyrinths. 

The pronoun / is present in Gothic and Latin in 
its naked simplicity, except that it has a declension. 
Gothic nominative and genitive singular is, accusative, 
singular and plural ina, ins, dative singular and 
plural imma, im. Latin is, id, ii, its, and the attenua- 
tions ea, ei, ejus, eum, &c., with suffixes isdem, idem, 
ita, itidem, i{s)pse, ihi, immo, enim ; there is, moreover, 
a permanent confusion with the form ya, ja, derived 
itself from i. From Latin it is permissible to pass to 
French to show the ancient pronoun / in a group where 
it is not easy to perceive it : the word meme, through 
the old form meisme, and the Italian medesimo, leads 
us to a supposed Latin form, met-ipse-timum, super- 
lative of metipse. Now in these compound forms the 
central significant part is the vanished pronoun /. / 
has not left in Zend or Sanscrit an isolated form, de- 
clined, but it appears in adjectives like idriga, such ; 
itara, other (Latin iterum), and the adverbs iha, idha, 
ithra, here ; itas, from here ; it-iham, iti ; Zend, itha, 
thus ; idanim, now ; tchet, for tcha-it, with the sense 
of yes, in contravention of a negative proposition (the 
French si) ; net (Zend, noid) for na-it, not, no-indeed, 
as one who should say, " Away with that." 

We have just mentioned a syllable famous above all 
others, since it has almost everywhere furnished the 
most energetic word in any language : English no, 
French non. It may cause surprise, perhaps, to learn 
that the root na and its neighbour ma, which have 
often taken a negative tense, have retained in most 



268 The Indo-European Organism. 

cases, either as isolated words or as suffixes, a distinctly- 
positive demonstrative value. Affirmation and nega- 
tion would, therefore, seem not to be primitive ideas, 
or else they were sufficiently rendered by gesture. It 
is only after a long lapse of time, and by a sort of 
detour, that simple or compound pronouns have been 
used for the expression of consent or refusal. It is, 
moreover, easy to show that in modern speech our 
affirmative and negative terms are merely indicative 
adverbs or pronouns : si in Italian is but sic, so ; the 
French oui, that that ; German and Breton ja and 
English yes have no other meaning than the Aryan, 
Greek, and Latin expressions, such as ita, na, ma, md, 
nee, nai, &c. No doubt we see in Greek, Sanscrit, 
in Armenian, under the forms me, ma, mi, in Sanscrit, 
Gothic, Slav, Borusse, Lithuanian, Zend, under the 
forms na, ni. Tie, naiy, our two pronominal syllables 
employed for inhibition and negation ; but we must 
not forget that fia tov Qeov in Greek means " by God, 
I call God to witness ;" that nai means " certainly ;" 
"that in the Veda we find na with the sense of sicut, 
as, so, and nana with the sense of much ; finally, that 
mxi and no, form nouns, adjectives, participles, super- 
latives of every kind. It is, therefore, only by custom, 
by a choice easier to verify than to explain, that ma 
and na, isolated or prefixed, have become negative 
signs. Yet another striking example, non, an abbre- 
viation of the forms Tienu, ne-unum, has the root na 
twice over, negative at the beginning of the word, 
simply demonstrative at the end of it. 

These curious anomalies throw light upon the pri- 
mitive equality of all those cries of summons or warning 
to which gesture, accent, intonation, in the first place, 
and afterwards custom and Ihe preferences of a tribe, 



Indo-European Roots. 269 

gave at last a precise value, a definite meaning. And 
here we are only dealing with independent words, 
composed of naked or declined demonstrative roots, 
and with compound words in which this wealth of 
roots takes the place of radical or central sj-llable. 
But wlien we come later to these same roots com- 
pounded with attributive roots, their original insigni- 
ficance will be yet more clear. There is no single 
vowel, there is no syllable formed of a consonant and 
a vowel, which, suffixed to a verbal root, cannot form 
part in the same way of a noun, an adjective, a 
participle, a verb, or an adverb. It is the need of 
clearness, the instinctive choice, variable according to 
vocal aptitude, according to intercourse with foreign 
peoples, which have more or less fixed upon a certain 
class of affixes {ta, ti, to, da, ma, na, ra, la, sa, pa, va, 
ka, ja, sja, &c.), the sense of agent, action, future, past. 
And in most cases to these sufiBxes, called primary, 
worn, altered, unrecognisable, it has been necessary to 
add one, two, or three others, neither more or less 
significant in themselves, to supply the exigencies of 
speech. Reckon in a word used by Lucretius, in-saf-i- 
a-bi-U-ter (insatiably), the syllables added to the root 
sat (found in the French saoul, from saturus, in rassasier, 
and in assez, from the Latin ad-satis). Leaving out 
the prefix in, which answers to the Greek and Sanscrit 
an and the Teuton un, and which has taken a negative 
sense, i, d, bi, It, and ter, for tara, are pronominal roots, 
atrophied and fused, and having at most a value of place; 
for we shall find them all, singly or in pairs — and that 
in all the idioms of the family — capable of forming 
nouns and verbs of every kind. But before giving 
other examples, we must define and class the attribu- 
tive roots, those which we tried in an earlier chapter to 



270 The Indo-European Organism. 

connect either with onomatopoeia or with a metaphori- 
cal imitation of a sensation, a movement, an object. 

An Indo-European attributive root is neither noun 
nor verb ; it becomes the one or the other by tlie 
addition of verbal or case endings. It is, like the 
Chinese monosyllable, an utterance capable of specify- 
ing either a class of things, beings, or phenomena, or 
a state, an action of the thinking subject or of the 
object under observation. It may include all the 
vowels and consonants which can be contained in one 
syllable. " It can always be shown," says Max Mliller, 
" that the roots composed of more than one syllable 
are themselves derivatives." And even in the true 
monosyllabic roots we must distinguish between primary, 
secondary, and tertiary roots. 

Primary roots are composed : ( I ) of a vowel, i, to 
go ; (2) of a vowel and a consonant, ad, to eat ; as, to 
breathe; ag, to lead; alt, to run, to pierce; (3) of a 
consonant and a vowel, or of a semi-vowel and a con- 
sonant, da, to give ; pa, to drink, to pasture, to protect ; 
ma, to create, to measure ; 6Am., to be, to grow ; ^■a, to 
blow ; hr, to do ; itir, to die ; jii,, to join. 

Secondary roots are composed: (i) of a consonant, 
a vowel, and a consonant, tud, tup, to strike ; hhar, to 
bear ; ruk, to shine ; vak, to speak ; sak, to follow ; 
yug, to join and to fight ; vid, to see, to know ; dvJc, 
to milk ; gan, to engender ; man, to tliink ; (2) of two 
consonants and a long vowel, hlira, to carry ; mTvd, to 
remember ; gna, to know ; plu, to flow ; klu, to hear ; 
stha, to stand ; (3) of a vowel and two consonants, 
ard, to wound. 

Tertiary roots are composed : (i) of two consonants, 
a vowel, and a consonant, spak, to look at ; tras, to 
tremble ; grabh, to seize, to hollow, to engrave ; star, 



Indo-European Roots. 271 

to stretcli ; livan, to make a noise, or vice-versa ; vart, 
to turn; cand, to shine; sarp, to glide; (2) of two 
consonants, a vowel, and two consonants, spand, to 
tremble ; shand, to ascend ; scalp, to hollow, to write, to 
carve. Initial groups of three consonants are also found. 

It is clear that the two first classes are the most 
ancient. In the second we already see traces of suf- 
fixation ; if we compare with yu, to join, the forms 
ytuf, to join (Lat. jugum, Gr. "Qjjov), yung and yu-na-j 
(Ij&\i. jung-ere), and yudh, to fight, you will perceive other 
suffixes, ga, dha, 71a, or a nasalisation of the secondary 
root. If to tud, to strike (Lat. con-tud-tus, contiisus), 
and to Utp (same sense, Gr. TUTrrai), we add ticnd (Lat. 
tund-ere), tump (Sans, tump-ati, he strikes), hibh and 
toih (Sans, tuhh-nati, tobh-ate) ; tuj, to strike, to ex- 
cite ; tur, to wound ; tuh, to afflict ; turv, to conquer ; 
we shall be led to connect all these variants with a 
primitive tu. So with ruk, ruksh, rut, rud, rub, ruth, 
clearness, colour. As for the tertiary roots, they result 
for the most part from very ancient agglutinations. 

Ghav^e, I think, was the first to propose the group- 
ing together of the roots which are related to each other; 
thus he constituted the families to blow, to make a 
noise, to shine, to burn, to strike, to measure, &c. 
The idea is valuable, and has been more or less accepted 
by all philologists. M. Paul Regnaud has followed it up 
with great ability, with the awowed intention of tracing 
back all roots to a single vague cry, from which ail 
are issued. This appears to me to force the theory 
too much, like Darwin's tendency to trace the whole 
living world back to a single cell of protoplasm. We 
have here a touch of monogenistic atavism. The 
theory of evolution accords perfectly with a great 
number of primary germs and a cerLain original variety 



272 The Indo-European Organism. 

in them. "We have recognised differences and variations 
in the specific cry of animals ; the voice of man had 
even more. And in its earliest state language must 
have had at its disposal a variety of sounds which it 
slowly developed and defined into vowels, aspirants, 
and finally into consonants. Indo-European is far 
from being a primitive language ; and however far we 
trace back the efforts which led it from agglutination 
to inflexion, it would seem to have possessed even 
then at least three pure vowels and several varieties 
of the consonant. 

However this may be, the attempt of M. Eegnaud 
is full of interest, from the boldness of his comparisons, 
which are always supported by proof or by scientific 
hypotheses, and we will borrow from him a few ex- 
amples. 

There is in Sanscrit a root harsh, which it is easy to 
recognise in the Latin hirs-utus, horreo (Jorhors-eo), her 
(Fr. herissorij hedgehog). The same root is found in 
Greek as )(op6s, -^aipia (for -^^apaw, ■^(appw), perhaps 
'Xpi-poi, pig ; but if this last word has kept something 
of the sense of bristle, " to be stiff," the two others, 
signifying dance, song, and to rejoice,have no connection 
with the usual meaning of the root Yet we must 
remark that in the Yeda, harsh has precisely and solely 
the sense of the Greek ■^a'lpo), to rejoice, and that it 
is difficult to separate from it the harifas, the trium- 
phant horses of the dawn, of which the Greeks have 
made their Graces or ■)(apiT>]i. How shall we reconcile 
these discrepancies ? First, is harsh the true form of 
the Indo-European root ? The jj of x/^ipw indicates 
a lost guttural, whence a more ancient variant, gharsh, 
which exists in Sanscrit itself in the adjectives gKrshii, 
gkrshvi, ardent, active, joyouB. The sibilant which 



Indo-European Roots. 273 

terminates the root may be the remnant of a suffix, 
and this is the more probable that a well-known group, 
ghar, char, kar, skar, with the general sense of heat 
and brilliancy, presides over a whole series of kindred 
terms. M. Regnaud thinks that ghar, gharsh, harsh, 
meant before everything to burn and to shine, and 
answered to the two most important senses, touch and 
sight. This opinion is the more probable that more 
than thirty groups of roots have retained more or less 
of these two primitive meanings. With to hum and 
to shine are closely associated to shake, to move, to spring 
forth, to tremble, to radiate, to vibrate, and to bristle. 
To bwn has its proper successors, such as to dry, to 
harden ; to shine has also its own, such as to rejoice 
and to be joyful. 

Among the numerous roots which belong to the 
category to burn, to shine, to shake, to harden, there are 
two which approach the first named very closely ; they 
only differ from it by the initial letter. Tarsh and 
tras, tar and star, have given to Zend taresh, to be hot, 
dry, or thirsty ; to the Greek, Oipcros, heat ; Tepcrofiai, 
to dry, to harden ; rapdcraw, to agitate ; rpew, to 
tremble ; to the Latin, torreo, to burn, and torrens, 
to be agitated ; probably also stella for star-u-la, the 
shining thing, the star; terreo, to cause to tremble, 
tremo, to shudder; teira for tersa, the dry land, the 
earth ; to the Teutonic, starr, stiff, hard, bristling. 

Bhresh, bhreksh, Ihraj, bhrajj, are rich in specimens of 
the three or four meanings attached to the preceding 
groups. Moreover, they evoke from heat, from flame 
even, through the intermediary of dryness and hard- 
ness, the name of shivering and cold. The Latin 
torreo = terreo leaves besides no doubt about the transi- 
tions which connect these two opposite ideas. 



2 74 T^^ Indo-European Organism. 

In the matter of heat and brilliancy we have the 
Greek (pXeyw, to burn {Phlegethon), (p\oys, flame ; the 
Latin flagro, I burn, flagma ( flamina), flagrum, and 
flagellum, (the whip or scourge, which seems to burn 
the skin), fulgor, fulgur, splendour, lightning. Agita- 
tion is expressed by the German spriessen, spriTigen, 
and by the Greek (pp'ia-a-ta ; hardness by this same 
word (ppla-ardd, and by piyeia ; by the Latin rigor, 
rigidus ; finally, the French word froid is but the 
Latin frigidum, frigus, frigere, of which rigen is a 
doublet. 

All these facts are ascertained, and if any doubt 
remain, it is because we have not yet been able 
to consider Indo-European phonetic laws. But M. 
Regnaud, considering the primitive unity (or rather 
indecision) of the three consonants, guttural, dental, 
and labial, whether aspirated, hard, or soft, proposes 
to assimilate, to reduce to a common type, the roots 
gharsh, marsh, and bharsh, with all their variants and 
transposed forms. He thinks also (judging from 
numerous and plausible indications) that a sibilant 
once preceded the initial consonant. In fact, a number 
of roots similar in form and meaning differ only by 
the presence or absence of an s. We should thus 
need to write above these three series a triple key, 
skar, star, sbar,- susceptible of a great number of 
attenuated or strengthened forms, and easily augmented 
by one or two atrophied pronominal suffixes, tch, dj, sh, 
kch, k, g. 

The three roots mentioned above are by no means 
the only ones which our learned friend desires to trace 
back to a common source ; and whether he succeeds 
or not, he deserves our gratitude for having in most 
cases led us back to the beginning, to the dawn of 



Indo-Eiiropean Roots. 275 

intelligence. He has contributed in great measure 
to the destruction of the old psychology, to the 
demonstration of the slow and unequal growth of 
ideas, of the progressive but unconscious ingenuity, 
whereby our ancestors attained to the endowing of 
barren and naked syllables, not only with abstract 
meaning, but with moral concepts. 

In truth, they are not numerous, the roots which 
man used in the earliest ages to express an action 
of the mind : three or four sufficed. And it is easy 
to trace them . back to a physical sense ; smar, to re- 
member, connected doubtless with mar, to die ; hudh, in 
Greek irvQ, ttvvO (doubtless to shine, to lighten) ; grad, 
to believe (from Jcru, to hear) ; gna, to know, to name, 
a contracted form of gan-a, to engender, to produce ; 
finally ma, which we have already mentioned, and 
which, either simple, short, or long, or with, added con- 
sonants, has representatives in all the parts of speech. 
We have seen to what uses it has been turned in the 
pronominal order, now personal, and now demonstra- 
tive pronoun, now affirmative or negative particle, now 
case or verbal ending. In the attributive order ma 
plays a yet more important part. 

We do not doubt that it was one of the earliest 
distinct sounds pronounced by the child, and retained 
by men and women, in the sense of creating, en- 
gendering, producing, and afterwards commanding, 
weighing, thinking. The word ma-tar has been 
adopted by all the Indo-Europeans to designate the 
mother ; but it has also been used in a masculine 
sense, matar, creator, whence the name of the Vedic 
god, Matarigvan. In this connection one may mention. 
mas, the male, the producer. To this general sense be- 
longs Lat. ma-nare, to flow, to emanate; materies, matter, 



276 The Indo-European Organism. 

the fertile substance, space ; and also, striking contrast, 
the famous m&yA, space, phenomenon, the illusion of 
the universe. The moon which traverses space, and 
whose phases are the measure of time, is mas, and its 
career is ma-sa, the months. The nasal sufSx adopted 
by the Greek ixypri, n-qv, by the Latin mensis, by the 
Gothic and Slav mena, menoth, by the Anglo-Saxon 
mona (moon), manadh, has not changed the value of 
these two correlative terms. 

A group of the family ma has remained in the 
direct and limited sense of measure. There are 
analogous terms in all our languages ; metiri, to 
measure, fierpov (whence metre), mensura, medius 
(whence nioyen, moitiS, mean), modus, modius (muid), 
magh, to grow, whence maha, fj-eya?, magnus, majestas. 

To pass now to the forms under which ma ex- 
pressed, before the separation of the idioms, actions 
of the human mind. It appears to us here in at least 
five different states — short, as it was originally ; long, 
with the nasal or the suffix na ; with the soft, hard, 
or aspirated dental ; with the causative suffix ya {ma, 
ma ; wMn and mn& ; mad, math, manth ; manya) ; 
and its meanings vary between thinking, knowing, 
remembering, hoping, desiring, loving. 

The naked form has been the least fertile : it has 
furnished in Sanscrit ma-tas and ma-ti, thougrht. 
Greek has retained it in the name of the goddess 
Metis, in the adjective ixtiTiera, an attribute of Zeus, 
and in the compounds PoXu^jjtj?, AvkuXo/hi^tij^, the 
ordinary surnames of Ulysses and Prometheus. 

Two plausible reasons have been suggested for 
the passage of ma into man : first, the very general 
tendency, even in Sanscrit, to nasalise the sound a; 
and secondly, the frequency of the suflSxes- na, na, ni. 



Indc-European Roots. 277 

mm,, no, which characterise several classes of verbs. 
The addition of the dentals and of ja may be ex- 
plained in the same manner. The root man, once 
constituted, it acquired, according to a constant law, 
a contracted form mna, a reduplicated form ma/man, 
mimfie, and a causative manya ; and entering into the 
frequentative mat, became prolonged into mant and 
manth. The variation of the vowels in the different 
dialects came to increase the richness of the primitive 
sound ; and we may expect to find everywhere else 
than in Sanscrit equivalent forms : men, mon, min, 
miin, main, meth, med, ment, mna, mne, &c. 

Sanscrit offers several terms of great interest ; first, 
the verb to think ; man and mana, manute (lie thinks), 
then manas, thought, mind. The manas has been 
reckoned by the Hindus among the senses ; it is the 
sixth and greatest. In the most recent of the Vedic 
hymns it is deified. Another name has had a yet 
finer destiny, manu, man, the thinking being, the 
legislator, deified also among the Germans of Tacitus 
and during the Brahmanic period. With the causative 
particle ja, man forms words which signify a move- 
ment of the mind ; for example, anger, in Sans, man- 
y-u ; whence in Zend, Angro-Mainyu, the spirit of 
anguish, of pain, the Evil Genius, Ahriman. We 
may also mention man-tra, the instrument of mind, 
which has come to mean sacred text, verse, litany, in- 
cantation, talisman. To the root mna a liturgical 
meaning has also attached itself; to repeat the Vedas 
to oneself. The variety math, munth, has produced a 
famous name, Pramantha, he who turns the fire-stick 
for the sacrifice and produces the spark. Here we 
recognise Prometheus, another form of the same type. 

The same elements have been fertile in Greek. The 
19 



278 The Indo-European Organism. 

name of man, manu, has been abandoned, but ft-evoi, 
thought, has formed numerous derivatives; such as 
wfi.evrj<s (Sans, vasumana, Zend vohumano) and ^ya-yuev^? 
(Sans, durmanas). Manju is directly connected with 
mania, fury, mainomai, to rave. With man and mew, 
fjiivce, fievedivo), to wait, to hope; Mentor, the sage Aga- 
memnon, for Aga-me-men-on, a name famous in the Iliad, 
and an epithet of Zeus, " the wise ; " Men-e-laos, desire 
of the people, or guardian of the people ; fidvTK, the 
diviner, the poet, of which the feminine fiavna or 
jxovria, has been regularly altered into ixoiaa, fuwaa, 
fiova-a. Hence the Muses and Mnemosyne, or memory. 

The sub-root mna, mne, has given fivijfi.^, fi-v^fiov, 
/nefxvtifjLOU, I think, and fxifivija-KW, I remember. With 
the addition of s, the remains of a suffix implying 
desire, it has produced a curious group, quite distinct, 
and isolated ; iJ.vaoiJ.ai, to think of a woman, to woo 
her; when fJvricrTos, fjvricrTuip, fivrjo-Trip, wooer, suitor; 
fAvtjiTTvs, an asking in marriage, fivrja-rpov, betrothal. 

From math the Greek has fiaQrifja, science, juavOavw, 
to learn. Prometheus, " the wise," the " far-seeing." 

Latin is hardly less rich. There is Menerva, Minerva, 
she who thinks or warns, the ancient divinity of Etruria 
and Latium, with which we should connect the old verb 
promenervare, discovered by Festus. Then moneo, I cause 
to think, I warn ; moneta, an epithet of Juno, which has 
. become our money ; monumentum, moTistrum, and ynon- 
strare ; maneo, I waitj and its derivatives ; imminere, 
eminere, minari, minitari, to menace ; Tnemini, remini- 
scor, comminiscor, I remember; Tiiens, mind; mentiri, to 
imagine, to lie. With mad are connected meditari,medi- 
tatio ; medeor, medicus, remedium, which, like the Zend 
,m£d, have to take the sense of to cure, doctor, remedy. 

For tbe Teutonic languages we will only cite in 



Indo-European Roots. 279 

Gothic, muns, mind, gemunan, to warn, gaminthi, re- 
membrance ; Old High Ger. minnia, minna, love, 
whence Minnesanger ;' finally, the generic term Mannu, 
the son of Tuisco, the god of the whole race, whence 
mannish, mensch, man, universal nonn and suffix. 

There is no difference in etymology between the 
classical languages and French ; it is easy to recognise 
in French numerous words derived from the roots ma, 
Triad, and mant. Greek has given manie, maniaque, 
mnimonique, matMmatique, muse., mUre ; Latin, the 
popular forms mois, rnaison, manant, monnaie, montre, 
ramentevoir, mensonge {mentitionia), mesure, m^ge, and 
numerous others, which are simply transferred from 
the Latin with very little change ; permMnent, immi- 
nent, Eminence, mental, ddm^nce, mention, commentaire, 
monstre, demonstration, monitewr, monument, reminis- 
cence. Lastly, French, like the other Eomance lan- 
guages, has made for itself an abundant resource by 
tlie formation of the adverbs in -ment. This suffix -ment 
is but the ablative of the Latin mens, which from words, 
logically formed, like honamente, malementc, ionnement, 
malement, has passed to the adjectives affreusement, 
splendidement, &c. Its origin is certain, and we note 
in confirmation that the sufiBx is always added to the 
feminine of the adjective, since mens is feminine. The 
use we make of the suflBx -ment has its parallels in the 
classic languages, the Sanscrit and Greek participles 
in manas, /mevos, the Latin terminations men, mentum. 
At the risk of puzzling the reader by quoting ex- 
amples in which the root is changed according to 
laws which we have not yet set forth, and surcharged 
with appendices, terminations, and even sometimes 
with prefixes, of which the presence is not yet suffi- 
ciently accounted for, I have wished to call attention 



28o The Indo-European Organism. 

fco the extreme importance of these roots, to their pre- 
ponderance in a hundred languages and dialects of 
which they are the framework or skeleton. In them 
we see the long evolution of thought, its origin, its 
progress, even its errors. From the impression of the 
senses, at first confused, then distinct, then varied, 
they lead us to the expression of the feelings, of the 
affections, of the passions, and of mental concepts. In 
following their successive changes and developments 
we seem to watch the travail of the brain, the birth of 
intelligence. 

All the early roots might be taken in like manner 
in illustration of the same general principle, and I 
hold some in reserve. But it must not be forgotten 
that the' attributive roots are not nouns, nor adjectives, 
nor verbs ; they need the terminations which show 
case or person. Here we find two alternatives : either 
the root is joined immediately to the termination, as 
in the Latin words rex, lex (for reg-s, legs), das, dot, 
fers, fert, and in the French est, or it is separated from 
it by one or more suffixes, as in the Sanscrit hhar-a-s, 
burthen, bearer, ihar-a-mi, biMr-a-ja-mi, hliav-i-shya- 
mi ; in Greek, (pik-eo-fjiai, SeU-vv-fjn ; in Latin, fer-i- 
mus, fer-e-hat, fra-ter-em, am-a-bilis, am-a-n-ti-bits, &c. 
The first and earlier method causes numerous curious 
changes by the influence of letters one upon another ; 
the second, more recent, but more convenient, was 
already the more common before tbe separation of the 
Indo-European languages. In the one, the root itself 
constitutes what is called the radical or indeclinable 
theme ; in the other, the theme is a simple or a com- 
plex derivative of the original root. In dic-i-mus, we 
say, in bhar-a-nti, they carry, the roots are die and 
bhar, the themes dici and bliara ; in tanumi, I stretch. 



Indo-European Roots. 281 

the root is ta or tan, the theme is tanu ; in datus the 
root is da, the theme is daiu ; in amare, root am, 
theme ama ; in amdbili-s, -tas, -ter, the theme is 
aniaiili. This notion of the radical or theme is im- 
portant ; it modifies and simplifies grammatical teach- 
ing. There are no longer declensions in us, os, as, es, 
a, i, but there are numerous varieties of themes to 
which are attached the terminations s, m, t, &c., of the 
nominative, accusative, and ablative. 

What, then, are these letters a, i, 0, e, u, these 
syllables nu, ya, isc, tu, abili, which are interposed 
between the root and the termination ? They are 
pronominal roots, those which, as we have seen, fur- 
nish, either independent or agglutinated, all our pro- 
nouns, and acquire by degrees distinct meanings. 
Subordinated to attributive roots, they acquire other 
meanings and values, which are generally determined 
by their function, some being usually annexed to 
nouns and adjectives, to" masculines, feminines, or 
neuters, others to verbs, moods, and tenses, but all 
susceptible of uses which vary according to the pre- 
ferences of races and the needs of speech. 

The accumulation of suffixes, the formation of themes 
which are already derived from earlier roots, and form 
in their turn generations of new words provided with 
various meanings, is one of the most striking and 
peculiar characteristics of the Aryan languages. It 
is by such methods that arise collateral series such 
as jiiste, ajuster, justesse, justice, justicier, justiciable, 
justifier, justification, jur6, jiiron, juridique, jiujer, judi- 
cieiac, judiciaire, jvdiciaireinent, or as lui, luire, luisant, 
lucide, Lucina, allumer, lumidre, lumineiix, luminaire, 
illumination. Of course these French products of the 
roots jus and luc are the issue of corresponding Latin 



282 The Indo-European Organism. 

forms, and similar facts may be reckoned by thousands 
in all the Indo-European idioms. But, as I have said, 
each language has made its choice, each has combined 
and amalgamated in its own way the suffixes taken 
from the common stock ; and herein it is distinguished 
from its kindred, hereby it displays its own genius 
and vitality. 

A good many ready-made themes may, however, be 
found throughout the entire family, more especially 
the simpler ones, those which have only a letter or a 
syllable, called formative, attached to the root. The 
origin of these snffixes, so slight and meaningless yet 
so persistent, is the subject of many hypotheses. Is 
the extra vowel a very early appeal to the attention of 
the auditor ? Is it a survival from a time when our 
idioms, like many Asiatic and African idioms, avoided 
final consonants ? In that case the dissyllables hhara, 
valca, iferi), voca, would be earlier than the contracted 
roots hilar, vak, fer, voc. It is possible, but more 
often, perhaps, the superfluous vowels are merely a 
rest for the voice, an easy liaison, what have since 
been called euphonic letters, whose use commonly 
escapes us. 

It is almost always in company with this vowel that 
the root forms part of a compound word : Theophilus, 
Philotheos, lover of God ; Patrocles, Cleopatra, pride of 
his (or her) father ; and a study of this great class 
of words often shows us the true theme of a noun or 
verb, preserved from attrition by the second term in 
the compound. 

Composition, properly so called — that is, the juxta- 
position of two or more themes, of which the second 
only is declinable — belongs to the inflected languages 
only. Sanscrit, Greek, and German have made great 



Indo-European Roots. 283 

use of this power, indeed even to abuse ; the Eomance 
languages, on the other hand, have almost lost it. 

In a-more general sense, the word may be applied 
to the prefixing of adverbs, prepositions, conjunctions, 
which do so much in the way of differentiating and 
altering the meaning of the root. Maleficus is a com- 
pound, just as are mMgnificun, pacijicus, cafnifex; nor 
should I exclude contradictio, intercipere, pro- or post- 
pi mere, profundus, from the category of compound worda 

The formation of Indo-European words is, then, the 
result of two processes : derivation, by means of suffixes, 
and composition, by the joining of two or more themes, 
and by the addition of prefixes. We have only now 
to consider the terminations which show the function 
and place of words in the sentence. 



CHAPTEE II. 

PARTS OF SPEECH— THE NOUN. 

Original identity of substantive and adjectiTe — Parallel formation of 
noun and verb — Deolenaion — Case-endings are postpositions of 
demonstrative suffixes — The attrition of terminations a proof of 
their great antiquity — Nominative, accusative, subject, and 
object in old French — Genitive, dative, vocative, and ablative — 
Insufficiency of oases — Gradual substitution of the preposition for 
the postposition — Degrees of comparison — The disappearance of 
the suffix before the adverb. 

Before we touch on the mechanism of the declension, 

we have to solve one or two preliminary problems, 

which offer a far greater interest than the nomenclature 

of the forms special to the various families of languages. 

What is a noun ? What is a verb ? Is the one class 

anterior to the other ? What are case and verbal 

endings ? what their na'^ure and their office ? 

The noun is the individual or generic designation of 

an object, in the most general sense of that word, of 

a sensible or intellectual object, concrete or abstract. 

So at least it now appears to us when we say lion, 

tiger, sheep, horse, house, or even pain, pleasure, 

feeling, idea. But although we apply to these words 

the term noun-substantive, although all hold the same 

place in the sentence, and are brought into relation 

with one another in the same manner, and with other 

parts of the proposition by the same methods, yet 

even a cursory examination will reveal to us shades of 

difference among them. Most of them are derived 

284 



Parts of Speech. — The Noun. 285 

from derivatives ; and it is through forgotten meta- 
phors and transpositions, through changes innumerable 
in form and meaning, that they have come to express 
for us the, so to speak, real and objective image of 
the being, thing, or quality with which we identify 
them. They represent no substance. But, by a useful 
illusion, eliminating more or less voluntarily all acces- 
sory or approximative meanings, we imagine that we 
conceive clearly the object designated. Are there now, 
were there ever in the past, true substantives ? Yes 
and no. That is to say, that in the earliest times, 
when man was yet unable to analyse his sensations, 
the meeting with an animal, the passage of a meteor, 
a blow received, pain or joy experienced, may have 
provoked cries, phonetic gestures, which answered ade- 
quately to the impression received ; and these sounds, 
these primitive names, may, by the merest chance, 
have come down to us. I am speaking here of the 
Indo-European languages only. For it cannot be 
doubted that a great number of Chinese monosyllables 
were signs, attached either to objects or to various 
aspects of objects, aspects considered as new objects. 
M. Michel Br^al is disposed to believe that naked 
roots like sr/>, like av, as, va, ap, or aqv, gvau, which 
are found in serpeiis, in ovis, in asu, in vata, in aqua, 
in bos, were essentially what I shall call raw substan- 
tives, and signified before anything else serpent, sheep, 
breath, ivind, water, ox, and by analogy only came to 
be applied to other beings, or the actions of other 
beinjjs, and took other meanings, such as to glide, 
undulating, to breathe, life, to run, agile, to wander, to 
walk, earth. This opinion is plausible, yet the haste 
with which speech seizes upon those raw substantives 
to express qualities or actions seems to prove that 



286 The Indo-European Organism. 

they themselves are the result of an unconscious 
analysis, and that they describe the most salient 
peculiarity of the object heard, seen, or touched. Our 
sensibility, in fact, having five ways of perceiving 
external things, is itself an instrument of abstraction. 
The senses co-operate and supplement each other, but 
it is always the one which is most directly affected 
which determines the impression on the brain and its 
expression by the voice ; and the vocal symbol neces- 
sarily differs according as it corresponds to an indication 
of sight, hearing, touch, smell, or taste. Hence the 
number of synonyms ultimately rejected, or reserved 
for approximate shades of meaning, or qualities per- 
ceived in the same object by the eye, the ear, or the 
hand. So that the noun, even when the most in- 
voluntary expression of the primitive impression, can 
only be the expression of a quality. 

There is, then, no original difference between the 
substantive and the adjective. Both are names which 
express a quality, a manner of existence, either gene- 
ralised and applicable to all the objects which possess 
it, or specialised and identified with the whole of the 
object, of which it really designates but one property. 
" All substantives," says M. Breal, " were, to begin 
with, adjectives taken substantively." 

How did the adjective come to be distinguished in 
the long-run not only in meaning but in form ? In 
the first place, the adjective, habitually used to signify 
an object, lost its qualifying value, and came to be 
solely the name of the object. For instance, deva, 
which means the shining, and which has still in 
Sanscrit the three degrees of comparison, ended by 
meaning the god. Sourya, the brilliant, became the 
name of the sun. Akva, the runner, became the 



Parts of Speech. — The Noun. 22, j 

name of the horse. Manu, the intelligent, signified 
man. The epithet was forgotten in the thing desig- 
nated hy it. Other words,- on the contrary, laghu 
(Gr. e\a-)(y<s, Lat. le(ff)tis, light), tanu, hrghu {tenuis, 
hre{cj)vis), nava, new, not being specially attached to 
any object, retaiaed, with their qualifying power, the 
faculty of taking the three genders, which the sub- 
stantives lost, and the comparative and superlative 
forms. Phonetic change, by obscuring the meaning 
of the roots, contributed also to separate the two 
classes of words. The Hindu, whose language is 
less modified, can perceive the relationship which 
exists between akva and agu, rapid. But what Greek 
would have guessed the affinity between (L/cyy and 
iTnroj ? Then a choice was made between the sufiixes, 
and that as early as the Indo-European period. If 
in Latin the suffix ti forms alike nouns and adjectives, 
pestis, vestis, fastis, mentis, forlis, mitis, tristis, there 
are others, such as mun and men (agman, the sky, 
nomen, documcn, foramen, exam^n, agmeri), such as tra, 
tro (yrXrJKTpov, rastrum, cuUrum, Trwnstrurri), which 
were used for nouns alone. We sometimes find this 
sorting process taking place in a single language. In 
the Vedic dialect, for instance, the suffix as still forms 
adjectives, tar-as, penetrating, ap-as, action ; but in 
Sanscrit it rarely forms anything but nouns, man-as, 
gan-as, as in Greek and Latin, fxevoi, gen-us, op-us. 

The verb, which we shall consider separately, is 
closely related with the nonn and the adjective. At 
first it expresses, like them, a state, a manner of 
being, an action ; it even borrows their form for its 
supines, participles, gerundives, and infinitives. Re- 
duced to its simplest elements, it is composed, like 
them, of an attributive root or theme, and of a 



288 The Indo-European Organism. 

demonstrative root. Bhara-s, the bearer, the burden, 
is the same as Ihara-ti, he bears ; for the two suffixes 
a and ti represent the same pronoun, of which the 
sound varies between sa and ta, and which replaces, 
announces, or recalls the subject expressed or under- 
stood. For the rest, the fundamental identity of noun 
and verb is proved to us by the moDosyllabic lan- 
guages ; the same word may be noun, adjective, or 
verb, according to its place in the sentence. But in 
the agglutinative idioms we see the beginning of the 
differences, which become more marked in inflected 
languages. Affixes placed before or after a root, 
which may turn into a verb, are pronominal, and not 
simply indicative ; already some rudimentary artifices 
add to the action the idea of present or of past time. 
The complicated edifice of the Indo-European conjuga- 
tion is raised on the same foundation, and by the aid 
of similar materials. Only the joints, which a ruder 
construction allowed us to see, have here disappeared, 
hidden by the fusion in which the theme, itself in- 
flected, contracted, lengthened, reduplicated, and the 
auxiliaries and terminations are welded together. 

In his ingenious essay on the chronology of language, 
Cuvtins rejects the hypothesis of a parallel develop- 
ment of the noun and of the verb. He believes that 
while the former remains as a naked root, or at least 
as an indeclinable theme, the latter acquired the six 
personal terminations, " invariable characteristic of all 
the Indo-European languages." The declension is 
even, in his view, of later date than the insertion 
between the root and the termination of the suffixes 
a, ja, nu, na., pa, ta, s^a, which give an intensive, causa- 
tive, frequentative, desiderative meaning to the verb, 
and later also than the period of auxiliary verbs and 



Parts of speech. — The Noun. 289 

compound tenses. For my part, I can see neither the 
reason nor the probability of these successive forma- 
tions. If the method of the conjugation has been 
longer preserved, it is because it corresponded longer 
to the needs of speech ; perhaps also because it was 
only just completed, and present to all memories at 
the time when the different tribes which composed 
the family began their slow exodus. In any case, we 
find the declension far more worn by time, an indica- 
tion of an antiquity at least as great ; even in Sanscrit, 
Zend, and Lithuanian, where we find it most complete, 
it can hardly account for some of its forms. The 
cases whose endings are worn and altered become 
confounded and mislead the auditor, so that recourse 
is had to prepositions. The Latins, Greeks, and Ger- 
mans lose two or three terminations as their language 
progresses, though some remnants still embarrass our 
modern idioms. 

Though it be impossible to recover the primitive 
form of the case-endings of nouns, it is none the less 
certain that they were equivalent to the prepositions 
which have reinforced or replaced them ; when they 
were joined to attributive roots, the mother-tongue 
admitted only of postpositions or enclitics ; these 
latter, perforce abridged in pronunciation, became con- 
tinually effaced, and numerous independent particles, 
already provided with a meaning and a form of de- 
clension, were used to supplement their failing and 
decrepid predecessors. As for Schlegel's mystical 
theory of the vegetation of words, each word throwing 
out, like certain insects or molluses, filaments, buds, 
ephemeral or durable excrescences, it has long been 
forgotten ; even if the testimony of the three first 
persons of the verbs in mi, si, ti, did not prove the 



290 The Indo-European Organism. 

pronominal origin of these terminations, it would be 
impossible to sepai-ate them from the suffixes. Now 
no suffix proceeds from the root ; all are added to it. 
This is equally evident for case-endings and personal 
terminations. 

The history of the noun and that of the verb can- 
not be separated ; they arose together, bound together 
by a common radical, and gradually distinguished by 
suffixes hardly different from each other, but increas- 
ingly used for special purposes. One of the functions 
of the verb being to put two nouns into relations, to 
mark the action of one upon another, the case-endings 
seemed appropriate to lend the phrase force and clear- 
ness. When the prepositions had acquired definite 
meaning, the verb and the noun both made use of them, 
the one to enable it to govern a given case, the other 
to enhance some weakened or doubtful termination. 
And in the meantime, the natural process of derivation 
was always producing verbs from nouns, and from the 
verbs verbal nouns and adjectives, which themselves 
produced new forms. 

The earliest known form of Indo-European declen- 
sion has eight cases : nominative, vocative, accusative, 
genitive, dative, ablative, locative, and instrumental; but 
it might admit, it has admitted, of many others; for there 
are other shades in the relations of words to each other, 
and even these cases express imperfectly the meanings 
of with^ for, through, on,, imd&r, &c. The Latin me-cum, 
vohis-cum, is a combination which is apparently a form 
of case ; and the Umbrian dialect offers examples of 
true cases which it has created for itself : anglu-to, 
from the angles ; anglonn-e, near the angle ; totam-e, 
asavi-e, near the city, near the altar ; tota-per, for the 
city ; asa-co, with the altar ; asam-ar (for ad), towards 



Parts of Speech. — The Noun. 2 9 1 

the altar. These examples are precious because they 
are still lucid. The Umbrians formed these cases iii 
iniitation of those they already possessed ; they did 
not innovate. Their method is that which gave rise 
to the declension ; and the suffixes they used are not 
different in kind from earlier terminations. 

Since each language has used the cases in its own 
way, whether it suppresses them, whether it com- 
pensates for their loss by lengthening or accentuating 
the syllable which should bear their sign, or whether 
it modifies their meaning and function, it is impossible 
to present a general table of the Indo-European declen- 
sions, unless it be in Sanscrit and Zend, in which all 
the forms are moulded on the same type. Tet even 
here examples of these two idioms, embracing the 
comparison in detail of the various masculine, feminine, 
and neuter themes, in the singular, dual, and plural, 
would overwhelm the reader with their innumerable 
variations, and would only leave a confused and 
fatiguing impression. This mechanism, too delicate, too 
minute to be solid, should be studied separately and at 
leisure in each language. 

Omitting the instrumental, which is peculiar to the 
Oriental and Letto-Sclavonic groups, we will give a 
few succinct notions, incomplete but clear, about the 
other cases, showing in especial certain agreements or 
curious divergences. 

If the declension, instead of being the unconscious 
work of time, had been invented by some judicious 
grammarian, it would seem that the nominative might 
have dispensed with any outward sign. And, in fact, 
many neuter and some masculine nouns in ar, as, i, 
and u, have no termination in the nominative. But 
in ffeneral, the naked theme has been reserved for the 



292 The Indo-European Oi-ganism. 

vocative, with abbreviation of the final vowel : aina, 
'i-Trire, domine, amice, horse, O master, O friend. 
In the great majority of nouns and adjectives, mas- 
culine and even feminine, whether the theme ends in 
a vowel or a consonant, the nominative termination is 
an s, the remnant of the demonstrative sa. The s is 
too familiar to those who have any knowledge of Latin 
and Greek to need examples ; it persists in the Scla- 
vonic languages and in Lithuanian ; retained in Gothic, 
it has become r in Scandinavian ; German has almost 
entirely lost it. But it is clear that at the time of the 
separation of the idioms this case-ending was already 
decaying. Before certain words, and in virtue of 
euphonic laws which are often strange, Sanscrit re- 
placed it by soft breathing. Zend compensated its 
loss by a change» in the final vowel : mazddo, for 
mazdas, just as the Provencal says chivao for chivals 
(whence the French chevau-leger). In Latin the s was 
faintly heard, and was elided before a consonant 
{omniiu, Jovi) from the time of Bnnius ; and while the 
classic Latin re-established and maintained it scrupu- 
lously, the popular speech cared little for it. There is 
no trace of it left in Italian, and very little in the 
other Romance languages. French retained it longest, 
using it, however, with very little sense of its meaning. 
As early as the fifth century, M. Bracket tells us, 
long before the appearance of written French, popular 
Latin reduced the number of cases to two, subject and 
object, and, in order to distinguish them, chose the 
two terminations which recurred the most frequently, 
bonus, bonu-m, muru-s, muru-m. French grammar, 
which is a continuation of Latin grammar, inherited 
this system in part ; it could not revive the final m 
of the accusative, nor the i of the plural niuri, which 



Parts of Speech. — The Noun. 293 

had dropped ; but it retained the s wherever tho 
Latin of tlie people pronounced it, and constituted a 
declension which was very useless, but simple and 
acceptable : in the singular li murs and le mur, plural 
li mw (illi muri), les murs (illos muros). In the 
thirteenth century, taking as type the second Latin 
declension, the s of the nominative singular was extended 
to those forms which had never possessed it ; and men 
wrote li pastres, the shepherd. This artificial con- 
struction ruined the declension which it was intended 
to confirm. Eejected by the people from the thirteenth 
century, and often neglected by the lettered class, the 
French declension died out in the fourteenth century. 
Thenceforward only one case was used for each number, 
and as the objective case was longer in Latin, and 
therefore better able to resist th^ strong tendency of 
modern French to contraction, this case prevailed 
and was chosen as the type. Murum became mur, 
muros, murs. Thus all our nominatives represent 
Latin accusatives. Some remains of the. ancient 
nominative singular may be found, however, in tie 
nine following words : fils,fonds, lacs, legs, lis (lilius), lez 
(latus),puits (jputeus), rets (retis), queiuc {coqmis) ; which 
were in the accusative : fil (whence filiation), fond 
(fundum), leg, li, U (still used in the phrase un U 
dlitoffe) ; latum, fuit {puteurri), ret (retem, rMaire), 
queu or coq (ship's cook). Such is the end, in no sense 
to be regretted, of the earliest and most inconvenient 
of case-endings, of that which engendered perpetual 
confusion between all the cases and both numbers. 

The general sign of the accusative singular, m, has 
everywhere, as we have seen, marked the direct object, 
with the primitive sense, doubtless, of movement to- 
wards an object, a place, a being : eo Romam, eo 
20 



294 ^'^^ Indo-European Organism. 

Zugdunum, I go towards Rome, towards Lyons. When 
the verb, at first intransitive, took the active sense, 
implying in itself movement influence on the object 
" governed," the original sense of the termination m 
became fainter, but the sign remained no less useful 
to distinguish the object, and mark its subordination 
to the subject of the sentence. Universally adopted, 
it perished nevertheless, and its disappearance began 
early in our Western languages. Greek, rejecting the 
sound of m at the end of a word, replaced it everywhere 
by a n, an equivalent nasal, which became labialised 
before a p or a 6. They pronounced toij. TroXe/xov, war, 
eh TOju TToXiv (whence Istamboul, towards the town). 
But in a number of forms this n dropped, preserving 
only the vowel a of the primitive theme, which would 
otherwise have been ^yeakened into e ; avSpa for avepav, 
Kvva for Kvvavi, dog. In Latin, in Umbrian, in the 
earliest inscriptions, the final m is no longer written, and 
even in the classical Latin of the Augustan age it was so 
little pronounced that its elision was the rule in poetry. 
It became completely obliterated from popular Latin 
and from the Romance languages ; so much so, that in 
order to pronounce it in reading Latin, an Italian is 
obliged to double it and add a mute e, sanctumme, like 
Gerusahmme. Retained in Gothic, tunthum, the tooth, 
it becomes lost in the German n, or rather it remains 
only as the sign of the dative {dcm, gutem), in the use- 
less pronominal declension of literary German. One 
of the reasons for the disappearance of this interesting 
m is again to be found in its double use. Not .only 
was it attached, we know not why, to the neuter 
nominative (danam, donum), but it also terminated the 
genitive plural, devasam, deorum, rosarum, omnium. 
Thus it lost its distinctive value and its vitality. It 



Parts of Speech. — The Noun. 295 

existed also, but followed and soon absorbed by an s, in 
the accusative plural : /xeyaXou?, domiiws, are contrac- 
tions of fxeydXovs, dominuTTis, a regular phenomenon 
of assimilation in Greek and Latin. 

The genitive or possessive ease is peculiar ; it is, so 
to speak, retroactive ; it seems governed by the nomina- 
tive, but it contains the true subject. Liher Petri, it 
is Peter who is • the possessor of the book. It would 
not therefore be surprising if the sign of the genitive 
were derived from an enclitic relative sufiBx, " The 
book, Peter, which," " the book which Peter " (under- 
stand possesses, or wrote) ; hence the frequent con- 
fusions with the ablative of or iy, " the object which 
is of, or made by, Peter." This case, the genitive, was 
represented by several terminations, especially sya for 
vowel themes, and as, os, is, for donsonantal themes. 
Greek used both, but the intermediate s has dropped 
as commonly happens between two vowels, and the 
Homeric 010 (deolo) for otrto has been contracted into 
ov — \6yov. The other case-ending has remained with- 
out alteration, (pXoyoi, of the flame (as also the plural, 
wv), avSpos, of the man, ^a(ri\e[F]a>g, A([f]o?. Latin 
has completely rejected sya, and replaced it by an i in 
three declensions : rosai, diei, domini. It has kept as 
under the archaic forms os {senatuos, magistratuos, 
maniios) and tos ( Venerus, Cererus) ; and finally under 
the classical form is (frafris,sororis),w}nch unfortunately 
created a confusion with the numerous nominative sin- 
gulars in is and with the ancient plurals frugiferenteis, 
parenteis. The Germanic languages still use the geni- 
tive in s, and it is one of the few traces of grammar 
retained in English. 

The comparative table of the dative shows the inde- 
cision of the various dialects between the modifications 



296 The Indo-European Organism. 

of the original form. I take my examples from Bopp. 
Sans, agvdya, Zend aspdi, Lat. equo, to the horse ; 
agvdydi, hisvaydi, eqiim, to the mare. A primitive 
dya, contracted into ai and e for consonantal themes, 
may be traced in these variants. The Latin eguce is 
clearly for equai ; but eqiio would be embarrassing 
were it not that the i reappears in all the datives in i, 
sorori, fratri, menti, and the archaic populoi, BomaTwi. 
0, like e, is in Latin the usual substitute for the Sans- 
crit a. Lithuanian has a form ashvai, to the mare, 
and gibai, to the gift (dono), in Gothic (side by side, 
it is true, with wulfa, to the wolf), tends to confirm 
the opinion of Bopp. We are tempted to assimilate 
the Greek dative singular in tS and in i pure {iraTpl, 
eXirlSi), but it is a temptation which must be resisted. 
Greek has substituted for the dative, as a rule, the 
forms of another case; in the plural the transposition 
is evident. The termination cri, crcri, which is equivalent 
to the Sanscrit su for sva, belongs to the locative, of 
which the proper meaning is evident in AQr/vriari, 
OXuyUTr/acrj, at Athens, at Olympia. Now in the sin- 
gular, the locative is simply mai-ked by a short i, the 
same doubtless as we find in the preposition in ; and 
it is this pure i which Greek preferred to the doubtful 
sound of the dative. 

The two cases are very nearly alike in meaning as 
in form, and one of them has rightly seemed super- 
fluous. Latin, like Greek, has been seduced Hythe 
simplicity of the locative singular, and while retaining 
probably the original dative in the singular long i, and 
in the plural hus (Sans, and Zend bhyas), it has em- 
ployed the locative not only in the apparent exceptions, 
humi, domi, Zugdttni, but in the office of the genitive. 
It had rejected the suffix sya, and' kept only the ter- 



Paris of speech. — The Noun. 297 

mination as, is ; and introduced the locative i into 
three of its declensions, thus creating confusion be- 
tween the genitive singular and the nominative plural 
domini, and among the three forms rosce, of the rose, 
to the rose, roses. 

The Indo-European ablative was marked hy a, t or 
a d preceded by a lengthening of the vowel of the 
theme. At least we must infer this from the agree- 
ment between several Sanscrit declensions and the 
ancient Iranian, Latin, and even Gothic forms. The 
Sanscrit agvat, vrkJiat, the Zend aspad, vehrlcad, by 
the horse, by the wolf, have been rightly compared 
with the old Samnite forms touta-d, by the people, 
suva-d (sua), preivatu-d (privatu), dolud malud (dolo 
malo), prcesentid, ligud {lege), co7iventionid, and the 
Latin forms found on the rostral column of Duilius, 
in the Senatus Consult of the Bacchanals and on the 
tombs of the Scipios ; navuled prcedad, in altod marid, 
summed didatored, Gnaivod patre natus (son of Cnaeus). 
But the sign d had already lost its force in the time 
of Plautus, who used ted and med in the accusative. 
The old ablative can still be traced in the adverbs in 
a and 0, contra, intra, extra, pro ; the latter has re- 
tained the dental in prodire, prodtst. But as a rule 
the d has been lost, in ancient Persian, in Latin, in 
German, and in Slav, abandoning the final vowel a, e, 
i, 0, u, to every species of alteration. Hence an in- 
tolerable and fatiguing confusion among a number 
of words which ought to occupy different offices in 
the logical construction. For the rest, the ablative is 
quite insufficient to render all the meanings which are 
required of it ; it originally signified the point of de- 
parture, and thereby drew near to the locative ; after- 
ward came the idea of cause, of effect, of dependence. 



298 The Indo-European Organism. 

and the perpetual interchange with the instrumen- 
tal (by means of, with, by), and with the genitive. 
When taken with a pronoun or an adjective (the 
ablative absolute, and the same construction is found 
in Greek with the genitive), expulso Tarquinio, illo 
dormienti, the ablative easily acquires a completely 
adverbial sense. Most Latin, Gothic, and even Slav 
adverbs {tamo, jamo, kamo ; illic, uU, quo), in 0, in e, 
in i, ia it, are ablatives. Greek, which has rejected 
this case from its declension, has kept it for this 
same use ; at least it has been surmised in the 
numerous adverbs in ws, such as aXrjGws, oixuk;, ovtws, 
<o?, TtiXiKwg, <roo(pp6vQ)i, for aXriOiioT, o/jloot, &c. (Greek 
not permitting the letter t at the end of a word). 

These few notes are far from exhaustiug the subject, 
which has occupied so many practised philologists. 
Bopp has devoted more than one volume to the 
analysis of the Indo-European declensions, and the 
German declension alone, one of the vices of that rich 
language, occupies as much space in the works of 
Jacob Grimm. In these delicate matters we must 
have regard to phonetic laws, both general and par- 
ticular, to the successive periods of each language, to 
all the reasons which may have suggested, maintained, 
modified, or condemned the various case-endings. We 
have only considered the singular of those classes 
and idioms which are most familiar to us ; but if we 
were to pass to the altered and complicated forms of 
the Slav languages, to the strange metamorphoses of 
Armenian ; if we were to examine the plural termina- 
tions, the an, on, or son (Lat. um and rum) of the 
genitive, the hhyas, hhis, is, which commonly serve 
the three cases, dative, locative, ablative, and even the 
instrumental, which is probably a pronoun already 



Parts of Speech. — The Noun. 299 

declined, and which may cerbainly be found again (in 
the singular) in tuihyam, Unibrian tefe, Lat. tibi, and 
in the Homeric forms in (pi — ^IrjCpi, by force ; if, lastly, 
we were to adventure ourselves on that superfluous 
doubling of the plural called the dual, of which the 
practical Latin has only retained the two useful words 
duo and amho (Sans, dvau and ubTiau), we should lose 
ourselves in an endless labyrinth. I may, however, 
without engaging myself too far, contest the most 
commonly received opinion on the origin of the dual. 
It is generally supposed to be a development of the 
plural, and even a late development. But why should 
this be so ? Does not the invention of the plural take 
from the dual its raison d'Stre ? We find the dual in 
the languages which are nearest to the common idiom, 
in Sanscrit, in Zend, in Greek, and among the Letto- 
Slavs. I should prefer to regard it as a survival from 
the time when our savage ancestors first distinguished 
two from one, when they perceived that man had two 
arms, two legs, two eyes, when the worship of genera- 
tion inaugurated the idea of the couple, animal, divine, 
or human. 

I will now sketch the evolution, which is relatively 
simple, of the degrees of comparison. Three words 
suflSce in French, plus, moiiis, derived from the Latin 
comparatives, and Ms, the Lat. trans, beyond. In 
the languages which are nearer than French to the 
Indo-European idiom we shall find suffixes placed after 
the theme, or between the root and the termination, 
and consequently subject to numerous alterations. 

Leaving on one side monosyllabic and agglutinative 
methods, such as the repetition of the word, or the 
addition of a term signifying quantity, which have 
again become familiar to our analytic languages, we 



300 The Indo-European Organism. 

find that the only means possessed by language for 
expressing difference in size, distance, or number was 
the postposition of a demonstrative root. Custom, 
use, would do the rest, would fix and define the sense. 
The general suSixes, ra, ta, ma, were used, suffixes 
which form adjectives, nouns, or participles, without 
having a very defined meaning in themselves ; then at 
a later period two compound suflBxes were employed, 
ja7is or yos, this latter very nearly allied to the present 
or perfect participles, cms, on, 6s, and os, and ishtJia, 
which seems to be connected with the desiderative 
verbs. 

I give a few examples of the two groups. Ba has 
only survived in a small number of words ; Sans. 
ava-ra, inferior ; wpa-ra, superior ; apa-ra, posterior ; 
probably pra for pa-ra ; in Gr. virep for inropa ; Lat. 
inferus, superus, infra, supra, super, pro, pros. We 
have not here a very Tnarked comparison, yet it marks 
differences of situation, behind, before, under, upon. 
The r of the French preposition sur is a relic of the 
old suffix ra. 

Ta is not more precise ; it is retained chiefly in 
the ordinal nurnbers ; Saps, tchaturiha (fourth), Gr. 
■wpwTOi, TpiTO's, SeKaro? ; Lat. quartus, quintus ; Goth. 
saihs-ta-n, ahhi-da-n, sixth, eighth ; and in certain 
adverbs, Gr. evro^, Lat. intus, suhtus, penitiis. 

Similarly ma signifies first of all the order of things; 
Sans, saptama, daqama, seventh, tenth ; Gr. e^So/jLos, 
seventh ; Lat. primus, decimus ; afterwards it expressed 
the superlative ; Sans, apama, the lowest ; Lat. 
sit(pymus, summus, minimus, infimv^ ; supremus, ivius, 
the last ; Goth, fruman, first ; auhuman, superior. 

These suffixes have been variously combined to give 
more force to the idea : tara, tero, ter, for the com- 



Parts of Speech. — The Noun. 301 

parative ; tania, timo, tato, for the superlative. The 
first has been adopted by Sanscrit, Greek, and the 
Germanic languages ; Sans, katara, which of two, 
antara, interior ; Goth, khathar, which, anthar, the 
other ; Gr. iroTepo^, Seurepos, KovcpoTspoi. Latin has 
kept this form in a number of words in ter — uter, 
noster, dexter, alter, ceteri, iterum, ultra, intra, citra ; 
but it is still more used to determine comparatives 
properly so called. 

Tama is one of the regular forms of the Sanscrit 
superlative : punjatama, the purest ; Goth, aflumun, 
the last ; Lat. optiinios, ultimus, mactimus (maximus), 
intimus, &c. In Greek there is only the doubtful 
form Ar-temi-s, the most noble ; tato is used by pre- 
ference : yXvKurarog, the sweetest, KovcpoTUTOi, &c. 

The other group of suffixes, well known in Sanscrit 
(hhu-yans, more numerous, mahishtha, garislitha, the 
largest, the heaviest), in Greek {^Slayv, fiaylcov [/xe/^mi'], 
apelwv, softer, larger, stronger), in Gothic and Teu- 
tonic (hauhista, the highest, best), has furnished Latin 
with all its comparatives ; but the nasal has dropped, 
and sometimes the y also; and the sibilant in the 
masculine and the feminine has been changed into an 
r (as in odor, odoris, arhor, arhoris, &c., for arhos, odos) ; 
it has per!5*isted in the neuter : melior, melius ; major, 
maj'us, magis (the Fr. mais), pluris, {Jor pleoris), plus ; 
minor, minus ; pejor, pejus. 

Lastly, the combination of the two groups has pro- 
duced double comparatives and superlatives ; Sans. 
papi-yastara, wickeder, panisthatama, very praise- 
worthy, surabisthal ama, very sweet-smelling, nedis- 
ihatama, the nearest ; Gr. XaXiarepoi, apia-repos, 
evSaifMovecTTepos ; Goth, aftumist, the last ; in Lat. in- 
terior, superior, minister, " the last of the least slaves,'' 



302 The Indo-European Organism. 

sinister, inagister ; and perhaps the superlatives in 
-issimus, doctissimus, felicissimus. 

Thus grammatical forms wear out and are reinforced 
by doubling them, or by other complications which do 
not preserve them from becoming obsolete. But their 
long life has not been inglorious ; they have closely 
followed or determined the progress of the human in- 
telligence towards order in the proposition, towards 
clearness, precision, towards variety also in the expres- 
sion of every shade of thought. Their use has at last 
rendered them useless. Prom reins they became 
fetters ; the apparent aid which they gave to syntax 
ended by impeding the logic of the sentence. The 
declension, of which we have given this hasty sketch, 
seemed at first to be sufiBcient for all the relations of 
the noun with the verb and the adjective ; but even 
before the separation of the idioms, the too narrow 
meaning of the case-endings, and the inevitable con- 
fusions between their worn-out forms, rendered neces- 
sary the employment of more varied and distinct 
prepositions. As the terminations implied nothing 
more than the sense given by these free particles, they 
gradually lost their ancient function. Greek, Latin, 
and the Germanic languages reduced their number 
and altered their meaning, and finally modern lan- 
guages are gradually losing all trace of them. German 
alone retains their vestiges without deriving any profit 
from their use. 



CHAPTER III. 
THE INDO-EUROPEAN VERB. 

Office of the verb, expressed or understood — The addition of the per- 
sonal pronoun after the root or theme constitutes the verb — The 
augment, reduplication, the alteration of the root, the insertion of 
auxiliary suffixes between the theme and the termination, express 
tense, mood, voice, and circumstance — Richness and complication 
of the Indo-European conjugation — Verbal evolution of the roots 
hha, kar, vid, svap, &c. — The middle voice — Free auxiliaries tend 
to replace suffixes of tense and mood — Persistence of the termina- 
tions signifying person — Verbal nouns and adjectives — The in- 
finitive and the participles — A nthropism of the noun and the verb. 

The verb, like the noun, expresses a manner of being, 
a quality, a fact, or an action, or, lastly, an abstract idea. 
But to tbe signs of case, that is to say, of gender, 
number, cause, of dependence, place and movement, 
which the declension grafts upon the root, conjugation 
adds person, tense, and mood. When the various 
elements of the sentence are in their respective places, 
ready to play the part assigned to them by the ter- 
minations, but still dumb and immovable, the verb 
intervenes to preside over their evolutions. " It is I," 
it says, indicating the subject, " it is thou, you, he, we, 
or they, who, in such a place or with such an object, 
with such means, endowed with such and such proper- 
ties, urge to-day, urged yesterday, shall or should urge, 
if it were necessary, them, him, us, you, myself, and 
all which in the phrase wears the livery of the accusative, 
of the complement.'' The imperative commands, and 
everything moves to execute the marches and the 



304 The Indo-European Organism. 

counter-marches, the feints, and the retreats, the 
peaceful or violent actions indicated by the verb. 
Thus declension forms, dresses, isolates, groups, sets 
in line and distributes the personages ; it gives the 
form ; the conjugation gives the illusion of life. 

No idiom, even the language of the dog or of the 
toad, can do without tlie verb, expressed or understood ; 
only this rudimentary verb is included in the cry 
itself, or supplied by the immediate action. In the 
monosyllabic phase, in which grammar is yet unde- 
veloped, in wliicli the functions of the words are not 
yet specialised, only the position in the phrase indi- 
cates the quality of miun or verb. Similarly in the 
animal world we find certain organisms in which all 
the fibres or cells accomplish indifferently and in turn 
all those functions which in higher organisms are dis- 
tributed among distinct organs, nutrition, circulation, 
sensibility, locomotion. The agglutinative languages 
have no clear conception of the shades of meaning 
expressed by mood, nor of the distinction, so indis- 
pensable to us, between the past and future ; but they 
have characterised the verb by the more or less close 
association with it of the personal pronouns, either as 
•subject or direct and indirect object. It is one of the 
points in which Basque resembles Chinese. You cannot 
say, " Peter will return the ox to the shepherd," but 
" Peter -he -it -to -him -give later ox to shepherd." 
Another very common artifice of the agglutinative 
dialects, from which Indo-European is not exempt, is 
the addition between the theme and the termination 
of numerous suflaxes, intensive, frequentative, de- 
siderative, causative, inchoative, &p., which are as 
destructive to the facility as to the clearness of a 
language. Yet we must not speak too ill of these 



The Indo-Eii-ropean Verb. 305 

characteristics, because we owe to them not only voice, 
tense, and mood, but also an extreme variety in deri- 
vation. The Indo-European has in the main retained 
from the agglutinative period only acceptable methods ; 
it has almost entirely rejected the incorporation of the 
direct and indirect objects. It has not loaded the 
verb with the whole meaning of the proposition; on the 
contrary, it has charged it with the conduct and' the 
development of the phrase, that is to say, of the thought. 

We shall have to point out certain regrettable con- 
sequences of the personification of words by the noun 
and by the verb ; but in the linguistic order the Indo- 
European conjugation should be considered, as a whole, 
a wonderful construction, an unequalled monument of 
what has been called collective genius, since it is im- 
possible to determine the share of the individual who 
first invented a given form, which, retained by his 
family and clan, was afterwards adopted and modified 
by the neighbouring tribes. 

At the time when the Indo-European dialects began 
to leave their common cradle, the declension was already 
in a decadent state, so worn-out that its weakened 
terminations needed the reinforcement of prepositions ; 
and the majority of the races, far from repairing the 
ruins and filling the gaps, allowed it gradually to fall 
to pieces and be forgotten. The conjugation, on the 
contrary, was, at least as regards the voices, tense, and 
mood, in full development. The personal terminations 
only had suffered, like all terminations, from the con- 
tacts to which final letters are exposed, and also from 
the law of the least effort, which tends to reduce to a 
minimum the signs which are known to all ; but their 
meaning was so clear and their use so convenient that 
they are everywhere maintained (except in the Ger- 



3o6 77/1? Indo-European Organism. 

manic languages, and especially in English), although 
no longer recognisable, together with the six pronouns 
of which they once took the place. The forms of 
mood and tense were multiplied with that wealth, 
that luxuriant variety, which denotes the enthusiasm 
and the vigour of youth : two parallel and often equi- 
valent voices, the active and the middle (or reflected), 
that is to say, two ways of conjugating every form ; 
four pasts and aorists, two or three futures, as many 
conditionals and substantives. The movement, begun 
before the separation of the idioms,, was long continued 
in five or six of the principal branches — in Sanscrit, 
in Zend, in Greek, and in the Letto-Slav languages. 
Instead of eliminating what might have appeared 
superfluous, Sanscrit and Greek cleverly made use of 
it to express the finest shades of tense and mood. 
They assigned slightly difierent functions to the 
aorists and futures which did double or triple duty ; 
they even added some new forms. Greek, without 
abandoning the middle voice, drew from it a passive, 
and transported into each mood the complete series of 
tenses; the optative, the infinitive, the gerundive, even 
the itnperative had futures, aorists, and perfects, like 
the indicative and the. subj unctive. 

It is impossible to admire without reserve this ex- 
cess of grammatical ingenuity, the equally rigorous 
and subtle arrangement of the Greek verb, more pro- 
fitable to refinement than to clearness, and from which 
the language and the thought gain less precision than 
a teasing ambiguity ; it will be permitted us to prefer 
the clear and strong Latin conjugation. The Latins 
chose the useful, and the Romance language retained 
it. But, as if to show that the creative energy was 
not extinct in them, they remodelled that which they 



The Indo-European Verb. 307 

borrowed, contracting and abridging terminations which 
had become useless, and having recourse at times to 
happy combinations of suffixes and incorporated auxili- 
aries to renew their future and their past subjunctive, 
and to create for themselves an original passive. In 
this matter also the Celtic languages are innovators. 
Gothic, it seems, had retained several vestiges of the 
Indo-European verb, but as sporadic exceptions, which 
have not been able to maintain themselves against the 
analytic tendency of the kindred languages, English 
and Piatt- Deutsch. German has almost entirely re- 
nounced the ancient forms, and the ingenious distinction 
which it has established between the present and the 
past, icli hind, ich hand, gebunden, I bind, I bound, 
bound, although it is connected with phonetic pheno- 
mena observed in Sanscrit and Greek, and also in the 
Semitic languages, is a German creation, and the latest 
example of the verbal instinct which is so remarkable 
in the Aryans of India and the Grseco-Latins. 

The Indo-Europeans possessed no other materials for 
the vast and complicated structure of their conjuga- 
tion than the constituent elements of the agglutinative 
languages, attributive roots, and demonstrative roots. 
The former furnished the body of the word and the 
more ancient of the auxiliaries, the verb to be, as. 
The others gave three kinds of suffixes : the copulative 
vowels or syllables, of which the sense is vague or nil ; 
secondly, the determinatives, such as ya, aya, paya, ta, 
sa, which give to the verb an intensive, causative, pas- 
sive, desiderative value; finally, the six terminations 
mi, si, ti, mas, tas, nti, of which the original form is 
unknown, but of which the meaning is certain, since 
they take the place of the six personal pronouns T 
thou, he, we, you, they. Add a termination dhi, Greek 61, 



3o8 , The Indo-European Organism. 

proper to tlie second person of the imperative; the 
contracted reduplication of the attributive root which 
characterises the past tense, the perfect of all the 
verbs, and some forms of the present and imperfect 
indicative ; a prefix signifying distance, in Sanscrit, 
in Greek, peculiar to the past, and we have all the 
elements of the Indo-European verb. 

In its simplest form the verb is composed of the 
root and the termination : Sans, hhami, hhasi, hhati, 
hhamas, hhatha, hhanti ; in Dorian Greek, (pajj-i, (pr^^ 
(pan, (poLfxris, (pare, (pdvri. This root, iha or pha, is 
curious ; its original signification is to shine, to give 
light ; Greek <pdoi, (pUs, (pwrog (whence photograph), 
and also a living one, one who sees the light ; it after- 
wards took the sense of to make clear, to speak, wliich 
Greek has preferred : (pdiTis, (fxavri, (pij-fAi (Latin, fari, 
fatum, fama, famosus, fabula, fahulari, whence the 
Spanish hahlar). But the meaning to give, light is 
still visible in the sufiixed form (palvw, I show, whence 
phenomenon, that which is shown, that which appears ; 
phantom, fantasy, fantastic. 

Among these primitive verbs, in which the termina- 
tion is immediately joined to the naked root, we may 
mention as, to breathe, to be, asmi, assi, esti, est; i, 
to go, imas, itha, yanti ; invm, itis, eunt ; ad, to eat 
(Latin, edd) ; vak, to speak (Latin, voc-s) ; in all, 
about seventy in Sanscrit. 

The reduplicated roots are equally ancient, such 
as (bhar) hibharmi, libhrmas ; {da, to give), dadd-mi, 
SlSoifii ; (dha, to place, to establish) dadha-mi, tithemi, 
which has produced so many words in Greek (theme, 
avMthema, Themis) ; (gan, to engender) ja-jan-mi, 
y'lyvofxai, Latin gigno ; but the Greek and Latin forms 
are already conjugated on a sufiix. 



The Indo-European Verb. 309 

The verbs just quoted shorten the root vowel before 
the plural terminations, and this character, which they 
have in common with the themes in wm, Greek §61kvv/ui.i, 
SeiKvOfies, ranges them in a class apart, which Bopp 
calls the second principal conjugation, and which it 
would be better to call the first ; for it is from this 
conjugation that Greek, for instance, received its most 
primitive verbs, those verbs in mi which were formerly 
regarded as exceptions. 

The other verbs, and they are the immense majority, 
are conjugated on a theme and not upon a root, and 
they insert between the termination and the theme a 
vowel a, variable in Greek and Latin, or a suffix termi- 
nated by a vowel. By this method we obtain such 
forms as hhara-a-mi, contracted into hhardmi, hharasi, 
hharamas, har-ayami, I cause to be done ; vedayami, 
I make known ; svapayami, I cause to sleep ; sthapa- 
yami, I cause to stand ; yuyutsyami, I wish to fight ; 
bubhusyami, I desire to be ; giynasyami, I desire to 
know ; and so on for the immense family of verbs 
derived from nouns, by means of the suffixes ya, sya, 
asya, whence come all the Latin verbs of the first, 
second, and fourth conjugations, amo for amao, moneo for 
moncyo, &c., and in Greek the contracted verbs (piXew, 
Ti/xdci), SriXoia, and the verbs in al^w, i^w, Xw, /aw, voo, pea. 

I pause a moment to show, in the few verbs already 
mentioned, words and formations which are familiar to 
us, and are only hidden by the complicated forms of 
Sanscrit. The reader will have already recognised 
the roots hhar, to bear ; kar, to do ; vid, to see, to 
know ; yna, to know ; stha, to stand ; svap, to sleep ; 
bhu, to grow, to be. Disregarding the various con- 
jugations into which all can enter, we shall find them 
everywhere with their fundamental meaning. 
21 



3IO The Indo-European Organism. 

Bhar and its contracted form hhrd have given in all 
Indo-European languages the name of brother, frater, 
(ppdrwp, hhrdtar, and hhartar ; to Greek and Latin 
their verb fero and its innumerable derivatives, and 
those powerful suffixes (popog, fer, which are still living 
in the Romance languages ; Nicephorus, the bearer of 
victory ; Phosphorus, the bearer of light ; Zucifer, the 
same sense ; Prolifer, the bearer of generations, &c. 

Kr or Icar exists in French only in its derivatives 
(crier, criateur, cirimonie) taken from the Latin ; but 
we find it in many mythological names : Chronos, the 
father of Zeus ; the Keres, or Destinies ; the Latin and 
Gallic Karanus ; the Ceri or genii, and Geres. 

Vid, besides a number of Sanscrit words, such as 
Veda, science, may be recognised in the Greek [f]'^, 
[fji'^^ej/, we know; ej^o?, .form (whence our anthro- 
poid, rhomboid, Sue), iSea ; in the Latin vid-ere (French 
voir, voyant, vision, visible, visiter, and their compounds). 

Gnd has given gnoseo, nosco, nomen ; connaitre, notion, 
nom, note, notaire, &c. 

Stha, stare, slator ; stdbilire (establish), statiie, station, 
stage, state ; it appears in the French participle of the 
verb to be — Mi. 

Svap (in Greek v-ttvo^, in Latin sopor, sopire, somnus 
for sopnus) is concealed in the French somnu, sommeil, 
assoupir (which has no connection with soupir, sub- 
spirium, deep breath). 

Bhu, in Greek and Latin fu, is of capital import- 
ance. The earth, that which causes growth, takes one 
of its names from this root, Sans, bhumi, Latin humus ; 
and very probably homo, humanus, man, the son of 
the earth ; cjiurov, plant, (pvaif, nature, whence physic, 
physiology, physiognomy, and metaphysics, are unques- 
tionably derived from it. The Latin fui, fuissem. 



The Indo-European Verb. 3 1 1 

fidiirus, fore, and their French descendants /e /its, ilfut, 
que je fusse, are tenses of the Indo-European verb hhu. 

Thus all these syllables, which at iirst appear to 
succeed each other at random, a mere babel of empty 
sounds, are our own patrimony; they live for the 
most part in all the languages which are anterior to 
us or contemporary with us; they have constituted for 
thousands of years the foundation of our languages 
and of our ideas. But what has become of all these 
accessory letters, these syllables inserted between the 
theme and the termination which embarrassed the 
ancient conjugation ? Bach language has freely, or 
rather under the guidance of varying circumstances, 
altered, atrophied, fused them, to such a degree that 
hundreds of learned men are employed in tracing 
their metamorphoses. If time had destroyed the last 
vestiges of them, their work would still subsist ; their 
native virtue has not perished, and is revealed in the 
infinite variety of forms and meanings. We should 
not have possessed the words established, stable, if a 
suffix pa, added to the root sta, had not received in 
Latin the suffixes li and re : sta, stop, stapili, stalnlire. 
We are able to form an indefinite number of verbs 
and nouns, such as moralise, modernise, dandyism, 
chemist, linguist, artist, because a suffix ya was com- 
bined in Greek first with themes in id : eknnS, hope ; 
v^piS, insult; iptS, quarrel; then by analogy with all 
sorts of roots, and produced the termination l^w, 
whence Icrna and to-T>?y. 

The personal terminations may also seem to be very 
far removed from the forms which we use ; mi, si, ti, 
mas, &c., appear perhaps to have no relation to our 
personal terminations. Yet we still use them. The 
first person, indeed, has disappeared from Latin, except 



312 The Indo-Ejcropean Organism. 

in sum and inquam. Greek, Lithuanian, and Slav 
have kept it in a few hundred verbs. But when we 
say tu es, il est, noiis sommes, vous etes, Us sont, tu fais, 
il fait, Turns faisons, vous faites, Us font, or again tu lis, 
il lit, Twus lisons, vous lisez, Us lisent, we still employ 
the Indo-European conjugation. We have the ancient 
forms as a faint but faithful echo. The second person 
plural faites, Mes, fltes, is even less obliterated than the 
cojTesponding Sanscrit form. The loss of the final i 
in tu fais, Us font, is not a inodern phenomenon ; it is 
seen first in Latin, often in Greek, and also in Sanscrit 
aud in Zend, everywhere except in the indicative pre- 
sent and in the subjunctive of the conjugation in mi. 

These terminations, which I have shown in the 
most ancient form in which they are known to ns, 
are themselves only relics ; and their growing weak- 
ness was of small importance if enough of them 
remained to mark the order of the persons. Mi, 
which doubtless stood for ma, became, for instance, 
in the optative, m: dbharam, a^yam, Greek echepov, 
e'lr/v. (We have seen that Greek does not admit of a 
final m or t.) Latin, more exact, has such forms as 
eram, siem (sim), duim, (derrC), (cf. dasyam, Swa-et>]v), 
legam, legerem, &c. Si, for sva, thva, tvani, kept only 
the sibilant, abharas, ecpepeg. Ti, except perhaps in 
Greek in the third persons (pepei, tutttsi, losing its 
vowel, gave the almost universal sign of the third 
persons : dat, fm% dicit, il tient, il fait, &c. This t 
reappears in French in such locutions as va-t-il, 
aime-t-il, and as a sibilant in the English he does, for 
the older form he doth (the English th generally cor- 
responds to primitive t). 

The plural mas (for masi, masa, I and he — that is, 
we), faithfully represented by the Lat. mus, is also repre- 



The Indo-European Verb. 3 1 3 

sen ted, in spite of the nasal termination, by the Greek 
(t.ev (Xvofiev, rvirTOfiev) for /we? (^olian, cpepofiLes:), and 
by the French noics aimdmes, nous eHmes. Tas or 
tJias (for dhvas, for twasas, thou and they) does not 
exist in Sanscrit properly speaking (ta, thds), nor in 
Greek, but Latin and French have retained it : legitis, 
essetis ; dites, faites, vous ites. The third person, nti, has 
been atrophied in various fashions ; it is almost un- 
altered in the present Sanscrit, Zend, Lithuanian, and 
even in the Greek ovtL, afterwards elcri, ova-i, wcri, acri ; 
it is still recognisable in the Latin sunt, amant, monent, 
audiunt, &c., in the Gothic bairand (ferunt), they bear, 
and sind, they are ; and in French in the silent termi- 
nation of the words awient, aimaient, aimeraient, no 
longer pronounced except in Normandy {Us ne peuvte 
pas, Us ne veulte pas) ; but the termination nti is gene- 
rally reduced to an n, sometimes suppressed or replaced 
(in Sanscrit) by a vague sound, uh. In Greek the 
letter n has a double use in the imperfect : eXd/i^avov, 
I took, they took ; in the aorist it has disappeared from 
the first person, eXvaa, I loosed. Italian has become 
fixed at the Greek stage, so7io, amano, amino, adding, 
however, an as a support. From these facts I con- 
clude that if the destiny of the suffixes and termina- 
tions excites our wonder, it is far less on account of 
alteration and loss, than of the extreme tenacity of 
these frail particles. The sign of the first person alone 
has almost entirely disappeared, doubtless because it 
was the least useful. Greek, Latin, and the Germanic 
languages, in adopting verbs ending in 0, u, and e have 
only followed the example set by the ancient Indo- 
European. 

In Sanscrit the middle voice has rejected the termi- 
nation of the first person, and, by a singular privilege, its 



314 The Indo-European Organism. 

forms are the more simple as they express in principle 
a more complex idea. This voice, which the Greek 
uses -without attaching a very definite meaning to it, 
and of which the agglutinative character reveals the 
antiquity, was designed to unite, if not the forms, at 
least the values of two pronouns, or rather of the same 
pronoun in the nominative and in the accusative : je 
me plais, je me souviens, tu f habitues, il se parte, il se 
plaint, nous nous concertons. But it has generally lost 
this meaning, and is used indifferently as a doublet of 
the simple form. The middle voice presents, there- 
fore, no sign of mi, and to discover the probable priini- 
tive termination, we must consult the Gr. (pepofiai, 
which leads us with a little trouble to an organic 
hliaramami, I bear myself, or I bear, condensed and 
abbreviated into hhar-S ; while to account for the 
Greek second person, (pepet, we must compare both the 
Sans. Iharase, and the Greek third person (peperat, and 
we may then assume almost with certainty a primitive 
phereai, pheresai, pheresasi. 

The termination of Greek and Latin verbs is 
hardly more embarrassing than this middle voice in 
the long e. Whether it be that mi early ceased to be 
"pronounced in verbs analogous to the Sanscrit krndmi, 
I do ; strndmi, I stretch ; tandmi, kardmi, rndnii, &c. ; 
or that thp usefulness of the developed form being no 
longer felt, the Greeks, and especially the Latins, were 
content to mark its place by a lengthening of the 
vowel of the theme ; at any rate, we may be sure that 
the innovation was the result of very gradual change. 
The transition stages have perished, but a few scattered 
indications such as inguam, possum, and the Sa:?Lon ieom, 
am, throw light on the methods and the progress of 
language. 



The Indo-European Verb. 315 

' One of the characters of the modem verb, with its 
analytic tendency, is the use of auxiliaries, which alone 
take the inflexion and the personal terminations, and 
lend to verbal nouns, such as the participles and the 
infinitive, the notions of tense and mood. These auxi- 
liaries, such as to be, to have, to be able, will, shall, 
should, could, &c., abound in the Germanic and Neo- 
Latin languages ; nor are they quite wanting in such 
synthetic languages as Sanscrit, Greek, and Latin ; it 
will be enough to mention the Sanscrit compound future, 
datarasmi, datasmi, datasi, datasmas, the perfect pas- 
sive in Greek and Latin, XeXi//u^iei/o£ etcri, they have been 
loosed ; amatus sum, I have been loved ; or the de- 
ponent profedi sunt, they have set out ; secutus, locutus 
est, he has followed, he has spoken. But these lan- 
guages have, as a detached auxiliary, only the verb to 
be ; locutions like dicere habeq, I have to say, I shall 
say, observed here and there in Cicero and in St Augus- 
tine, which explain the Provencal future, dir vos ai, and 
the French futures and conditionals faurai, je saurai, 
fe serai, je serais {fai, j'avais, d avoir, cc itre), these 
locutions were not a part of the regular Latin conju- 
gation. But as we look closely into the Indo-Euro- 
pean verbal organisation we si 1 all see that it already 
required the aid of the verb "to be," and of a few 
others, and used them by incorporation, by suflixation, 
by agglutination in its aorists, futures, and passives. 
We have just seen the same method employed in 
French for the formation of its futures and condi- 
tionals. Bopp conjectures with much probability that 
the future dasyati, he will give, is a disused potential 
or future of the verb as : syam, syas, syat, and he com- 
pares with dasyamas, dasyatha, the Lithuanian dusime, 
diisiie, we shall give, you will give. The Dorian 



3i6 The Indo-European Organism. 

■7rpa^L0iJ.ev, we shall do, and the classic Greek Swtrw, 
SwcTOfiev, Select), Sei^ofxev, the old Latin accepso, and with 
a change of the s into an r, accepero, acceperim, belong 
to the same class. There' is a relationship from this 
point of view between the future and the desidera- 
tives like pipasami for pipasjami, I wish to drink. 
The verb substantive contributes also to the formation 
of several preterites : Sans, adiksham, adiJcshi, Lat. dissi, 
scripsi, Gr. eSei^a, eXvcra. 

Another auxiliary, dha, to put, seems to have given 
to Greek its passive aorist and future, eXvOr/i, XvOtja-ofiai; 
to Zend an imperative like yaus-dath-ani, may I purify ; 
to Lithuanian some infinitives : dim-deht, to sound ; 
bai-deht, to frighten (bhi, to fear) ; shJcum-deht, to 
sadden ; to Slav the future hundun, I shall be ; the 
imperative butulemu, may we be ; to the Gothic habai- 
dedum, we had. This root dha (Gr. 6t]) becomes in 
German thum, to do ; in English, do and did. In 
Latin, as in the Low German dialects, it has naturally 
been confused with da, to give, because Latin has no 
aspirates ; it is nevertheless recognisable in subdere, to 
place beneath ; credere, to place in the heart {erad-dha 
in Sanscrit), or to place, as understood, from a root p-u, 
to understand, to believe ; vendere, to offer for sale. 

Finally, in the Latin terminations of the imperfect 
and future bam and bo, we must recognise the auxiliary 
bhu, to grow, to be ; and it has been thought that this 
same verb may be traced in the perfects in m and ui 
for fui : crevit, gemuit, obstupuit. 

These rapid and fragmentary indications leave on 
one side many curious problems, many interesting 
comparisons ; the subject of the verb would not be 
exhausted in a year of methodical study. But I 
must be content if I have convinced the reader that 



The Indo-European Verb. 3 1 7 

the Indo-European tongues proceed from a single 
type, that all descend from it, none return towards it, 
and that, in the thousand variations which these lan- 
guages have introduced into the vocabularies and into 
the grammatical forms, they have worked with the 
same elements and by the same methods as our intel- 
lectual ancestors in primitive times. 

To complete our sketch of the verb, we will pass in 
review certain accessories which are properly adjectives 
and nouns, but which it is right to rank in the cate- 
gory of verbs, since they belong to it by two acquired 
characters ; some by the privilege which they have of 
governing the same cases as the verb ; all by the 
differences of tense, the active and passive value which 
use has given them. These are the infinitives and the 
participles. The substantive origin of the infinitive 
may be seen in many of our expressions : mirabile visu, 
facile d, retenir, le boire et le manger, le Men dire, le 
repentir, le plaisir (Latin plaeere). The relationship of 
the participles with the adjectives is yet more striking. 

The infinitive, in the various languages of the family, 
pi-esents a great variety. All its forms : datum, iroieiv, 
StSovai, Xvcrai, amare, fieri, &c., come from the common 
stock ; but our languages, says M. Michel Br^al, " had 
at their disposal, to form nouns of this description, a 
great number of sufi]xes,and after having long hesitated, 
they only fixed their choice after the separation of the 
idioms. The older the language the more clearly 
defined is the distinction between the verbal noun and 
the substantive proper ; but in the beginning the divid- 
ing line is not clearly marked." In the Vedas words in 
tar, datur, giver, are constructed with the accusative : 
data maghani, dator divitias, " giver of riches." Plautus 
has quid tibi haiic curatio'st rem. ? In the Vedas 



3i8 The Indo-European Organism. 

certain words which afterwards become infinitives are 
declined: Vritrdya hantavS, "for Vritra, for murder;" to 
kill Vritra. In SakuntaU there is an infinitive associated 
with an abstract compound : hdMJcsMpan rodituntcha 
pravritra ; " beginning the opening of the arms and to 
weep," beginning to extend the arms and to weep. 

Bopp gives infinitives in every case, but petrified as 
it were in these cases, such as datum, sthatum, and 
the Latin supines datum and statum, to give, to stand, 
in the neuter accusative ; datu, visu, in the ablative 
or the dative, for visui ; in the instrumental, the ger- 
undives paktvd, attvd, after having cooked, after having 
eaten ; in the locative, grahane, to take (in the grasp) ; 
in the dative, piladhydi, for to drink, for drinking, 
mddayadhydi, to rejoice. 

We will pass to Greek and Latin, which interest us 
most nearly. Nothing seems harder to explain than 
the syllable re : amare, dicere, the source of the French 
infinitives. Yet several explanations have been pro- 
posed, which are more or less satisfactory. First of all, 
the presence of an r between the two vowels indicates, as 
a rule, a primitive s, as in posse, esse, velle for velse, ferre 
for ferse. This is the form of many Sanscrit infinitives : 
dd-se, to give, sthd-se, to stand, vah-se, to drive, givase, 
to live, kchahchase, to see. The Latin form appears 
then to be the usual weakening of an ancient dalive. 
As the long e in Sanscrit always represents a diphthong 
ai, there is, it will be seen, a close analogy between 
posse, for instance, for possai, or ferre for fersal, and 
the Greek infinitive aorists Tvcpcrai, \vaai, XvicrOai, in 
which the terminations differ from that of SiSovai, 
TiOepat, efijuLevat. That is, it is everywhere a case-ending. 
There remains the s of the sufiix ; is it a relic of the 
pronoun, or the remains of the auxiliary " to be " ? 



The Indo-European Verb. 3 1 9 

Assuredly iu the Greek aorist infinitive it is the verb 
as; it is impossible to separate Xi/crat from the parti- 
ciple \\icra<s and from the preterite eXvcra. In Latin it 
is not so easy.to determine, any more than in the case 
of the infinitive of the middle voice, dicier, laudarier, 
audirier. Bopp, guided by the analogy of the forms 
amor for amo-se, amamur for amamus-se, connects the 
termination er in this case with a reflexive pronoun. 
I should rather see in it a relic of the verb esse : to le to 
love, to le to praise, would satisfy me better than to love 
oneself, which suits the sense of the middle voice, it is 
true, but not the far more frequent passive meaning. 

The dative value of ai in the Greek infinitives in vat, 
fj-evat, efifjLevai, by abbreviation efiev, eev, eiv (Germanic 
an, en), appears clearly in such common locutions as 
avdpwTTOS TritpvKe cf>iX>ja-ai, man is born to love ; eSuyKe 
TOUTO (popijaai, he gives that to carry ; ^6e ^ijrijo-ai, 
he came to seek. The syllables or letters which pre- 
cede this ai alone suggest a doubt. What are the eaO 
of ea-Oai, efiev, fxev, the ^olian r/v, the Dorian ev, the 
classic etv ? EcrOai would seem to unite the two 
auxiliaries as and tha. For the other forms, in spite 
of some ingenious conjectures of Bopp, I believe them 
to be one of the most usual sufiixes of the participle : 
a-mana (ofxevo), mdna (XeXuiU-yueVos), ama, which forms 
so many Latin derivatives, Octavianus, Julianus, &c. ; 
the naked forms e/xev, eiv, ev, would be in the accusative, 
and not in the dative. 

These sufiixes ana, mana, lead us to one of the 
most interesting forms of the verbal noun, not only 
because it has furnished to Greek and Sanscrit innu- 
merable participles of the passive and middle voice, 
both present and aorist, which may be found in the 
grammars, but because Latin, though rejecting them 



320 The Indo-European Organism. 

from its conjugation, has retained a few examples as' 
curious as they are rare. Those of us who have had 
to learn our Latin verbs will remember the unex- 
pected forms of the second person plural of the passive, 
amamini, ervdimini. They are now no longer mysteri- 
ous; the singalsbT prcefamino has been discovered. These 
are participles ; the verb "to be" is understood. Greek 
presents exactly the same phenomenon : XeXv/u./uLei'oi, 
TcQufi/xevoi ela-'i, they have been loosed, struck. This 
is not all ; words like terminus, Picumnus, Voliummis, 
Vertumnus, perhaps Tellumo, are believed to be con- 
nected with this class of derivatives, while Pomona, 
Fortuna, Portunus, Kephmus, show the sufBx aTia 
combined with the vowel of the theme. 

Past participles in ta are very numerous in Sanscrit, 
in Latin, and in almost all the languages ; in Greek 
ttiase suffixes have only formed adjectives, which are, 
however, similar to participles. There is a complete 
pp.rallelism hQtwe&u gndtas, yvmto^.gnotus, known ; datas 
(Zend datd), Sotos, datus, given ; yvJdas, l^evKTOi, junc- 
tus, yoked, joined ; grutas, /cXdto'?, chdus, heard ; strtas, 
crrpaTOi, stratus, stretched. In Lithuanian the Sans- 
crit ta reappears without any change : sektas (Lat. 
secutus), junldas (j'unctiis), Jceptas (Trewro^, coctus). Slav, 
though it does use tu and ta, prefers he, la : bulu, hula, 
hulo, where Sanscrit hfis bhutas, bhuta, bhutam, Greek 
(puTov. Elsewhere it is da or d, or na, n, en, which 
fills the same office. The different simple sufiixes, 
combined with others, already compound and atrophied, 
have furnished a variety of terminations which we 
rarely find pure except in Sanscrit : viant, vant, mat, 
vat. They are easily seen in the familiar forms 
amans, amctntis,amandus, Xvaiv, Xuto<tov, XJcra?, XeXvKw^, 
XvOev, XvdevTOS, and are analogous to the English 



The Indo-European Verb. 321 

present participle and verbal nouns in ing, loving, saying, 
where the g is but a nasal reinforcement, and to the 
German participles in and, end. French owes to this 
compound suffix, first, all its present participles, and 
secondly, a great number of words such as pi'ovidence, 
circonstance, doquence, presence, and many others formed 
by analogy, which had already in Latin passed from 
plural neuter participles to be abstract feminine sub- 
stantives. 

We owe not less to a suffix of the future participle, 
a suffix which Latin appropriated, but of which the 
origin is very ancient ; I mean that root tar, perhaps 
tvar, then tvara, turns, which appears in the word 
katvar, katour, quatour ; which has Icsit its t in the 
I'reton fewar and the English /owr, and its semi-vowel 
in the innumerable names of agents, Sanscrit and 
Greek, tvashtar, daiar, dhatar, ganitar, matar, ^w-rrip, 
yeverrip, firiTrip, all words which are known to us. 
Often when Latin has dropped a primitive v it has 
vocalised it into an or a u, e.g., svapnas, somnus ; 
^vagura, socer, svasri, soror ; akvas, equus. It is this 
practice wliich has given to Latin so many words in 
tor, victor, actor, spectator, domitor, dormitor, so many 
nouns and adjectives in torius, toria, and to the French 
so many words in teur and eur, in toire and oir or aire : 
lutteur, domptew, chanteur, empereur for empereor (im- 
peratorem), &c., coviptoir, miroir, histoire, icritoire, 
nnangeoire (manducatoria). The verb has obtained 
from this same suffix tvar, tor, the fine participle in 
turns, amaturus, profecturus, adepturus, of which the 
feminine, junctura, fractura, candidatura, profectura, 
struciura, may still be found on every page of our 
dictionaries. When no longer part of the verb, through 
which it has nevertheless come, this termination tura. 



32 2 The Indo-European Organism. 

ture, has lost the special sense which it bore as part of 
the conjugation, except perhaps in natiira, that which 
shall be, may be born. It has resumed the vague and 
general signification of the old suffix tvar or tar, which 
has also put on other forms and adopted other functions; 
we have seen it as the sign of the comparative, repo^ 
in Greek, taras in Sanscrit, ter in Latin, and in the 
Germanic languages. In its contracted form, tra, iron, ■ 
trum, it has produced names of instruments, rostrum, 
cultrum, aratrum, claustrum, and derived verbs, mons- 
trare, claustrare, intrare, penetrare, administrare, and 
again new verbal nouns, claustration, penetration, admi- 
nistration. 

As the reader will have noticed, in the Indo-Euro- 
pean languages, roots and suffixes, nouns and verbs, 
have continually lent each other mutual aid, engender- 
ing the one the other with marvellous ease, plasticity, 
and variety of forms, all issued, like themselves, from 
the first stammering speech of humanity. The history 
of the verb alone traverses all the phases of the lin- 
guistic circle. The naked root is in the monosyllabic, 
stage ; agglutination brings to it the declinable or con- 
jugable theme, and reinforces it with suffixes, intensive, 
causative, desiderative, and lastly personal. To the in- 
flected stage belong the fusion of these various elements, 
the choice of the shades which correspond to the present 
or the future, to the expression of , mood and voice. 
Finally, the analytical stage drops what has become 
superfluous, throws light on forms obscured by con- 
tractions and by the wear and tear of- centuries, and 
replaces them by independent auxiliaries, which are at 
once more convenient and more precise. 

We have said how much this elaboration was due 
to the progressive development of the intelligence, of 



The Indo-European Verb. 323 

industry, of the arts and social institations, and pointed 
out also what a constant aid it has been to thought. 
More than once, however, we have needed to put the 
reader on his guard against a too entire admiration for 
the alliance, for the unfortunately indissoluble cor- 
respondence, between the idea and the sound which 
represents it, between reason and its necessary instru- 
ment. Language was born before science ; the elements 
which compose it date from an age when rudimentary 
observation was controlled by no experience ; objects 
were named, classed, compared, by means of unreason- 
ing and erroneous analogies and metaphors ; as reason 
developed, it could only express the slowly acquired 
truths by using vicious locutions, rectified indeed, and 
amended in the measure of the possible by the intro- 
duction of new meanings and of less inexact figures. 
But it could not remodel e/ery day (nor was such a 
thing ever dreamed of), the archaic tool of which the 
use had been the necessary condition of progress. Hence 
inevitable confusion and delay in the clear conception 
of things, in the succession of ideas, which were falsi- 
fied by the signs which represented them. To this 
imperfection of his speech man adds a vice inherent in 
his very nature. Man refers all to himself, and creates 
all in his image. This anthropism, which lends to 
things and beings, real or imaginary, a human person- 
ality and purpose — ^this illusion is mingled with the 
first attempts of language and has survived them. 

The noun and the verb have communicated to ob- 
jects a truly human existence and activity. Not only 
those objects determined by a quasi-individual form, 
animals, plants, stones, stars, but even the place and 
the aspects of the place where they occur, then the 
categories in which they are classed by reason, relations. 



324 The Indo-European Orgamsm. 

and ideas, are endowed with sex, gender, life, and by 
the verb, act after the manner of man and woman, 
moving, rising, going to bed, trembling, running, loving, 
reproducing their kind ; even entities — that is to say, 
qualities considered as apart from their real subject — 
light, heat, fecundity, beauty, pleasure, pain, vice and 
virtue, good and evil, took a personal existence, became 
the subject or the object of propositions implying action 
or will. It was forgotten that these words express 
only states, durable or ephemeral, of hot or light- 
giving bodies, and the resultants of particular organisms; 
men saw in them the pre-existing cause of facts of 
which they are but the general or analytic expression ; 
they invented virtualities, forces, illusory powers, which 
have no other origin than instinctive anthropism 
and the metaphorical essence of language. Numina, 
nomina. The noun created the gods ; the verb de- 
veloped myths. 



CHAPTER IV. 
THE COMPOUNDS— THE INDECLINABLE WORDS. 

Ancient character of compmnd words — Rarity of the declensinn of 
the first term in a. compound — Determinative compounds ; com- 
poimds of dependence ; possessive compounds : examples — Verbal 
compounds in French — Piirticlea, conjunctions, prepositions — 
Placed after the theme, they have produced the declension ; placed 
before, they vary nd infinitum the meaning of verbs and nouns ; 
free, they are the ligaments of the sentence, and supply advan- 
tageously forgotten terminations — Notes on the indeclinables sa, 
saha, sama ; abhi, afupi, and — Original or acquired diversity of the 
vocal organs. 

In our rapid sketch of the grammatical forms, we have 
never lost sight of the distant origin of the elements, 
either primitive, or the result of numerous contrac- 
tions, which agglutination has combined into suffixed 
themes, and inflexion has fused into declined and conju- 
gated words, into nouns and verbs. Tliis succession of 
the stages of language is more especially evident iu 
a large class of terms called compound words. The 
framing of compound words is a familiar expedient of 
the isolating idioms, applied to themes already formed 
by agglutination, and finally polished by grammar. 
We are speaking, be it noted, of the most ancient and 
most correct of these compounds ; analogy, local usage, 
individual fancy have here, as everywhere else, played 
their part, and disturbed an order which the rigorists 
among philologists believe they can trace among the 
smallest accidents of language. 

22 3=5 



326 The Indo-Eu?opeatt Organism. 

" Tlie characteristic of the true compound," says M. 
Br^al, " is the union of two terms of which the first 
has no case-ending," even though it is in close 
grammatical relation, either of subject or object, of 
adjective or substantive, with the second. This ab- 
sence of the sign of case justifies us in thinking that 
the formation of the word dates from an age anterior 
to grammar. Similar juxtapositions abound in Chinese, 
Japanese, and Malay. But when the declension was 
established it naturally affected the compounds. Some- 
times it even penetrated to the first member of the 
double word, but usually the compound was regarded 
as a single whole, lacking only the termination which 
marked its value in the phrase. 

The Latin compound pronouns show some traces of 
declension in the first term ; thus we find eapse as 
well as ipsa for the feminine of ipse. But the chief 
example is furnished by certain Sanscrit copulatives 
called compound dvandvas, such as Mitra and Varuna, 
heaven and earth, night and day, in which the termina- 
tion of the dual affects both nouns : Mitrd- Varundu, 
Agni-Somdu (Agni and Soma), Indrd- Varundu, Dyavd- 
Prithivt, Pitard-Mdtardu, the iather and mother; it 
even happens that one of the two names is not ex- 
pressed, and that MitrS, alone should be understood 
to mean Mitra and Varuna. These constructions are 
peculiar to the Vedas. Tlie very rare compounds of 
this nature in Latin and Greek inflect only the second 
term : the adjectives, Xei//co-yueXas, black and white, 
sacro-sanctus, sacred and holy, and the substantive 
siMvitaurilia, the sacrifice oF a pig, a sheep, and a bull. 
Perhaps, also, we may compare the formation of the 
words CrTOBCo- Roman, Austro- Hungarian, saltpetre, beet- 
root, aigre-doux, douce-amdre, clair-dbscur. 



Compounds. — Indeclinable Words. 327 

The most interesting classes of compounds are the 
determinative compounds, the compounds of depen- 
dence and the compounds of possession. In the first 
the two terms are related to each other as a noun, 
an adjective, or a verb is related to -an adverb or an 
epithet. Sanscrit : maha-hula, great family ; sat-suta, 
good son ; ghana-cydma, black as a cloud. Greek : 
avSponrati, child-man, child who shows the courage of 
a man ; KaKoirdpOevos, unhappy girl ; KaKoSalfiwv, evil 
genius ; Acropolis, the high town. Slav : Bielhog, 
Cernoiog, white god, black god. In Latin : decemvir, 
semideus, peninsula, primordium (primum ordium), 
leneficium-, henevolus, semijustus, altitonans. French 
can also form compounds of this nature : saupoudrer, 
colporter,. maintenir, pjimevdre, vif-argent, printemps, 
auMpine, sat if -conduit, sauve- garde, hanlieue ; but 
especially by the aid of adverbs and prepositions : 
hienveillance, Menfaisance, malapprls, demi-heure, mi- 
clos, milieu, minuit, midi, contre-indication, surf ait, 
surencMre, surtaxe, surhumain. But they abound 
especially in the Germanic languages. Gothic : junga- 
lauths, young man ; langamodei, from mods, now muth, 
longanimity. German : Vollmond, full moon ; Gross- 
that, noble deed ; Wundermann, wonderful man ; 
kohlschwartz, coal-black ; bildschon, fair as a picture ; 
schneeweiss, snow-white; spiegelhdl, sonnenhell, silberklar, 
clear as a mirror, as sunshine, as silver. 

In the compounds of dependence, the first term is 
governed directly or indirectly by the second. Sanscrit: 
Brahmavit, be who knows Brahma ; graddhap4ta, puri- 
fied by faith ; pitrisadriga, like his father ; naustha, 
which is contained in the vessel ; Greek : /jlovo/xu^os, 
he who fights single-handed ; avSpo^pws, avSpoSoKOs, 
who devours, who welcomes men, ropvevToXvpaa-iriSo- 



328 The Indo-Europ£an Organism. 

TTijyoi, turner of lyres and shields ; TroSavnrr^p, foot- 
bath ; TToSwKrii, light-footed ; av§paSe\(p6g, husband's 
brother ; avSpaKairriKog, seller of men ; avSpoTre7r^9, 
proper for man ; otKocpvXa^, guardian of the house ; 
apyupcovriTO?, bought with a price ; cppevoXvTTOS, 
possessed in mind (mad) ; aKavOo-rrXn^, wounded by a 
thorn ; 6Xivr]KTrig, sea-swimmer ; vcorocpopoi;, bearing 
on the back ; aippoyevi^i, foam -born ; SetirvoKXijTccp, 
he who invites to a feast. Most proper names in 
Sanscrit, Persian, and Greek are compounds of this 
class : Hippolytus, llipparchus, Hippocles, Hippodamas, 
the looser of horses, the ruler of horses, famous for 
horses, the tamer of horses. 

Latin makes great use of this artifice : remcx, judex, 
pontiff carnifex, aurifex, aurifaber, aurifar, ignivomus, 
carnivorus, fiammiger, opifer, dapifer, haruspex, augur, 
cmlicola, muricida, herhigradus. Such words have a 
double value ; they show us attributive roots changed 
into true suffixes, and preserve for us, under their 
simplest form, roots which are often no longer found 
in an independent state. Fur exists in the sense of 
thief, AwA faber with the sense of smith, but it is not 
so with ex, he who pushes ; spex, he who looks ; dex, 
he who indicates or who says (a syllable which is 
almost effaced in the French word ju^e), or with fex, 
he who makes, with cola, he who cultivates or inhabits, 
or gur, who tries {au-gur, who consults birds ; Sans. 
djush, Gr. yevu), Goth. Jciusan, to choose), the root of a 
disused verb, gusere, gurere, and of the substantive 
gustus ; or finally with vortts, vomus, gradus, fcr, ger, 
- hj means of which Latin can form an indefinite 
number of . substantives and adjectives. Modern 
French has a certain number of words of this class, 
and uses with freedom the suffixes cole, vome, vore. 



Compounds. — Indeclinable Words. 329 

grade, cide, gire, and fire; e.g. agricole, ignivome, fumi- 
vore, centigrade, homicide, Iwnigdre, conifire.. The Ger- 
mans compose any number of words like Finger-gold 
(ring), Opfer-tisch, offering- table (altar), Eichter-stuhl, 
judge's seat, rothgesclilafen, in which the construc- 
tion is reversed, " who has slept red." ' German and, in 
a less degree, English, have the power of juxtaposing 
as many words as can be pronounced in a breath, and 
providing them with a common suffix, to form com- 
pounds similar to those of the Esquimaux or the 
Algonquins ; in this they do but follow the example 
of classical Sanscrit, in which the compounds form true 
propositions without verb or grammatical sign. 

Among, the words of which we have mentioned a 
few, just as they occurred to tlie memory, the majority 
can be easily turned into compounds of possession ; 
they only require to be made to agree in gender, 
number, and person with a substantive. The termina- 
tion attributes to a given subject the qualities included 
in the compound ; nigahegaqmagrunakhds, he who has 
short nails, hair, and beard. It is a very concise, 
mode of expression: ^avOoKOfios, ^avOodpi^, fair-haired ; 
Kvvocppwv, with the soul of a dog ; ^ovKe(pa\o9, with 
the forehead of an ox {Front-de-hoeuf) \ aeWoTro.Sijs, 
aeWoTTOug, with the feet of the tempest ; avSpo^ovXos, 
wise as a man ; ato\6/j.opcl)o?, KopvdaioXos, of changeful 
form, helmet with varied reflections ; eTTTacrTO/uos, 
seven-gated ; deoTrvpog, filled with divine fire ; 6e6- 
crocbos, OeocpiXo^, OeoSuipos, OeoSoTOS ; fxeyadufios, great- 
hearted ; fieyuKXeij^, much i-enowned ; XevKOWTepos, 
white-winged; /Sa^JcrTejOi'Of, deep-chested; rawVeTrXo?, 
richly clad. The first term may of course be a pro- 
noun, svayamprcibhas, shining by its own light, 
avToSlSoKTOi, self-taught ; or a numeral, tritchakra. 



330 The Indo-European Organism. 

three-wheeled, TerpaKVKXoi, four-wheeled; or an ad- 
verb or preposition, tathavidhas, so made; saddgatis, 
always moving, rapid ; ae'iKapirog, aenraOrii, ever fer- 
tile, ever suffering; amala, abala, abhaya, stainless, 
strengthless, fearless ; " a-n-ais, airovs, a(j)o^09, avoiKO<;, 
childless, footless, fearless, homeless. We shall return 
to this class of words when we treat of prepositions. 
Possessive compounds are not rare in Latin : misericors, 
lidens, Ufrons, hicorpor, tripedorus, multigenus, multi- 
formis, magnanimus, loncfanimus, alipes, loiigipes.fissipes, 
anguipes, qiiadrupes, acupedius, versicolor, piidoricolor, 
pulcrico7tius, grandiloquus, &c. The Germanic lan- 
guages form these compounds with extreme facility 
by means of the suflSx ig, equivalent to the Gr. j/co?, 
to the Sans. Jca ; hochherzig, high-hearted, i.e. proud, 
rothhaarig, red-haired ; formerly it was hatih-hairts, 
roth-haars, and even now in the case of nicknames 
the terminations are suppressed ; lang-ohr, long-ear ; 
dicJc-kopf, thick- head, schwartz-hopf, black-cap ; roth- 
hals, red-neck, roth-kehlchen, red-breast. 

French still forms, if not possessive compounds, at 
least possessive combinations, rouge-gorge, chauve-souris, 
rouge-bord, pied-hot, higueule ; they exist in hundreds as 
proper names, Testemdde, GrossetMe, Francmur, Lortge- 
pied, Blanchecotte, Barberousse, Barbaroux, Bonvin, 
Gmndval, Gharmolue, Malapert, Mauvoisin, &c. Tet 
these are but the remains of a faculty which is almost 
extinct among us. Latin, though rich in compounds 
of all sorts, already preferred derivation ; where Greek 
has (piXoyvveia, a\o')(0(pi\os, 6>]pioiJ.d')(09, Latin has 
suffixes, mulierositas, uxorius, bestiarius. The Romance 
languages have followed this tendency, and the more 
easily that in losing the declension they have lost the 
notion of root, theme, and termination. 



Compounds. — Indeclinable Words. 331 

It would, however, be unjust to omit an original 
form of compound substantives of which French makes 
considerable use ; the first term is in reality the third 
person singular of the indicative present, but so to 
speak neutralised, and playing the part of an inde- 
clinable theme : vaurien, faineant, apj)ui-main, passa- 
vant (a sailing term), passe -montagne, passe-pied, 
passe-poil, passe-temps, passe-droit, chauffe-doux, chausse- 
pied, ronge-lard, ronge-maille, pince-maille, vide-gousset, 
v'ide-poche, coupe-chovx, coupe-jarret, ampe-t&te, gdte-sauce, 
pousse-caillou, couvre-joint, couvre-chef, hrise-tout, hrise- 
lames, hache-paille, mde-tout, songe-crctuE, iire-hottc, 
tourne-vis, tourne-iroche, tire-halle, tire-iouclion , gratte- 
papier, porte-halle, porte-pluine, porle-coton, serre-file, 
serre-tik, garde-fou, garde-feu, perce-pierre, perce-neige, 
rahat-joie, tranche-lard, grippe-sous, essuie-main, passe- 
port, guide-due. If we forget for a moment the verbal 
origin of the first term to consider 'the interior order 
of these compounds, we note first of all that they differ 
from the ordinary type in that the second term is 
governed by the first ; but this irregularity cannot be 
called an exception. Sanscrit, Zend, and especially 
Greek, sometimes reverse the order of the terms ; 
vidad-vasu, kshayad-vira, varedat-gaetha, finding riches, 
killing warriors, making the world prosper, are con- 
structed like tue-mouches and hrule-tout. In Greek it 
will be sufficient to contrast ^iXodeog, ^iXoXoyo?, 
$iXoXa09, ^iXqSri/xos, AiapoOeoi, NiKOcrTjOaToy, ttoio- 
vofioi, cpvyofj.a'^fof, (pepeKapiro^, with Qe6(pi\oi, 0e6- 
ScDooi, Ai]fxo(j)i\oi, vofioTTOios, (TTparoviKOi, Kapiroipopoi. 
Greek uses composition with full license ; it is never 
weary of varying the form and the meaning of words, 
and of creating a verbal wealth which is sometimes 
equivocal and superfluous. Its fecundity, hardly 



332 The Indo-European Organism. 

equalled by that of German, excites more admiration 
than envy, and the advantages of the facility of com- 
pounding words cannot hide its defects, the ambiguity 
of the phrase, and the monotony of the style. 

The power of framing compoundsis, in the analytic 
stage of language, a survival which must not be' under- 
valued, but neither must it be abused. This is why -we 
have placed it after the noun and the verb, as a method , 
of word-formation which is complementary to the others. 

I pass now to the particles known as invariable, the 
conjunctions and prepositions, which have regulated — 
the word is not too strong — the evolution of Indo- 
European grammar ; they furnished it in the first 
instance with the elements which formed the de- 
clension and the conjugation, then the shades which 
differentiate meanings, then the links and copulatives 
which join together the parts of speech and frame 
the; logic of the sentence; and, lastly, they survive 
the combinations which they engendered and end by 
superseding. 

• They constructed, and they have in the end de- 
stroyed, the synthetic system. Though employed in 
the machinery, yet they kept their independence, and 
have used it to ruin the edifice which they had built 
up. For what are these particles ? Nothing else but 
pronominal and demonstrative roots, tjje first vague 
i;idications of speech, which, by the aid of gesture and 
intonation, signified distance, movement, number, and 
the identity of persons and things not already speci- 
fied by the attribution of certain qualities. 

We have already considered some of them, either 
simple, like a, i, sa, ta, ya, ma, ga, bhi, or aggluti- 
nated, such as ana, ima, tava, sva, sma, sya, and noted 
how. they take the value of demonstratives and of 



Compounds. — Indeclinable Words. 333 

personal pronouns, and, joined together, are declined ; 
others, still indefinite in meaning, often identical with 
the first in sound, await a use. From these the 
attributive roots borrowed those extra vowels, those 
incorporated syllables which lengthen and vary the 
forms, changing, e.g., gan into gnd, ihar into bhrd, 
man into mnd, yu into yud, yug, yung, tu into tvd, tup, 
tuk, tund, &c., gradually furnishing the naked root 
with thematic letters, and with intensive, causative, 
temporal, and modal suffixes. From the declined 
pronouns and demonstratives the nouns and the verbs 
borrow their terminations of person and of case.' 
This is not all ; tbe syllables which have remained 
free, sometimes taking rudimentary case-endings, 
acquire a more precise meaniiig, pf movement, place, 
distance, of relative position ; they gave to the adjec- 
tives their suffixes of comparison, and to nouns and 
verbs those numerous prefixes which give to form 
and meaning ever new varieties : incipere, decipere, 
concipere, accipere, excipere, prcecipere, suscipere, inter- 
cipere (capere) ; initium, prodire, proire, adire, abire, 
exire, suhire, circuire, amhire, amhitio, coire, dbire, perire, 
interire (ire). Multiply these ten or twelve variants 
by five hundred, and you will be able to judge how 
much onr vocabularies, and our thought itself, owe 
to prefixation. 

It might have seemed that as prepositions and as 
postpositions the pronominal roots had rendered all 
the service of which they were capable ; that their 
independent existence was no longer of any use; that 
case and person marked sufficiently the relations of 
words to each other ; that suffixation and prefixation 
had reabsorbed, as it were, the free particles, as mag- 
netised bodies attract iron filings. But it was not so ; 



334 T^^ Indo-European Organism. 

the force of the particles was by no means exhausfced.- 
Neither pronouns nor demonstrative terms perished, 
nor the little words. Syllables almost devoid of 
meaning, expletives, enclitics, continued to have their 
place in the sentence, as if to lighten it and render it 
more supple. We know how they abound in Homer, 
in classic Greek and Sanscrit, in German and English ; 
the reader is wearied by these small parasites, /or, 
then, how, &c., by the prepositions detached from the 
verb which encumber the Iliad, the Mahahhdrata, and 
the best writers of Germany and England. But 
excess does not condemn use ; and the persistence of 
the demonstratives has been the necessary condition, 
as we said just now, of the passage of languages into 
the analytic stage ; they inherited the office of the 
disused terminations, and a function simplified by the 
work they had already accomplished, that of binding 
together the different members of the phrase, at length 
freed from the leading-strings of declension. 

" When we take," says M. Michel Br^al, " the 
ancient prepositions of our family of languages, we 
note a remarkable similarity of form, with a consider- 
able divergence of signification." Often the same 
word will be an adverb or a preposition in one lan- 
guage, and a conjunction in another, or even merely a 
prefix. Apa, airo, ah, af, of, have pretty much the 
same meaning, but anti becomes und, and and. Ati, 
on, above, in Sanscrit, becomes 'in again in Greek ; it 
is the Latin and French et ; Gothic ath, but ; Slav, at, 
anew. Often, too, the sense varies within the limits 
of the same language. In Sans, api (Gr. e7r/) means 
now towards, against, now also, now however. In Latin 
cum is both conjunction and preposition. This is not 
surprising when we consider how vague was the original 



Compounds. — Indeclinable Words. 335 

meaning of these particles, how completely the dis- 
tinction, now so usefully determined, between the 
adverb, the preposition, and tlie conjunction, were 
unknown to the unconscious inventors of language, 
how the pronouns that are now used as relatives were 
employed for every use ; and finally, that it is the 
position of the indicative syllables between declined 
words, subjects and objects, verbs and substantives, 
between coupled or contrasted propositions, between 
principal or dependent sentences, which has sooner or 
later, before, and especially after, the separation of the 
idioms, determined the fate and the use of these incon- 
stant auxiliaries. There are a few, however, and these 
among the most primitive, of which the meaning has 
never varied, such as a, privative ; 'ipra, forward ; and 
Ahea or d/oa, the origin of the numeral two, of the pro- 
noun thou, of the duplicatives dis and his (twice), and 
of the separatives vi in Sanscrit and ^t? in Greek. 
Many others do not belong to these original roots ; they 
are forms rejected by the declension of nouns or pro- 
nouns, ablatives, locatives, datives, without a use, they 
have come to swell the number of indeclinable words. 
We have already mentioned a few of this kind ; the 
greater number of adverbs and conjunctions have been 
formed in this manner. They continue to bear, of 
course, the meaning they had as part of a declension. 
Thus in Greek words like irpiv for irpiov, like aWa, 
neuter of aXXoy, mean necessarily before and otherwise 
or lut ; t]v, neuter of the Homeric (yi'y, gentle {vasu, 
good, in Sanscrit), could only mean well ; in Latin hodie, 
diu, noctu, explain themselves ; circum, circa, idcirco, 
cannot be detached from circMS (doubtless allied to 
kvkXos, Sans. tchaJcra, wheel). But these are indeclin- 
ables of secondary or tertiary formation, 



336 The Indo-European Organism. 

We will give a more detailed study to a few older 
and more obscure words, and follow up their changes 
in form and meaning. 

Sa, which we know as a demonstrative, and as the 
origin of the termination of the nominative case, is 
prefixed to adjectives as if to reinforce them, to indi- 
cate their intimate relation with the noun to which 
they belong. In Sanscrit a married woman is sa-dhava, 
provided with a husband, in opposition to vi-dhava, 
widow, deprived of a husband. From this the transi- 
tion to the sense of with is easy ; sa-k6]pa, vi\.t\i anger, 
as opposed to a-lcopa, without anger ; and further to 
the implication of a common character : sa-varna, of 
the same colour or caste ; sa-vaj'as, of the same age, 
contemporary ; sa-udara, of the same breast, brother 
(Lat. sodalis ?) ; sa-gotra-, of the same stable, same 
family ; sarupa, samana, of the same form, same mea- 
sure. Sa is not found in this sense when standing 
alone, because the use of its homophone, the demon- 
strative sa, would have involved perpetual confusions ; 
hence it has been combined with other syllables, e.g., 
ma, ha, and we have the prepositions sama (sam) and 
saha, with. 

But to follow up our inquiries upon the naked root. 
Zend, which aspirates an initial s, serves as the tran- 
sition to Greek. I will give only two examples : 
ha-zaodha; ■who has the same will, who is in agreement 
with any one ; ha-merena, a dying together, a battle. 
' In Greek sa' becomes an aspirate, and according to 
Herodian the hard breathing existed in adpoos, crowded, 
but it has disappeared except in dfAU, o/xo, for sama, 
together ; Hamadryades, the sylvan comrades, the wood 
riymphs. All trace of the aspirate which is equivalent 
to the primitive s has disappeared, but the meaning 



Compounds. — Indeclinable Woi'ds. 2)o7 

remains in oKo-jfo^, aKoiTrj^, uKotrig, bed-fellow, hnsband, 
wife ; aSeXcpeoi, aSeXipos (Sans, sagarhhja), of the same 
breast, brother ; a-yaXa/cre?, who have the same milk ; 
aire^o?, on the same level, equal ; araXai/TO?, of the 
same weight ; o takes the place of a in OTrarwp, who 
has the same father ; otv^, under a common yoke ; 
odpi^, of the same hair or dress. 

The compound forms, which are more persistent, 
oiFer more interest ; they can be found independent. 
Sama {ma is the suffix of the superlative) is in San- 
scrit an adjective, and signifies m?ne ; as a substantive 
it means all : detached cases, sama, samaj/d, saman, 
may be translated with, together, equallj/, entirely. 
In Gr. biJ.0 and ofio? correspond letter for letter with 
sama, samas : OfMOCppoov, 6/ji.oSe\<p6g, ofiOKonK, Ojuo^v^, 
of the same thought, same mother, same bed, same 
yoke ; and the independent forms ofioO, ofi)}, ofioOev, 
6/jiop-e, ofMuis, in the same manner, on the same side, 
towards the same place. The word is famous in Church 
history : Homoousion, Homoiousion, Are the Father and 
the Son of the same substance ? Is the Son of a sub- 
stance similar to that of the Father ? How much blood 
and ink were expended over this subtle distinction ! 

Similis has taken the place of sama in Latin com- 
position ; yet we still find in Plautus simitu, at the 
same time, and on an inscription simitus (compare noctu 
and subtus). Another form is simulis, whence simul, 
together ; simulare, to imitate, and simultas, doublet 
of similita^. The inscriptions give also semuJ, semol, 
and Plautus uses semel also in the sense of " at the same 
time" (Aulul. iv. 3, i): radebat pedibus terram et voce 
crucibat sua, " (the crow) at the same time raked the 
earth with its feet, and tortured the ears with its 
croaking." Semper, continuously, always. From at once 



338 The Indo-European Organism. 

the idea passes to all at once, once only. Simplex, 
that which has only one fold. From the idea of simili- 
tude, of community, we pass to the idea of nearly, half ; 
sami, hemi, semi; semivir, sinciput (semi-caput). Forms 
like se7n and sim bring us back to the Sanscrit prefix 
sam, which is either the accusative of sa or the abbre- 
viation of sama : sam-prati, now ; sam-'iidra, the gather- 
ing of the waters, the sea; sam -gam (samgatchati), to 
gather together, to assemble ; samhitd, collection of 
hymns ; santchaya, heap, multitude ; sandhi, bond, 
euphony ; samantdt, from all sides ; sanfamas, com- 
plete obscurity. Note that sam corresponds closely to 
C7im in Latin ; the latter bears the same relation to the 
relative Jca, qui, as sawi to the demonstrative sa. 

In Zend ham or hdm, together, which is sometimes 
found alone, corresponds exactly to the Sanscrit sam. 
A comparison with the Greek occurs naturally to the 
mind, and in spite of certain difficulties I am inclined 
to identify the two forms. Initial s does not always 
disappear in Greek ; moreover, it may come here from 
a primitive t, as the root sa alternates with ta in the 
declension. I may mention the double forms tv and av, 
Selene and Helene, Selloi and Hellenes (the Hellenes of 
B pirns called themselves Selles), Ss, crw, the wild boar. 
The Attic and Ionian ^vv points to an effort to preserve 
the primitive sibilant. The Sanscrit sam and the Greek 
avv have exactly the same office as prefixes : crJju/ia^os', 
companion of the strife ; a-vfi^alvw, to walk together. 
As an isolated preposition, since the final m is rejected 
in Greek, <tvv could only terminate in n. 

The Germanic languages have made a considerable 
use of the forms sama, sum, som, and sam. Gothic, so- 
sama, so-samo, ihat-samM,the same; samakMns(pnoyevriis), 
of the same family ; samaleiks, of the same body, the 



Compounds. — Indeclinable Words. 339 

same aspect, similar. Old High German : den samuni 
(eumdem, the same), der selps samo, the self-same ; 
sama/iA, like; savianlcunft, assembly; samana, together; 
zu-sam-ana {zusammen), together; samanon (sammehi), 
to assemble ; saman, with ; niit saman iu, together with 
you. Old Norse declines samr, som, savit ; English 
same, some (Anglo-Saxon sv,me, Gothic sums, suma 
sumata). With a dental suffix (neuter accusative), 
the Gothic samalh (samathgegangan, samathrinnan, to 
run together or towards a same place) recurs in Old 
High German, samant, samet, samt, and in German, 
sammt, sammtlich. Sammt is sometimes an adverb, 
and sometimes a preposition governing the dative. 
Tbis is very common ; the preposition being only an 
adverb endowed with transitive force. 

Sam, despoiled of almost all meaning, has furnished 
the Germanic languages with an adjectival and verbal 
suffix. Old High German, anc-sam, anxious, arpeit- 
sam, painful (arbeit-sam, laborious), leid-sam, loath- 
some ; modern High German, muot-sam, courageous, 
gruoz-sam, horrible, lob-sam, honourable. In modern 
German these forms are multiplied, and sam may be 
joined to all sorts of nouns, and even to verbs. 

The combination of sd with dha and Tia, with tra, 
Tea, or tcha, with na, leads to similar developments ; 
but it then takes, as a rule, the sense already noted in 
semper, always, universal, eternal ; German, sinngriln, 
evergreen ; Anglo-Saxon, sinbirnende, ever burning. 
But the meaning of sin is no longer understood, and 
sin-vluot, the deluge, has been replaced by suTid- 
Jluth, the sin-flood. 

Such has been the fortune of a pronominal root 
which has answered to the ideas of unity, equality, 
resemblance, communitj"^, coincidence. 



340 The Indo-European Organism. 

I shall choose one more example, which will lead us 
from meaning to meaning and from idiom to idiom. 
A-hhi, which will be more readily recognised under 
the nasalised form am/phi, is already composed or 
declined by the aid of a root hhi, which has furnished 
terminations to tibi, ihi, and sibi, and which, either 
isolated- or prefixed, still lives in the hei, hy, he, of the 
Teutonic languages : hei Gott, helieve, hefore (it is the 
prefix of a number of words, reduced sometimes to the 
single letter h — hut, for he-out, hange, fear, harm-hertzig, 
full of compassion for the poor, &c.). As for a, it has 
served as the poiut of depairtare for so many meanings, 
that I will not stay to consider it. Returning to abhi 
suffix and prefix, adverb and pfeposition, we find it 
implies, in the first instance, movement towards some 
object, and is opposed to apa, apo, ah, off. Rig- Veda, i. 
123,7: ajpa any ad eti, abhi any ad eti (obit aliud, adit, 
venit aliud), "one thing goes, another comes." But 
it is also used in ihe sense of against, upon, in, for. 
While ahhi-mukha, " turned towards," means favour- 
able, ahhimdti means ambush, enemy ; abhikram, 
abhitchar, abhijug, abhikr, to walk, to act against, to 
attack ; abhibhu, to be above, abhirasJitra, he who 
rules afar ; sa maniishir abhi vigo bhati, he shines on or 
in the dwellings of men ; abhi lomani, in the hair ; 
abhi subhagam, for riches. This comprehensive word 
has also the sense of around : abhi-tas, on either hand, 
from all sides. Tam abhitas dstnds, being, seated all 
round him. SarvS Pradjapatim abhita(s) upaviganii, 
all approacli Pradjapati from all sides. 

Zend and Persian have retained abhi under the 
forms aibi, aiwi, ahis, with the meaning towards, 
upon. Greek has adopted afxdl, aficpli, and with the 
dual a/xcpui, in, for, round, afar, from both sides, both ; 



Compounds. — Indeclinable Words. 341 

afx<^i and aficpk are sometimes prepositions and some- 
times adverbs : a/xcpiKTiovtjs, those who dwell in the 
same house ; a/x(^/ KTa/uievtjs e\a<poio, for the possession 
of a stag. It would seem that the sense " around " 
had need of commentary. Homer often strengthens 
afKpl with Trepl : afx,(j)i irepi Kpijvrjv, round about the 
spring ; yet he employs it alone, as a rule : afi(pl Se 
XeifjLwv, the marsh extends all round ; ofAcfn §e eraipoi 
eSSov, the comrades slept round about ; afi(pi Se 
HapOeviov, near the Parthenios ; afx<plvoof, irresolute ; 
afjiipidoXog, who strikes or is stricken on all sides, 
whence aiLitpi^oXoyia, ambiguity of phrase. 

It is curious that afxcjils should pass from around to 
far from, from the idea of contiguity to that of distance : 
/Sapvi Se TapTapo^ cifipii, and the deep Tartarus around, 
at a distance ; apyaXeov, Tocrov yjpovov afjL<pts eovra, 
eiTrevai, difficult to say, fcr one who has been so long 
at a distance ; but this notion of distance is still very 
vague : ^aXe kukXu afxipli, she placed the wheels at 
either side ; tjuyov ci/x^i? eepyei, the yoke parts them ; 
here we have the idea of separation. The columns of 
the Atlas, says the Odyssey, support and separate earth 
and heaven, yaiavre koi ovpavov afjL(f)Li e-^ovariv. The 
horses left the track, a/j.(pis oSov Spa/jieT^v. Athene 
and Hera stood apart from Zeus, Aioy aix<piq. To hold 
different opinions, aficph (ppoveiv. 

The inquiry leads to even more curious results in 
Latin, because the use of circa, circum, having prevailed, 
the traces of the primitive abhi, and even of the first 
terra a or am, must be sought in words which are often 
but ill explained. The most developed form appears 
to us at once in ambo, both, amhivium, amiidens ; in 
amhire, ambitus, ambitio, to go round, circuit, ambition ; 
ambigere, ambiguus, ambages, to tergiversate, ambiguous, 
23 



342 The Indo-European Organism. 

ambages ; amhtihts, ambulum, formed like circulus, 
whence prceambulum, little introduction ; funambidus, 
rope-dancer ; ambulare, to c»me and go, to walk ; 
amptruare, redam/ptvmre, to dance in a circle ; amhesus, 
gnawed on all sides ; amiarvalis, victim carried round 
the fields ; amhurhium, procession round the city. What 
of aniplus ? Is it for ampulus ? More probably the 
sufBx is the same as in duplus, simplus. In any case, 
the sense of am is not doubtful in the word ample, 
nor in am-plector, I embrace, in am-putare, to cut 
round, am-terminum (in Cato the Elder), around the 
boundary; amicire, aynictus, to wrap, mantle. In 
ampulla we hardly detect a/ACpi, but Latin has bor- 
rowed it from the Greek afji<pt(popa, vessel which is 
borne round, amphora, amporula, ampulla, little am- 
phora, little flask. We should not expect to find the 
word in the French an, anneau ; yet the Oscian am- 
nud is the ablative of a noun amnus, annus, the 
circle, of which the diminutive ammt^MS has retained the 
meaning. The year is the circle of days. We may 
mention also ancisus, cut in a circle, and ancile, cir- 
cular shield; anceps, two-headed, uncertain, amsegetes, 
whose harvests border the road ; and in Virgil amsancti 
valles, sacred valleys on all sides. 

The Celtic dialects have kept this am (Breton) in 
Irish imm, imme, around ; and it would seem that we 
have it in the name of an ancient Gallic king, Ambigat, 
and in anibactus, a slave walked up and down for 
sale, then servant and messenger, the humble proto- 
type of our ambassadors (ambadiator). 

Amb is wanting in Gothic and English, which have 
preferred the simple form bhi, but it has taken many 
forms in the other Teutonic dialects. Old High Ger- 
man, and even Middle High German, liave adopted it : 



Compounds. — Indeclinable Words. 343 

umpiseJien, to see round; tiinbi-hanc, veil; umhi-hrinc, 
circle ; umhe-gang, circuit, frequentation. (German um- 
sehen, umhang, umgaTig). The Saxon ynibe belongs 
to the same times. In Dutch omm, in Danish om, in 
Norse and in German iim, at once adverb, prefix, and 
preposition, represent in modern times the ancient 
dbh, dbhi, and correspond exactly to the Latin am; 
erum, around ; darum, therefore. Er geht um, he 
makes a circuit ; umgehen, to go round, to elude ; 
tomschreiben, to paraphrase, to transcribe ; Umlaut, 
change, deviation of the sound of a vowel ; umdrehen, 
to reverse ; imJcommen, to perish. Es ist um uns 
gesehen, it is all up with us. 

The history of ana, nearly allied to ahhi, is yet 
more varied, especially if we consider the various uses 
of the second member na, affirmative, interrogative, 
and negative : Gr. vai, vv, vvv ; Lat. nam, num, ne, nunc, 
nempe, enim, non, nisi, the suffix of so many verbs, 
nouns, adjectives, or adverbs (u-nurus, do-nu-m, do- 
na-re, po-ne, si-ne). Ana, a demonstrative which is 
declined in Sanscrit, in Zend, in Lithuanian, in Polish, 
has, combined with the diminutive lo, given to Latin 
unu-lu-s, ullus, olle, ul-tra, ul-timus, olim ; to the 
Celtic, Persian, and Armenian languages the pronoun 
an, the article an and n ; Welsh, an cu, the dog ; 
Breton, aimn n'apstil, the name of the apostle ; Ar- 
menian, mart'n, the man, the mortal. Its scattered 
cases, become invariable, have taken the meanings : 
certainly (Sanscrit), on, upwards, toioards, vnth, in, 
through, while, hy turns, again, backwards. It is 
more especially in Greek that these variations are 
multiplied ; we find the forms ava, avtn, av, evi, ivg 
(«?) even (V (in Arcadian and Cypriote). Homer has 
aWava, "up, then;" av S'OSvcrcrevi -TroXvfxtiTis 



344 The Indo-European Organism. 

avlerTUTO, "then rose the wise Odysseus;" /xeXai/e? 
S'ava ^orpues ^erav, " above hung the black grapes ; " 
avaQriij.a (ex-voto), an object dedicated to a god ; ava 
veux; jSaiveiv, to go on board ; ■^(pvtTew ava aKe-rrrpw, 
on or with a golden sceptre ; ava TroTafiov TrXeeiv, to 
go up the river ; ava^aaris, a climb in high land ; ava 
iracrav ^/nepav, throughout the day ; avu jnepoi, by 
turns ; avairnrpria-Kw, to sell again ; avapao/u-ai, to re- 
tract. Latin has only an-helare, to take breath again ; 
but G-othic has anafotmis, on the feet ; High German 
ana-sikt, view ; an, towards, on, as ; e.rkomm an, he 
comes, he arrives. . English, on. The combination of 
upa with ana (Gothic infana, afana) has produced 
the English wpon, OHG. fana, fona, the German von, 
and the Dutch van. 

Under the forms ein, en, in, the ancient ana has 
caused the gradual disappearance of the locative in 
Greek, Latin, and the Germanic languages ; it has ex- 
pressed all shades of movement towards or against, and 
thence of rest in, a place. It has been questioned how 
ei? could be derived from en. Evy, the Cretan form, is 
to en as ex to eh, aps to apo, pros to pro, and in Latin 
ahs to ab, subs to sub ; it may be an accusative plural 
or a contraction of evTo^. We know that the Greek v 
readily gave place to an i {evTi = elcri) ; for the rest, the 
identity is complete ; so much so, that the Dorians used 
ev with the accusative in the sense of movement to- 
wards ; etj is a later variant which the language has 
made use of. 

But it is time to draw some conclusion from these 
lists. All show us the reciprocal action of the form 
and the thought, the precision of the meaning increasing 
with the number of variants. Whether we have simple 
prepositions, par, pour, de, ct, or compounds, sur, en 



Compounds. — Indeclinable Words. 345 

(inde), dans (de intus), sous (subtus), &c., or conjunctions, 
either relative or dubitative, dass, denn, when, oh, if, que, 
quand, si, wbether copulative or disjunctive, at, et, que, 
und, and, everywhere we find a primitive pronominal 
root, or group of roots, of indetermined meaning ; these 
syllables, which served for suffixing and to form the 
declension, were also capable of independent life. As, 
by inflexion, they were gradually atrophied in the body 
and at the end of words, the freedom, or the more ex- 
tended use, of prefixation retained their forms while 
it varied them and accentuated their diverse meanings. 
They were thus prepared to supply with advantage the 
place of the worn-out terminations. As they gave 
to the phrase and to thought greater exactness and 
elasticity, they broke the fetters of grammatical syn- 
thesis, and urged language forward along the open 
road of analysis. Auxiliaries simplified and developed 
the verb, invariable words broke up the declension. 
Philologists are sometimes inclined to regret the in- 
genious mechanism which has been unable to resist 
the slow action of these dissolving particles ; but how- 
ever great our admiration for the stately Latin phrase, 
for the luxuriant wealth of Greek and German, we 
cannot see that the language of Rabelais and of Ron- 
sard, of La Fontaine, of Molifere, of Voltaire, of Meri- 
mee, or of Victor Hugo, that the tongue of Cervantes 
or of Ariosto, or finally that of Shakespeare, Swift, 
Byron, Shelley, and Dickens, need fear comparison 
with the famous idioms of Homer, -^schylus, Aristo- 
phanes, Lucretius, Virgil, Cicero, Horace, and Tacitus, 
or with the languages of Schiller and Goethe, of Tolstoi 
or Mickiewicz. Regrets, moreover, are unavailing. 
Language resists all efforts of the will ; it has fatally, 
unconsciously, followed its destined path. 



346 The Indo-European Organism. 

In the course of these researches mimerous com- 
parisons have shown what metamorphoses, what accre- 
tions, what mutilations, what changes in form and 
sense are produced in the elements and the combina- 
tions of Indo-European speech by the preferences, the 
aptitudes, the decadence, and tlie progress of the various 
nationswhich have received an Aryan education. Among 
the circumstances, the natural causes, which have most 
contributed to the differentiation, and thus to the re- 
spective originality of ihe idioms, the most powerful 
was undoubtedly the original or acquired difference in 
the vocal organs. Why does the Frenchman say tnois 
and voir where the Latin pronounced mensis and videre? 
Why do the English say tooth, the Germans zahn, and 
the Latins dens? Why have we irevre in Greek, 
quinque in Latin ; Tecra-apes instead of qyiMuor or 
tchatvaras ; vttvos for svapna and somnus ? Because 
at a given time among different peoples the larynx, 
the teeth, the lips, and the third frontal circonvolution 
of the brain worked differently. These initial diver- 
gences cannot be detected by physiological examination, 
because the observation could only be taken after death ; 
but if it is almost impossible to discover them in 
themselves, it is interesting and comparatively easy to 
describe their effects and to classify the results. This 
comparative study of the different pronunciations of a 
same vowel or consonant, or of the same original vocal 
group, is the object of phonetics. Its importance will 
be seen if I indicate beforehand the conclusion arrived 
at : it is a general rule that, except in the case of 
borrowed words, or of accidental similarity, or of par- 
ticular affinities, the same word cannot exist in the 
•same form throughout the series of kindred idioms. 
This is the first principle of scientific etymology. 



CHAPTER V. 
INDO-EUROPEAN PHONETICS. 

I. THE COXTINUOCS LETTERS. 

Dialectic variation — The primitive vowels — Tlie metamorphoses of the 
o— Bopp's theories on the weight of vowels — The variants of i 
and ou {ii, i, y) — Contraction and abbreviation of Latin — A word 
about spelling — R and I, vowels ; guttural and dental r— Vowel 
scale — The semi-vowels y and v in Sanscrit and in Latin, in Zend 
and in Greek — The sibilant s and its transformations in Greek 
and ill Latin ; its affinities with the hard breathing and with the 
liquid )• — The nasal n both liquid and dental — The labial nasal m. 

The respective individuality of languages which have 
a common vocabulary and a common structure results 
from two distinct though connected phenomena : the 
varied use of attributive or demonstrative roots and 
of grammatical artifices; the different pronunciation 
of the phonetic elements. These are the two factors 
in dialectal variation ; they have acted together, and 
under the rule of circumstances which are as little 
understood as they are evident : the progressive 
distance and isolation of the various groups imbued 
with the primitive Indo-European culture, contact and 
crossing with foreign groups, unequal or varying 
development of the variously proportioned mixtures 
which have constituted the nations as we now know 
them, the influence of climate, of new needs and 
interests. These are causes of an historical nature, 
yet of which the history is often unknown to us, even 



348 The Indo-European Organism. 

in the case of languages sucli as the Neo-Latin or 
English, which have grown up as it were under our 
eyes. But whatever share, and it is a considerable 
one, we must attribute to these historical causes in 
the growing separation oF Sanscrit and Greek, of Latin 
and Slav, of Teutonic and Persian, we must seek the 
starting-point of these divergences in cerebral and 
vocal aptitudes. It may even be doubted if there 
ever was a time when the common speech was pro- 
nounced in the same manner by the seven or eight 
tribes destined to scatter it over the world ; whether 
the children of the first chief who first pronounced 
distinctly a few Indo-European syllables did not 
modify them from the beginning, one inclining to a 
thickness of speech, another to a lisp, another harden- 
ing or suppressing the aspirate, or perhaps unable to 
distinguish the r from the I, the v or the s from the 
hard or soft aspirate. The comparison, letter by 
letter, of a number of words, of which the identity is 
no less obvious than their differences, of such forms, 
for instance, as padas, tto^o's, pedis, fotus, foot ; or 
again hrdaya, KupSla, cordis, Jiairts, heart, herz ; gunai, 
/cucoy, canis, hund ; djanu, jovv, genii,, hniu, knee, has 
shown that the observed changes are constant, and 
thus the phonetic laws peculiar to each idiom have 
been established. And while the general regularity 
of the corresponding changes threw into relief the 
physiognomy of each language, it was itself an evi- 
dence of the primaeval unity. Thus the deviations 
proper t) Prencli, to Provenpal, and to Italian bring 
us back to an original Latin form wliich dominates 
and throws light upon them all. The r61e which 
belongs to Latin in the phonetics of the Eomance 
languages is therefore precisely that of the lost mother- 



Indo-European Phonetics. 349 

tongue in Indo-European phonetics, and this lost idiom 
has been restored, according to the indications given, 
by the agreements and the divergences of the derived 
idioms. Each variation may be traced back to a 
vowel or a consonant of an original alphabet, which 
serves as the point of comparison and the common 
measure. The thousand details noted by a patient 
and intelligent observation have their place ia a 
natural classification, and the study of phonetics has 
become the science of the evolution of articulated 
sounds. Like every other science, it formulates laws, 
the summary of observed phenomena — laws which 
record, but do not govern. These laws allow us to 
supply by legitimate hypothesis the lacunae of experi- 
ence ; but they yield before that which resists them, 
and do not pretend to embrace more than they can 
fairly hold. They are no imperious rulers, but con- 
venient data, the instruments of investigation. This 
is the case in every department of knowledge; but 
here, in the domain of language, in which reigns the 
mobility of life, here more than anywhere it is desir- 
able to recall the empirical origin and the relative 
character of laws. 

We now approach the history of the vowels, the 
simplest of the continuous sounds, if not the most pri- 
mitive, which were uttered by man. There is reason 
to believe that certain other sounds, also continuous 
— the sibilants, the aspirates, and other confused 
groups from which the consonants were afterwards 
developed — belong also to the earliest attempts at 
human speech. But Indo-European is far removed 
from the origin of speech, and its phonetic system is 
already clearly constituted. It has three vowels, short 
or long, a, i, u, and two diphthongs, ai, au, which tend 



350 The Indo-European Organism. 

already to be pronounced ^ and 6. A species of trill, 
T or I, is reckoned by the grammarians of India in the 
number of the vowels, but Sanscrit alone makes use of it. 

The absence of two familiar letters, the short e and 
0, will be noticed. It has caused wonder that these 
two sounds, so easy to utter and so widely spread 
among the nations, should have been wanting among 
■the Aryans. They were, in fact, unknown to Sanscrit 
and Gothic. Zend has only the e. It has been said 
that Sanscrit pronounced them without writing them ; 
but why, then, have they no signs in the Devangari 
alphabet, which notes the least shades in the conso- 
nants, and which was adopted in India when Sanscrit 
survived only as the speech of the lettered ? Why 
did not Ulphilas, who created in the fourth century 
the Gothic alphabet, borrow the e and the o from 
Greek and Latin ? Why do e and o in all the lan- 
guages which possess those sounds invariably corre- 
spond to a Sanscrit a ? Why do they alternate with 
this a, of which they appear to be doublets or varie- 
ties ? Why, when it modifies an a, does Sanscrit 
change it into % or ij? Because the intermediate 
sounds were wanting. Lastly, a comparison which 
has its value : the Semites, when they invented their 
alphabet, fifteen or sixteen hundred years before our 
era, did not distinguish the e or the o ; they assigned 
special characters only to the afep/i, the iod, and the 
vav, and even these they confounded with the soft 
breathing, and with the semi-vowels y and u. They 
had not even a clear conception of a, i, and ii. A 
fortiori, then, we may admit that Indo-European only 
possessed these fundamental sounds. 

A, which is the most frequent sound in Sanscrit, 
has rarely remained pure in Greek and Latin, as in 



Indo-European Phonetics. 351 

aga, ayeiv, agere ; agra, aypos, agar ; aksha, a^aiv, axis ; 
ang,angor; TraTrip, pater ; agrou, Swcpu, lacruvia. More 
often the primitive a is represented by a short e. Root, 
gan, to engender ; yevos, genus ; dalcan, Sexa, decern ; 
daksha, Se^irepos, dexter ; sarp/epirio, serpo ; hhar,fero. 
Gothic, more faithful, still has vasti, dress ; avi, sheep. 

The change to 0, which is also very common {avi, 
OK, ovis ; pati, -ttoitk, potis ; dktau, octo), is especially 
common in terminations; dama, domos, domus; the 
Latin terminal being blunted into u. Greek and 
Latin are sometimes agreed in the change in the a, as 
if this phenomenon had begun among them before the 
division of the Hellenes and the Latias ; but as they 
diverged, each claimed its liberty, using each one e or 
at will. The one has Satrv^, thick ; ^pa-)(iii}, short ; 
eXai^i;?, eKurov, the other densus, bre(g)vis, levis, centum ; 
or again, magnus and caput correspond to /jiiyas and 
Ke(f)a\ij, Safxav to domare, KupSia to C07-di, sapiens to 
aocpog, genu to yoi/v, pedis to ttoSos, e/jLetv to vomere, 
i/e'[/:]o? to navus, \eyovT09 to legcntis. In Greek 
sometimes the fall of the nasal has preserved the a : 
adikshaw., eSei^a, eXua-a, \i\vKa, &c., irarepa, iroSa, 
SeKU, e-Trra, where the Latins have preferred patrem, 
pedem, decern, septem. If Greek, like Latin, has adopted 
the e in irevTe, in quinque, it is because the final nasal 
had already fallen from panhan, Sanscrit, pantcha. 
Latin especially affects the when the primitive a is 
followed or preceded by a v ; but there are exceptions : 
nOivan becomes novem ; navas, novus ; svap, sopor ; vam, 
vomere; vak, voc-s. 

In fiae, the e and the proceed alike from a and 
represent it equally ; and the proof is that we find 
them alternately ia the same language, in the same 
word : yevoi, for example, or /cXe'oy, reproduce ganas 



352 The Indo-European Organism. 

and kravas, but their o may be changed into an e, 
yeveos for yeuecros (ganasas). In the nominative \oyos 
and the vocative Xo'-ye the o and the e are equivalent; 
so also in the imperfect eXnrov, eXiire. In (pepeiv the e is 
the same letter as the a of (paperpa, as the o of (popog ; 
^6v9o9 and ^a6o9, depth, TraOos, emotion, and Trej/^o?, 
pain, are two forms of the same word. The vowel 
varies according to the dialect, but not regularly. 
Dorian usually prefers the a, in a-rpacpw, to turn, 
Tpaj(w, to run, lapo?, sacred, compared with the classic 
(TTpe(pa), Tpeyai, lepos, but it has epcrt]V, Tecrarepei, and 
TSTTopes for apcTijv and recrcrapes. 

These variations are a source of great wealth to 
the Greek language ; from the single class of the 
causative verbs in aja Greek derives its three con- 
tracted conjugations (biXeoj, Ti/jLaw, StiXow • from a root 
trap, to turn, it derives the present and the imperfect 
Tpeiroo, eTpe-TTov, the aorist eTpairov, the perfect TSTpoira, 
the noun t^otto'?. The variants serve to indicate cer- 
tain cases or tenses, and facilitate derivation. Latin 
uses them in the same manner and less rigorously ;. 
compare the forms genus and gcnerare, scelus and 
scdestus, seeleratibs,pars smA portio, ager ajiA peregrinus, 
terra and extorris, velim and vohmtas, verto and vorto, 
prex and prncus, scqui and socius, tegere and toga, 
pendere and pondus ; facio, infectus, inficio ; patrare 
and perpetrare, fallo and fefelli, cano and cecini. 

The Teutonic languages have adopted analogous 
methods in their conjugations, reserving to the present, 
the perfect, the subjunctive, and the past participle 
respectively some one of the variants of the primitive 
root vowel, e, a, a, o, u : ich bind, ich band, gebunden, &c. 

The long a which is vei-y common in Greek and 
Latin, is often the result of the juxtaposition of two 



Indo-European Photietics. 353 

similar or different vowels, of the presence of two con- 
sonants, or of the dropping of a consonant ; thus rtyua? 
for Tifidii, Kepa for Kepaa, Kepara, horns ; fieXav for 
fA.e\avi, amasti for amavisli. A was originally long 
in the feminine terminations ^fiepa, Qvpa, (piXla ; 
in Ennius's time the final a of aquila was still con- 
sidered long, but it has usually been modified into 
a long e or 0. In the choice of the vowel there is no 
very marked agreement between Greek and Latin ; 
often also it is represented by a short a, e, or 0. Dhd, 
to place, gives in Greek riOifxt and O^ctk, in Latin 
dere, siihdere ; dd gives SlSoi)fji.i, Suipov, Soa-is, in Latin 
donum, dare ; sthd, ^a-Ttifxi, stare, stetit ; dicu, rapid, 
ujKvi, Latin ocior ; sdmi. ^/ni, semi ; pd, to drink, TrUofia, 
potus. The Greek ft-rp-rip is the Latin mater ; /Bpa-^vTrjS, 
hrevitas ; (f)>'inri, fari, fama ; hhrdtar, (ppuTcop, f rater ; 
yeverrip, genitor. The Greek dialects interchange the 
long 0, a, and e : fiovcrwv and fiovirdv (genitive of yuowira 
or fiova-t]) ; (piXia and (f>i\trj, Trpwrog and irpaTOi, first ; 
(^aiJ.1 and (prifii, Troi/uiav and Trotfiijv, shepherd. Gothic 
keeps the long a ; Old High German takes an e or 
a modified a : yar, yare, year ; blasou, hlesa, I suffei-. 
Nothing is more capricious than this distribution of 
the a and its variants. 

The vowels i and ou short are primitive or secondary. 
In giv, vivere, /^/foy, to live, life, in imas, imus, we go, 
in tchid, to cut, the Greek o'X^^'^ l Latin ahscidit, in 
the innumerable uses of the suffix ti, mati, M't^^, 
mentis, thought, devatati, deity, voluntatis ; in the 
terminations of bharanti, (pepovrt, ea-rl, there is no 
doubt that we have the primitive i. So with mi (u) 
in the supines sthatum, datum, visum, in svagrti and 
socrus, mother-in-law, baJui and pechus, arm, elbow, 
rule and hiJcs, light. 



354 The Indo-European Oi'ganism. 

But Mta, the participle of the verb dhA, hm-rtias, 
first person plural of kar, to do, reduplications such 
as tishthami, SiSoyfii, show us these letters already as 
secondary vowels. They conceal an earlier a, of which 
they are the weakened form. This is sufficiently 
proved hy a comparison between staiut and institution, 
between capio, accipio, and aiocupium, between sam and 
the Greek (Tvv,with,hetween gani (Sans. ^am),wife, Old 
Prussian gannan, Greek yvvri, and the English queen. 

Bopp, who is the true founder of the science of 
phonetics, has tried to establish among the vowels a 
gradation of weight, which pronunciation is ever seek- 
ing to lighten. A being the fullest sound, the vari- 
ous languages attenuate it into e and d, and shorten 
it as much as possible into a and i in compound 
words. This would be an application of tlie law of 
the least eifort, which has so many exceptions, such 
as the strengthening of the Teutonic consonants, and 
of the Greek aspirates, or the excessive development 
of the Greek passives. This ingenious observation is 
supported by examples ; it agrees with the effect of 
accentuation. Accentuation subordinates the whole 
word to one more weighty syllable, and forces the 
others to contract or disappear. This idea is fami- 
liar to those who have compared in any degree the 
Romance languages with each other ; these, as it is 
well known, are derived from Low Latin under the 
influence of the Latin tonic accent, and differ from 
each other according to the unequal effect which this 
accent has had upon their respective vocabularies. 
Unfortunately nothing is more variable or more arbi- 
trary than accent; if it is very i-egular in Latin, it 
is also purely artificial, depending entirely upon the 
quantity of the penultimate syllable. In Greek and 



Indo-European Phonetics. 355 

Sanscrit, if it appears sometimes to fall upon the root 
syllable, sometimes upon the suffix which most deter- 
termines the meaning, yet the application of these two 
contradictory principles is so entirely unsystematic, or 
at any rate determined by laws so complicated, so 
subtle, and so fantastic, that the ablest specialists are 
often at a loss. We readily believe that there is a 
relation of weight among the vowels, but this relative 
weight appears to vary according to the caprices of 
accent ; and we can hardly say that i or ow are more 
or less light than e or 0. Bopp himself recognises 
that e, heavier than i in legere as compared with legimus 
and eligere, is lighter in admirdhile compared with 
admirabilis, in sequere compared with sequeris. 

Admitting, then, in a certain degree, the law of 
least effort and the relative weight of vowels, combined 
with the caprices of accent, we would add to these 
causes of variation two causes more general in their 
application, the one physical and the other intellectual. 
The first is so simple that we hardly dare to propound 
it : it is the want of definiteness of the primitive 
vowels and the reciprocal nearness of their variants. 
The second is the instinctive and growing need of 
fixing the significant variations and grammatical values 
as new sounds were developed and noted, a need which 
found its expression in a use gradually fixed by custom, 
assigning in Latin, for instance, e to past tenses, i to 
roots of which the meaning had been modified by 
prefixes, and u to the terminations of certain nomina- 
tives and accusatives, or to certain links between root 
and suffix. 

But I would insist upon the primitive confusion of 
vowel sounds, and on the insensible transitions which 
connect the variants, U stands half way between a 



356 The Indo-European Organism. 

and i ; occupies the same place between a and u 
(ou). Between e and o we have ev, ; between i and 
ou, u (u) ; the slightest movement of the veil of the 
palate is enough to change a into e, i into o, ou into u 
and i; but a keeps the summit of the double ladder; 
it may be exchanged for one of the other vowels, but 
itself takes the place of none. We may conclude 
from this fact, that a is the earliest vowel in the Indo- 
European language, and even in pre- Aryan speech. 

Greek, says Bopp, is less sensitive than are San- 
scrit, Latin, and Teutonic to the relative weight of the 
vowels, and offers no change of a into u or i, which 
appears at first sight to be regular. We have already 
given a few reduplications above, SlSwfM, TiOtj/iii ; we 
may add wc/coy (tTTTro?), for akva ; Xiyvvi (Lat. lignum), 
from the root dah, dagh, to burn ; the Homeric form 
TTiarvpe?, for katwaras ; vv^, vvktu (Lith. Tvaktis, Goth. 
naht), for naktam, ; ovv^ for nakhas, nail (Lith. naga), 

yvvrj^ CTVV. 

I and ou are not less interesting, both from the 
point of view of origin and of pronunciation. The 
latter has passed from the sound ou to ii and i, and 
at the same time it became a semi-vowel v before a 
consonant, or between two vowels (efuyyeXiov, ^aa-i- 
Xe/ry) ; the former, the i, has frequently taken the 
place of the semi-vowel y or j. Greek has not this 
sound at the beginning of a word (where it replaces it 
by a consonant, yugam = Xvyov), and in the middle of 
a word it transcribes it by an i. Nothing is more 
strange than this partial repugnance for a sound which 
is, it would seem, so easy to pronounce ; but similar 
rejections may be observed in all languages. Such 
forms then as (^iXoa-ofpla, (jiiXia, iraTpiov, d'ytoy, (piKoiriv, 
Oeoio, apiwv, ^Siwv, jill include the suffix /a or ?/« ; and 



Indo-European Phonetics. 357 

this is true also of the Latin patrms, siem, sief, durior, 
melius. Greek goes farther, weakening i into e : 
TrXevcreofiev, (pev^eofiai, for TrKevaloixev, (bevycrio/iai. 
The Sans. Tcunja becomes Keveog, empty (whence the 
-^olian Kevvog and the Classic (cei/o'y, which we have in 
cenotaph) ; a-repeos, solid, ^olian aTeppoi, corresponds 
to an ancient starya (from star, to extend, a-Topevvvfii). 
In the proper name Meocles we recognise the Sans. 
satyagravas (owning a true glory), satya, that which is. 
When i takes the place of y after a liquid (r, I, n), it 
readily passes into the preceding syllable, as in the 
French faillir from fallire, despmdller from despoliare. 
This is the explanation of doublets like fxeXavia and 
fieXaiva, iroTvia and Mcnroiva. The Sans, tanyami, 
tanyamas, should give in Gr. Tevlw, revlofiev, but we 
have Te'wu), Teivo/nep, (pQeipw, KTelvw, ocpeiXo), fxalvofiai 
for (pTepiQ), KTevLW, 6(peKiw, fxavioixai ; in the Homeric 
and ^olian forms 6(peKKw, Krevvai, (pQeppw, the trans- 
posed i has disappeared by assimilation. 

It is not surprising that i should take the place of 
the semi-vowel y, but we should hardly expect to 
find it in the place of an n. Yet this is what happens 
in ela'i, TiOeicri, XvOets, XvOetcra, [xoiara, for evri, rlQevri, 
\vQevi, XvOevTia, fiovTia ; similarly in our French words 
gaulois, frangois, provinois, mois, prise, from the Latin 
words gallensis, francensis, pruvinensis, mensis, prehensa. 
It has been said that this substitution is more apparent 
than real, that it is explained by a compensation for 
the loss of the letter n; but the compensation in its 
turn is only explained by a physiological affinity of i 
with the nasal. In pronouncing i, the veil of the palate 
is raised towards the nasal cavities, whence a peculiar 
resonaiice, already sensible in an e, a little weaker in 
an a or an pronounced from the throat ; thence 
24 



358 The Indo-European Organism. 

the frequent phenomenon of nasalisation, common to 
all languages, and so difficult to distinguish from the 
insertion of the primitive consonant n. With the 
sounds i and ou the leaning to the nasal approaches 
identity. We have already observed that the accusa- 
tives Toh, \6yovf, ^fiepa^, irarkpa^, stand for roi^y, 
Xoyovs, &c. ; the terminations of Xeyovai, irotovcri, 
apxovcri, are for Xeyovri, &c. If we seek the original 
of the Greek diphthong, we shall find a primitive a 
replaced by an 0, and an n supplied by u : bharanti, 
ferunt, (pepovari. There is no doubt that the sound 
was simple and answered to the Indo-European vowel 
ou or u. But the Greeks being used to write it with 
two letters, like the Beotian yXouKov, soft, tov, thou, 
the upsilon lost its ancient pronunciation, gradually 
narrowed down to u, and towards the first centuries 
of our era became confounded with i ; certain dialects 
said Xeyoicri, cjiEpowi, instead of Xeyovcri. The pro- 
nunciation V in the words quoted above (eu, j3aa-iXevi) 
is to-day the only vestige of the original sound, for 
■D is to M asy to i. Still we should not pronounce 
ancient Greek authors in the modem fashion : in 
Homer and Sophocles, in Plato and Callimachus, 
upsilon was never pronounced i. It was the first ou, 
then u, the French u. 

Latin had this to, intermediary between ou and i, 
slighter than the one, stouter than the other, says 
Marcus Victorinus ; nearly resembling the ipsilon, says 
Priscian. The Emperor Glaudian wished to distinguish 
it by a special sign, which may be seen on a few in- 
scriptions, and Quintilian recognises its existence. 
But it was in the end confounded with i, as was the 
ipsilon, which became our y. But the Latin letter u 
regained and kept the sound ou, more or less akin to 



Indo-European Phonetics. 359 

the short 0, and the substitute of a primitive a. We 
note in passing that the diphthong (m, retained in 
Greek, was lost in Latin; the ancient forms /ows, doucere, 
pious, were contracted into jus, ducere, plus ; and this 
is why the letter u, after having hesitated between the 
sounds 0, u, and i, returned to its primitive value. 

To give a few examples of this variation of the 
Latin tt ; From the time of Plantus and Ennius it was 
pronounced in verbal terminations and case-endings 
and in certain suffixes ; men said servos, the slave, 
dominos, the master, pravos, wicked ; makistratuos, 
senatuos, domuos (inscriptions) ; Twminos, Veneros, 
Gereros, partes (written with a w in the inscriptions) ; 
poJcolom, drinking-cup (Osc), oraclom, qucesumos, legi- 
mos, volumes, sumos. They pronounced septumes (Gr. 
e^Sofjios), dehumes, optumes. Caesar preferred dekimus ; 
Augustus from archaism affected optoumov^,inaxeumeus. 
The alternation between and i, i and u, i and ou, is 
visible in oUe, ille, and ultra ; incola, inquilinus, cultus ; 
silva, and the Greek vKri, ovKri ; exsul and exsilium, 
facultas and facilitas, cornu and cerniger, arcus and 
arcitenens, manus and manihus, menumentum and 
monimentum, mancupium and mancipium (taken with 
the hand), luhet and libet, existumat and existimat, in- 
clutus and inclitus (Greek /cXuto?, Sanscrit qroutas). 
But whateveT was the pronunciation finally adopted, 
it is easy to see what. a variety, what a wealth of forms 
was assured to Latin by these diverse transformations 
of the Indo-European a. From salio, to leap, are 
derived insilire, insultare; from capere, incipere, in- 
ceptus, aucv/pium, occupare, nuncupare ; from tango, 
tetigi, attingere, tactus ; and each root had thus its 
family of derivatives, in which the words are counted 
by ten and twenty. 



360 The Indo-European Organism. 

Sanscrit has made a similar use of its long e and 0, 
which it could strengthen into ai and an. These 
latter diphthongs are peculiar to Sanscrit. The long 
e and were not yet simple sounds when the Grseco- 
Latins separated from the common cradle. For the 
rest, their composite nature is revealed, even in San- 
scrit, by their perpetual changes into ai and av before 
a vowel. The Greek renders the former by ai : (peperai 
for the Sanscrit iharatS, and also by 01 and ei ; Latin 
by ae and oe, contracted into u long ; by ei, contracted 
into i and e long ; av, ev, ov, in Greek, au and u in 
Latin correspond to the Sanscrit long 0. There are 
analogous gradations in the Teutonic ai, ei, iu, ie. 
Sanscrit Smi, Mi, Greek eim, Latin ire, to go. Gd, for 
gau, Greek ySoi/j, Latin dative hubus. Root 6g, to shine, 
Greek avyri, brilliancy. It is curious to remark in 
Latin the transformation of ai, ai (materai, terrai, 
rosai) into ae (teirce, rosce) ; of oi into oe and u, 
Greek iromi, Latin poena, punire (Fr. peine, punir) ; 
mcerus (Po-mceriwn), murus; mcenia, inunire; Tncenera, 
munera, eo-itus, coetus ; Greek (pom^, Latin pcenus, 
punicus. Elsewhere oe becomes e ; foetvs, fetus. 

Classical Latin is a contracted language, in which 
the diphthongs are almost all reduced to long vowels, 
excepting aut (Gr. avQiis), autem, aufero (but au is for 
ab, av, au), with a tendency to abbreviation. Virtutei, 
which we read on the tomb of the Scipios, becomes 
virtuti and virtute. Ennius still wrote ei, or at any 
rate made the e long: "Turn cava sub montei late 
specus intu' patebat." 

The forms mihei, tibei (tlmbrian, te-fei) are lightened 
into mihi, tibi, ubi, ibi ; domoi into domt The Oscan 
svai, Latin, svei, sei, si, is abbreviated in stquidem, sicubi ; 
jous, jousare, jurare, gives pejerare, to perjure oneself ; 



Indo-European Phonetics. 361 

noscere, nottis, co-gnitiis. Dederunt steterunt. Ennius, 
speaking of Prometheus, says: "Vulturius miserum 
mandebat hemonum." Hemonum Las become hominem. 
Moneo represents mdna-jdmi. Final liquids shorten 
the preceding vowel : laudator, longior {or represents 
OTis) ; nihil for ni-filum (not a thread, nothing) ; 
rogasne, rogan ? T has the same influence : amat, 
monSt, audit. Ennius reckoned esset as two longs, 
init a short and a long : " Alter init oculos. Omnibu' 
cura viris uter esset induperator." Virgil still has 
revocdbat. " Cum clamore Gyas revocabat ; ecce 
Cluentus ; " but this was then an archaism. 

We seem to see the hesitation in the quantity of 
the finals. Beside pulchre, valde, rede, of which the 
last syllable is still long, we have male, bene, probe, 
superne ; beside videre we have vide, jube, cave, vale ; 
beside Cnam for Gnaivod, domino, /ago, we find modo, 
cito, illico. It is no longer doubted that homo, virgo, 
ind.ignatio, ego, were originally homon, virgon, indig- 
TMtion, egon. The quantity of ago, volo, veto, is at most 
doubtful ; in scio, eo, it is shortened by the influence 
of the previous vowel. We find the two quantities in 
a verse of Horace : " Omnia legisti, credo, scio, gaudeo." 
The grammarian Diomedes, finding amo in Virgil with 
the long, concluded that the poet had lengthened the 
vowel in imitation of the Greek verbs, and his obser- 
vation proves that all the finals in the verbs were short 
in his time. Prom the Augustan age we feel that 
Latin was altering rapidly ; the great break-up from 
which the Eomance languages at last emerged had 
already begun; the permanence of the spelling alone 
conceals from us the progress of the change. It is 
very certain that, with the exception of a few errors, 
we read a far more correct Latin than that which was 



o 



62 The Indo-European Organism. 



spoken, I do not say in the provinces and among the 
soldiers, but at Rome itself, where cavneas meant 
either " do not go " {cave ne eas) or " figs to sell " {fi^s 
from Oaunas). 

When in French we say faime, we only continue 
the tradition of centuries which has mutilated into 
(k)amao, amo, the original kamajama, modified the 
primitive a into e and 0, and then gradually dropped 
the final syllable. Here an objector might urge the 
forms in ifnent, in on, el, ail, ait, at, aut, ot or e, grogne- 
ment, poison, poitrail, fait, mai, seigneur, dti, 6tat, 
grelot, faux, where the final is precisely the long- 
syllable. But this is an illusion; in the feminine 
terminations the mute e no longer counts, but in the 
masculine it is not written, and this is the only differ- 
ence ; it is enough to compare the analogous Italian 
forms, mxm/mento, pozione, faito, signore, stato, falso, in 
which the final is still sounded. The fact is that in 
the Romance languages quantity being confounded 
with accent, the accentuated syllables are the only 
true longs. Now in French, the most strongly 
accented of the Neo-Latin languages, the tonic accent, 
having contracted what precedes and devoured what 
follows it, falls with the quantity on the last syllable 
which is pronounced, and which sometimes seems to 
be the penultimate, but only in versification. 

It is sometimes asked why we continue to write 
what is no longer pronounced. Simply in order not 
to take away all the rhythm from the poetry of 
Corneille, Racine, La Fontaine, Andr^ Chenier, De 
Musset, and Victor Hugo. This is to me an all- 
sufficient reason, and renders me very sceptical about 
the proposed reforms in spelling. There are others. 
In many cases the e mute could not be removed with- 



Indo-European Phonetics. 363 

out being replaced by an apostrophe, or how should 
we distinguish eharmant and charmanie, or Sem, the 
son of Noah, from sdme ? For the rest, French spelling, 
which I am far from defending, be it observed, is 
capable of defending itself by the mere force of inertia, 
and if it allows an occasional useless letter to be 
dropped, it will never, being the result of long use, of 
ancient custom, yield all at once to the laws of a 
narrow reason. Spelling reform is advocated with 
excellent sense, but it is none the less an empty 
hobby. 

There remains one vowel to be considered, a doubt- 
ful vowel, employed only by Sanscrit, but it is at 
any rate one of the continued sounds which hold the 
debatable land between the vowels and the consonants. 
This is the liquid r, doubled with an I in Sanscrit, 
which is also regarded as a vowel. Most philologists 
consider these letters to be the contraction of a syllable 
ar, iri, ere ; they point out that in all Indo-European 
languages r is replaced by one of these groups, and 
that even in Sanscrit it generally disappears from 
most of the cases of the nouns in tr : pitri, pitaras ; 
mdtri, mdtaram ; and that in bibhrmas it takes the 
place of ara : habharamas. Yet, as there are several 
manners of pronouncing r, it is probable that this r 
answered to something particular. It, says the master 
of M. Jourdain, is pronounced by " carrying the tip of 
the tongue to the roof of the palate ; so that, being 
pressed by the air as it is emitted with force, it yields, 
and returns always to the same place, producing a 
species of trembling r, rrra." This r is the dental r 
of singers, who learn to pronounce it by repeating 
td, td. But there is also a guttural r, very difficult 
to pronounce at the end of words, soir, gloire, art, 



364 The Indo-European Organism. 

fiqueur, monsieur (whence monsieu and piqueux). 
The r which figures as a vowel in the Sanscrit alphabet 
appears to be closely related to a ; it is dropped from 
the nominative of the nouns in tr : pita, mata, Ihrata, 
and only reappears when a vowel follows it. It is 
represented in Prakrit by an i, and not by the group 
ar : hidaya, heart, for hrdaya. 

In any case, the vowel r corresponds in Greek and 
Latin to er, or, ar, ur, and ra. Bhrtas, carried, is the 
Greek (pepros (in a(pepTos) ; drshtha, seen, SepKTOs (in 
aSepKToi) ; str-na-mi, crTopw/u, I extend ; mrtas, death, 
jSjOOToy for fj-opros; rksha, bear, apKTOs; jalcrt, rjirap, 
liver ; pitrshu, TraTpdcri. Latin shows fertis, fertilis, 
sterno, niortuus, Jecur ; vermis for krmis, cord for hrd, 
mordeo for mrd, to crush, stratus for strtas. 

The terminations in tr, tar, which abound in Sans- 
crit, play a great parb in Greek and Latin. Greek 
has them under the two forms Tnp, Twp, long and 
short. Sorrip, giver ; ^OTijp, shepherd ; olvoiroTt'ip, 
wine-drinker; p-^rt] p a,nd pijTwp; yucijo-T?/jO, bridegroom ; 
OTTTijp, spy ; a(ppjjT(i)p, inhuman (not a brother) ; 
mrjTpoiraTcop, maternal grandfather ; 'IcrToup, witness 
(whence historian) ; aXeKTwp, cock, and ^XeKrcop, the 
sun ; oKprjTddp, archer ; akau-Twp, avenger. Latin has 
comparatively few words'in ter: frater, arbiter, magister; 
but the number of words in tor, such as dator and 
stator, monitor, bellator, pastor, gumstor, is considerable ; 
from these proceed, as we have seen, the future par- 
ticiple, futurus, daturus, natura, and the desideratives, 
pasturire, esurire, emptuire, &c. 

Zend and Persia,n are also without the vowel r, and 
present the same phenomenon as our two classical 
languages. In the inscriptions of Persepolis, harta, 
Jcarta, correspond to hhrta and krta. Tarsno is trshna, 



Indo-European Phonetics. 365 

thirst ; farst is prshtha, the back. Zend prefers ere, 
are. Sukrta, well made, becomes hu-kereta, hu-kareta ; 
hhrta, bereta ; Vrtra, Verethra ; prastrta, extended, 
Jraqtareta ; hrdaya, heart, zaredaya ; prthu, wide (Gr. 
TrXarus), perethu, whence Huperethu, Hufratu, the 
Euphrates, the wide river. 

The vowel I was invented by the grammarians of 
India for symmetry ; it exists only in one Vedic word, 
kip, to succeed (perfect tchaklpe), participle klpta, well 
made, constituted ; whence kalpa, rule, order, one of 
the great periods of the Brahmanic universe. This 
single root, kip, will account for one of our most useful 
and most frequently employed words. 

Since the two liquids I and r were not clearly dis- 
tinguished the one from the other before the separation 
of the idioms, we have more chance of finding the 
root kip under the form karp or kerep in the Indo- 
European languages, and the Veda itself presents it 
to us under this form : uttisMha daivyd krpa, " arise 
in thy divine form." What is this p joined to kr ? 
Probably the remains of a suffix of causation, pa ; kr 
{creare), to do ; krpa, to fashion ; whence the Zend 
hu-kerep ta, well made, beautiful ; karefs or kerefs, the 
body, genitive kehrpo, kehrpam, accusative kehrpem. 
Old High German has href, Anglo-Saxon hrif, the 
body, the womb (cf. midi-rif). We have here the 
original of the Latin corpus, an isolated word, which 
has no allies in Greek, nor even in Latin, and which 
has come down to us with its numerous derivatives, 
corporeal, incorporate, corporation, corpitscle, &c. 

The Indo-European vowel system may be represented 
by a triangle, of which a, short and long, occupies the 
summit, i and ou the two lower angles ; beside the two 
last are placed the two diphthongs ai and au, which 



366 The Indo-European Organism. 

have given to Sanscrit its long i, and 6. The short a 
has furnished to all the idioms, except Sanscrit and 
Gothic, the short « and ; from the long a arose the 
long S and 6 of all the Western languages. Side by- 
side with the primitive i and cm, a secondary i and ou 
were developed as weaker forms of the a, by the inter- 
mediary of the short e and 0; a sound u (upsilon) 
served as link between oil, and i. Two semi-vowels, 
the labial v, the dento-guttural y or /, issued from ou 
and *; and perhaps a liquid r was the result of a 
guttural pronunciation of the a. Finally, a naso-dental 
closely connected with i, e, and oio, modified from the 
beginning the timbre of the vowels. The few examples 
I have been able to give will have shown what Greek, 
Latin, and Zend owe to the various uses of these 
resources. 

Letters, says Moliere, are divided into vowels, so 
called because they express the voice ; and consonants, 
because these sound with the vowels and serve to mark 
the various articulations of the voice. But the science 
of phonetics brings an amendment to this summary 
definition ; it points out that the vowels, or pure con- 
tinuous sounds, are connected with the explosives by 
mixed sounds, which may be at once and by turns 
continuous and explosive, the semi-vowels, liquids, and 
spirants, which partake of both natures. 

In dealing with the vowel i, we have seen that in 
Greek and Latin, in the middle of a word, it frequently 
represented the sound y (ie), which belongs to both 
vowels and consonants. Distinctly retained at the 
beginning of words in Zend, Latin, and Teutonic, this 
y has undergone various alterations in Greek which 
have changed it, now into an aspirate, now into a soft 
sibilant, now into a dental. Finally it became con- 



Indo-European Phonetics. 367 

founded with, a softened, palatal guttural, genus, genre, 
and took in Prench the value and the sound of/. These 
changes are partly determined by the proxinfity of 
certain soft consonants; but some can only be explained 
by a defect of pronunciation proper to the Greeks. 
Where Sanscrit has yuhsma, we, Greek has va-fie^, 
vfjLfies (classical vfji.ecs) ; yas, yd, yat, the relative pro- 
noun, becomes oy, )?', o ; the root yag, to sacrifice, to 
sanctify, whence the Persians have derived Yaxata 
(Ized), the name of the divinities who form the escort 
to Ahuramazda, furnishes to Sanscrit the verbal ad- 
jective yagyas, worthy to be sacrificed ; this same word 
must be recognised in the Greek dfyjoj, saint. Aspira- 
tion is the great resource of the larynx when embarrassed. 
Greek, in using it here and in many other circumstances, 
as we shall see, does not differ from Spanish and 
Plorentine, of which the one replaces the y and the x 
by the/oto, Quijote, Quexada, Jeres, or the / by an h, 
hijo, humo, Jiahlar ; the other, initial c or ^ by an h, 
hasa, house, hrazie, thank you. 

Z is another sound which imitates a certain number 
of allied sounds ; half way between the gutturals and 
the liquid dentals, z is as near as possible to the semi- 
vowel y. In Zend, Yama, the god of death, has 
become Djem-schid from Yima-Kshaeta, as if a c^ had 
placed itself before the inconvenient syllable. This is 
pretty much what has happened in Greek in the case 
oftevyw/M, ^vyov, e^yyi;!/, compared with yugam, jugum, 
jungere ; with pw, I boil, corresponding to the Sanscrit 
yasdmi; with barley, from yava. The Greek trans- 
cription of the Latin words Julia, Jesus, Maia, conjux, 
was very often ZovXla, Zijcrv, Ma^a, ko^ov^. 

What was the exact pronunciation of this z? Slightly 
dental no doubt, more like dz than z sibilant. This is 



368 The Indo-European Organism. 

indicated by the change of Byaus into Zevs, and of all 
the verbal terminations in dy6 into ^w. A slight 
effort of the tongue applied to the palate produces a 
prosthetic d; from the confusion of the two sounds 
arose dz ; the reverse order appears to have the same 
effect : when the d is preceded by an s, the same z 
replaces the group sd : a-vpiCw for crvptaSw. Perhaps 
the group sd represented the simple z, itself the result 
of very various origins. In any case, the primitive 
sound y (the agreement of Zend, Latin, and the 
Teutonic languages witnesses to its antiquity) is not 
less related to the consonant d, perhaps to the con- 
sonant g, than to the vowel i. 

V, which is rightly called consonantal u, offers even 
more striking examples of these transitions between 
vowel and consonantal sounds. It presents itself 
either as the substitute of u before a vowel : divas, 
genitive of dyu ; or as the link between u and a 
termination : gravas, dya, vas ; dhenva, with a cow ; 
or in a group beginning with a sibilant : sva, a 
guttural ; gva, leva, or a dental ; tva, dva ; or finally 
in the pure state: va enclitic, or va suffix (sarva, 
vigva, all ; nava, now), vant, termination of Sanscrit 
adjectives and participles ; vd, to blow ; whence vdta, 
the wind ; vid, to know, to see ; vagh, to drive ; varn, 
to vomit ; vak, to speak ; vart, to turn ; avi, sheep. 

Classic Greek has no character to represent v, but 
it had existed. It was the Phoenician vav, which 
remains, for the rest, in Greek numeration, where it 
represents six. Grammarians have called this letter 
digamma, because its form resembles a double gamma ; 
it may still be read on a few inscriptions : polSa, foiKOi- 
for oiSa, oiKOi (Latin vdcus, vicTis, Sans. viga). The 
Beotian, Laconian, and ^olian dialects long wrote 



Indo-European Phonetics. 369 

and pronounced it : fepyov (work, root varg). The 
digamma has disappeared from the Homeric texts as 
they were compiled at the time of Pisistratup, but it 
certainly existed in the older songs from which the 
editors of the sixth century composed the Hiad and 
the Odyssey. Many verses would be incorrect if it 
were not supplied. Here are two : — 

Tov Se ap iivoSpa ISuiv Tro'^a? w/cw 'A.-)(tWevf. 
ImQwv re priTrjp efievai TrpijKT^pa re epywv. 

To restore the rhythm we must read fi^wv, fprjT^p, 
fepywv. Philologists have noted hundreds of similar 
passages. 

The letter at last fell into desuetude, and the sound 
with it. It is very singular, when we recollect that 
Greek pronunciation had no repugnance to the sound 
V ; far from it, the 6 and u have very often that 
sound : faa-iKefs. One would expect at least to find the 
primitive v represented by one or other of these 
characters ; but this is rare : we have fiovKoixai, ^ovkri, 
compared to the Sanscrit val, and with mlo and 
voluntas, and a few Laconian forms such as ^epyov, 
jSiSeiv, /3epyi\ioi. In yovvos from lyow, knee ; Soupos, 
from Sopv, lance ; vevpos, -Travpo^ (compared with 
parvus and nervus), u seems to be the equivalent of v. 
F" is M in Svo, in ^ovg, and in virvos, for svapna. 

It is probable that the digamma was abandoned 
because b and u were commonly pronounced like it; 
but its loss has none the less materially altered the 
physiognomy of the language, and separated Greek 
from its congeners, Latin and Sanscrit. Sometimes 
the soft breathing, as in e/xw (emetic) for vam, e;(5, 
for vagh ; the hard breathing in evvvfii, ea-Oiis (vestis, 



370 The Indo-European Organism. 

vestire), in hrcop, witness, historian, in ea-irepof (vesper, 
the evening), takes the place of the v ; sometimes it 
disappears completely : SmSsku, duodecim ; onrXovf, 
duplies ; ^069, lovis; vrjos, navis ; yios, novus ; evvea, 
novem ; ok, ovis ; pew, irKew, veco, eTSoii, (cXeo?, and 
the suffix K\t]i (grava), instead of /cXe/roy, efiSov, pefw, 
TrXlfft), vef^w. It is true that in the case of the three 
last the forms TrXevoo (pluo, pluviiis), pevdo, veuce, present 
equivalents. But who would recognise in b\oi for 
iTo'Xfoy the Latin salvus, in Xaio's the Latin Icevus, in 
(TKaios, sccevus, in ^Sus, suavis, in vXij for auXfr], silva, 
in 0? for avoi, the theme sva, the Latin suus ? Pro- 
bably the V was not precisely either h or u; it was 
the w, as is shown by the transcription OuepytXios 
(Virgilius) ; and this sound was not familiar to Greek. 

Latin is here much more faithful to the original 
type. It has vinum, vicus, verto, vigor, valere, volere, 
videre, veJiere, vir, vis (Greek t's), nervus, parvus, salvus, 
novus, curviis, eqvus, loves, novem, divits, Jovis (Atoy), 
arvum, alvus, , that which is ploughed, from arare, that 
which nourishes (cf. alu-rrmus). Yet two special 
causes have often in Latin brought about the modifica- 
tion or the loss of the v. The first is the confusion 
with the sound and the letter u ; the second the 
tendency to contraction. 

Latin had attributed to the spirant / the vav of 
the Phcenician alphabet, and therefore had but one 
character to represent the ou vowel and the ou con- 
sonant. Claudian, who had a few valuable ideas, had 
proposed to represent the pure v hj & reversed /, 
which is found on a few inscriptions of his time ; but 
custom was too strong for his reform, and we do not 
know whether quatuor, vacims, reliquus, fatiius, eqwus, 
were pronounced quatvor, reliqvos, fatvos, egvos, or 



Indo-European Phonetics. 37 1 

Icatour, vacous, reliqous, &c. In duo and duplex, the v 
was a true ou, as in iouni, iubus, for lovum and hovibus. 
In fe, se, stJi, <^zo, <?m, rfi'cs, the ■« had simply dis- 
appeared ; a species of friction had obliterated it. 
Sometimes it is hidden in a contracted syllable : nolo 
for non volo, nuper for novumper, nauta for navita, 
upilio, shepherd, for ompilio (organic, avipalayan), 
prudens, for providers, (the prudent man is he who 
foresees) ; seorsum, seversum ; rursum, anew, for re- 
versum ; elsewhere it leaves no trace ; compare retror- 
sum with retroversum ; sursum with suhversum ; concio 
with conventio; ditior, junior, nonus -with dives, j'uvenis, 
novem, ; malo, commorunt, petii, probai, with magvolo 
(which is partly recovered in mavult), commoverunt, 
petivi, prohavi. A primitive aiva-s, time, from which 
the Greek derives atoov, alei, gives in Latin cevum; the 
addition of the suffixes tas, ter, nus, lis, cus, lengthens 
it into cevitas, ceviternus, ceviternalis, ceviternitas, cevita- 
ticum, of which we know the Latin and French forms, 
cetas, OBtemus, ceternalis, iternel, iterniU, idage, iage, 
aage, and finally dge. Thus age and eternity are 
without a doubt derived from the same root. In the 
second of these words the initial e still represents the 
first syllable of cetas ; but in dge, the circumflex accent 
alone reveals the strong contraction which has devoured 
the significant parts of the word. 

Zend uses v largely ; and though it often changes 
this letter into u (dceum for dcevam, the demon), it 
sometimes retains it where Sanscrit alters it : vavatcha, 
"he speaks," is nearer the primitive form than the Sans- 
crit uvatcha; vaz (vagh), viz (vid), vig (inhabitant), 
haurva (sarva) havami, vayu {vayou), the wind, are, as 
far as the retention of the v is concerned, faithful to 
the organic form ; w in the accusative thwam, thee, 



372 The Indo-European Organism. 

in rathwo, rathwe, of the master, to the master, is only 
a reinforcement for the sake of euphony. 

There are other metamorphoses which we shall en- 
counter here and there in Greek and Latin, dependent 
on the consonant with which v is associated in a sort 
of indissoluble group. Kva, gva, sva, dva, are usually 
considered as a species of national caprice, a pheno- 
menon attributable to a more or less tardy preference, 
and presenting interesting but entirely secondary varia- 
tions, as, for instance, in Wilhelm, William, Guillaume, 
Gilles. I am rather disposed to recognise in these 
groups very ancient transitions between the semi-vowels 
and the consonants. 

If we compare the relatives has (Sanscrit), ttoi? or 
/cto? (Greek), with the Umbrian po-ei, with -the Latin 
qui and quis, with the German kea, " who," we shall see 
that they only differ by the abandonment or the feten- 
tion of the semi-vowel v. The original theme appears 
to have been Jcva. The sam.e observation applies to 
the forms kvwv, canis, gvan, Zend cpan. The root kvit, 
gvU, to shine, retained in the Sanscrit gvetas, Gothic 
hveits, German weiss, English white, gives in Zend 
gpaeto, brilliant. Whence are these w, v, p, if the v, 
vocalised into u or hardened into p, be nob part of the 
original root ? The ^ of the Zend and Sanscrit words 
qvan and gpan always represents a primitive Tc, as in 
agu, rapid, Greek owcyy, aqva, Latin eqaus. The Zend 
aqpa, vigpa, all, helps us to understand 'hnros. The 
Greeks received the horse from Asia, where it was 
already called aspas or ispas (Ispahan) ; the s, a weak 
representative of the primitive k, became assimilated 
to the p, in which the Greeks no longer distinguished 
the original v; certain dialects, which confused the 
Bounds p and k, adopted the form uckos, which corre- 



Indo-European Phonetics. 373 

spends only in appearance with equus. The two words 
are in truth derived from the same root, but they 
have a different descent. But the point of departure 
of these changes is the presence of a v in the root. So 
with quatuor, tchatvar, Tecrcrap, irlcrvp, patur, fidvor, 
the type is hvatrar. 

The soft guttural g, often palatalised into the Sans- 
crit dj, frequently corresponds in Greek and Latin to a 5 
or to&v: gaus is ;8oi/j, hos; ga (Sans. gam),to go, becomes 
in Greek /3a/i/co, in Latin va-do ; djiv, to live, becomes 
/3/oy, for ^Ifo^, vivere, vita. Zend has zbayemi, I in- 
voke, for ghvaydmi, Sans, hvaydmi (whence hotar, 
the priest, in Zend zaotar) ; in the same way the 
differences between our classical languages reveal to 
us the primitive forms gvaus, gvam, gviv. The first 
has preferred the guttural, the others the v pure, or 
hardened into I. 

The example of the groups tv, dv, lends support to 
this explanation. We will not recur to the changes 
of tvam into te, into tibi (for tve, tubi), except to show 
that vos (Siins. vas), probably the remains of a plural 
tvas, is connected with it. But dv has a special in- 
terest ; the Zend dbis, to hate, for dvish, bae for dvS, 
show us the evolution of v into b; much more so 
Latin : compare duo, duellum.^ dvxinus, duplex, with 
bis, helium, bonus (Ital. buono, Span, bueno), hidens. 
Greek, rejecting the u, has kept the dental : SuiSeKa, 
Si^, for SvoSsKU, Su'l's. Thus these are forms of the 
same root, not because d changes into b, but because, 
when two sounds are combined to make one group, the 
different languages have made a different use of them. 

The group sva has an equal number of variants. 
It remains pure in Sanscrit : svan, to resound ; svar, 
to shine, the heavens ; svagura, father-in-law ; svasar, 
25 



374 The Indo-European Organism. 

sister; sxa, he; svadu, sweet; svapna, sleep. Latin 
lias rarely retained it (suadvis, suadere, suus) ; it is 
generally contracted into so : sonus, soTUire (note the 
Italian suotw) ; sol, socer, socrus, soror (for sosor), somnus. 
Greek has made an effort to imitate the sound in crcpoi, 
(r(perepoi, but it has abandoned the v, and never likes 
the initial s, which it changes into the hard breathing ; 
consequently this sound is usually represented by e, !?, 
V, 6 ; jJXtoy, eXX^vrj, for svarya, surya, svaranu, the 
brilliant ; eKvpos, for svagura, socer ; i/ttj/o?, for svapna ; 
^Sui, for svadu. 

But it is in Zend that the difference is most consi- 
derable. The sibilant is rejected at the beginning, 
and even in the middle of words, and is supplied by 
a very marked aspirate, sometimes accompanied by a 
nasal ; hasanhreni, for instance, represents the Sanscrit 
sahasram, a thousand. Sva, hvare, correspond to sva, 
svar. The initial aspirate is often so strong that it is 
equivalent to a hard guttural ; in that case all trace 
of the V disappears, and the sibilant becomes q. Qa 
represents sva. Qddata, which has been modified into 
Koda, a god, Qddata, the eternal, he who finds in 
himself his law, his basis, is the Sanscrit sva-dliata, 
and might be found in Greek under the form trcpederoi, 
auToderos. Qafna \s, svapna, v'wvo's, somnus ; qap, sopire ; 
qar, to shine, and to eat, from svar, svorare, vorare ; 
qan, to resound (cf. the Latin canere, canorus) ; qanhar, 
svasar, soror ; Haraqaiti, land of rivers, Sarasvati. The 
Greek transcription, Arachosia, allows us to surmise 
that a, V or an. was still heard after the aspirate. 

After these various examples, no one would venture 
to assert that the labials b and p are not a hardening 
of the 1), which itself is issued from the vowel ou, 
and that the gutturals and their aspirates are not 



Indo-European Phonetics. 375 

the result of the effort made to vary or to utter the 
sound V ; this effort would have produced at first a 
confused sound, gradually growing more precise as the 
friction of agglutination obliterated the semi-vowel. 
This is perhaps more than a hypothesis. In any case, 
these considerations will not appear out of place in a 
philosophical study of language and the formation of 
sounds. They seem to grasp those physiological causes, 
so long ignored, of the deviations which concealed from 
the Latins, Persians, and Teutons the fundamental 
identity of their vocabularies. 

The pure sibilant, s, which has been so differently 
treated by all peoples according to its position, whether 
initial, final, or in the body of a word, and according 
to the letters, vowels, or consonants with which it 
found itself in contact, may be ranked among the first 
elements of language ; it is common to man and many 
animals, such as the goose and the snake ; it is even 
one of the sounds of inanimate nature. Human breath 
produces it in the same manner as the breath of the 
wind. S is sometimes, and rightly, classed among the 
dentals ; through the z it is allied to d, pure or aspi- 
rated ; it constantly permutes with t; but ch connects 
it also with q, with the palatal tch, and through these 
with the gutturals. Lastly, with the Persians and the 
Greeks it is confounded with the aspirate ; with the 
Latins, with the liquid r. These manifold affinities 
render its history most complicated. 

We will be content with a few rapid notes. In 
Sanscrit the primitive s is sometimes replaced by the 
palatal g : gushlca, Latin siccus, Zend htoshJca ; gvagura, 
Latin soeer ; by sh : usli, to burn, Latin cestus, ustus 
{aurora, atinim, irom urere) ; tarsh, to dry, Greek Tepaoo, 
Latin tersa, terra, the dry land. Often before certain 



376 The Indo-Etiropean Organism. 

consonants, and according to euphonic laws proper to 
Sanscrit, it is dropped in the terminations, and its 
place taken by a slight aspirate ; the a which precedes 
it becomes o ; divas, divah, and ' div6 ; this pheno- 
menon is common in Zend : MazdaS for Mazdas. It is 
frequently dropped in Greek, and in the older Latin 
poetry : ovmibu', intu', in Ennius. Otherwise it keeps 
its original value in Sanscrit. Many roots begin with 
the sibilant : sad, to sit, whence sadas (Greek, eSo^) and 
sadanam ; satch (for the organic sah), to follow ; su and 
sil, to water, to bring forth, whence soma, the sacred 
liquid, and sunu, son ; sru, to flow ; svid, to sweat ; 
stha, to stand ; star, to stretch ; smar, to remember, &c. 
S persists also (except in the cases above mentioned) 
in the middle and at the end of substantives and 
verbs : asti, hharSs, ahharas, svasar, vahchas (yocis\ 
manas. 

We have already seen what changes initial s suffers 
in Zend and Greek. According to a constant law, e§, 
e^oy, eSpa, correspond to sad, sadas, so that in the 
French words chaise, chaire, chaydre,. derived from 
KuOeSpa, nothing recalls the root sad. Saptan gives 
eiTTu, whence h^bdomadaire, which as little resembles 
semaine ; sah becomes ex, 'iirofiai ; sru, pv ; vw, to rain ; 
and uioy, son, are in the same relation as su and sunus. 
The disappearance of s between two vowels has thrown 
confusion, into the declension, and even into the con- 
jugation, aypov, \6yov, Oeov, conceal aypoio, a'^pocrio, 
Xoyoa-io, 6eo(tio; fiivovs, yevous, fieveoi, fievea-os, yeve(ros; 
elt]V, e(Tit]v ; (pept], (pepeai, (prjpea-ai ; a>v, " being," ewv, 
ea-wv ; XeXuKvia, XeXvKocria. S falls before and after r : 
pew, Tranjp, )(eip ; it is often substituted for t, rea-a-apes, 
QaXcuraa, /xeKurcra, iroiovcri, (pepovcri, " they do," " they 
carry," xoVi? (pati), (picric (Sans, hhuti), fioOa-a (montia), 



Indo-European Phonetics. 377 

all the ancient ablatives in cot are become adverbs in 
ft)? : ofiwi, TrjKiKWi, a\>]6wg, aoipwg, &c. 

It is a curious fact that, in spite of these anomalies, 
s is still in Greek one of the letters which occurs most 
frequently. Sometimes it has been preserved at the 
beginning of a word by the consonant or vowel which 
follows it : crdevos, force ; a-Tevog, groan ; crT^aw, I will 
place ; arTopvu/jii, I stretch ; aKwp, crKaroi, whence 
scoria and scatology (which must not be confounded 
with eschatology, a science consecrated to metaphysics), 
criTos, bread (perhaps ^veta, white), creX^vi], crelpioi, for 
svar ; lastly, all the words with the prefix cruv. These 
are departures from the more general use ; but we 
must admit that the pronunciation changed in process 
of time, and that the double forms Selene, Helene ; Sellos, 
Hellen; Seirios, Helios, belong to different periods of 
the language. Sometimes s is maintained in the body 
of the word or in the terminations : ecrofiai, \vcra), 
eXiKTa, \vcras, Xvcrav. Finally, it abounds, in general, 
in various cases of the declension, notably in the 
nominative and genitive singular, and in the nomina- 
tive and accusative plural : Sofiog, veavlas, iroSos, ottos, 
TTo'Xeos ; Trare^ey, avSpai, \6yovi. 

Latin, which has none of these aversions to certain 
sounds, especially sounds so easy to utter as the y, the 
V, or the s, has preserved better than other languages, 
even than Sanscrit, the Indo-European pronunciation. 
It allows the s at the beginning of the word, as well as 
y or j, unless a too great accumulation of consonants 
necessitates the abandonment of some of them. Locus, 
the grammarians tell us, stands for stlocus ; lis, trial, for 
stlis (from a root sir, Ger. streit). But s pure,- or accom- 
panied with a moderate number of other sounds, remains 
unaltered in all the words which we have quoted : sedes, 



378 The Indo-European Organism. 

sedeo {vooi sad, Gr. e^), sudare, sudor {root svid, Gr. vSayp), 
sonare (svan), sequi, septem, serpere (ep-rrw, epireTov), 
servus, scdmos {sarva, oXos), and soUios, sum, &c. ; and 
also everywhere that we find it in the Greek termina- 
tions : dominm, dominos, dominis, genus, tempus, manus, 
manihus ; fratris, fratres, fratrihus, dies, materies, &c. 
Yet, as we have said, there is a tendency to efface the s 
in the terminations. I will add a few examples of this 
to those already given: vita dignu' locoque ; omnihii' 
princeps ; on old inscriptions : Furio, Terentio, for 
Furius, Terentius ; even in a more dignified style, 
exoriare aliquis nostris ex ossihus idtor ; amare, sequere ; 
for exoriaris, amaris, sequeris ; mage for Tnagis (magius), 
pote for potis. Messala, says Quintilian (in the time 
of Augustus), wished to omit the final s in writing. 
It has disappeared from the middle of a few words, 
such as pono (posno) compared to posui, positus, repostum; 
remus, the oar, for resmus (Gr. epedfxoi) ; Camillus for 
Casmillos ; camena for casmena (root fas, gans, to sing). 
But the most interesting fact about s in Latin is its 
affinity with r : nasus, naris ; carmen for casmen ; ara 
for asa. In a whole class of words between two vowels, 
of which the last is an i or an e, sometimes an a, 
the primitive s is constantly changed into r. This 
phenomenon affects, for instance, all the words of the 
third declension which end in lis short ; in Us, in as, 
and in os. In corptis, genus, onus, pecus, opus, latus, 
vetus, scelus, Venus, jus, mus, mas, glis (dormouse), which 
correspond to Greek words in oy, to Sanscrit words 
in as, s is part of the root, and is found in such deri- 
vatives as mu^culus, mascidus, corpusculum, onustus, 
venustus, vetustus, scelestus, Justus ; so with arhos, 
arhustum ; Jionos, honestum ; roiur, roiustus ; flos, 
Jlosculum ; but all the oblique cases are declined upon. 



Indo-European Phonetics. 379 

a theme in or, er, or ur : muris, maris, glires, corporis, 
pecoris, juris, generis, operis, lateris, veteris, sceleris, 
veneris, roborisj and in the second series, arboris, 
honoris, floris, odoris, coloris, laboris, &c., and these 
forms, reacting on the nominative, have in almost all 
cases changed os into or, arbor, honor, color, odor, labor, 
sapor, amor, robur. The nominatives in us have per- 
sisted ; but r has triumphed in the derivatives, and 
has given us a whole number of verbs, nouns, and 
adjectives which are found in all the Romance lan- 
guages. Opera, oeurre, operare, ouvrer, and operate, 
operarius, ouvrier, urere, aurora, for ihsere and ausosa ; 
jurare, juratus, perjerare, generare, generosus, sceleratus, 
veternum, venerari, onerare, onerosus. I may mention 
also the French words arbre, arbricole, honneur, laheur, 
odeur, couleur, honorer, honorable, colorer, corroborer, in- 
corporer, odorant, veteran, laborieiix, liqtooreux, savourer, 
savourevx, all of which are taken straight from the 
Latin. The Latin conjugation has been no less in- 
fluenced by this permutation of the s into r, if indeed 
it be true that the re is everywhere the substitute for 
se in the infinitives. This is my opinion. It seems 
hardly possible to separate esse, posse, fuisse, amavisse, 
legisse, audiisse, from legere, audirc, amare, fore, gignere, 
especially when we compare them with amore, genere ; 
the same cause alone could produce effects so similar. 
The passive is entirely founded {amor, amaris, amamttr, 
legor, legeris) on the same metamorphosis of the s. 
The attraction of the r must have been very powerful 
to have lightened the suffix j'ans, ion, of the compara- 
tive into or ; major, pejor, suavior, &c. (corresponding 
to the Greek neiQov, ^Slwv), which have passed through 
intermediate forms such as magions, magios, majosis, 
majus ; the earlier s is still heard in mc^cstas. No 



380 The Indo-European Organism.. 

accident of phonetics, except perhaps the rejection of 
the aspirates, more clearly separates Latin from its 
Indo-European kindred. The substitution of the liquid 
for the sibilant remains one of the principal charac- 
teristics of this language, and of the Eomance lan- 
guages derived from it. The Latin r is not only allied 
to the sibilant, it is also akin to the soft dental : irieri- 
dies, stands for medi-dies ; we find arvcna for advena. 

The liquid seems to have as many affinities as the 
sibilant ; both may be at once guttural and dental. 
In the Semitic languages the double guttural aspirate 
is confounded with r : GhadamAs is pronounced Rhada- 
TnAs. In Greek it is easy to distinguish two pronuncia- 
tions of the r, aspirated at the beginning of words 
{pew, priyvvfii, &c.), soft in the suflBxes pa or po, 
which occur so frequently in our languages, ^fiepa, 
Xvpa, avKvpa, la~^up6s ; its influence even adds an 
aspirate to the preceding consonant, epvOpoi, red, 
avdpwTTOi for avSpocnros^ (ppuTws, as in the Sanscrit 
hhratar; when two r's come together, the first is marked 
with a hard breathing, the second by a soft breathing. 
This aspirate is constantly attached to the initial r, 
which is always announced by a labial breathing ; 
whence the dialectic variants ^poSov, the rose ; ^pijTwp, 
the orator. 

The primitive liquid was a sound which hesitated 
between r and I. In the Veda the Sanscrit root lith, 
Greek \e[-^oo, Latin lingo, French lecher, is still written 
rih ; flu (vrXeo), pleure, Jluere) appears as pru. Zend 
and Persian are without the letter I. In other lan- 
guages there is a perfect equivalence between the two 
liquids ; at least, in the middle of a word euphonic 
reasons alone determine the use of the one or the 
other ; Greek has Ke(f)a\apyia, XeOapyla from aXyos, 



Indo-European Phonetics. 381 

pain (Latin algere, algidus), to vary the sound ; in the 
same way Latin uses indifferently the suffix aris or 
alis : salutaris, floralis, militaris, mortalis, normalis, 
jocularis. But at the beginning of a word, Sanscrit, 
Latin, Greek, and the Teutonic tongues have generally 
adopted either Z or r once for all, and use it con- 
sistently. Sanscrit, for instance, has chosen lump for 
the meaning to break, Latin rump ; Sanscrit has rik, 
ritch, to abandon ; Greek Xiir, Xeliro), \0t7r0y ; Latin 
liv, reliquus, relictus, linquere. " To shine " is in Sanscrit 
ruk, rutch, in Greek Xey/c, \vk, XevKOs, white, Ay/cetoy, 
the Lycian Apollo (the shining one), in Latin luc : 
lucere, lucidus, Lucina, lu{c)na, lu(c)men, lusciniola, the 
nightingale, which sings towards the day ; here French 
has shown a tendency to revert to the r ; scandalum 
gives esclandre, apostolus, apotre. 

N, which we have already encountered as the nasal 
utterance of the vowels, is reckoned among the liquids, 
for its independent existence is anterior to the separa- 
tion of the idioms ; it is none the less nearly related 
to the i, which often takes its place in Greek termina- 
tions ; one of the most visible traces of this affinity is 
the addition of n to many Greek terminations in i and 
e : diCKovcriv, they love, etroLrjo-ev, he made. Greek 
makes almost as much use of n as German. As a 
liquid, n is rarely confounded with I or r; yet \v/ji.(f)a 
and vufjLCpa are clearly the same word, and donum and 
planum are the equivalent Latin forms for Swpov, 
TrKrjpov, in Greek. In ullus for unidus, hellus for 
bonulns, the 71 has been assimilated to the I. Finally, 
the French diacre for diaconus, the Proven9al canorgm 
for canonicus, is a further proof of the relationship 
among the three liquids. iV, as a purely initial consonant, 
persists in a great number of words, nac, to kill, necis, 



382 The Indo-European Organism. 

necare, veKVs, corpse ; Tiam, to take, to rank, vefieiv, 
vo/j-oi ; nava, novus ; nara, nero, avijp, man ; but more 
frequently it is the remains of a guttural-nasal group, 
gn : yvtjTOs, gnatus ; yvarros, gnotus, notvs, nomen, 
noscere, Eng. know. 

M, at once liquid and labial, preferred in termina- 
tions by Zend, Sanscrit, and Latin, familiar also to the 
Teutonic languages, alternates with n in compound 
words, according to euphonic laws peculiar to each 
tongue ; it replaces n before labials : emmenei; embon- 
point, oifi^poTo?, immortalis. At the beginning of 
words it is simply a soft labial or the remains of the 
group sm : smar, to remember, smemor. I give a few 
examples : mar, to die, mors ; marta and mard, man 
in Zend ; Ariomardas, the famous mortal, Gayomareta, 
mortal life, the first man (GaiomMrz in Pehlevi); marg, to 
mark ; Harms, the distinguished ; margo, mark, margin, 
march (border) ; mrdj, to knead, to soften ; midcere, 
mulgere, mollis (for moldis) ; mardjara, the Sanscrit 
name for the cat, that which is always licking or 
polishing itself ; man, to measure, to think, Tnanu, man ; 
lastly, ma, me, the famous ego which has played so 
great a part in history, in social life, and in philosophy. 
But here I pause ; we have already studied the root 
Tna and its derivatives. 



OHAPTEK VI. 
PHONETICS— 11. THE EXPLOSIVES. 

Consonantal cjiphthongs — Indecision and variations in the primitive 
consonants — Attempt of M. Regnaud to trace back all articulate 
sounds to a group, sh or keh — Numerous examples — The aspirate 
is the link between the consonantal group and the pure consonant 
— The palatals in Zend, in Sanscrit, and the Romance languages — 
The substitution of consonants in the Teutonic languages — First 
stage : Gothic and Low German — Second stage : High German — ■ 
The three periods of German — Numerous exceptions to the law of 
permutation — A sixteenth-century English lesson. 

We have shown by what insensible transitions the 
semi-vowels, the spirants, the sibilants, the liquids, 
and the nasals have detached themselves from the 
continuous sounds, and attained the position of inde- 
pendent consonants ; we have reached articulation, but 
it has yet to frame the explosives, sounds which cannot 
be pronounced without a supporting vowel or con- 
sonant preceding or following them. 

But it seems that a stage, that of consonantal diph- 
thongs, still separates us from this last state. We have 
already considered a few of these in the groups kv, hv, 
dv, sv, sr, gn, but there are others, which are often 
explained as contracted syllables, and the juxtaposi- 
tion of explosives which are already distinct. This is 
often true, and it is easy to prove and to admit that 
suflBxing continually brings the different classes of 
consonants into relations with each other, and forces 
them into combinations which are modified according 

383 



384 The Indo-European Organism. 

to the euphonic laws of each idiom. Yet when alliances 
of this nature occur in the root itself under its most 
primitive form, when the groups issued from this con- 
course of consonants are differently heard and pro- 
nounced by the different idioms; when we note, moreover, 
that there have existed, from the time of the common 
Indo-European speech, numerous forms derived from 
these by the loss or weakening of some one of their 
elements, there is reason in the inquiry, first, whether 
these groups are not earlier than the utterance of the 
pure consonant ; secondly, whether they do not belong 
to the age when the primitive gutturals, dentals, and 
labials were still confounded; lastly, whether they do 
not allow of the reduction to a hundred, even to 
twenty types, of the fifteen hundred Sanscrit roots, or 
the five hundred called Indo-European ? Such is the 
theory, if I do not mistake him, of the eminent philo- 
logist, M. Paul Regnaud. He argues it with an ardour 
which is most attractive, if not entirely convincing; 
and, while leaving to specialists the task of controlling 
and checking in detail the numerous comparisons sug- 
gested and ably defended by M. Regnaud, we cannot 
refuse a general assent to his doctrine and his method. 
We have already quoted several of the series which 
he gives, where, alternating with ingenious boldness 
the consonants of the three orders, he follows at the 
same time the phonetic variation and the fluctuating 
evolution of the idea. Here, for example, in the 
famous class hrtlUr, briller (to burn, to shine), is what 
may be called a species derived from a type kshar, a 
type which M. Regnaud recognises under twenty dis- 
guises : kshar, kshir, krch, kar, hra, Jiar ; gharsh or 
ghrsh, ghar, ghrd, ganr ; tchar ; djurv, djUr, djval ; 
darg ; star, tarsh, trish ; sphurj, sphur, sphid, sphut 



Phonetics. — The Explosives. 385 

(for sphurt) ; sphar ; priksh, prish ; plush ; pal ; hhru ; 
ihradj, hharg, varJc, ark ; smarg, raarg, vnardj, mark, 
•mar ; svar, sHr, &c. And each of tliese roots, as it 
recedes from the type, gives birth to a family of words, 
all bearing some relation to the notions of burning, 
heat, or light. We leave Sanscrit, but Greek and 
Latin, which are nearer to us, owe to tliese variants a 
number of terms among which we can trace at least a 
collateral relationship. Thus we should rank under 
the first or guttural sub-group such words as ^i/troy, 
yjxtii, golden and colour, j^m^to, y^pujvwiJLi, to colour ; 
■)(\wp6i, green, pale; 'X^orj, turf, foliage; /caXXo? (Kapaoi), 
brilliancy, beauty; Kepavvos, lightning; -yXaii/coy, shiny, 
green ; yKavcra-O), to shine. Compare Xeva-a-w, Xyj^i/oy, 
find, from the root rukch, yKay, yaXwc, milk, and 
-yaXa^/ay, milky, yaXepos, laughing, yaXijvti, serenity, 
yXvKUi, yXevKo^, sweet, sweet wine, yX^voi, yXijvij, light, 
eyeball. In Latin, croc-us, corusc-us, col-or, cal-or, 
calidus, h(T)aridus, ardor, areo, area, arena ; clarus ; 
carlo : galbus, gilvus, yellow, green ; (^g)viridis ; (g)lact. 
In the dental sub-group : ddXXo}, OaXiroi, dripta, Oep/xof, 
SaiSaXXw (for SapSaptrai), to make shining, to embellish ; 
crre'Wft), to shine, to sparkle ; SepKOfiai, to see ; Teptrw, 
to warm, to dry ; Tpav^g, brilliant, piercing ; Tepa^, 
star, prodigy, portent ; torreo, torridios, tersus, sterla, 
Stella, astrum. To the labial variety belong irvpa-o^, 
torch, Trvpp6<!, burning red, irvp, fire, irvpsTO's, fever ; 
TToXio'j, wLite ; Latin, splendor, pallor, palam, pareo, 
puipura, purus, purg-o, pulcher, pol-io, harrhus ; Greek, 
cbpuyw, to grill, (pXo^, flame, (piXeya, iropcpvpa, irop- 
<puptt}, to die red, to redden, to tremble, TrpjiOw, to 
burn, irpriCTTrip, meteor, storm, Trifnrptiiu.i, to set fire to. 
Latin, fulg-or, fnlg-ur, ful{g)men, fulvus, formiis (hot), 
fornus, fornax ; frondis, flos, flagro, Jlagma, flagmen 



386 The Indo-European Organism. 

(who sets fire) ; flavus,far, &vA farina, haru (lightning); 
Greek, anapaySos, (TfjivpiXw, to polish, fjLapyapo?, mother 
of pearl, frnpimapo^, shining, marble, fiapiXt], Jiot cin- 
ders ; Latin, margarita, pearl, marmor, merits, mirus, 
brilliant, admirable ; lastly, all those words we know 
so well : ^€ipt09, creXay, lightning, fire, aeXijvt], eiprivri 
(for arueipijvrf), peace, Hermes (for SvevTnes), HdenR, 
the fair, &c. ; serenus (cf. Sirene) ; sol, Soraete (shining 
summit). 

This ramification admits, no doubt, of more than one 
foreign member, yet it would seem that all the branches 
are really connected with each other. M. Paul 
Regnaud's work has extended to thousands of words, 
each form being only ranked in a given category after 
it has been examined in every way, analysed and recon- 
structed. As a result of these researches, he has caught 
sight, with the faith of the seeker, of the primary group 
from which all consonants are issued. This micro- 
cosm is ska, a very common group in all our languages, 
shand, to mount, slcarp, to hollow, to engrave, and 
which often alternates with kcha, qa, gha. M. Regnaud, 
of course, considers that sk = st = sp = Jcs = ts= ps. We 
are not altogether able to share this faith ; but the 
antiquity of the vague group ska, kscha, or tchu, is 
manifest, and this ancient syllable has had the most 
varied fortunes. 

There exists in the Romance languages a word 
pecho, petto, poitrine, poitrail, which is obviously con- 
nected with pectus, pectorina, pedorale. All the Indo- 
European languages have lost this word except Sanscrit, 
where it appears under the form pahchas, flank, side, 
wing (of a bird), wing (of an army), party, opinion. 
It is remarkable how near in this instance the Spanish 
is to the Sanscrit form, both languages having pala- 



Phonetics. — The Explosives. 387 

talised one of the elements of the complex syllable in 
which Latin has distinguished a guttural and a dental. 
The same group evidently existed in the original type 
of the Greek ireKTwv, builder, mason, of the Latin 
tectum and tegvla, tile, words nearly allied to the great 
Vedic god Tvashtar, the architect, or the smith, only here 
Sanscrit has obliterated the guttural ; it has retained 
it, on the contrary, in rkchas, Greek apKTOs, Latin ursus. 
By a yet unexplained confusion, rkchas, very near to 
the Sanscrit arkas, the sun (cf. the Greek hero Arkas 
son of Kallisto, the nymph changed into the Great 
Bear), rkchas signifies both bear, and the constellation 
of which Plautns invoked the final star, Arcturus, " the 
tail of the bear ; " a softening of the root, ricM, has 
added to the confusion, so that Bear and the seven 
Richis, or wise men, are names for the same group of 
seven stars : septem striones or triones, the seven oxen 
or stars of our north. Similar coincidences are not 
rare : urka, the wolf, another probable variant, owing 
to its Greek name Xvko?, has been associated with 
Phoebus Apollo, the god of light. 

Nakch, night, vakch, voice, akch, the eye, are in close 
relations with nox, vvktov, with vox, Greek 6(^, with oc 
of oculus, and with 6(j)d of 6(p9a\ju.6i. 

Kchan, gans, qas, ghan, ghna, dhan, and han, in 
Sanscrit, signifying to kill, to cut, and analogous 
forms in Zend, Verethra-ha, Verethra-gna, slayer of 
Vritra, are sufficient to connect the Greek word Krelvw 
with ddvuTOi, OvijTOi, 6vt](TKw. Sanscrit kcha, kchdm, 
kchama, kchma, djma, gham, gam, gau, the earth, has 
its equivalents in the Greek Kaiw, to bum, x^'^"' 
^OafiaXos, X"M«') yaia, 7«. jn, ^V (iii Demeter, mother 
earth) ; in the Zend zdm, zdo, and zem, which is still 
the Persian word for the earth, the world (a prince of 



388 The Indo-European Organism. 

the "Arabian Nights " bears the name of Oamaralzaman, 
the moon of the world). In Latin, humus, humi, homo, 
previously referred to the root bhu (bhumi, the earth), 
may equally be attributed to gham, ghumus, ghumdn 
(the meaning is the same). 

When by the fall or effacement of one of the ele- 
ments of the diphthong,' the explosive, whether guttural, 
dental, or labial, remained alone, the aspirate took the 
place of the lost sound, marking the weakened effort of 
the throat, or as it were the surprise of man before the 
simple pure consonant to which his tongue at last gave 
utterance. This character of the aspirate escapes us 
at first, because different languages have kept certain 
aspirates side by side with the explosives, or rather 
have invented them to vary their vocabulary. Thus 
Sanscrit, not content with the soft aspirates gh, dh, bh, 
which seem to be earlier than the separation of the 
idioms, has developed a symmetric series of rough aspi- 
rates; while Zend, after rejecting the primitive aspirates, 
has adopted new ones before v and r : thwam for tvam, 
Mithra for Mitra. Greek, Latin, and the Teutonic 
languages have had recourse to similar artifices proper 
to each idiom. But when we remember that the ori- 
ginal aspirates permute or alternate either with each 
other or with pure consonants of the same order, and 
that use is constantly abandoned to dialectal or even 
to individual caprice, when we note, moreover, that h, 
the soft labial, did not yet exist, but was included in 
the aspirate bh, the transitional part which we attribute 
to the aspirates will be more readily admitted. They 
are intermediate between the groups of consonants and 
the distinct utterance of the explosives. 

We will consider them in Greek, where they are still 
marked; in Latin, where they tend to disappear, chang- 



Phonetics. — The Explosives. 389 

ing, as in Sanscrit and in Zend, into continuous sounds 
such as h and /. 

Greek replaces the soft aspirate by the rough breath- 
ing. Primitive : stigh, to walk, becomes (rreiyui ; vagh, 
to bear, to lead, 'i^w, o)(oi, chariot ; Ugh, to lick, Xe/^w; 
migh, to water, viegha, cloud, o/xi-^ew, /uoi)(o?, adulterer, 
the soiler ; avgh, to stifle, av)(o} ; ghima, winter, snow 
(Him§,laya, the abode of snow), ^eifia, yeiixwv, yj-wv ; 
axjhi, serpent, e;^tj ; Irghu, short, baghu, thick, laghu, 
light, ^pa)(ys, irayu's, eXa^w ; g/iansa, goose, x>?i'. 

In the dentals we have : dha, to place ; in Greek, 
TiOrifii, Qeros ; dha, to give suck ; QyjcrQai, rid^vr], nurse ; 
O^vs, feminine ; dhvar, door ; Oupa ; hidh, to know ; 
■iruvQavoixai, TlvOayopa^ ; dhuma, smoke ; Ovoo, to sacri- 
fice ; Ovfios, soul ; rudhira, red ; epvOpos ; dharsh, to 
dare ; Oapcros, 6paeru9 ; idh, to burn ; aiQw. 

So with the labials ; we know almost all these roots : 
hlia, to shine, to speak; c^ctoy, (pdcris, (pifJ-l; hhar, to 
bear ; (pepu), (popos, (ppaTpia ; bhrit. eyebrow ; 6(ppvs ; 
libhau; afKpoo; hhu, to grow ; (pvw, <piai^; nabhas, c\oud ; 
vicpos, veipeXt] ; grabh ; ypacpw, to engrave, to write. 

Sometimes, especially in the case of the guttural 
aspirate, Greek employs the soft explosive : e'ytoi', for 
the primitive ogham ; yews, cheek, for ghanu ; fxeyas 
for maghant, great. This proves that the Greek as[ii- 
rates were not so marked as the written character 
would seem to indicate, and that their true affinity is 
with the soft consonants ; and this explains their sub- 
sequent softening into ch, th, f. But it is not a matter 
of doubt that before the time of Alexander they were 
pronounced as double letters, and were long so written. 
The Latins of the classical epoch transcribed j^ctoy by 
cAoos, 'Aj(jXXew? by Achilles, ^Opcpevf by Orfevs, and 
even ^iXiTnros, AtcplXos, by Pilipos, Diphilos, and pre- 
26 



390 The Indo-Eiiropean Organism. 

cisely because the Latin/ a simple spirant, did not 
correspond in sound to the Greek ^. In classical 
Latin there are no longer any true aspirates ; the dental 
aspirate lias left no trace, or is confounded with the 
labial aspirate ; the latter is represented, as we shall see, 
by /; the guttural aspirate by a faint sibilant, soon 
neutralised, h. But in most cases' it is the soft ex- 
plosive, d, g, I, which has taken the place of the 
primitive aspirate. Bare, condere, suldere, root dha ; 
credidi, graddadhami ; medius, mad/ijas ; cedes, hearth, 
temple, house, root idh ; ad, adhi ; sometimes a t corre- 
sponds to a th : latere (XavQavw) ; pati, to suffer, Greek 
eTraOov; rutilus, Sanscrit rudliira. The primitive gh 
has disappeared in ego, magnus, lingo, and ligurio, mingo, 
angor, angusius, anguis, vecsi, vectura, gratus (kharis), 
pi.ngtiis, legvis, hregvis ; the ih in amho, nubes, nebida, 
nimbus, luhet, libido (cf, Eng. love, from luhh), ibi, uhi, 
partibus, amabo, amdbam, scribere, &c. 

In the beginning, as we have said, the h took the 
place of the aspirate gh, as in vehere, hiems, hesternus, 
heri (from x^^?)) fiumi (■)(a/j.ai) ; / of bh, as in flavima, 
fcro, fui, futurus, fabula,fari, fateor, frater. But this 
Latin / came soon to be pronounced in a manner 
peculiar to that people, so much so, that a Greek could 
not pronounce fundanius correctly, and this /, which 
Quintilian and others define as a breath, served, as in. 
Oscan {mffiai viai, for mediae vice), to render the dental 
aspirate; cf. inferios, infimus, with adhara, adhamas ; 
fumus with dhicma, Greek OviuLOi,fumus ; /era with d^p; 
fores with 6vpa ; rufus with epuOpos ; fetus, fecundus, 
felix, femina, with Qrja-Qai and S^Xv? ; farstv^, fastus, 
with Oapcroi. Then came the turn of the guttural : 
gharma, hot, was represented hy formats ; ^oXoy hjfel; 
yku) by fusus, fundere. Finally, the confusion between 



Phonetics. — The Explosives. 391 

/ and h became complete, people wrote indifferently 
filufn and hilum, folus and holus, fostis and hostis (Ger. 
gast, Eng. guest), fcedus and haedus, fordeum and hor- 
deum, farerw, and harena, faba and haba. They went 
even further, adding an h where this letter had nought 
to do : haudoritas, haditus, humerus (from c5/xoy, Sans. 
amsas), and omitted it where it might have been ex- 
pected : nemo for nehomo, 'prendere for preheTtdere, debeo, 
prcebeo, for dehibeo, prcehibeo, asta, lance, for hasta, ireus, 
he-:goat, for hircus, omini for homini, oc for hoc, anser 
for hanser (xw, gang, goose). 

The Latin aspirates are thus in process of decay, and 
this disintegration continues in the Romance languages, 
resulting, in Spanish, in the abuse of the sibilant h : hijo, 
hdblar, hacer, hambre (Lat. fames), hombre, hierro, henTW 
(femina), &c., and in the loss of the aspirate in Italian : 
uomo, onore, umore. Wallachian has hoblu for fabulari, 
hiliu fovjilius, hern iov ferrum. Portuguese keeps the/. 

Nor are these variations limited to the Latin group ; 
Russian and Polish differ by the employment of the th 
and the/; Timofeo, Feodor. Even in Homer we find 
^r'jp, the Centaur, for Qrip ; in Sappho, iroiKiXocppovos 
for irotKiXodpovoi ; in ^olian, bpn^ps for opviQo^ ; e^ii 
and o(f>is (serpent) are doublets of the same root. And 
in Sanscrit it has not been sufficiently noted how fre- 
quent are the alterations and permutations ; sometimes 
the aspirate stands alone, Mta for dhata, deros ; pa-hi 
(protect) ior padhi; sometimes it is of another class: 
lidhe, I lick, ledhi, he licks, compared with leksjasi, he 
will lick, root liffh, Greek \ei)((D. 

In sketching the changes of the consonantal groups 
and of the aspirates, we have already in great part 
traced the history of the Indo-European pure explosives, 
which are five in number : two gutturals, k and g, 



392 The Indo-European Organism. 

which are not very clearly distinguished in Italy for 
want of distinct characters ; two dentals, t and d ; one 
labial, p. Before pursuing their metamorphoses in the 
Germanic languages, where they offer the most remark- 
able results, it will be well to recapitulate generally the 
diverse forms or causes of phonetic changes. They 
result almost always from the need of simplification, of 
.softening of the sound ; it is what has been called the 
law of the least effort. Adjndpajatu auryah, may the 
Lord permit, said the Brahmans ; Aiiabedu adjo, said 
in Prahkrit the ladies of- the court. Imperatrix, the 
Latins said ; the English have reduced the word to 
empress. So with all the languages of the family, the 
words are contracted, letters dropped, syllables omitted. 

We have seen how the loss or the assimilation of 
the s, V, and y has altered the physiognomy of Greek. 
We have also to take into account the reciprocal 
influence of letters, which modify, compensate, and de- 
vour each other. Sometimes an i or a tt replaces an n 
or an s : euri, (biKovtri ; eifii, opeivoi for ecrfu, opeiTvo<s ; 
sometimes a dental is effaced : irocri for iroBari ; a-w/xaa-i 
for ard/ULaTcri ; sometimes a semi-vowel disappears before 
a doubled consonant : Terrace?, iTTTro? ; aXXoy (Sans. 
anyas), ktevvui, Kpeia-acov (KpaTuav), eXaarcrajv, orrcre (phje); 
or a dental changes into an s before another dental : 
TretOTOs (TrtcTTOg) ; aSreov, to be sung, a(TTeov. 

This last phenomenon occurs constantly in Latin. 
Rodere, to gnaw, gives rod-trum, the gnawing instru- 
ment, rostrum, the beak ; claudere, clavdtrum, clmtstrum; 
in codcstis, palustris, in potestas, egestas, the theme 
is coslit, pahid, potent, agent. Assimilation is also com- 
mon : summus (supmus), flamma {fiagma), sella {sedla), 
lapillus (lapidhis), asellus (asinlics), stella (stei-la), villa 
(yicla), vanus (vacnus), puella {fuerla), corolla (coronla). 



Phonetics. — The Explosives. 393 

Latiu is a much-contracted language : it abridges 
diphthongs, drops final letters (we have already pointed 
out these characters), and especially omits from the 
middle of the word letters, and even whole syllables : 
colwmna (from colus, shaft), alumnus, Vertumnus, ha\e 
lost an e; eo gigno, privignus, maMgnus. Ohjurgare, 
purgare, are for dbjurigure, purigare; retrorsum ibr 
reiroverswm; costce for compositce, placed together; quinus 
for quintnus; consuetudo, veneficus, nutrix for nutritrix, 
venenificus, consuetitudo. 

On the other hand, euphony has introduced into the 
words a number of adventitious letters : u in sum, 
sumus, volumus, oraculum, poculum, singularis, jEscula- 
pius (Greek ^Ao-KXtjirlos); i in aridus, frigidus ; e in 
humerus, Numerius {Numsius) ; in ager, niger, teter, 
ruber, &c. ; p in sumpsi, redemptus. 

The same phenomenon occurs in Greek: avSpoi, 
aV/SjOOToy, for avepo?, afxporo^ ; another, very common 
in Greek, the prostliesis of a vowel {ovo/na, oSou^, sfie, 
ocbpvs, e'lKoa-i (e-fiKQcri), epvOpoi, eXa-^us aa-rrip), is 
foreign to classical Latin. It is only in the fifth cen- 
tury of our era that we begin to find in inscriptions 
istafua, ispiritus. Nothing is more common in the 
Romance languages : esprit, este, escole, estable, espaule 
(scapula), echelle ($cala), estatue, &c. Examples are not 
wanting in Persian, Celtic, Slav, nor even in the 
Sanscrit of the Vedas (uloka, the world, iradjati, he 
reigns). 

This reappearance of transitory fashions of pronun- 
ciation after a lapse of ten centuries is not rare. Such 
is the so-called palatal pronunciation of the explosives 
and the aspirates, the ancient practice of the speakers 
of Zend and Sanscrit, unknown to classical Latin, and 
now so common in all our modern idioms ; so that the 
tch, ch, Q, c, dj. ./, of English, French, and Italian : Tchit- 



394 ^-^^ Indo-European Organism. 

cherone, Ciceron, cheval, prigione, prison, istituzione, 
institution, giorno, jour, gender, &c., are nearer the 
ancient Indo-European pronunciation than the Latin 
Kikero, caballus, hinis, diurnus, genus, whence the 
Italian and French vocabularies are directly derived. 

Such are a few of the innumerable phonetic pheno- 
mena which, together with differences of accent, the free 
choice of suffixes, of grammatical forms, of methods of 
composition and of derivation, have determined the 
direction, the respective physiognomy, the individual 
or original character of each of the branches growing 
from the same stock, of each smaller branch of those 
which grow from the same limb. 

To all these causes of divei-gence the Teutonic lan- 
guages add a species of perversion of sounds, Lautver- 
schieiung, which has, among them, altered, destroyed 
the impress of the common type. It is not that the 
substitution of one consonant for another is a fact new 
to us ; far from it, since it dominates all Indo-European 
phonetics, since it has been the object of these labori- 
ous summaries. But it here presents itself with a 
regularity which might appear to be the effect of a pre- 
conceived system, if it were not purely physiological 
and involuntary. The Germanic alphabet includes all 
the hard, soft, and aspirated consonants which are known 
to us, and the German pronounces them easily; but 
the German ear does not hear, nor the German throat 
utter, the sound which strikes a Latin ear, which the 
Latin lips pronounce. The German says blanche 
planche for plan^he blanche; projet hecomes brocket ; bon, 
pon; danser, tamer; cceur, guier. This habit is so in- 
veterate that it extends even to sounds fixed by writing, 
turning v into/; von, vater; changing Berlin s.nd Bis- 
marck into Perlinn and Bismarck. It is thanks to the 
regularity of these permutations that philologists such 



Phonetics. — The Explosives. 395 

as Grimm and Bopp have been able to trace the original 
roots of which the Latin or Sanscrit form was more fami- 
liar to them, and to formulate laws, to which, however, 
the caprice of languages occasions many exceptions. 

A difficulty presents itself at the outset : of the two 
great divisions of the Germanic family, Low and High 
German, the first alone obeys the general law as a 
whole ; the second introduces a partial amendment, 
which is very important, as it constitutes the principal 
distinction between the two groups ; there are thus two 
degrees, two stages in the substitution, and the ques- 
tion arose whether the second proceeded from the first, 
or if the one and the other grew up simultaneously and 
independently. Now, except for a few quoted words, 
which seem to prove that in the first century the first 
substitution was already accomplished, we do not know 
the primitive state of the Teutonic idioms. Although, 
no doubt, nations belonging to the High German group 
occupied from the time of Marius the mouths of the 
Elbe and of the Weser, and the banks of the Rhine at 
the time of Caesar — the Teutons beaten at Aix and 
Verceil, the Suevi or Snabians of Ariovistus driven back 
by the conqueror of Ganl — yet we know nothing of their 
dialects: neither Tacitus nor Marcus Aurelius gives us 
any information about the speech of the Cherusci, of 
the Marcomans, or of the Quadi. The best example 
which we have of ancient German is the Gothic of 
Moesia, dating from the fourth century. Gothic is the 
eldest of the Low German dialects, which include: 
Scandinavian, Norse or Icelandic, and Danish, Swedish, 
Norwegian, Piatt Deutsch, spoken in North Germany 
between Cologne, Cassel, Magdeburg and the sea, 
ancient Anglo-Saxon, Frisian, Dutch, and Flemish; 
these last intermediate between German and English. 



396 The Indo-European Organism. 

We will briefly recall the most aBcient documents 
left by these languages, which all participate in the 
first substitution of consonants (first, that is, in chro- 
nological order). Gothic of the fourth century is 
known to us from the Bible of Ulphilas. This man, 
of Cappadocian origin, born among the Goths of Chris- 
tian parents, lived from 311 to 381. Bishop and 
Arian apostle of the Goths, he dwelt at Constantinople ; 
lie knew Greek and had some acquaintance with Latin; 
Fragments of his translation of the Gospels, the Epistles 
of St. Paul, the books of Esdras and of Nehemiah re- 
main to us. The principal manuscript, purple, with 
letters of silver and gold, originally from the convent 
at Bobbio, transported to Werden at the time of 
Charlemagne, then to Prague,' then to Holland, and 
finally preserved at Upsala, dates from the fifth cen- 
tury. We also possess in this language, which became 
extinct in the eighth or ninth century, two contracts 
of sale, a calendar, and some alphabets. Gothic was 
spoken in Italy by Alaric and Theodoric, at Toulouse 
by Ataulf and Thorismund, at Toledo by Eecarede and 
Totila, perhaps at Aix by the Austrasian Franks. 

The poem of Beowulf (seventh century), and the 
Reljand (eighth), attributed to Louis le D^bonnaire, 
make known to us the language of Hengist and Horsa, 
and that of Witikind, the brave adversary of Charle- 
magne — the language which constitutes the ground- 
work of modern English. Finally, the two Bddas of 
Iceland (twelfth and thirteenth centuries) have pre- 
served for us ancient Scandinavian as it was spoken 
in Norway in the ninth century, when king Harald 
Harfagar drove into exile the nobles who had revolted 
against his power and against the Christian propa- 
ganda. 



Phonetics. — The Explosives. 397 

The alteration of the explosives throughout this great, 
group affects the three orders of consonants, working, 
by regular stages ; there are, of course, exceptions. 
The Indo-European soft consonant is changed into the 
hard consonant, the hard into an aspirate, the aspirate 
into the soft consonant. It is not, as has been said, a 
mere strengthening of the sound, since the aspirate is 
weakened. I will give a few examples among the 
gutturals, which are, in the Indo-European alphabet, 
g, k, gh. 

With gan, to engender (ffigricre, genus), and with gan, 
gnd, to know (yiyvwcrKw, gnoscere), are connected kuni, 
kind (Goth.), race, son; Anglo- Sax. did; Eng. child; 
Norse, kgnd, kan, know ; Ger. ICtinst. With gnd 
(pronounce dj'na, woman, Gr. yvvij) correspond quino, 
queen, Soand. kona; cf. the well-known masculine 
forms Konig and king. Genu, gonu, give the Gothic 
kniu ; gula, kele, gala (gelu, gelidus), kalds, Eng. 
cold ; gau, kuh, cow ; gam, kommen, come ; gush, to 
try ; kiusan, to choose ; gust, taste ; kustus, trial ; gena, 
cheek, Scand. kinn; reg-ere, reg-em, rcgula, reiks, 
kingdom ; varg (epyov), waurkjan, work ; agros, akr. 
The original soft consonant has become hard. 

We have already mentioned the words cor, cordis, 
KapSla, grad (in the verb grad-dadhami). What will 
be the Gothic form ? . Evidently hairto, Eng. heart ; 
the Sanscrit hrd is already German. Collum becomes 
hal-z ; calamus, halm; canis, Mmths ; cornu, haurn, 
Scand. horn; quis, hvas, hvo, Eng. who, luhat ; kveta 
(Sans, gveta), hvit, white ; copio, hafjan, to take ; 
kKvw, hlu, ear, hlium/i ; hlut, heard, Ger. laut, sound, 
letter; caput, heafod, Eng. head; precari, fraihan ; 
dico, taihan, teach ; pecu, faihu ; Kpeas, Scand. krae, 
corpse ; cutis (skin), Scand. hud ; Goth, svaihra ; dekem, 



398 The Indo-European Organism. 

taihun, ten ; centum (Gr. eKUTOv), hundred. The original 
strong consonant becomes an aspirate. 

The original aspirate gh is represented by the Greek 
letter, by the Latin h ; we shall see that it is replaced 
in the Teutonic family by the corresponding soft con- 
sonant. Stigh, Gr. aTe[-)((t), to advance, to march, 
a-T!)(fj, file, rank, becomes in Goth, steigon ; vagh, Goth. 
gawagjan, Scand. ega (cf. Gr. jJ^*^) '■> %^' ^Ho^'kion, to 
lick ; kku {fundere), Goth, giutan, Dan. gyder, I over- 
turn ; x^^^' ^'^^ 5'%^*' hesternus, Goth, geistra, gesterh, 
yesterday ; xoXo^, bile, Dan. gali, Bng. gall ; x^" 
(hansa), ganz, goose. We find homo under the form 
guman in Goth., Ger. braiMigam; hostis, for ghostis, 
becomes geist, ghost, and guest, the stranger. 

The same phenomenon occurs in tlie dental con- 
sonants. D: duo, dvplus, tioei, tweifels, tvro; drus, 
Goth, triu, Eng. tree; dom/ire, tamjan, Scand. tamr, 
tame ; Seipw, to cut (whence Sepfia, thin slice, skin) ; 
tairan, to tear ; dahru, taihr, tear ; ducere, tiuhan ; 
Scand. dantas, Lat. dentis, Goth, tunthus, tooth ; suadvis, 
Goth, suts, sweet ; ad, to eat ; sedere, sitan, set, sit. 
T: to, turn, tad, Goth, thata, the, that; treis, threis, 
three; tu, Goth, thu, thou; ts'lvw, Goth, thanjan ; 
tonitru, thunder ; tenuis, thin ; tarsh, to burn ; tharsjan, 
thirst ; tauta, town, people, Lat. tot%is, Goth. Thiodisc, 
Ger. Deutsch ; mater, mother. Bh: dha, to place, 
Gr. Qri, Eng. do ; doms, doom ; dhars, Gr. Odptros, 
Goth, ga-daursan, to dare ; rudh, red ; Oijp, Anglo- Sax. 
deor, wild beast, Eng. deer ; dvpa, door, Dan. dyr. 

The labials p and hh remain to be considered (h is 
not primitive). P: pater, father; primus, Goih. fru- 
man, Eng. first; ttovs, ttoSos, Goth, fotiis, foot; priya 
(deav), Goth, fi-ijon, friend; pur, fire; pellis, Goth, fel, 
leather; ireimve, Jimf, funf, five; iroXvf, Goth, filu, 



Phonetics. — The Explosives. 399 

numerous ; TrXeos, full ; super, ufar, over. £h : hhar, 
Goth, hairand ; bhratar, brother; hhu ((puce), ich bin, 
Anglo-Sax. beom, to be, (prjyoi, fagus, Goth, hohi, Dan. 
log, beech ; frangere, fregi, brukjan, to break ; 6(ppvi, 
(ej'ebrows), braue, brow; ve(pe\rj, nebl. 

This substitution presents itself in the Gothic stage 
with great regularity; but the exceptions (which I 
have omitted) throw some shadow over the picture. 
Thus a secondary law preserves from all change the 
hard consonant preceded by an s or an aspirate. Fislcs 
corresponds to piskis, beside the English form Jish, 
of which the h appears more regular ; ist, and not isth, 
corresponds to asti ; stairno, and not sthair, to star a, 
star ; skaida, speva, staiga, I separate, I spit, I mount, 
to scindo, sptco, a-Te'i-^w. Elsewhere, but without suffi- 
cient regularity to form a law, the aspirate is wanting 
to the initial consonant ; for instance, in band, I bound, 
budum, we offered, gridus, hunger, gavi, country, grip, 
to take, dauthar, daughter, daur, door, ails, party ; 
and in the body of the words fadar, father, fidvor, 
four, sibun, seven (Anglo-Sax. seofon), biuga, I bend 
(root bhug), skadtts, shadow, slepa, I sleep, the substi- 
tution does not take place. , Sometimes Anglo-Saxon, 
or even German, are more regular than Gothic. There 
is nothing to astonish in these uncertainties, which are 
common in all languages, and would be very easily 
explained if we knew all the -circumstances of place 
and time which influenced pronunciation. 

It is the general opinion that the first or simple 
substitution, to which all the dead or living branches 
of the Low German group have remained more or less 
faithful, was up to the seventh or eighth century 
common to all Teutonic nations. Towards that epoch 
Old High German applied to the Low German conso- 



400 The Indo-European Organism. 

nants the same modifications which Gothic had applied 
to the Indo-European consonants, changing the soft 
Gothic consonant (originally an aspirate) into a strong 
one, the Gothic strong or hard (originally a soft) into 
an aspirate; finally, the Gothic aspirate (originally hard) 
into a soft consonant. Or, in other words, the Indo- 
European soft consonants, d, g, h, become German 
aspirates or sibilants, z (ts), ch or h, f or pf ; the 
Indo-European hard consonants, t, h, and p become soft, 
d, ff,b ; the Indo-European aspirates d/i,gh, 6A, become, 
hard, t, k, p. 

Before showing how far this new change is real and, 
how far it is imaginary, we will mention the most, 
ancient documents which remain to us in" Old High 
German. ■ ; 

The Frain-ks, the Suabians, the Alamans, the Austro-, 
Bavarians' had no Bishop Ulphilas, and their troubled 
existence in the fifth and sixth centuries scarcely 
allowed them to become conscious of themselves. 
Harried and invaded by the Huns and the Slavs, 
urged across Europe by the invading hordes, they 
form no true German state before the reign of the 
Garlovingian Louis the Germanic. The Pranks alone 
succeeded at the end of the fifth century in including 
the greater part of Gaul in the domain of Clovis, but 
it was soon divided among his descendants. Too few 
in number to impose their language upon populations 
which for four centuries had spoken Low Latin, they 
were forced to have their own Salic law drawn up in 
the .idiom of their subjects. The Frankish language 
was spoken only at the court of the kings of Austrasia. 
In fact, nothing has remained of the ancient poems of 
which Tacitus had heard ; therei is no doubt, however, 
that the Franks knew and recited a great number ; 



Phonetics. — The Explosives. 401 

and Cliarleraagne, who loved his mother-tongue (or 
rather his father's, for his mother was a Gallo-Eoman), 
undertook to have them collected. Of this treasure, 
which would have been of inestimable value, nothino- 
remains but a fine fragment, Hildehralit and Eade- 
h'aht, the terrible combat between a father and son. 
This, with the obscure commentary on the Salic Law, 
is the only Frankish text which is as early as the 
■eighth century. A description of the sea {Merigarto, 
Mer-garten, the garden of the sea), a poem in honour 
of Louis III., conqueror of the Normans, the Liidwigs- 
lied, a prayer in verse, and a rhymed Life of Jesus are 
attributed to the ninth and tenth centuries ; finally, 
the numerous prose translations of an industrious monk, 
Notker, prolong into the twelfth century the life of 
Old High German. Middle High German, the dialect 
of the Suabians, flourished from the twelfth to the 
fifteenth century, under the brilliant dynasty of the 
Hohenstaufen, the Suabian emperors ; it is the idiom 
of the " Book of Heroes," a collection, unfortunately 
small, of precious legends, which relate the exploits of 
Btzel and of Dietrich of Berne (Attila and Theodoric), 
and the famous battle of Ravenna, in which fell the 
empire of Otokar (Odoacer) ; it is the tongue of the 
Nihelungenlied and of Gudrum ; the two national 
epics, and of the Minnesanger, the poets of chivalry. 
With Luther's translation of the Bible the modern 
period begins, that of the New High German, an 
artificial composite language, borrowed by the great 
reformer from the style of the lawyers of Mediaeval 
Germany, especially of Saxony, in which the various 
elements of Suabian High German, of the old Lovv 
German Saxon, and a few reversions to the Gothic 
pronunciation are all confounded. 



402 The Indo-European Organism. 

From the point of view of philology, the distinguish- 
ing marks of the various phases of German are, first, 
the gradual muffling of the sonorous vowels, the loss 
of the «, a, and o of the terminations, and the abuse 
of the finals in w; and secondly, the more or less 
ricrorous use of the second substitution. All the 
dialects of High German, says M. Hovelacque, have 
changed into t, z, and d the d, t, and th of the 
Germanic idioms of the first phase. In this they are 
all strictly High German. But the other two ordta-s 
of consonants do not follow the rule so rigorously ; 
only part of the High German idioms altered the k 
and g, the p and h of the first or Gothic stage. Thus 
the first or Gothic change remains in Kinn (chin, 
kinnus), kann, can, Hund, dog. Gad, guest, gebe, give, 
fange, take, Vieh (Goth, faihu), cattle, hiTide, I bind, 
hiege, I bow. On the other hand, Pracht (splendour), 
whence Berthe, Albrecht, Dagdbert (brightness of day), 
&c., corresponds to the Goth, bairhts, Anglo-Sax. 
beorht, Eng. bright, and to Sans, bhrag, to shine, 
(pXeyu!, flagro, fulgeo ; breche, I break, Jlehe, I implore, 
fiHige, I ask, hnnge, I hang, lecke, I lick, schlafe, I 
sleep, laufe, I run, b-leibe, I stay. Joke, yoke, correspond 
to the Gothic brika, fleka, fraihna, haha, laigo, sepa, 
hlaupa, ccf-li/nan, jog ; ich pirn was developed correctly 
from ik bin, prinnan from brinnan, chunnan from 
kunnan, kilih from galeihs ; but the German of Luther 
returned to biyi, brennen, kennen, gleich. 

The dental evolution alone is complete and per- 
sistent. The Gothic t becomes a, tz, sz, ss; tivau, 
zwei ; taihun, zehn ; to, zu ; taihan, Ziehen ; tunthus, 
Zahn; fotus, fuoz ; tamjan {damayami, to tame), 
zamom; ita, to eat, izzu; Tuesday, Zientag; taihr,Zehre; 
vet (wise), weiss, wissen ; sitan, sitzen. The Gothic d 



■Phonetics. — The Explosives. 403 

becomes < ; do,tuom,; day, Tag ; ga-dars, ge-tar ; rod, 
red, rot; deer, Tier. The Gothic th becomes d: 
irother, Bruder ; thou, dv, ; three, drei ; thin, dunn ; 
the, der ; that, dass ; thorn. Darn; through, durch; 
thunar,Donar; thirst. Durst ; thorp (trapa, tribus), Dorf. 

No other group in the Indo-European family has 
had such vicissitudes, nor is sa far removed (in 
appearance) from the primary type from which it 
springs. I say in appearance, for there is no Aryan 
idiom richer in ancient roots and elements. The 
probable cause of the Teutonic modifications has been 
sought in the contact, perhaps the mixture of race, 
with the Finns. But it is simpler to attribute them 
to one of those physiological phenomena which are 
peculiar to each race. Another problem is whether 
the two substitutions were produced independently of 
each other. Max Miiller inclines to believe that they 
were ; but it is difficult to explain in that case the 
period when the first change was common to all the 
idioms, and the tardy and partial development of the 
second, or to understand why the latter always took 
the former, even in its exceptions, as basis and point 
of departure. The root grabh ought to give krab in 
Gothic and chrap in High German ; yet Gothic has 
greipan, German, greifen. Gothic has fadar, where by 
the rule it should have fathar ; German takes fadar 
as its point of departure, and changing the soft into 
a hard consonant, has vater. This ingenious observa- 
tion was made by M. Michel Br^al. 

We must admit that the two series of modifications 
arose from the same desire of strengthening the sound, 
or of change, and that the second being the con- 
sequence of the first, would have taken place with 
more or less uniformity in all the Teutonic languages, 



404 The Indo-European Organism. 

if Gothic had not perished, if the idioms of Scandinavia 
and of England had not had, owing to isolation and 
other historical causes, an independent career. This 
may be assumed from two facts : the dental evolution 
has taken place even in Flemish and Dutch, whei'e the 
is de. Even in English the soft aspirate th maintains 
itself with difficulty ; the ih of them, is often dropped 
in popular speech. Shakespeare, in the scene of 
Henri/ V. where Catherine of France learns English 
from her waiting-maid, represents th by d, the hand, 
de hand, &c. The whole scene is curious, but we 
need only here retain a valuable hint on phonetics. 



CHAPTEK VII. 

'Two ANALYTICAL LANGUAGES. 

The Saxons and the Angles — Formation of Anglo-Saxon : Saxon and 
Anglian texts — Anglian has felt the influence of Danish — Low 
Latin and rural Latin : formation of French — The Oath of Stras- 
burg, &e. — The Song of Roland — French introduced into England 
by the Normans and by the Angeviiis (Plantagenets) — The French 
vocabulary permeates and disorganises Anglo-Saxon (twelfth to 
the fifteenth centuries) — Old and Middle English — Two languages 
in English — The two stages of French : popular and learned 
French — Lament of the Komanibts — Doublets — French words 
borrowed from Italian, Spanish, German, and English — Greek 
siaiiixes — Vitality of derivation in French, of juxtaposition in 
English — Coneluaion. 

We have nearly reacbed the end of this philosophical 
review of language. After having connected human 
speech with the cry of animals, we established the 
co-existence of monosyllabic, agglutinative, and inflected 
idioms, and we accepted as probable the succession 
of three periods, corresponding to these three stages 
of languages : we regarded inflexion as the fusion of 
agglutinated syllables. But we have not been guided 
by any preconceived system, by any prej udice ; we 
have adopted with perfect indifference the order which 
seems most in accordance with the facts observed. We 
are not of those who lament the loss of the Sans- 
crit instrumental, of the .^olian digamma, or of the 
two cases of Old Frencli. We do not in the least 
believe in the decadence of language ; we see in the 
thousand phonetic substitutions and modifications 
27 ""s 



4o6 The Indo-European Organism. 

adaptations of speech to the temperament of the 
various peoples, and to the growing complexity of in- 
tellectual needs. 

We have a great admiration for the synthetic idioms 
which mark by varying terminations the relations of 
one word to another; but we prefer the simplified 
languages in which the word is even more the willing 
servant of the thought, in which prepositions and 
auxiliaries supply with greater subtlety and precision 
the vanished forms of the declension and conjugation. 

Among the European tongues which have attained 
more or less completely to the analytic state, we will 
choose two, those which have the greatest interest for 
us, the one Teutonic, the other Latin, very different 
and yet inseparable, English and French. 

Towards the middle of the fifth century the tribes 
who spoke the Low German language formed in the 
south the vanguard of the Teutonic body, and in the 
north its rearguard. The half-civilised Visigoth of 
Ataulf and Buric learnt Latin, and forgot their own 
idiom at Toulouse, at Narbonne, and soon beyond the 
Pyrenees ; but the Scandinavians, the Saxons, the 
Frisians, and the Dutch, confined towards the east by 
the Lombards, the Eugii, and the Heruli, on the south 
by the Alamans, separated by the Franks from the 
Rhine, the Mease, and the Scheldt, were driven out, 
partly into the great peninsula of the north, or into 
Jutland and the Danish isles, partly into the district 
lying between the Elbe and the Weser, and towards 
the mouths of the Ems, the Eider, the Rhine, and the 
Meuse. They were a seafaring people, pirates and 
adventurers, constantly visiting the British coasts. 
In 448 a band of this hardy race, Saxons and Jutes, 
established themselves on the coast of Kent, in the 



Two Analytical Languages. 407 

Isle of Thanet ; their chiefs, Hengist and Horsa, who 
were at first engaged iu the service of a British prince 
to aid him against the incursions of the northern tribes, 
were not long in establishing a claim to territory. 
Horsa was killed, but reinforcements of Saxons and 
Jutes gradually drove the ancient inhabitants, disputing 
every inch of the ground, as far as the Severn and the 
Welsh mountains, and in less than a century had con- 
quered all the south of England and the basin of the 
Thames. Towards 5 60 the Angles who inhabited the 
south-east of Schleswig established themselves in great 
numbers in the neighbourhood of the Humber and on 
the marches of the Cambrians and the Scots. By 586 
the Celtic element was completely expelled, and the 
little kings of the Heptarchy were eagerly disputing 
the suzerainty one with another ; those of Wessex 
in the ninth century had succeeded in bringing them 
all into subjection, when the Teutonic invasion, which 
had ceased for a space, was renewed with great ferocity. 
Other Northmen, those who from Friesland, Denmark, 
and Norway fell upon the Carlovingian empire and 
took possession of Normandy, founded in the land of 
the Angles a kingdom of North umbria. Then, in spite 
of the victories of Alfred and of Athelstan (871—941), 
Danish kings, Sweyn and Knut, became masters of 
England. A lucky rebellion in 1 03 5 re-established 
for thirty years Saxon autonomy, but the Danes did 
not readily renounce their pretensions, and at the very 
time when William the Conqueror disembarked at 
Pevensey (1066), Harold, the last national king, was 
occupied in the north in destroying a Danish and 
Norwegian army. 

The Anglo-Saxon language, if we omit a few varia- 
tions of dialect, was sufficiently strongly constituted in 



4o8 The Indo-European Organis7?i. 

ibur centuries (500-900) to reduce to small importance 
the mixture of the Danish vocabulary, which was, 
moreover, almost identical with its own ; it had only- 
retained a very few traces of Celtic, names of places, 
for instance, and about fifty words such as Mw, crook, 
wicket, dan, claymore,, &c. This disappearance of Celtic 
is at first sight surprising; it had survived the four 
centuries of Roman dominion, and it disappeared before 
an idiom as little cultivated as itself. The reason is 
that the Saxon invasion was the substitution of one 
race for another ; the conquered British fled into 
Armorica, or to the west, into the territory of the 
Cambrians, "Wales, and Cumberland. 

The Roman occupation had been intermittent and 
superficial ; when, in 412, the legions withdrew from 
Britain, Latin went with them. English soil, indeed, 
retained the ruins of cities and traces of Roman roads, 
but in primitive Anglo-Saxon there are only three 
words of Latin origin : coin, Lincoln (Lindi colonia) ; 
caestre, castra, and straet, strata. The establishment 
of Christianity in the seventh century introduced a 
certain number of terms which have been transmitted 
by Anglo-Saxon to modern English : ancor (anachoret), 
apostol, postol, hiscop (bishop), aelmcasse (alms), colic, 
candel, clustor and claustre, discipul, deofol, deacon, engel, 
mynster, pistol (epistola), predicyan, profost (propositus), 
purpur, sand, ymn, culture {columhd), castell, douiher 
(doctor], gigant, meregreot (margarita), pund, plant, 
yip (vulpes), yncia (uncia). Latin letters were very 
little cultivated in the time of the Heptarchy. Bede 
urged the Archbishop Egbert to have the Lord's Prayer 
and the Creed translated by the most learned of the 
clergy or laity of his diocese. In Alfred's time not a 
single priest understood the mass. The Norman inva- 



Two Analytical Languages. 409 

sion raised the level of classical studies ; Latin became 
the language of the law, of theology, and of public 
proclamations. But we have to say a few words of 
Anglo-Saxon itself, the language which is the founda- 
tion of modern English. It is, as its name indicates, 
the fusion of two groups of the Low German dialects, 
of which we ' studied the phonetic system in the last 
chapter. It must be noted that these idioms, far from 
being derived from German, are the elder brothers of 
the oldest High German, that though they have the 
same grammar, and essentially the same vocabulary, 
they are not its descendants, and that in the consonantal 
permutations the most modern English lias generally 
remained at the Gothic and Scandinavian stage. 

Among the Saxon dialects that of Wessex, of which 
Winchester was the capital, spoken under Egbert and 
Alfred, naturally held the first place, as the kingdom 
of Wessex gradually absorbed the other states of the 
Heptarchy. To it belong the most important texts 
that we possess : the poems of Beowulf, Cffidmon's para- 
phrase of parts of the Bible, Andreas and Elene, the 
Codex Exonieiisis, the Cronicon Saxonicum, King Alfred's 
translation of the historian Rosius, the vernacular 
version of Bede's " Ecclesiastical History,'' the Anglo- 
Saxon Laws, and ^Ifric's translation of the Gospel and 
Homilies. 

The Anglian dialects are represented by two MSS. 
of the Cronicon Saxonicnm, which are believed to be 
from Mercia, and by the Durham Book, which was 
drawn up in Northumbrian in the time of Knut. In 
the latter text we find the most frequent traces of 
Danish influence, such words as afledd, begotten (Scand. 
afla, to engender), agede, lasciviousness (Scand. agceti), 
heggse, bitter (Scand. heiskr), hulc, bull (Scand. holi), 



4IO The Indo-European Organism. 

Uunt, stupid (Scand. Uunda, to sleep), Inlaxt, hatchet 
(Scand. boloxi, cf. TreXe/ci;?), Jcide, kid, are of Danish 
origin. The suffixes or terminations, legge, sunnd, 
agg, egg, eimde (pres. part.), inn and enn (third person 
plural), are attributed to Scandinavian. The charac- 
teristic of the Anglian dialects is the indistinctness 
and comparative uniformity of the vowel sounds. West 
Saxon, on the contrary, the type family, possesses an 
elaborate and complicated vowel system, even more 
varied than that of Gothic. Some traces of reduplica- 
tion remain in the verbs, and the present and past tenses 
are often distinguished by a modification of the vowel of 
the root, of which the so-called irregular verbs in modern 
English offer numerous examples : do, did, done ; hear, 
hare or hore, born. ; begin, began, begun. This method 
of internal modification is for a language almost 
reduced to the monosyllabic state a source of infinite 
wealth ; a simple change of the vowel makes of the 
same word of four or five letters a verb, a preterite, a 
participle, a noun, an adjective, an adverb. What has 
already in Anglo-Saxon suffered most is the declension ; 
like the German declension, it is thrown into confusion 
and tends to disappear. Prepositions take the place 
of cases. But nothing, so far, distinguishes the English 
of the future from the other Teutonic idioms. 

It is doubtful whether Anglo-Saxon alone could 
have developed into the rich and delicate language 
which we now know as English ; it might, indeed, 
have given it the strength and poetry in which it 
abounds, but hardly the simplicity of structure and 
the subtlety and variety in the means of expression 
which make it the type of the analytical languages. 
It was the infusion of the French blood and ways of 
thought which made English what it is, freed it from 



Two Analytical Languages. 41 1 

its grammar and complicated inflexions, and completely 
disengaged it from its German swaddling-clothes. 

For French was arising, full of the vigour and 
promise of youth, just at the time when the Teutonic 
idioms were vegetating on the Continent, and when the 
Anglo-Saxon element, both language and race, was 
languishing in complete intellectual and political stag- 
nation. The leaders of the Visigoths, Burgundians, 
Saliens and Sicambri, more fortunate than Hengist 
and the Anglian Idda, had found in Gaul a population 
far too numerous to be destroyed, or even absorbed, by 
its half-savage conquerors, and still possessed of suffi- 
cient cultivation to civilise to a certain extent their 
barbarian masters. In spite of their sincere disdain 
for the conquered race and of the self-content wliich 
they display so naively in the preamble to the Salic 
Law, Clotaire (Clotachar, the illustrious), Hilperic 
(mighty protector), Dagobert (brightness of day), sur- 
rounded by astute and politic bishops, had to learn 
Latin, and to have their laws and proclamations drawn 
up in the language of those who knew how to write. 
For five centuries Latin had been spoken everywhere 
in Gaul, except in Armorica and in the Basque country ; 
the official and administrative Latin in the cities, the 
Latin of the people and of the camp in the country 
districts and in the neighbourhood of the garrisons. 
Frankish was spoken only at the Austrasian court, in 
a few districts of Hainault, and along the Rhine, just 
where the invaders formed sufficiently compact groups. 
The rise of the Carlovingians, Franco-Belgians, with a 
sti'ong infusion of Roman blood and culture, did not 
change the habit of centuries." Though Charlemagne 
had the excellent idea of collecting the songs of his 
ancestors, he does not seem to have dreamed of estab- 



412 The Indo-European Organism. 

lishing anywhere a Germanic sctool. Latin alone 
was spoken and written at his court, but a thousand 
Teutonic words, Latinised for the most part, came to 
be added in current speech to a small number of Gallic 
words, and the Teutonic pronunciation, exaggerating 
the tonic accent, modified yet further the Latin words, 
already transformed by Celtic utterance. An insensible 
change, begun in the seventh century, developed from 
Latin a new language, an embryonic French, which all, 
counts and villeins, priests and laity, spoke, but which 
Was not yet written. In 660 a certain Mummolin 
was elected bishop because "he excelled in the Eomance 
language;" in 750, Adhalard, abbot of Corbie, was 
" clever in Romance, brilliant in Tudesque, elegant in 
Latin ;" thus the three languages were very distinct. 
This Adhalard, adds his biographer, preached in the 
vulgar tongue with an abundance full of sweetness. 
In 768, the year of Charlemagne's accession, we find 
manatees for minas, helmo for galea, solament for sing%i- 
lariter, macioni for cwmentarii. The Church hastened 
to consecrate this patois, which was strongly recom- 
mended by the Councils (813, 842, 851). Preaching 
was everywhere in French, and there can be no better 
proof that the common people no longer understood 
either Latin or German. There is also documentary 
proof, the text of the Oath of Strasburg (842), pre- 
served by Nithard (843). 

The eldest son of Louis le D^bonnaire, Lothair, 
invested with the imperial title, finding himself too 
much confined in his long, narrow kingdom, which in- 
cluded Italy, Provence, Alsace, Lorraine, and the pro- 
vinces of the left bank of the Rhine — territories which 
did not belong in any sense to Germany — Lothair 
wished to enlarge his boundaries at the expense of 



Two Analytical Languages. 413 

his brothers, Charles, king of the Pranks, and Louis, 
king of the Germans. These latter, having met at 
Strasbnrg, took an oath of alliance in the presence 
of their assembled troops, and Louis pronounced the 
following words: ''^Pro deo amur, et pro christian poblo 
et nostra commun salvament, d'ist di en avant, inquant 
Beus savir ct podir me dunat, si salvarai cist meon 
fradre Karlo." (For the love of God, and for the com- 
mon safety of the Christian people of our own, from 
this day forward, as God shall give me knowledge and 
power, so will I Fafeguard my brother Charles), "et 
ah L^idher nul plaid numquam prindrai qui, meon vol, 
cist meon fradre Karle in damno sit " (and with Lothair 
no accord will ever take, which, of my will, shall bring 
harm to my brother Charles). 

This is the most ancient document not only in 
French, but of the whole Romance family. Poor as it is, 
it is yet so precious that we might pass hours in analys- 
ing it word by word. I will only note the presence of 
the Romance future, salvarai, prindrai, one of the crea- 
tions of our languages (fai d sauver j'ai (t prendre), and 
of a form proper to French, meon vol, " will," of which 
there are three hundred examples (viol, vol, dol, recel, 
recul, recueil, accord, pli, affront), happily conceived 
abbreviations of the indicative or of the infinitive, which 
might have spared us many words in ion and ment. 

French was soon cultivated as a written language. 
Without pausing to consider the naive complaint of 
Sainte Eulalie (ninth century), the short poems of the 
I'assion and of Saint Leger (tenth century), or the too 
long Saint Alexis (eleventh century) texts, which are 
valuable to specialists, we find as early as the eleventh 
century the Chanson de Roland, certainly the earliest, 
and perhaps the best, of modern epics. According to 



414 'rf^^ Indo-European Organism. 

tradition, it was sung, in a form which we no longer 
possess, at the battle of Hastings. What is certain 
is that the most ancient text of the Chanson (that 
at Oxford) dates from the first half of the twelfth 
century ; that all the details of manners, the descrip- 
tions of arms, of dress, &c., carry us back to the end 
of the preceding century, and that we must attribute 
the poem to some Franco-Norman who accompanied 
William the Conqueror to England. This origin is 
curious from more than one point of view, and we 
shall return to it. But in the first place, the Chanson 
de Roland gives rise to certain observations. It is 
documentary evidence against the prejudice which 
denies to the French the epic genius ; it teaches 
historians that Charlemagne, so constantly claimed by 
the nations beyond the Ehine, of whom he was the 
bitter and unrelenting enemy, founded not the German, 
but the French nationality, which was nearly destroyed 
after his time by the feudal system ; it reveals to critics 
how popular epics are formed, before history or along- 
side of it, round real events and persons, transfigured 
by distance. The central event here is the strife of 
Christianity against Islam, a strife jvhich did not end 
at Poitiers, but was prolonged into the ninth century 
in France and among the Pyrenees ; the dramatis 
personm are a nation, France, and a man, the great 
Karl, the emperor with the flowing beard, who strove 
to constitute the civilised West, attacked on the one 
hand by barbarians, belated in the East, and on the 
other by the Saracens pressing in from the south, and 
who, driving back in each case the invading flood, 
restored the equilibrium of Germany and forced Islam 
into retreat. These memories were not very ancient; 
hardly three centuries separated the trouvere from the. 



Two Analytical Languages. 415 

great Charlemagne ; but the whole civil order had 
changed ; there was a great gulf between the centralised 
empire conceived by Charlemagne and the feudal system 
of parcelled-out territory ; and from the depths of this 
new chaos, which was so slowly and painfully reduced 
anew to unity by the persevering ambition of the little 
kings of the Isle of France, the men and the events of 
the eighth century appeared gigantic and clad in a 
magic glory. Some have sought to find a Germanic 
character in the " Song of Eoland ; " there is no trace 
of it, unless it be found in the crudity of manners and 
of the thought ; but everywhere the Prance which is no 
more is invoked by the France which is not yet formed. 
The name recurs a hundred times, accompanied always 
by a Homeric epithet : '^France la doulce." Even Aix- 
la-Chapelle is Ais en France ; the generals are pairs 
de France, the soldiers are Frangais ; Roland, striking 
Marsile, cries, " Que didce France par nuz ne seii himie ! " 
and weeping over the dead : " Terre de France, moult 
Sfes doiKC pays ! " Oliver, lying on the ground in his 
death-agony, prays for Charles, for douce France. The 
Emperor faints on the body of Roland, and on coming 
to himself, tears his beard and cries, " In Prance 
strangers will ask, where is the captain ? I shall tell 
them he died in Spain, and a hundred Frenchmen 
mourned with him." The style, sober and strong, is 
not unworthy of the subject : — 

" Ce Sarrazin semble moult heretique ; 
Plutot mourir que de ne pas I'occire ! 
Oncques n'aimai les couards ni la couardise. . . . 
Sempre ferrai de Durandal grans colps ! 
Sanglans en iert li brans entresque a I'or." 

What a difference between this simple style and the 
insipid padding with which the jongleurs of the 



41 6 The Indo-European Organism. 

tliirteeath, fourteenth, and fifteenth centuries filled 
out the old Chanson, in order to change the assonance 
into rhyme and the ten-syllabled verse into an alex- 
andrine ! But the important point is the European 
renown of this epic, the germ of innumerable lesser 
epics which arose in the twelfth century ; translated 
and imitated in all languages, German, Icelandic, 
Dutch, Spanish, Italian, it carried the fame of the 
French name to all courts and all nations ; it had at 
length the honour to inspire the Orlando of Ariosto. 
Because of this poem French was in the Middle Ages 
the universal language, written by Marco Polo, by 
Brunetto Latini, the master of Dante, " because French 
is the most delightful and the most general speech." 
Lastly, French was for nearly three hundred years the 
official language in England ; and our tongue reckons 
among its earliest texts the laws of William the Con- 
queror, published in 1069. Here is the preamble: 
" Cessount les his et les custumes que le rei Willams 
(jrentat d, tui le puple. de EngUterre aprds le conqimst de 
la terre." 

Why was the Boland composed in the west of 
France ? Because the first author wished to celebrate 
the ancestors of Norman and Angevin, Lord Geofiroi 
GrisegoneUe and Eichard the Fearless. Why is Roland, 
llruodlandus, who is hardly mentioned by the historian 
Eginhard, become the hero of the poem and the nephew 
of the Emperor? Because he was prefect of the 
marches of Brittany, and his memory was kept green 
in the country of the unknown poet. This poet, or 
his successor, followed the troops of William the Con- 
queror into England, where he wrote in French for 
his companions in arms ; for French only was spoken 
in the Continental estates of William; so much so, that 



Two A nalytical Langzuiges. 4 1 7 

the Norman dialect was, with Bargandian, Picard, and 
the speech of the Isle of France, one of the principal 
elements of the French language. 

The Scandinavian origin of the famous bastard was 
one source of his ambitious designs on England, which 
had been so often pillaged by t^e renowned Vikings 
or kings of the sea, the rivals of his own ancestors ; 
but he was quite ignorant of the language of the 
North, which his subjects had never learned. When 
RoUo, the founder of the duchy of Normandy, came, 
in the beginning of the tenth century, to pay homage 
to Charles the Simple, he had hardly pronounced the 
first words of the oath, " By Got" than the whole Car- 
lovingian court, Germanic also in its origin, burst out 
into laughter. He was not understood; it was the 
same at Rouen and at Falaise, and the Scandinavian 
tongue fell into disuse. 

A similar lot would have befallen French in Eng- 
land if William had come to govern the English, 
but he had only conquered the country in order to 
dispose of its wealth. He distributed land, offices, and 
honours to his vassals so systematically that the whole 
country was enveloped in a close network of which 
every knot represented a barony, a county, a Norman 
domain. The accession of the Plantagenets introduced 
a new influx of greedy strangers, the honoured ances- 
tors of the noblest families. In each town, each rural 
district, the proprietor, the judge, the civil or military 
chief and his vassals constituted a French centre. 

But in creating a dominant caste, the Conquest really 
reanimated Anglo-Saxon, of which the decadence is 
apparent in the eleventh century under Edward the 
Confessor. The speech of the vanquished resisted 
energetically the language of the conquerors; they 



4 1 8 The Indo-Europeati Orgamsm. 

subsisted side by side without mingling for about two 
centuries. The barons, the monks, the soldiers from 
the Continent were obliged to learn some Saxon in 
order to speak to their vassals and serfs, who refused 
to learn French. In 1205 the conquest of Normandy 
by Philip Augustus,, by isolating Great Britain from 
France, arrested for a time the foreign immigration 
and created the English nationality. Henry III., in 
1248, was forced to publish the deliberations of Par- 
liament in the vulgar tongue. From the year 1350, 
according to Chaucer, the use of French was confined 
to the court and to the public acts. In 1363 Edward 
III. commanded that English was to be used at all 
tribunals. From that time every one spoke English ; 
yet Henry V. still considered French to be his mother- 
tongue. 

Saxon, however, had but a partial victory. Towards 
the end of the twelfth century it had accepted and 
made its own a few French words. In the fifty-eight 
thousand lines of Layamon's Brut there are eighty 
words of French origin. At the end of the thirteenth 
we find a hundred in the first five, hundred lines of 
Robert of Gloucester. Nor is this all ; the grammatical 
structure is modified. The case-endings are beginning 
to disappear ; the final vowels, changed into e mute, 
are no longer pronounced. These features are most 
sensible in the Ormulum, a Northumbrian text of the 
thirteenth century. At this period we pass from semi- 
Saxon to Old English, represented by the Chronicles 
of Robert of Gloucester, by Langtof, and by an ancient 
Psalter. The conjugation, already poor, as in all 
Teutonic languages, loses a few more of its shades. 
The internal changes in the verbs diminish ; Infode or 
lufede and gelufod are confounded in loved; and the 



Two Analytical Languages. 4 1 9 

old forms, of which a great number have survived to 
modern days, come to be considered as irregularities 
and exceptions. The plur.J terminations of persons 
are dropped. The so-called regular verbs have now 
only three persons in all, the first, second, and third 
person singular of the indicative, and two forms, the 
infinitive and the past. The noun loses the plurals 
formed by the weakening of the vowel, such as feet, 
teeth, men, mice, its genitive plurals in ene ; the genitive 
singular in is, es, soon reduced to s, alternates with the 
prepositions 0/ and to. The adjective is still declined, 
and often forms its plural in e, but the insignificance 
of this termination causes it to be soon lost. The 
finals have no longer any meaning in a language which 
ever throws the accent back towards the beginning of 
the word ; this marked feature of English pronunciation 
was already appearing, and it has been one of the 
factors of the language. It is this remarkable volu- 
bility of the accent, joined to a strange guttural effort, 
which, slurring over the rest of the word for the sake 
of the first syllables, has produced those violent con- 
tractions of which the spelling preserves the traces. 

This Old English period was naturally a crisis of 
disorganisation, full of the new life. In the succeed- 
ing period, that of Middle English, Saxon accepted its 
losses and its gains, abandoning the shreds of grammar 
which have no longer any use, and adopted with its 
own accentuation and pronunciation the Norman and 
Latin vocabulary. Wycliffe's translation of the Bible 
(i 324-1 384), Maundeville's "Travels" (i 300-1 371), 
"The Vision of Piers Plowman," by Langland, and 
especially the poems of Chaucer, are the principal 
works written in Middle English. Chaucer is the 
father of English poetry ; his " Canterbury Tales " will 



420 TJie Indo-European Orgajiisni. 

ever be famous for the perfect ease and freshness of 
their style. 

The tendency to analysis continues in the English 
of the Tudors and of the Renaissance, and a great 
number of Latin terms are introduced, both by tbe clas- 
sical revival, which extended even to the women, and 
by the theological mania, which prevailed in England 
fully as much as in France and Germany. From the 
sixteenth century English literature took a wonderful 
development, of which in this rapid sketch we cannot 
even give an idea. It is the language, moreover, 
which concerns us. The undecided spelling, the 
presence of numerous final mute «'s, the use of the th 
for thi s, seem at the first glance to be the only 
distinguishing marks between the English of Surrey, 
Thomas More, Spenser, and Shakespeare, and that of 
Bacon, Milton, Locke, Pope, and Swift. But we scon 
perceive that the authors of the age of Elizabeth are 
distinguished by a freedom and originality, by an abun- 
dance of strength and vigoir, not without a certain 
extravagance, which, however, is not unsuited to youth. 
Their successors feel the need of the greater correct- 
ness, of that studied elegance which has given a certaiu 
monotony to our own classical literature. Finally, the 
nineteenth century seems to have restored to English, 
as to French, freedom, ease, and variety. 

It is interesting to know in what proportions the 
Teutonic and the Latin elements are mixed in modern 
English, what sacrifices Saxon had to make to French, 
what resources it drew from its rival. These calcula- 
tions have been made. Out of 1069 words used by 
King Alfred, Turner has counted 230, or more than a 
fifth, which are no longer in use. He also reckons 
that the proportion of Norman to Saxon words in the 

\ 



Two Analytical Languages., 421 

vocabulary is as two to three. And in counting all 
the words in the dictionaries of Webster and Robert- 
son, M. Thomttierel has shown that out of 43,566 
words, 29,853 come from the classical languages, 
13,230 from the Teutonic. Now of these 30,000 
French or Latin words, which for the rest are entirely 
■ Anglicised in pronunciation, more than a third are but 
doubles of Saxon words. Hence there is an infinite 
wealth of synonyms — that is, of expressions applicable 
to the subtlest variations of the' same idea. The Teutonic 
element predominates in all that relates to the pro- 
ducts of nature, minerals, plants, living animals, to the 
structure of the human body, to temperature, and to 
atmospherical phenomena, in the names of utensils, 
Jurniture, and tools, AH which marks the relation of 
words to each other, articles, prepositions, conjunctions, 
is also Saxon. Politics, law, social functions, wealth, 
honours, philosophy, art, science, trades, and cooking 
derive their terms from French and Latin. Poetry 
uses Saxon words by preference, and this it is which 
renders it so difficult for foreigners to understand ; 
there are two languages in English, and he must know 
them both who would read Shakespeare and Byron. 

English, in short, received in the eleventh century 
rfrom a more advanced civilisation, and has con- 
tinued to borrow from it, the expression of those ideas 
which it had not had time to acquire. French presents 
a similar phenomenon; only it was from the same 
source as it had drawn the elements of its popular 
-form that it sought the new terms required by the 
progress of science and thought Having exhausted 
the Latin of the people, it borrowed freely from classical 
Latin. It therefore remains far more homogeneous 
than English; but it has drawn similar advantages 
28 



422 The Indo-European Organism. 

from a parallel method. All the other Komance 
idioms have done the like ; they have continually in- 
creased their vocabulary by borrowing from Latin, 
their common source, and from the sister languages 
issued from Latin. If philologists have not found in 
Spanish or Italian this so-called learned formation 
which they complain of in French, it is because these 
languages are less contracted and have from the be- 
ginning followed more closely the original type. They 
are astonished and aggrieved that Frenchmen have not 
treated th« Latin of Cicero as the Gallo-Franks treated 
the Latin of the people ; that the new acquisitions were 
not passed through the popular channel, and thus 
did not suffer that insensible transformation which, I 
readily allow, has given to our tongue its proper char- 
acter and aspect. I do what I can to share in these 
regrets, but I cannot prefer Villehardouin, Joinville, 
Beaumanoir, to Froissart, Commines, Villon, much less 
to Rabelais, Montaigne, or Ronsard, to La Fontaine, 
Molifere, Voltaire, Musset, or Victor Hugo. 

I will give in a few words (for Littr^, Gaston Paris, 
and Brachet have made an exhaustive study of this 
matter) in what consists the popular formation, and 
in what the learned. 

■ Four fundamental laws governed the development 
of French: (i) the persistence of the Latin tonic accent 
and the alteration of the accented vowel ; (2) the sup- 
pression of the short vowel before the accent when 
this is not the initial vowel ; (3) the loss of the middle 
consonant; (4) the contraction or the loss of the ter- 
mination. By these marks we recognise the words 
of the first formation, of popular origin : claritatem, 
clarU; comitatum, comU; simulare, sembler; videre,voir; 
audire, ouir ; magistrum, maifre ; viearius, viguier ; 



Two Analytical Languages. 423 

advocatus, avcyuA ; augustus, aoUt ; tepidwm, tiede ; male- 
aptum, malade ; mansion, maison ; regerni, roi, &c. 

This popular derivation constitutes the foundation 
and the substance of the French language : as it suf- 
ficed for the mental condition of the people, confined 
in the narrow limits of Christianity and feudalism, the 
people allowed towards the thirteenth century this 
method, of which it no longer felt the need, to fall 
into disuse. But with the revival of learning new 
words were required, either to translate the ancient 
authors, or to render more complex thought, to express 
a greater variety of ideas. 

In the fourteenth century, therefore, began a new 
formation, called learned, because it was the work of 
the lettered class, who, without considering the laws 
of internal contraction and phonetic alteration, simply 
adopted the Latin word. To this formation we owe 
so many of the doublets which enrich our tongue. 
When we compare terms such as incUnaison and in- 
clination, poison and potiori, avoui and avocat, esclandre 
and scandale, dimanclie and dominiqiie, chez and case, 
eombler and cumuler, &c., we recognise that of these 
double forms issued from the same Latin word, the 
first is of popular formation, the second the invention 
of scholars. To these two main sources of the French 
vocabulary we must add numerous terms borrowed from 
Greek, Italian, and Spanish (sixteenth and seventeenth 
centuries),from modern English and German (eighteenth 
and nineteenth centuries), the whole scientific and 
technical terminology, an interesting class of Eastern 
words, and finally the Greek sufBxes mane, Tiiorphe, gone, 
graphic, logic, ismc, iste, iser, of which the abuse is 
common, and we have an approximative idea of the 
history of the French tongue. From this composite 



424 The Indo-European Organism. 

material we are continually producing, by derivation 
and juxtaposition, new forms. For French, no less 
than English, is full of life, as Darmestetter has shown, 
and is continually drawing farther away, with all respect, 
from the classical models, which are a little stiff in 
their solemnity. 

Which of these two languages, French or English, 
both disengaged from the trammels of grammar, both 
arrived at the term of the linguistic cycle, shall obtain 
the prize ? It is not for us to say. English, justly 
proud of its wealth and of its immense expansion ; 
French, with its ancieat origin, its early literary de^ 
velopment, its lucidity, its pronunciation, clear without 
undue emphasis, have both had a career which need 
fear no comparisons ; both have been employed in evert 
style, for the expression of the deepest thought and of 
the subtlest ideas. 

And now, rciaking our excuses to interesting and 
beautiful languages which our narrow limits do not 
allow us to enter upon, Persian, Italian, Spanish, we 
leave with regret this vast domain of language, where 
we have everywhere found speech in exact correspond- 
ence with the intellectual and moral needs of man. 
The evolution of language is parallel to the evolution 
of humanity. The descendant of the animal cry, it 
has raised man above the brutes ; at once factor and 
instrument in our progress, creator of the conscience 
and of science, it is the link between nature and history, 
between physiological and moral anthropology. 



D. APPLETON & CO/S PUBLICATIONS. 

Recent Volumes of the International Scientific Series. 
A HISTOR Y OF CR USTACEA. By Rev. Thomas 

■^^ R. R. Stebbing, M. a., author of " The Challenger Amphipoda," 
etc. With numerous Illustrations. i2mo. Cloth, $2.00. 

*' Mr. Stebbing's, account of \ Recent Malacostraca' (soft-shelled animals) is practi- 
cally complete, and is based upon the solid foundations of science. 'I'he sstuaifhing 
development of knowledge in this branch of natural history is due to the extension of 
marine research, the perfecting of the microscope, and the general diffusion of informa- 
tion regarding what has been ascertained concerning the origin of species. . . . This 
volume is fully illustrated, and coniains useful references to important authorities. It 
is an able and meritorious survey of recent Crustacea." — Jhkiiadelpkia Ledger. 

*' In all respects an admirable piece of work," — The Churchman. 

"One of the most valuable and entertaining volumes in the series. . . . The author 
is master of an engaging style, and offers words of cheer and counsel to the beginner 
who may be dismayed lay the bewildering riches of the crustacean world. Every branch 
of the subject treated is presented in the most interesting and significant light." — Lcn- 
don Saturday Review. 

TTANDBOOK OF GREEK AND LATIN FA- 
J^ LMOGRAPHY, By Edward Maunde Thompson, D.C, L., 
Principal Librarian of the British Museum. With numerous 
Illustrations. l2mo. Cloth, $2.00. 

" Mr. Thompson, as principal librarian of the British Museum, has of course had 
very exceptional advantag;es for preparing his book. . . . Probably all teachers of the 
classics, as well as specialists in palseography, will find something of value in this sys- 
tematic treatise upon a rather unusual and difficult study." — Review 0/ Revieuus, 

" A, well-arranged manual from the hands of a competent authority. ... Of the 
nineteen chapters contained in the volume, seven deal with preliminary topics, as the 
history of the Greek and the Latin alphabets, writing materials, the forms of books, 
punctuation, measurement of lines, shorthand, abbreviations, and contractions; five 
are devoted to Greek paleography, seven^to Latin." — The Critic. 

" Covering as this volume docs such a vast period of time, from the beginning of the 
alphabet and the ways of writing down to the seventeenth century, the wonder is how, 
within three hundred and thirty-three pages, so much that is of practical usefulness has 
been brought together."— iVifw VorM Times. 

]\/r^N AND THE GLACIAL PERIOD, By G. 

■^ '^ Frederick Wright, D. D., LL. D., author of "The Ice Age in 

North America," " Logic of Christian Evidences," etc. With 

numerous Illustrations. i2mo. Cloth, $1.75. 

'* The author is himself an independent student and thinker whose competence and 

authority are undisputed."— iVl?w York Sun. 

"_ It may be described in a word as the best summary of scientific conclusions con- 
cerning the question of man's antiquity as affected by his known relations to geological 
t\m&."~~Phi(adelphix Press. 

" The earlier chapters describing glacial action, and the traces of it in North Amer- 
ica—especially the defining of its limits, such as the terminal moraine of the great move- 
ment itself— are of great interest and value. The maps and diagrams are of much as- 
sistance in enabling the reader to grasp the vast extent of the movement." — London 
Spectator, 

New York : D, APPLETON & CO.. 72 Fifth Avenue. 



D. APPLETON & CO.'S PUBLICATIONS. 



EVOLUTION SERIES, NOS. i TO 17. 

Popular Lectures and Discussions before the Brooklyn Ethical Association, 



E 



VOLUTION IN SCIENCE, PHILOSOPHY^ 

AND ART, With 3 Portraits. Large 12 mo. Cloth, $2.00. 

CONTENTS. 



Alfred Russet Wallace. By Edward D. 

Cope, Ph.D. 
Ernst Haeckel, By Thaddeus B. 

Wakeman. 
TJte Scientific Method. By Fhancis E, 

Abbott, Ph. D. 
Herbert Spencer^ s Synthetic Philosophy. 

By Benjamin F, Underwood. 
Evolution ofCheriiistry. By Robert G. 

Ecci.es, M. D. 
Evolution of Electric and Magnetic 

Physics. ■ By Arthur E. Ken- 
nelly. 
Evolution of Botany. By Fred J. 
- Wulling,* Ph. 0.. 
Zoology 0s related to Evolution. By 

Rev. John C. Kimball. 



Fonn and Color in Nature. By Wil- 
liam Potts. 
Optics as related to Evolution. By L. A. 

W. Alleman, M. D. 
Evolution of A rt. By John A. Taylor. 
Evolutio7i of Architecture. By Rev.' 

John W. Chadwick. 
Evolution of Sculpture. J3y Prof. Thomas 

Davidson. 
EvohiHon of Painting. By Forrest P. 

Rundell. 
Evolution of Music. By Z. Sidney 

Sampson. 
Life as a Fine Art. By Lewis G. 

Janes, M. D. 
The Doctrine of Evolution : its Scope and 

Influence. By Prof. John Fiske. 



*' The addresses include some of the most important presentations and epitomes pub- 
lished in'America. They are all upon important subjects, are prepared with great care, 
and are delivered for the most part by highly eminent authorities." — Public Opinion. 

EVOLUTION SERIES, NOS. 18 TO 34. 
'AN AND THE STATE, Studies in Applied 

Sociology. With Index. Large i2mo. Cloth, $2.00. 

By William 



J/: 



The Duty of a Public Spirit. ijy x 
Benjamin Andrews, D. D., LL. D. 

The Study of Applied Sociology. By 
Robert G. Eccles, M. D. 

Pepresentative Govemmeyit. By Edwin 
D. Mead. 

Stiffrage.and the Ballot. 5y Daniel S. 
' Remsen. 

The Land Problem. By Prof. Otis T. 
Mason. 

The Problem, of City Government. By 
Dr. Lewis G. Janes. 

Taxation and Revenue : The Free- 
Trade^ Vieiv. By Thomas G. 
Shearman. 

Taxation and Revenue : The Protec- 
iionist View, By Prof. Geokge 
Gunton. 



CONTENTS. 
By E, 



The Monetary Problejn. 

Potts. 

The Immigration Problem. By Z. Sid- 
ney Sampson. 
Evolution of the Afric-A 7nerican. By 

Rev. Samuel J. Barrows. 
The Race Problem in the South. By 

Prof. Joseph Le Conte. 
Education a?id Citizenship. By Rev. 

John W. Chadwick 
The Democratic Party. By Edward M. 

Shepard. 
The Republican Party. By Hon. Ros- 

well G. Horr. 
The Independent in Politics. Ey John 

A. Taylor. 
Moral Questions in Politics. By Rev. 

John C. Kimball. 



CinJmmiiTim^^'stt^^^^^^ sociology are exceptionally interesting in their field. "— 

Chr^ide^TT^^^^h^^ attention of the progressive student of politics."— P/Y&iwr^ 

Separate Lectures from either volume ^ 10 cents each. 

New York : D. APPLETON & CO.. 72 Fifth Avenue. 



D. APPLETON & CO.'S PUBLICATIONS. 

EVOLUTION SERIES, NOS. 35 TO 48. 

CACTORS IN AMERICAN CIVILIZATION: 

■L STUDIES IN APPLIED SOCIOLOGY. Popular Lectures 

and Discussions before the Brooklyn Ethical Associat*on. 

l2nio. Cloth, $2.00. Separate Lectures, in Pamphlet Form, 

10 cents each. 

This volume is uniform with the two previous volumes of the 

series, entitled respectively " Evolution in Science and Art " and 

"Man and the State." 

CONTENTS. 

35. Thi Nation's Place in Civilization. By CHARLES De Garmo, 

Ph. D., President of Swarthmore College. 

36. Natural Factors in American Civilization. By Rev. John C. 

Kimball. 

37. What America Owes to the Old World. By A. Emerson Palmer. 

38. War and Progress. By Dr. Lewis G. Janes. 
3g. Interstate Commetce. By Robert W. Tayler. 

40. Foreign Commerce. By Hon. William J. Coombs. 

41. The Social and Political Status oj Woman. By Rev. John W. 

Chad WICK. 

42. The Economic Position of Woman. By Miss Caroline B. Le 

Row. 

43. Evolution of Penal Methods and Institutions. By James Mc- 

Keen. 

44. Evolution of Charities and Charitable Institutions. By Prof. 

Amos G. Warner, Ph. D., Superintendent of Public Charities, 
Washington, D. C. 

45. The Drink Problem. By T. D. Crothers, M. D., Editor of the 

" Quarterly Journal of Inebriety." 

46. The Labor Problem. By Rev. Nicholas P. Oilman, Editor of 

the " New World." 

47. Political A-ifects of the Labor Problem. By Jeremiah W. Sul- 

livan. 

48. The Philosophy of History. By Rev. E. P. Powell, Author of 

" Our Heredity from God," etc. 

"One can hardly speak too highly of the work which is heing done by the 
Brooklyn Ethical Association. Its plan is to bring within definite compass and 
knowledge some of the largest subjects which can occupy the minds of thoughtful 
men. It has found students and thinkers who are equal to this task, and here we liave 
some of the best work on subjects of the highest meaning that has been done by 
Americans." — Boston Herald. 



New York: D. APPLETON & CO., 72 Fifth Avenue. 



D. APPLErON-& CO.^S PUBLICATIONS. 
IDLE DAYS IN PATAGONIA, By W. H. Hud- 

-* SON, C. M. Z. S., author of " The Naturalist in La Plata," etc. 
With 27" Illustrations. 8vo. Cloth. $4.00. 

*jOf all modem bonks of travel it is certajnly one of the most original, and many, 
we are sure, will also find it one of the most interesting and suggestive." — Neiv York 
Tribune. 

" Mr. Hudson's remarks on color and expression of eyes in man and animals are re- 
served for a second chapter, 'Concerning Eyes.' He is eloquent upon ihe pleasures 
afforded by ' Bird Music in South America,' and relates some romantic tales of white 
men in captivity to savages. But it makes very little difference what is the topic when 
Mr. Hudson writes. He calls up bright images of things unseen, and is a thoroughly 
agreeable coTapa.mon."-~'PAzladel/Aza Ledger 

7^HE NATURALIST IN LA PLATA. By W. H. 
Hudson, C. M. Z. S., author of " Idle Days in Patagonia," and 
joint author of " Argentine Ornithology." With 27 Illustra- 
tions. 8vo. Cloth, $4.00. 

" Mr. Hudson is riot only a clever naturalist, but he possesses the rare gift of in- . 
teresting his readers in whatever attracts him, and of being dissatisfied with mere ob- 
servation unless it enables him to philosophize as well. With his lucid accounts of 
bird, beast, and insect, ,no one will fail to be delighted." — London Academy. 

"A notably clear and interesting account of scientific observation arid research. 
Mr. Hudson has a keen eye for the phenomena with which the naturalist is concemedf 
and a lucid and delightful way of writing about them, so that any reader may be 
charmed by the narrative and the reflections here set forth. It is easy to follow him, 
and we get our information agreeably as he conducts us over the desert pampas, and 
makes us acquainted with the results of his studies of animals, insects, and birds.'' — 
New York Sun. 

n^HE NATURALIST ON THE RIVER 
J- AMAZONS, By Henry Walter Bates, F. R. S., late Assist- 
ant Secretary of the Royal Geographical Society. With a 
Memoir of the Author, by Edward Clodd. With Map and 
numerous Illustrations. 8vo. Cloth, $5.00. 
" This famous work is a natural history classic." — London Literary World. 
" More than thirty years have passed since the first appearance of • The Naturalist 
on the River Amazons,' which Darwin unhesitatingly pronounced the best book on 
natural history which ever appeared in England. The work still retains its prime in- 
terest, and in rereading it one can not but be impressed by the way in which the pro- 
phetic theories, disputed and ridiculed at the time, have since been accepted. Such is 
the common expenence of those who keep a few paces in advance of their generation. 
Bates was a 'bom' naturalist." — Philadelphia Ledger. 

"No man was better prepared or gave himself up more thoroughly to the task of 
studying an almost unknown fauna, or showed a zeal more inde&tigable in prosecuting 
his researches, than Bates. As a collector alone his reputation would be second to 
none, but there is a great deal more than sheer industry to be cited. The naturalist of 
the Amazons is,j*ar excellence ^ possessed of a happy literary style. He is always clear 
and distinct. He tells of the wonders of tropical growth so that you can understand 
them all." — New York Times. 



New York: D. APPLETON & CO., 72 Fifth Avenue. 



D. APPLETON & GO'S PUBLICATIONS. 



CAMP-FIRES OF A NATURALIST. From the 
Field Notes of Lewis Lindsay Dyche, A. M., M. S., Professor 
of Zoology and Curator of Birds and Mammals in the Kansas 
State University. The Story of Fourteen Expeditions after 
North American Mammals. By Clarence E. Edwords. 
With numerous Illustrations. i2mo. Cloth, $1,50. 

" It is not always that a professor of zoSlogy is so enthusiastic a sportsman as Prof. 
Dyche. His hunting exploits are as varied as those of Gordon Gumming, for example, 
ill bOLith Africa. His grizzly bear is as dangerous as the lion, and his mountain sheep 
and goats more difficult to stalk and shoot than any creatures of the torrid zone. Evi- 
dently he came by his tastes as a hunter from lifelong experience." — New York 
Tribune. 

" The book has no dull pages, and is often excitingly interestmg, and fully in- 
structive as to the habits, haunts, and nature of wild beasts." — Chicago Inter-Ocean. 

" There is abundance of interesting incident in addition to the scientific element, 
and the illustrations are numerous and highly graphic as 10 the big game met by the 
hunters, and the hardships cheerfully undertaken." — Brooklyn Eagle. 

"The narrative is simple and manly and full of the freedom of forests. . . . This 
record of his work ought to awaken the interest of the generation growing up, if only 
by the contrast of his active experience of the resources of Nature and of savage life 
with the background of culture and the environment of educational advantages that 
are being rapidly formed for the students of the United States. Prof. Dyche seems, 
from this account of him, to have thought no personal hardship or exertion wasted in 
his attempt to collect facts, that the naturalist of the future may be provided with com- 
plete and verified ideas as to species which will soon be extinct. This is good work — 
work that we need and that posterity will recognize with gratitude. The illustrations 
of the book are interesting, and the type is clear." — New York Times. 

" The adventures are simply told, but some of them are thrilling of necessity, how- 
ever modestly the narrator does his work. Prof. Dyche has had about as mHny expe- 
riences in the way of hunting for science as fall to the lot of the mcst fortunate, and 
this recountal of them is most interesting. The camps from which he worked ranged 
from the Lake of the Woods to Arizona, and northwest to British Columbia, and in 
every region he was successful in securing rare specimens for his museum." — Chicago 
Times, 

** The literary construction is refreshing The reader is carried into the midst of 
the very scenes of which the author tells, not by elaborateness of description but by the 
directness and vividness of every sentence. He is given no opportunity to abandon 
the companions with which the book has provided him, for incident is made to follow 
incident with no intervening literary padding. In fact, the book is all action." — Kansas 
City Journal. 

"As an outdoor book of camping and hunting this book possesses a timely) 
interest, but it also has the merit of scientific exactness in -the descriptions of the 
habits, peculiarities, and haunts of wild animals." — Philadelphia Press. 

" But what is most important of all in a narrative of this kind— for it seems to uil 
that 'Camp-Fires of a Naturalist' was written first of all for entertainment— these 
notes neither have been ' dressed up ' and their accuracy thereby impaired, nor yet re- 
tailed in a dry and statistical manner. The book, in a word', is a plain narrative of 
adventures among the larger American ^T^\m2\&." —Philadelphia Bulletin. 

" We recommend it most heartily to old and young alike, and suggest it as a bcauti 
ful souvenir volume for those who have seen the wonderful display of mounted animal? 
at the World's Pair."— Topeka Capital. 



New York : D. APPLETON & CO., 72 Fifth Avenue, 



D. APPLETON & GO'S PUBLICATIONS. 



r 



'HE GILDED MAN {EL DORADO), and other 

Pictures of the Spanish Occupancy of America. By A. F- 
Bandelier. i2mo. Cloth, $1.50. 

"Every paper in this volume is wonderfully ijiteresting, and the collection is of 
such historical value as to make it a necessary part of every library in which American 
history is represented." — Boston Herald. 

"One of the most entertaining of recent historical works, and, besides its no^^ty 
and freshness has the great merit of being original historical research." — Phlladel- 
ptiia Times. 

"Mr. Bandelier's work under the auspices of the Archaeological Institute of 
America and on the Hemenway Survey entitles him to rank as the leading docu- 
mentary historian of the Southwest. . . . The book possesses genume historical value, 
and is a necessary part of the annals of our country." — Fhiladetphia Ledger. 

" Just such a work as Mr. Bandelier has done has long been needed. ... A con- 
tribution of the first order of value to a part of American history that deserves to be 
more fully studied." — Literary World, 

TJZARRIORS OF THE CRESCENT. By W. 
1^ t^ H. Davenport Adams, author of " Battle Stori^es from Eng- 
lish History/' etc. i2mo. Cloth, $1.50. 

"A work without a rival in its particular field. . , . All ihe gorgeousness of the 
Ibarbaric East invests this glowing pageant of kings and conquerors. . . . This is a re- 
markably able book in thought and in manner of presentatipn.'*— /%//fflrff^/«Vr Ledger. 

" A lively, carefully prepared chronicle of the careers of quite a number of the Mo- 
hammedan rulers in Asian regions who made their marks, one way or another, in the 
dsvelopment of the peculiar civilization of the East. , . . This author has selected from 
the long chronicle the salients likely to be most interesting, and has obviously taken 
much pains to sift the fact carefully out oi the rather confused mass of fact and fable in 
the Musiem chronicles."— A^grw VorM Co mviercial Advertiser. 

" Nowhere in history are there to be found such records of conquest, such frightful 
tales of blood, such overwhelming defeats or victories, as in the Irves of the Asiatic 
sovereigns. . . . The author is a hiitorinn who tells his siory and stops. He has done 
his work faithfully and well." — Cincinnati Cojninercial Gazette. 

PICTURES FROM ROMAN LIFE AND STOR Y, 

^ By Professor A. J. Church, author of " Stories from Homer, " 

" ^tories from Virgil," etc. Illustrated. i2mo. Cloth, $1.50. 

*' Prof. Church is a tried and approved master of the art ofinterestingyoung people 
ill historical themes. The present work, while too thoughtful to be called strictlyjnve- 
nile, treats of the great emperors and families of Rome in a simple narrative style cer- 
tain to captivate youth and older people fond of historic lore." — 'J he Chautauguan. 

" The material for these sketches is drawn partly from the inexhaustible riches of 
Plutarch, partly from contemporaneous history, and partly from letters, edicts, etc.: 
and, well chosen and briefly related, are interesting, whetting the appetite of the stu- 
diously inclined, . . . Various illustrations add to the interest of the work.' — Sfring. 
Ji^ld Republican. 

" Each of the chapters presents some striking scene or personality in the period from 
Augustus to Marcus Aurelius. . . . Several of the chapters are thrown into the form 
of contemporary letters. The plan of the book is well conceived, and the subjects are 
those of general human interest." — New York Critic. 



New York • D. APPLETON & CO., 72 Fifth Avenue.' 



D. APPLETON & CO.'S PUBLICATIONS. 



L 



IFB IN ANCIENT EGYPT AND ASSYRIA, 

By G. MASpiRO, late Director of Archaeology in Egypt, and 
Member of the Institute of France. Translated by Alice 
Morton. With i88 Illustrations. i2mo. Cloth, $1.50. 

** A lucid sketch, at once popular and learned, of daily life in Egypt in the time of 
Rameses 1 1, and of Assyria in that of Assurhanipal. . . . As an Orienialist, M, Mas- 
p6ro stands in the front rank, and his Jearning is so well digested and sc admirably sub- 
dued to the service of popular exposition, that it nowhere overwhelms and always in- 
terests the reader." — London Times, 

" Only a writer who had distinguished himself as a student of Egyptian and As- 
syrian antiquities could have produced this work, which has none of the features of a 
modem book of travels in the East, but is an ai tempt to deal with ancient life as if one 
had been a contemporary with the people whose civilization and social usages are 
very largely restored." — Boston Herald, 

A most interesting and instructive book. Excellent and most impressive idea?, 
also, of the architecture of the two countries and of the oiher rude but powerful art of 
the Assyrians, are to be got from it." — Brooklyn Eagie4 

** The ancient artists are copied with the utmost fidelity, and verify the narrative so 
attractively presented." — Cincinnati Times-Star, 

Z'HE THREE PROPHETS: Chinese Gordon; 
Mohammed-Ahmed ; Araby Pasha. Events before, during, 
and after the Bombardment of Alexandria. By Colonel 
Chaille-Long, ex-Chief of Staff to Gordon in Africa, ex- 
United States Consular Agent in Alexandria, etc., etc. With 
Portraits, i6mo. Paper, 50 cents. 

" Comprises the observations of a man who, by reason of his own military ex- 
perience in Egypt, ought to know whereof he speaks."— Washington Post. 

** The book contains a vivid account of the massacres and the bombardment of Alex- 
andria. As throwing light upon the darkened problem of Egypt, this American 
contribution is both a useful reminder of recent facts and an estimate of present situa- 
tions." — Fkiladeipkia Public Ledger. 

"Throws an entirely n?w light upon the troubles which have so long agitated 
Egypt, and upon their real significance." — Chicago 'Jimes. 

Z^HE MEMOIRS OF AN ARABIAN PRIN- 
CESS. By Emily Ruete, n^c Princess of Oman and Zanzi- 
bar. Translated from the German. i2mo, Cloth, 75 cents. 
The author of this amusing- autobiography i^ half-sister to the late Sul- 
tan of Zanzibar, who some years ago married a German merchant and settled 
at Hamburg. 

**A remarkably interesting little vohime. . . . As a picture of Oriental court life, 
and manners and customs in the Orient, by one who is to the manner bom, the book is 
prolific in entertainment and edification." — Boston Gazette, 

*'The interest of the book centers chiefly in its minute description of the daily life 
of the household from the time of rising until the time of retiring, giving the most com- 
plete details of dress, meals, ceremonies, feasts, weddings, funerals, education, 
slave service, amusements, in fact everything connected with the daily and yearly 
routine of life." — Utica (N, V,) Herald. 



New York: D. APPLET0:N" & CO., 72 Fifth Avenue. 



D. APPLETON & CO.'S PUBLICATIONS. 



--nHE SYNTHETIC PHILOSOPHY OF HER- 

J- BERT SPENCER. In nine volumes. i2mo. Cloth, $2.00 
per volume. The titles of the several volumes are as follows ; 

(i.) FIRST PRINCIPLES. 

I. The Unknowable. II- Laws of the Knowablc. 

(2.) THE PRINCIPLES OF BIOLOGY. Vol.1. .^^ , , .. . „. , 

I. The Data of Biology. II. The Inductions of Biology. 

111. The Evolution of Life. 
(3.) THE PUINCIPLES OF BIOLOGY. Vol. II. 

IV. Morphological Development. V. Physiological Development. 

VI. Laws of Multiplication, 
(4.) THE PRINCIPLES OF PSYCHOLOGY. Vol. I. . 

I. The Data of Psychology. Ill- General Synthesis. 

II. The Inductions of Psychology. IV. Special Synthesis. 

V. Physical Synthesis. 
(S.) THE PKINCIPLES OF PSYCHOLOGY. Vol. II. 

VI. Special Analysis. , VIII. Congruities. 

VII. General .Analysis. IX. Corollanes. 

(6.) THE PRINCIPLES OF SOCIOLOGY. Vol. I. 

I. The Data of Sociology. II. The Inductions of Sociology. 

III. The Domestic Relations. 
(7.) THE PRINCIPLES OF SOCIOLOGY. Vol. II. 

IV. Ceremonial Institutions. V. Political Institutions. 

VI. Ecclesiastical Institutions. 
(8.) THE PRINCIPLES OF SOCIOLOGY. VoL III. 

# H< !(: # 

(9.) THE PRINCIPLES OF ETHICS. Vol. I. 

I. The Data of Ethics. II. The Inductions of Ethics. 

III. The Ethics of Individual Life. 
(10.) THE PRINCIPLES OF El'HICS. Vol. II. 
IV. The Ethics of Social Life: Justice. 
V. The Ethics of Social Life : Negative Beneficence. 
VI. The Ethics of Social Life: Positive Beneficence. 

r\ESCRIPTIVE SOCIOLOGY. A Cyclopedia of 

•*—^ Social Facts. Representing the Constitution of Every Type 
and Grade of Human Society, Past and Present, Stationary and 
Progressive. By Herbert Spencer. Eight Nos., Royal Folio. 

No. I. ENGLISH $4 ix> 

No. II. MEXICANS.'CENTRAL AMERICANS, CHIBCHAS, and PE- 
RUVIANS 400 

No. III. LOWEST RACES, NEGRITO RACES, and MALAYO-POLY- 

NESIAN RACES 4 00 

No. IV. AFRICAN RACES 4 00 

No. V. ASIATIC RACES . 4 00 

No. VI. AMERICAN RACES 4 00 

No. VII. HEBREWS and PHCENICIANS 4 co 

No. VIII. FRENCH (Double Number) 7 00 



New York : D. APPLETON & CO., 72 Fifth Avenue. 



D. APPLETON & CO.'S PUBLICATIONS. 



NEW EDITION OF SPENCER'S ESSAYS. 

TJ'SSA YS : Scientific, Felitical, and Speculative. By 

■ * ^ Herbert Spencer. A new edition, unifonn with Mr. Spencer's 

other works, including Seven New Essays. Three volunnes, 

l2mo, 1,460 pages, with full Subject-Index of twenty-four pages. 

Cloth, $6.00, 

CONTENTS OF VOLUME I. 



The Development Hypothesis. 

Progress : its Law a^id Cause. 

Transcendental Physiology. 

The Nebular Hypothesis. 

Illogical Geology. 

Bain on the Emotions and the Will 



The Social Organism. 

The Origin of Animal Worship. 

Morals and Moral Sentiments 

The Comparative Psychology of Man. 

Mr. Martineau on Evolution. 

The Factors of Organic Evolution.* 



CONTENTS OF VOLUME II. 



The Genesis of Science 

The Classification of the Sciences. 

Reasons for dissenting from the Phi- 
losophy of M. Comte. 

On Laws in General, and the Order 
of their Discovery. 

The Valuation of Evidence. 

What is Electricity ? 

Mill versus Hamilton — The Test of 
Truth. 

CONTENTS OF 

Manners and Fashion. 
Railway Morals and Railway 

Pohcy. 
The Morals of Trade. 
Prison-Ethics. 
The Ethics of Kant. 
Absolute Political Ethics. 
Over-Legislation. 
Representative Government — 

What is it good tor ? 



Replies to Criticisms. 

Prof. Green's Explanations. 

The Philosophy of Style.t 

Use and Beauty. 

The Sources of Architectural Types 

Gracefulness. 

Personal Beauty. 

The Origin and Function of Music. 

The Physiology of Laughter. 

VOLUME III. 

State-Tampering with Money and 

Banks 
Parliamentary Reform : the Dangers 

and the Safeguards. 
"The Collective Wisdom." 
Political Fetichism. 
Specialized Administration. 
From Freedom to Bondage. 
The .Americans.! 
Index. 



* Also published separately. i2mo. Cloth, 75 cents. 
t Also published separately. lamo. Cloth, 50 cents. 
X Also published separately. i2mo. Paper, 10 cents. 



New Yorlv : D. APPLETON & CO., 72 Fifth Avenue. 



D. APPLETON & CO.'S PUBLICATIONS. 



NEW EDITION OF.J'ROF. HUXLEY'S ESSAYS. 

COLLECTED ESSA YS., By Thomas H. Huxley. 
New complete edition, with revisions, the Essays being grouped 
according to general subject. In nine volumes, a new Intro- 
duction accompanying each -volume. i2mo. Cloth, $1.25 per 
volume. 

Vol. I.— method AND RESULTS. 

Vol. II.— DARWINIANA. 

Vol. III.— science AND EDUCATION. 

Vol. IV.— SCIENCE AND HEBREW TRADITION. 

Vol. v.— SCIENCE AND CHRISTIAN TRADITION. 

Vol. VI.— HUME. 

Vol. VII.— MAN'S PLACE IN NATURE. 

Vol. VIII.— DISCOURSES, BIOLOGICAL AND GEOLOGICAL. 

Vol. IX.-EVOLUTION AND ETHICS, AND OTHER ESSAYS. 

" Mr. Huxley has covered a vast variety of topics during the last quarter of a 
century. . It gives one an agreeable surprise to look over the tables ot contents and 
note the immense territory which he has explored. To read these books carefully 
and studiously is to become thoroughly acquainted with the most advanced thought 
on a large number of topics." — New York Herald. 

"The series will be a welcome one. There are few writings on tlie more abstruse 
problems of science better adapted to reading bj; the general public, and in this form 
the books will be well in the reach of the investigator. . . . The revisions are the last 
expected to be made by the author, and his introductions are none of earlier date 
than a few months ago [1893!, so they may be considered his final and most authorita- 
tive utterances. — Chicago Times. 

"It was inevitable that his essays should be called for in a completed form, and they 
Will be a source of delight and profit to all who read them. He has always commanded 
a hearing, and as a master of the literary style in writing scientific essays he is worthy 
of a place among the great English essayists of the day. This edition of his essays 
will be widely read, and gives his scientific work a permanent iarm." —Boston Herald. 

"A man whose brilliancy is so constant as that of Prof. Huxley will always com- 
mand readers; and the utterances which are here collected are not the least in weight 
and luminous beauty of those with which the author has long delighted the reading 
■iiaAi.."— Philadelphia Press. = e. " "s 

f u" "^V 5°»"j"'«^'' .""f'^g^™™,' of *e essays which their reissue permits brings into 
fuller relief Mr. Huxley s masterly powers of exposition. Sweeping the subject-malter 
clear of all logomachies, he lets the light of common day fall upon it. He shows ihat 
the place of hypothesis in science, as the starting point of verification of the phenomena 
to be explained, IS but an extension of the assumptions which underlie actiiors in every, 
day affairs; and that the method of scienlific investigation is only the method which 
rules the ordinary business of Uk."— London Chronicle. 



New York: D. APPLETON & CO., 72 Fifth Avenue. 



E 



D. APPLETON & CO.'S PUBLICATIONS. 



MEMOIRS OF PROF. E. K YOUMANS. 

DWARD LIVINGSTON YOUMANS, Inter- 
preter of Science for the People. ~ A Sketch of his Life, with 
Selections from his Published Writings, and Extracts from his 
Correspondence with Spencer, Huxley, Tyndall, and others. 
By John Fiske. "With Two Portraits. i2mo. Cloth, $2.00. 

"Whether as a memorial ot a noteworthy man, or as a record 01 a most important 
phase of intellectual life in our own time, the volume is entirely admirable, and must 
be given a high place in the honorable list of recent biography." — Philadelphia 
Times. 

" His life was at once inspiring and interesting. His career gave to manhood in 
America an ornament as well as a potent example. While he lived, he helped to 
enrich thousands of lives. Now that he is gone. Prof. .Fiske's beautiful biography 
not only shows us how noble the man himself was, but how great w as the public 
loss, and how precious must remain the possession of such a memory." — Aeiv 3 or/e 
Times. 

"It was eminently proper that the biography of Mr. Youmans should be written, 
and certainly there could not have been chosen a fitter man than Mr, P iske to write 
it. An acquaintance dating back thirty years is itself a qualification, and when to this 
are added Mr. Fiske's ability and the lucid method which characterizes his work, the 
elements for a satisfactory memoir are all present.."— Philacieiphza Bul.etin. 

" To enumerate Youmans's achievements in the dissemination and interpretation 
of scientific truth is to sum up the record of an epoch from the view-point of the 
gradual enlightenment of the American people. When Mr. Fiske remmds us that 
the discovery and propagation of truth are funciions seldom united in one person, and 
that science, like religion, must have its apostles, he speaks as one having experience 
and authority ; and no one will dispute his competence to define and applaud the 
services which his friend rendered In the capacity of a breaker of the bread of science 
to the multitude." — Neiv York Sun. 

"The selection of Prof. John Fiske as the biographer of the late Prof. Youmans 
was the best thing that could be mads. Prof. Youmanshas done more for the dis- 
semination of scientific information, and the cultivation of a taste for such knowledge, 
than any other American of his day." — Cleveland Plain Dealer. 

" We shall not be misunderstood as agreeing with all the views recorded here by 
Prof. Youmans, from whom we were often compelled to differ while he lived, when we 
say that we have read the book with great interest, and are thankful that one who 
truly and unselfishly labored in the cause of popular science has so worthy a memo- 
rial." — New York Observer. 

" He had the broad democratic spirit, and the absolute unselfishness which it 
reveals at every moment and in every act of his life ; and Mr. Fiske has written a biog- 
raphy which is tender and true, and rich and strong. To it are appended some of his 
writings which have a fitting place here, and fully illustrate his mental gifts and con- 
victions." — Boston Herald. 

" Edward Livingston Youmans was a remarkable character, and the world could 
ill afford to lack a history of his life. Fortunately, the best biographer possible has 
undertaken to write that history, and all thoughtful leaders may rejoice thereat ; for 
John Fiske came to this task well fitted in every way by his intimate personal acquaint- 
ance wiih Mr. Youmans, extending through many years." — Chicago Inier-Ocean. 

"Prof. John Fiske has performed a labor of love for the firiend whose name is its 
title, and one of whose closest intimates he was. The volume is a good example of 
friendly but not unwholesomeiy laudatory biography." — Boston Congregationalist. 



New York : D. APPLETON & CO., 72 Fifth Avenue. 



D. APPLETON & CO.'S PUBLICATIONS. 

'J^HE UNITED STATES OF AMERICA. A 
-* Study of the American Commonwealth, its Natural Resources, 
People, Industries, Manufactures, Commerce, and its Work in 
Literature, Science, Education, and Self-Govemment. Edited 
^ — ^ by Nathaniel S. Shalkr, S. D., Professor of Geology in Har- 
vard University. In two volumes, royal 8vo. With Maps, and 
150 full-page Illustrations. Cloth, $10.00. 

In this work the publishers offer something which is not furnished by 
histories or encyclopaedias, namely, a succinct but comprehensive expert 
account of our country at the present day. The very extent of America and 
American industries renders it difficult to appreciate the true meaning of the 
United States of America. In this work the American citizen can survey the 
land upon which he lives, and the industrial, social, political, and other 
environments of himself and his fellow-citizens. The best knowledge and 
the best efforts of experts, editor, and publishers have gone to the preparation 
of a standard book dedicated to the America of the present day; and the 
publishers believe that these efforts will be appreciated by these who desire 
to inform themselves regarding the America of the end of the century. 

LIST OF CONTRIBUTORS. 

Hon. WILLIAM L. WILSON, Chairman of the Ways and Means Committee, 

Fifty-third Congress. 
Hon. J. R. SOLEY, formerly Assistant Secretary of the Naw. 
EDWARD ATKINSON, LL. D , Ph. D. 
CoL. T A, DODGE, U. S. A. 
Col. GEORGE E. WARING, Jr. 

i'J'; ^t'^KAIS™' ^■■"fessor of History in the University of Pennsylvania. 
CHARLES DUDLEY WARNER, LL D. 
Majoh J. W. POWELL, Director of the U. S. Geological Survey and the Bureau of 

Ethnology. ' 

L^'.WAN'A3B0TTfD!D.'^^°' "^ ^^ Commissioner of Education. 
5i5i>5M%^^^'T^Sft?r°f,"^^''^= ^="5 of the Pacific Coast." 
^,t J^xhomII ^S^^?.?.i Jf?'? °==" "f *= Colleges, Univer ity of Chicago. 
Judge THOMAS M. COOLEY, formerly Chairman of die Interstate Commerce 

Commission. 
CHARLES FRANCIS ADAMS. 

°' tersft^^^'^'^' **■ °' °'''"'" "'■ ""= Hemenway Gymnasium, Harvard Uni- 
CHARLKS HORTON COOLEY. 
A. E. KENNELLY, Assistant to Thomas A. Fdison 

F. (). MILLET, formerly Vice-President of the Nahor,=l A „., J • r t-, ■ 
F. W. TAUSSIG, Professor of PoUtical Economv ?n Har "h n " ■ °J!''^"- 
HENRY VAN BRUNT '^conom.y in Harvard University. 

H. P. FAIK FIELD. 

'^Tufe'^ts^- ^^^°"' ^- °- ^--'^-^ °f "'^ State .oard of Health. Massa- 
N. S. SHALER. 

Sold only by t^^^^Il^^^^^^^^^^^^^^^^r-^^^^ and 
New York: D, APPLETO^T^^,^ piftj^ Avenue. 




THK 



'^0^4. 






SCIENTIFIC SERIES