f LIBRARY 
 
 I UNIVERSITY OF 
 
 cr-[ irni?NV. 
 SAN OKT,f>
 
 THE UNIVERSITY LIBKWI 
 SITY 6F CALIFORNIA, 
 U JWAJL*.
 
 TWO LECTURES 
 
 ON THE 
 SCIENCE OF LANGUAGE
 
 SonDon: C. J. CLAY AND SONS, 
 CAMBRIDGE UNIVERSITY PRESS WAREHOUSE, 
 
 AYE MARIA LANE. 
 (assom: 50, WELLINGTON STREET. 
 
 {All Rights reserved}
 
 TWO LECTURES 
 
 ON THE 
 
 SCIENCE OF LANGUAGE 
 
 by 
 JAMES HOPE MOULTON, M.A., D.LiT. (LOND.) 
 
 Late Fellow of King's College, Cambridge, 
 
 Tutor at the Didsbury Theological College, Manchester, 
 
 Late Senior Classical Master at The Leys School, Cambridge. 
 
 Cambridge 
 at the University Press 
 
 1903
 
 PRINTED BY J. AND C. F. CLAY, 
 AT THE UNIVERSITY PRESS.
 
 lOjANNI PEILE, Litt.D. 
 TP04>EIA
 
 PREFACE. 
 
 ^" V HIS little book consists of two lectures delivered 
 on August 1 9th and 2Oth, 1902, to students of the 
 University Extension at Cambridge. In deference to 
 the kindly expressed wish of many in my audience, 
 and in view of the non-existence (so far as I know) of 
 anything in English giving a purely popular intro- 
 duction to the Science of Language in its latest 
 developments, I ventured to offer the lectures to the 
 University Press for publication, hoping that they 
 misjht serve to stimulate interest in a most fascinating 
 
 o o 
 
 study, sadly neglected in this country. I was en- 
 couraged in this resolution by my friend Professor 
 Ridgeway, who very kindly read the lectures in MS. 
 and helped me with many suggestions. The first 
 lecture is printed nearly as delivered ; the second, 
 which was given extempore, was written out im- 
 mediately after, and follows the general lines of my 
 notes. I have added a brief Bibliography for those
 
 viii Preface 
 
 who may be tempted to pursue the study. It seemed 
 best to preserve the lecture form, which to some 
 extent mitigates the apparent absurdity of putting 
 such a title as "The Science of Language" over a 
 booklet of fifty or sixty pages. There are many 
 things here which would be out of place in a scientific 
 summary : there are many things absent which even 
 an article for a small encyclopaedia ought to contain. 
 Popular lectures will only be expected to include 
 what will rouse interest and lead to further reading. 
 As such I venture to put forth what is almost the 
 only published product of my sixteen years' teaching 
 in Cambridge, so far as the general subject is con- 
 cerned. Writing from a new sphere, where Hellenistic 
 Greek will claim yet more rigorously the time that 
 might have been given to Comparative Philology, 
 I feel as if I were hanging a votiva tabula in the temple 
 of Aius Locutius if that shadowy divinity may be 
 persuaded to take under his patronage a subject so 
 clearly appropriate to him. 
 
 It only remains to express my gratitude to some 
 of my old Cambridge friends for obligations very 
 deeply felt. The Master of Christ's, who has kindly 
 allowed me to inscribe these lectures to him, was the 
 teacher to whom, in undergraduate days, I owed my 
 introduction to the "New Grammarians," then very
 
 Preface ix 
 
 new. Aius Locutius has long received Dr Peile's 
 votive tablet, to the sincere regret of all his old pupils, 
 who will not forget the lucidity, wide knowledge, and 
 unfailing judgment which always informed his lec- 
 tures. His successor in the Readership of Comparative 
 Philology, Mr Giles, has most kindly read my proofs 
 and helped me with a number of suggestions. His 
 learning and acuteness have been an invaluable help 
 to me, as most of my pages would show if space 
 permitted separate mention of the modifications due 
 to his criticism. I need not say, however, that the 
 responsibility for statements made here remains 
 wholly my own. I have also been helped, not for 
 the first time, by my old friend and colleague, Mr 
 E. E. Kellctt, of The Leys, who has carefully read 
 the proofs. The last acknowledgement, alas ! is one 
 which its recipient is no longer here to see. Professor 
 Cowell, with whom I had the privilege of reading for 
 a short time in Sanskrit, and for some fifteen years in 
 Zend, leaves a venerated memory behind for all who 
 received out of his boundless stores. What he knew 
 not was not knowledge, in Aryan subjects certainly, 
 and in many other fields ; but his pupils always had 
 to struggle with the impression that they were there 
 really to impart information to him. The man in 
 the street knows of him as the " onlie begetter " of
 
 x Preface 
 
 Fitzgerald's Omar. Happily there are scholars enough 
 left to preserve in grateful memory more solid titles 
 to the fame of the greatest English Orientalist of his 
 time. 
 
 I should like to have named in closing at least 
 two other great scholars whose friendship, though 
 their work lies in very different fields, has been a 
 powerful stimulus to me. But since their influence 
 on this little book is only indirect, it seems hardly 
 fair to make them apparent contributories. I must 
 be content with merging these debts in the com- 
 prehensive acknowledgement to the genius loci, whose 
 influence is realised most keenly when a long resi- 
 dence in Cambridge is just closed. I would that 
 my parting tribute were worthier of the shrine. 
 
 J. H. M. 
 
 DlDSBURY, 
 
 March, 1903.
 
 I. THE SCIENCE OF LANGUAGE. 
 
 THERE are very few sciences for which the Nine- 
 teenth Century did as much as it did for the Science 
 of Language. It is indeed a question whether there 
 was such a thing as a science of language till the eve 
 of the " Wonderful Century/' unless the stage of rudi- 
 mentary guesswork in which this like other sciences 
 began is to be called "science" by anticipation. In 
 the eighteenth century etymology was defined as a 
 science in which the vowels mattered nothing at all, 
 and the consonants very little. Now, we are no longer 
 allowed to indulge in wild guesses when we seek the 
 history of a familiar word. We have to bind our- 
 selves rigidly within the laws of an exact scientific 
 method, and the science is the more complicated and 
 exacting in that it cannot confine itself to mechanical 
 processes which may be measured and analysed like 
 those of chemical or astronomical phenomena. The 
 Science of Language, as established by the labours of 
 the nineteenth century, combines the methods of the 
 natural and the moral sciences. On one side it deals
 
 2 The Science 
 
 with a purely natural evolution, on the other it studies 
 the workings of the human mind, which crosses the 
 stream of mechanical development and imperiously 
 turns it in directions which only the psychologist can 
 reduce to rule. 
 
 Perhaps I have said enough to suggest that the 
 Science of Language has a peculiar value as an 
 educating force. It may be fairly claimed that it 
 combines all the elements which are most necessary 
 for a really perfect educator. It is a science, and it 
 demands in the highest degree those methods of 
 exactness, of rigid investigation of facts and collection 
 of material, of precise and logical deduction, which we 
 associate with the physical sciences. But at the same 
 time it takes its material very largely from literary 
 sources ; and even where it deals with colloquial idiom 
 or non-literary dialects, the careful analysis of the 
 forms of speech cannot avoid the constant application 
 of principles which form the very basis of literary 
 composition. Our science therefore lies on both sides 
 of the frontier which divides the two great fields of 
 human study, and it is admirably adapted to correct 
 the narrowness which is often seen in those whose 
 training is purely literary or purely scientific. 
 
 The side of our science which presents itself to the 
 ordinary educated person is Etymology. No one can 
 fail to feel interested by a dip into a dictionary, which 
 tells us by what devious and lengthy paths words 
 have come to the meanings and forms they now show. 
 The dictionary of course only gives us results, which
 
 may stimulate us to seek for processes to establish 
 conclusions often paradoxical. When we are told 
 that Easter is akin to the Latin Aurora, and uncouth 
 to ingens, that sooth, (prc)scnt and suttee all come from 
 the participle of the verb " to be " as it shows itself in 
 three cognate languages, with onto(logy) depending on 
 a corresponding form in a fourth, we are easily con- 
 vinced that the ways of words arc peculiar. And when 
 we trace the development of a word like nice back to the 
 Latin nescius, " ignorant," or find in an old poem Christ 
 described as a " silly knave," the words then meaning 
 " holy boy," we can see that the laws by which words 
 change their meaning are complex enough to give a 
 science which examines them plenty of work to do. 
 
 The foundations of the science which changed 
 etymology from mere random guessing into a sound 
 process of reasoning were laid when, mainly through 
 the labours of our great countryman, Sir William 
 Jones, the Western world became possessed of the 
 ancient language of India. That the classical lan- 
 guages of Greece and Rome were very closely 
 connected had been always taken for granted : indeed 
 their nearness to one another was greatly exaggerated. 
 But that they formed only a part of a gigantic system 
 of related languages, spoken by races scattered over 
 the lands lying between India and Iceland, was never 
 dreamt of till the obvious identity between the San- 
 skrit noun and verb systems and those of Greek and 
 Latin was presented to the Western scholar's eye. 
 It was Sir William Jones himself who first drew the
 
 4 The Science 
 
 momentous inference, in words which well deserve 
 quoting : " The Sanskrit language, whatever may be 
 its antiquity, is of wonderful structure ; more perfect 
 than the Greek, more copious than the Latin, and 
 more exquisitely refined than either, yet bearing to 
 both of them a stronger affinity, both in the roots of 
 verbs and in the forms of grammar, than could have 
 been produced by accident ; so strong that no 
 philologer could examine all the three without be- 
 lieving them to have sprung from some common 
 source which, perhaps, no longer exists. There is 
 a similar reason, though not quite so forcible, for sup- 
 posing that both the Gothic and the Celtic, though 
 blended with a different idiom, had the same origin 
 with the Sanskrit." This brilliant discovery, declared 
 in the year 1786, practically lies at the root of all 
 linguistic science. Our science is not, of course, solely 
 concerned with the languages of our own great family 
 of speech, but the principles of the science have been 
 built up exclusively through the study of this family, 
 and no really scientific investigation of alien languages 
 could possibly be carried on without the tools which 
 we ultimately owe to the impulse given by the founder 
 of the Royal Asiatic Society. 
 
 It was early in the nineteenth century when the 
 English scholar's brilliant apcrcu was taken up by the 
 Germans, who developed it into a scientific fact, and 
 have largely kept the study to themselves as a close 
 preserve of German industry and thoroughness up to 
 the present time. PInglish genius has led the world in
 
 of Language 5 
 
 mathematics and physical science ; while our literature 
 for five hundred years has been without a rival among 
 the literatures of Europe. In the study of the ancient 
 classics we at least hold our own ; but we do not 
 seem to care to study the treasures of the English 
 language, which in the hands of German students 
 
 o o ' 
 
 afford material for two periodicals exclusively devoted 
 to them. And in the science of language we have 
 only supplied occasional rivulets to swell the stream 
 of progress ; while our editors of classical texts are 
 still too often content if in their etymological excur- 
 sions they lag no more than twenty years behind the 
 science of the day. 
 
 I must return from this digression, pleading in 
 excuse of it the necessity of accounting in advance 
 for the foreign names which will mark every step of 
 the advance recorded in a brief sketch of a science 
 born and matured within the nineteenth century. We 
 begin then with the year 1816, thirty years after 
 Sir William Jones's far-sighted announcement, when 
 Franz Bopp published the first of a series of works in 
 which he systcmatised the doctrine of the common 
 origin of the languages of our family, and examined 
 the history of their forms. His life-work may be 
 said to have defined for us, practically on lines which 
 we still follow, the limits and constituents of the Indo- 
 germanic or Aryan family of languages. Before I go 
 further, it may therefore be well to give some short 
 description of the field as left by Bopp's labours, with 
 very slight modification from later research. W T e
 
 have eight main languages (apart from a few that are 
 only known by fragments), which descend from a 
 single approximately homogeneous original, long ago 
 lost. Arranged geographically as on the dial of a 
 clock, they will stand thus. (i) Lithuanian, still 
 spoken on the eastern shore of the Baltic ; and 
 Slavonic, embracing Russian and other dialects of the 
 Slav nations. These, like the next two to be named, 
 are shown to be so closely akin that we must reckon 
 them as one branch rather than two. (2) Iranian, the 
 language of Persia ; and Indian, by which we mean 
 Vedic and the classical Sanskrit, with its descendants 
 Hindi, Bengali, and others. The Indian and Iranian 
 branches are combined under the common title Aryan, 
 by which both peoples knew themselves in the earliest 
 times. (3) Armenian ; and (4) Albanian two less 
 important branches, whose original position on the 
 dial is not quite certain. We are now, from the 
 results of recent investigations, able to class these four 
 together as the eastern section of the family ; the four 
 western branches will occupy the left-hand half of our 
 dial. These are (5) Greek, ancient and modern ; (6) 
 Italic, including Latin and certain minor dialects of 
 ancient Italy, together with the Romance languages 
 of to-day, descendants of colloquial Latin ; (7) Keltic^, 
 which preserves a rather precarious vitality in Brittany, 
 Wales and Ireland, and even less than this in Scotland 
 
 1 Italic and Keltic are so closely bound together by important 
 phonetic and morphological affinities that they arc sometimes .spoken 
 of as one branch.
 
 of Language 7 
 
 and Man ; and lastly (8) Germanic, the dominant 
 language of all the lands of Western Europe which 
 are not washed by the Mediterranean. We must 
 allow at least the fourth quarter of our dial to this 
 prolific member of the family, which at the top of the 
 dial touches the first of the eastern branches, Lithu- 
 anian. A name has to be found which will con- 
 veniently represent the whole. German scholars insist 
 on Indo- Germanic, a name combining the extreme 
 east and extreme west of the language area. Far less 
 cumbrous is the name Aryan, popularised in England 
 by Max Miiller, and plausibly supported by the 
 etymology which traces the word in Erin a fact 
 which, if proved, would have gone far to show that 
 the undivided people called themselves Aryans in pre- 
 historic times. But since Aryan is a name undeni- 
 ably appropriated by the ancient Indians and Iranians, 
 it is safer to restrict it to the second of the eight main 
 branches just described, and use for the whole family 
 the title Indogermanic, which, if clumsy, is at any rate 
 free from ambiguity. 
 
 Pursuing our historical order, we come next to the 
 great name of Jacob Grimm. We all become familiar 
 with that name in childhood through the great collec- 
 tion of folklore stories, in which the anthropologist 
 and the small boy are equally at home. Later on, 
 the sound of "Grimm's Law" forces itself on our 
 attention, and the great principle therein laid down 
 may very possibly be to this day the sole possession 
 we hold in the realm of Comparative Philology. The 
 M. 2
 
 8 77/6? Science 
 
 Law was enunciated in 1822, and may fairly be set 
 down to the account of the great scholar whose name 
 it bears, although the idea of it had been announced 
 before. Since it is obviously impossible in this lecture 
 even to sketch in the briefest manner the whole field 
 of Indogermanic philology, I shall probably lay out 
 my time to most advantage if I take up one or two 
 salient points and show their bearing on the principles 
 of the science as a whole Grimm's Law is certainly 
 the best possible point from which to begin, for I may 
 fairly assume it to be generally known, and it is at 
 the same time of immense importance in the history 
 of linguistic study. Its importance is indeed utterly 
 out of proportion to the field which it immediately 
 affects. We who speak English can easily realise the 
 significance of a law which must be considered almost 
 every time when we seek Latin or Greek cognates for 
 words in our own language. a law which in its further 
 development rules the relations between Dutch or 
 English and the literary language of Germain-. But, 
 after all, there are other civilised languages besides 
 German, Dutch, or Norse, and even besides English, 
 and we may find ourselves asking whether Grimm's 
 Law would have quite the same perspective if we 
 were Frenchmen or Russians or Hindoos. Practically, 
 the answer would be yes. Grimm's Law is not merely 
 a convenience whereby we may scientifically equate 
 our word brotJicr with the Latin /rater, the Greek 
 (frpdrjjp, the Sanskrit bhratar, and again the German 
 Brudcr, or deny the identity of call with the Greek
 
 of Language 9 
 
 Ka\.dj, for all their nearness of sound and meaning. It 
 has proved in experience the great educator in the 
 science of language. Its presence has perpetually 
 reminded amateur etymologists and it is astonishing 
 how universally people feel themselves qualified to 
 tackle an etymology, however innocent of special know- 
 ledge they may be that there are laws governing 
 the changes of human speech, which can only be set 
 aside by the presence of other factors known to the 
 expert alone. And even among experts, the wide 
 extent of its operations and the sureness with which 
 it works have done more than anything else, perhaps, 
 to evolve the conviction that phonetic changes are 
 exempt from mere caprice, and so to place our science 
 upon the firm basis which it occupies to-day. 
 
 For the present I propose to develop the history 
 of scientific method in terms of Grimm's Law, aban- 
 doning the strictly chronological order with which we 
 began. How did this far-reaching change originate, 
 and by what steps did it arrive at its present wonderful 
 uniformity? In those fascinating Lectures on tJic 
 Science of Language, by which the late Professor Max 
 Muller did so much to popularise linguistic study in 
 our country, an account is given which raises all at 
 once the question of the nature of phonetic change. 
 Practically it comes to this. The Germans found 
 themselves no longer able to pronounce the difficult 
 sounds b/i, dJi and gh which they had inherited from 
 their Indogermanic forefathers, and (like several other 
 members of the family) came to say /;, d and g instead.
 
 io The Science 
 
 But this involved confusion with words which had 
 a b, d or g already. Therefore, with a conscientious- 
 ness lacking in those other Indogermans, who did not 
 mind the confusion, they replaced b, d and g by p, t 
 and k. This, however, was thoughtless of them, for 
 these sounds likewise were appropriated. Having 
 committed themselves too far to go back, they had to 
 bring in a new set of sounds,/", th and /i, which accord- 
 ingly took up the old /, t and k, and the " sound- 
 shifting " was complete. We should have to postulate 
 a somewhat similar process when, about a thousand 
 years after the first sound-shifting, the High Germans 
 started a second, by which the existing Germanic 
 b,p and /"were shifted on further to p, j^f and b, with 
 similar changes for the dentals and gutturals. You 
 will probably anticipate the fatal objection against 
 any such explanation. It postulates a conscious 
 change, simultaneously adopted by a whole people, 
 and the briefest reflexion will show that such things 
 do not and cannot happen. Phonetic changes are not 
 determined by committees. Speech is unconscious, 
 except when we are trying to conform our pronuncia- 
 tion to that of our neighbours. The realisation of this 
 point will prepare us for the study of the latest phase 
 of enquiry upon which our science has entered. I 
 cannot enter now on the solution of the interesting 
 question as to the causes from which the " sound- 
 shiftings " arose. Suffice it if I observe that no expla- 
 nation will suit the phenomena of language which 
 does not recognise the unconscious and independent
 
 of Language 1 1 
 
 character of the changes. Some may have changed by 
 the imperfect efforts of natives to catch the pronuncia- 
 tion of foreigners. In others, an imperceptible variation 
 beginning on one kind of sound alone, and presumably 
 only under definite conditions such as combination 
 with other sounds, or appearance at the beginning or 
 the end of a word gradually spread till the change 
 was complete for that set of sounds. Meanwhile 
 another set would independently begin to suffer 
 change, till after a few generations the process was 
 accomplished without any of the speakers of the lan- 
 guage knowing how far they had come. 
 
 I have not yet done with Grimm's Law, but for 
 a few minutes I must relapse into history to show the 
 background on which modern science is set forth. 
 For this purpose we must pass over nearly half a 
 century of laborious collection and ordering of facts, 
 and neglect entirely the work of some of the greatest 
 masters in Indogermanic philology. I pause on the 
 names of Max Miiller and Curtius, the more readily 
 as they seem to represent the latest stage of science 
 as conceived by many English scholars when obliged 
 to venture on the unfamiliar ground of comparative 
 philology. Both were champions of law and order in 
 the realm of language, but their systems of law 
 allowed room for a carnival, in which ordinary princi- 
 ples were suspended. Max Miiller devoted himself to 
 the special study of comparative mythology, and 
 collected a large number of fairly similar names in 
 Sanskrit and Greek, which he paraded as historically
 
 12 The Science 
 
 connected words. A complex fabric of primitive Indo- 
 germanic mythology was thus constructed, mostly 
 centring on manifestations of the Dawn-goddess. 
 Alas ! this pretty theory has long since vanished into 
 air, into thin air, for hardly one of the innumerable 
 equations will hold when examined by more rigid 
 methods. Curtius attempted to hold the carnival 
 under a semblance of restraint. He laid it down that 
 \vhen sounds began to become difficult, in the speech 
 of any nation, they passed generally into some 
 "regular" representative sound, but also "sporadically" 
 into others. Thus the w sound in Greek, which we 
 call digamma, " regularly " disappeared in the course 
 of development, but " sporadically " metamorphosed 
 itself into b, g, h, m, ph, r, o and u. Most of these 
 were supported only by two or three examples, which 
 were regarded as proof specimens on very arbitrary 
 grounds. It was clear that although the realm of 
 chance and caprice in language had been very greatly 
 narrowed, there was still much to be done before any- 
 thing like an exact science could emerge. 
 
 Between 1870 and 1880 certain brilliant discoveries 
 were made which at one stroke reduced to order a 
 large proportion of the irregularities left by earlier 
 investigators into Indogermanic phonetics. The effect 
 of such discoveries is rather like that of the discovery 
 of Neptune in the realm of astronomy : serious irregu- 
 larities traced down to some hitherto unsuspected new 
 factor, the presence of which makes everything orderly, 
 form a most impressive argument for the universal
 
 of Language 13 
 
 reign of law. It is fitting that among these discoveries 
 should stand out an explanation of irregularities in 
 the working of Grimm's Law. Verner's Law, as the 
 new principle is called from its discoverer, deals with 
 cases in which original /, t and /', instead of passing 
 into f, tli and // as Grimm's Law demands, become 
 b, d and g. Verner showed that this depended on the 
 position of the accent, and that the accent thus evi- 
 denced for the primitive Germanic was identical with 
 that still preserved in Vedic Sanskrit and to some 
 extent in Greek. This discovery gave a great impulse 
 to the growing sense of regularity in language. At 
 the same time it went far beyond Grimm's Law in 
 the light it threw on the conditions of primeval Indo- 
 germanic speech, for the coincidence of the accent in 
 the two most widely-severed branches of the family 
 proved what the accent was in the original language 
 from which both were descended. Verner's Law was 
 accompanied b}* other discoveries which entirely trans- 
 formed our conceptions of this original Indogermanic 
 language. Schleicher, the great pioneer whose work 
 marks the first decided advance from the standpoint 
 of Bopp and Grimm, reconstructed the parent lan- 
 guage as an exceedingly simple organism, with only 
 three vowels, a, i and //, and consonants cut down 
 to a small figure. The discoveries of later years, in 
 which the name of Karl Brugmann holds the place of 
 honour, turn this reconstruction into something far 
 more complex. The simple vowels are extended to 
 include c and o and others ; by the side of / and //
 
 14 The Science 
 
 stand r, /, m and n as sharing their power of becoming 
 vowels or consonants at will ; the gutturals are turned 
 into three series instead of one, and on their behaviour 
 depends the allocation of any given dialect to the 
 Eastern or the Western side of the Indogermanic 
 family. In addition to this there is revealed a com- 
 plicated system of stress and pitch accents. The 
 result is that if anyone learnt to speak the Indo- 
 germanic language as it stands to-day, he could not 
 possibly make himself understood by one who had 
 similarly learnt the language according to Schleicher. 
 Five for Schleicher was kankan, for us pt'wq^e 1 
 (Western) or penqc (Eastern) : horse for him was 
 akwas, for us echivos 2 , and so on. The difference may 
 be a useful warning if we are in any clanger of regarding 
 our scientific reconstructions of the parent language as 
 the definite discovery of a dialect which was spoken at 
 one particular time in the dim and distant past. I shall 
 have to deal with this caution in the next lecture, in 
 which I shall try to show what help Language can give 
 us in unveiling the life and civilisation of those primeval 
 men from whom we are partially descended, and to 
 
 1 The symbol TJ represents our tig, the guttural nasal : the parasite 
 w (it is the consonantal u) is closely attached to the consonant pcng-que 
 (not pcnk-ive] would repre.-ent the pronunciation. 
 
 - The li palatal k " (k in Brugmann's notation) was probably pro- 
 nounced like the Scotch ch in loch. This will account for the Eastern 
 sh or s, and the Western k alike: the former change is paralleled by 
 the South German pronunciation of nicht as nish(, the latter by the 
 "Lock Lomond" which the steamer officials endure from so many 
 Southron lips in the tourist season.
 
 of Language 15 
 
 whom we owe our speech. Meanwhile it will be 
 enough to remind you that the forms which appear 
 in scientific books, as due to the parent Indogermanic 
 language, are only convenient formulae to show what 
 we have learnt of the history of words extant in 
 ancient or modern languages of our group. They 
 may represent words actually spoken by prehistoric 
 men, perhaps however at intervals of some centuries 
 from one another. Or they may be as far from the 
 words actually spoken as were Schleicher's recon- 
 structions from those in vogue to-day ; for it can 
 hardly be that this science will stand still in genera- 
 tions of research yet to come. 
 
 I pass on, then, to the brief enunciation and illus- 
 tration of the principles of our science as we under- 
 stand them now. What I have been saying will 
 prepare you for the latest development of our theory. 
 Exceptions to phonetic laws have been reduced 
 enormously by the successive establishment of new 
 laws covering every part of Inclogermanic speech, and 
 the natural result is that scholars have been drawn to 
 go a step further and declare that phonetic laws, as 
 such, admit of no exception. Since this bold declara- 
 tion was first made by Leskien, in 18/6, it has been 
 furiously debated, and it may perhaps be questioned 
 whether on grounds of theory alone it has been 
 conclusively established. But some of the most 
 contemptuous critics of the " Xeo-grammarians," as 
 they are called, have led the way in discovering new 
 phonetic laws, and therefore in reducing further the
 
 1 6 The Science 
 
 number of words which would have to be classed as 
 " irregular." Whether therefore we are or are not 
 prepared to assert as a matter of theory that there 
 cannot be irregularities in language caused by the 
 capricious action of phonetic law, we may certainly 
 use as a working principle the doctrine of the fixity of 
 law in human speech. For example, if anyone should 
 tell us that of course the Latin and Greek words 
 for God, dcus and 6eos, must have a common origin, 
 because they sound so nearly alike and have the same 
 meaning, we promptly deny the identity, because 
 phonetic laws stand in the way. A Greek tJi cannot 
 answer to a Latin d at the beginning of a word, and 
 we prove our point by citing a number of words in 
 which Greek tJi and Latin initial/ stand in clear rela- 
 tion to one another, while we challenge the objector to 
 produce any other example in which a Latin d has 
 ousted the regular f. He declares that this is an 
 exception, and denies our right to assert that such 
 exceptions are inadmissible. It is quite unnecessary 
 for us to fall back on a general theory that Language 
 knows nothing of exceptions, acting always with the 
 precision of a law of nature. It is enough to say 
 that our opponent is bound to show cause why the 
 originals of our deism and theism must necessarily be 
 the same. Meaning and similarity of sound count 
 for nothing, for coincidences of the kind can be pro- 
 duced by the score. The Hebrew kapJiar means cover, 
 but no one out of Bedlam thinks the identity a suffi- 
 cient proof of our descent from the Lost Ten Tribes !
 
 of Language 1 7 
 
 I proceed to show in what sense the dictum of 
 the invariability of phonetic law is to be understood. 
 I find it convenient to state the principle in the 
 following form, to which I ask your special attention 
 as covering the whole field in short compass. 
 
 " The same original sound cannot, in the same 
 period of the same dialect, pass under the same con- 
 ditions into two different sounds." 
 
 You will observe the four sanies of this state- 
 ment. Eliminate any one of them and you get what 
 seems an irregularity. It will I think be helpful to 
 give one or two illustrations under each of these 
 heads. 
 
 First, then, " tlie same original sound" Examples 
 under this head may be supplied from all the novelties 
 in linguistic discovery which I have been trying to 
 sketch this morning. Take the English words guest 
 (German Cast} and warm. The initials g and w are 
 found alike in Sanskrit as gh, and it was formerly 
 assumed that in such cases the Indogermanic sound 
 was gh, which was differentiated later in Western 
 languages into two sounds, the latter containing a w. 
 
 O O O 
 
 We now know that the guttural in question was 
 not the same in the two cases, the latter belonging 
 to a series in which the ii> was strongly developed 
 throughout the Western languages, and dropped in 
 the Eastern. An illustration of a different kind may 
 be seen in the English -ougli words, which supply so 
 powerful an argument to the advocates of phonetic 
 spelling, and so strong an irritant to the foreigner
 
 1 8 The Science 
 
 trying to learn our language. In plough and bough 
 alike the ough goes back to an Old English oh, while 
 dough is from dak, and Lough is a Keltic word, coming 
 under our third head below. Speaking generally, we 
 may say that modern science tends to seek different 
 originals when a wide-spread irregularity appears in 
 the representation of what has been taken to be the 
 same sound. To simplify the parent language, by 
 reducing the number of sounds in it, is no longer 
 felt to be obligatory. A very early language may 
 be very complex in its sounds, and progress is at 
 least as likely to weed out unnecessary sounds as to 
 invent new ones. 
 
 We pass to our second heading, "in the same 
 period'.' Phonetic laws must not be supposed to be 
 permanent : they came into being slowly and un- 
 consciously, and slowly and unconsciously they die 
 away. It is always therefore vital that we should 
 know of a given law at what period, as well as in 
 what area, it worked. Grimm's Law will supply ex- 
 cellent examples. What we call the ''first sound- 
 shifting" ceased to act in the Germanic languages 
 some centuries before Christ. It gave us words like 
 father, where the /was undeniably sounded by our rude 
 ancestors who roamed over Northern Europe before 
 the time when Caesar was invading Britain. It had 
 ceased to act long before the Saxons followed him 
 thither. Consequently, when the Saxons borrowed 
 Latin words like strata they had no instinct leading 
 them to change the t sound, and it has remained in
 
 of Language 19 
 
 the English word street for a thousand years. It 
 would not have so remained had the Saxons been 
 infected with the new tendency which in upper 
 Germany was beginning to shift all these sounds 
 afresh. The High Germans made strata into Strasse, 
 pondus into Pfund, etc.; but this tendency in its turn 
 died away, and when Modern German borrows a 
 Latin word \\ViQ praedico, German prcdigc, it leaves the 
 / alone, just as English does \n preach. 
 
 Thirdly, "of the same dialect" The colossal irregu- 
 larities of English are very largely due to dialect 
 mixture. A thousand years ago English included 
 many dialects, all with equal possibilities in the 
 struggle for survival as the ultimate literary language 
 of our country. The dialect which was spoken most 
 typically not far west of Cambridgeshire 1 finally 
 won the day; but words and forms from other dia- 
 lects became imbedded in the standard language, 
 so that to-day it presents a bewildering medley of 
 inconsistencies. To a less extent, but still con- 
 siderably, Latin similarly absorbed dialectic peculi- 
 arities. One example I may give, as a very pretty 
 specimen of the way in which irregularities are cleared 
 off in modern research. There are a number of Latin 
 words in which an original ^becomes /. Lingua is for 
 dingua, as the English tongue will show when Grimm's 
 Law has been applied. Olcrc, "to smell," is clearly 
 connected with odor, " odour." Larix (our larch} is 
 
 1 See Professor Skoal's interesting pamphlet on ihe place-names of 
 Cambridgeshire (Cambridge Antiquarian Society).
 
 2O The Science 
 
 akin to the English word tree. Proceeding on the 
 assumption that dialect mixture was the probable 
 cause of this irregularity, a distinguished philologist, 
 Professor Con way, began to look for an Italian dialect 
 in which all initial </s, or ds between vowels, became /. 
 The remains of the dialect on which he fixed are 
 extremely scanty. But it happens that Horace speaks 
 of a small river called Digoitia which flowed past his 
 country-house. Its modern name is Licenza, and 
 Dr Conway showed that its name would be pro- 
 nounced with the / by the Sabines, through whose 
 district it flowed, but that when it emerged into Latin 
 territory the older d would be heard. Modern Italian 
 has here preserved the Sabine form, as Latin did in 
 such words as I mentioned just now. It is therefore 
 no longer necessary to say that "d remains d in Latin, 
 but the rule is broken in a few words where / ap- 
 pears." On the contrary, d always remains d under 
 these circumstances, and the / of lingua and oleo is 
 simply due to borrowing from the Sabine dialect. 
 
 Finally, " under tJie same conditions!' A great 
 feature of modern research has been the emphasis 
 laid on the extent to which we change the pronunci- 
 ation of our words in different surroundings. We 
 say right-e-ous, pcr-haps, sup-pose, when we talk de- 
 liberately, richus, p'raps, spose, when we are in a 
 hurry. The / in right remains / unless there is a kind 
 of y sound following it, as in the righteous of the 
 educated man in a hurry, or the right-you-are (ri-chu- 
 arc) of the more slangy individual. An immense
 
 of Language 21 
 
 variety of differences are produced by the shifting of 
 accent. The word accent itself, when a noun, would 
 be correctly written dksnt, the second syllable being 
 reduced to a mere vocalic ;/ by the stress on the first 
 syllable; when a verb, accent, it has the full en. In 
 general, we can never say positively that "the same 
 original sound at the same period of the same dialect" 
 will produce the same resultant sound in two different 
 words, until we have examined into the effects of 
 accentual conditions, neighbouring sounds, rapidity 
 of pronunciation, and any other possible differences of 
 condition which may affect the ultimate form of the 
 words in question. 
 
 So much then for the changes in Language which 
 are due exclusively to phonetic development. A few 
 words should be added to describe the manner in 
 which all these changes arise. It must be remembered 
 that speech is transmitted entirely by the reproduc- 
 tion of sounds and words we have heard from others. 
 An English infant, placed from the first in the care 
 of Russians, Persians, or Zulus, would speak their 
 language, and would not have the slightest inherited 
 predilection in favour of English 1 . Speech is a joint 
 function of the ears and the vocal organs, and both 
 may fail to catch the sound correctly. So long as the 
 variation is not serious enough to be noticed by those 
 from whom he learns, the child will go on pronouncing
 
 22 The Science 
 
 in his own way; and by slow changes, accumulating 
 from generation to generation, the dialect will pro- 
 gressively alter. But this is assuming that the com- 
 munity speaking the dialect is compact enough for its 
 individuals to be in constant communication with each 
 other. It is communication which preserves speech 
 from change; and within such compact communities 
 the rate at which pronunciation changes will depend 
 entirely on their sensitiveness to variety in sound. 
 If they are acute of ear, they will soon notice and 
 check the variations introduced by the children who 
 are learning to speak their tongue ; if not, they will 
 pass over the children's mispronunciations, and the 
 language will change rapidly. Suppose however that 
 the community is not compact, that people living 
 at its extreme ends never meet. The result will be 
 a gradual shading off of dialect from one end to the 
 other; and if the area is sufficiently large, it may well 
 be that the extremes are mutually unintelligible, 
 though neighbours all along the line can understand 
 each other easily. Now suppose that an agricultural 
 people, scattered evenly over a large area, gradually 
 change their habits and concentrate in towns. Clearly 
 the result will be that each town will have a dialect 
 of its own, to an extent depending on the amount of 
 intercommunication with its neighbours ; and when 
 these communities become separated by migration 
 beyond seas or rivers or mountains, barriers preventing 
 intercourse, each dialect, pursuing its own development, 
 will draw away from those which were once all but
 
 of Language 23 
 
 identical with it, till at last the limit of intelligibility 
 is passed and a new language is established. Finally, 
 there is the disturbing force of foreign languages to be 
 reckoned with. A small body of warriors invades and 
 conquers a large but weaker population. It generally 
 follows that in a few generations the conquerors have 
 been absorbed and speak the language of the con- 
 quered; but the invaders, learning the new language 
 late in life, bring their own pronunciation, and many of 
 their own words. I have no time to enlarge upon this 
 subject of speech mixture, but if you follow up the 
 science you will soon realise how great are its possi- 
 bilities of influence upon the development of language. 
 Time forbids any adequate attempt to describe 
 the immense province of speech which separates our 
 science from the physical sciences and joins it to 
 those in which the human mind is the object of 
 study. On the purely phonetic side, as we have seen, 
 language develops with almost machine-like regu- 
 larity. But we perpetually find that phonetic laws 
 are being crossed by forces which are almost as 
 conscious and deliberate as the act of coining a new 
 word, or applying an old one to a new meaning. 
 These are the forces of Analogy, which assimilates 
 the forms of words that are brought into frequent 
 association with each other. The principle used to be 
 known as " False Analogy," but there is really no reason 
 thus to stigmatise a highly respectable and influential 
 factor in the development of speech. We are under 
 no sort of obligation to maintain the inherited forms
 
 24 The Science 
 
 of language when we can save our memories by 
 bringing a minority of words into conformity with 
 a majority. If we choose to say sorry instead of sory 
 (from sore}, we have introduced a " false" element, in 
 that sorry and sorrow are not really akin. But since 
 the words have come to be associated in meaning, it 
 is more convenient that they should be brought near 
 in form : the dictionary will keep us right on the 
 etymology. Or, to take an example from the largest 
 field of analogy's operations, when our ancestors gave 
 up saying rangJit as the past tense of reach, was it 
 a "false"' analogy which made them realise that a 
 past tense normally differs from a present by adding 
 </? They discarded forms which once had a meaning 
 in favour of forms which have a clear meaning still, 
 and we can only regret the caprice which failed to 
 insist on tcachcd and btscccjicd as well, or, if that could 
 not be, at least to make us keep the old forms and 
 bring in prangJit to match them! 
 
 It is usual to classify the operations of analogy 
 under the heads formal and logical. Formal ana- 
 logy assimilates the declension or conjugation of 
 words originally belonging to different categories. 
 Material, or logical, analog}- clears away variations no 
 longer intelligible in the forms belonging to one 
 word, after which formal analogy often steps in and 
 uses the convenient innovation for the benefit of other 
 words. A few examples will sufficiently illustrate 
 the extent of these operations. Old English inherited 
 from the Indogermanic period a system of vowel-
 
 of Language 25 
 
 gradation, originally the automatic result of accent. 
 This dominated the forms of the past tense very 
 largely, as it did those of the perfect in Greek. The 
 first and third persons singular had the forms sang, 
 wrote, sat. the second person singular and the whole 
 plural showed sung, writ, sit. As late as Chaucer we 
 still find / wot, we witen, in which same word Greek 
 likewise presented the identical vowel distinction (olSa, 
 i<rp.v). Obviously the distinction served no useful 
 purpose, and logical analogy cleared it away : later 
 English said we wot, and Hellenistic Greek said 
 oiBapev. In French the regular forms evolved from 
 the vulgar Latin parabola, plural paraboldnius, were 
 (je) parole, (nous) parlous, those from dmo, aindmns, 
 were ainic, anwns : modern French has levelled these 
 tenses, and only a few survivals like j'ai, nous avons, 
 je viens, nous venous, intrude themselves upon the un- 
 willing English schoolboy. Old English, like modern 
 German, insisted on modifying the vowel of a word 
 like long when the comparative suffix er was added, 
 a necessity which no longer ^ lenger" they would say) 
 appeals to us. It was the same principle which made 
 feet the plural of foot, and what would have now become 
 bcek the plural of book. From this last grammatical 
 abortion formal analog)' delivered us, by calling in 
 the aid of the multitude of nouns which made their 
 plural by the simple method of adding s or es to the 
 singular. The same beneficent process destroyed a 
 great number of highly interesting but highly incon- 
 venient "strong perfects," in favour of the simple past
 
 26 The Science 
 
 in d. In some cases it worked in the opposite direc- 
 tion : common types like tore from tear and bore from 
 bear naturally gave rise to wore for weared, and I 
 have heard scrape and brung produced on the same 
 principle by speakers more logical than educated. 
 
 The examples of Analogy which I have been 
 describing are cases affecting a whole grammatical 
 category. We may feel fairly certain that most of 
 the new tense-formations or noun-cases, etc., which 
 characterise the various dialects of the Indogermanic 
 or other families of speech, owed their birth to this 
 fertile principle. A single example, if it produces 
 something convenient and " meets a felt want," as the 
 advertisers say, is often quite enough to create a new 
 class of forms. Take for example the suffix -isc, 
 which in English, as in the Greek whence it was 
 borrowed, can be added so freely to other words to 
 make a nc\v verb. It started in Greek mainly from 
 an extremely small class of nouns with a stem in -18-, 
 from which a verb in -t'o> came by regular rule. But 
 the form was so convenient that it spread at an ex- 
 traordinary rate in Greek, was borrowed by Latin. 
 passed on to I-'rench and finally to English, so that 
 if a Mr MacAdam invented a new way of paving 
 roads the language was ready to coin " macadamise," 
 just as Greek could sum up in the opprobrious verb 
 Mcdisc the traitorous action of Greeks who helped the 
 Medes in their attempt to destroy Greek liberty. 
 
 A very large class of analogy forms has nothing to 
 do with grammatical categories, but consists of types
 
 of Language 27 
 
 which have affected a few words associated with them 
 by likeness or contrast of meaning. In all languages 
 such groups as numerals, colours, parts of the body, 
 trees, points of the compass, etc., have influenced one 
 another, especially in their endings. Norther loses 
 its suffix to match south, but northern produces 
 southern. The initial f of four is plausibly explained 
 by association \\\\h. five. Such associations will often 
 start new suffixes, appropriated by use to a particular 
 class of words. In Latin, presumably from one or two 
 examples, a whole series of tree-names has arisen with 
 a suffix -snus (as al(s}nus, pi(ts}nus). Names for 
 groves of trees usually end in -etuin. In Latin and in 
 Germanic the suffix -wos has become associated with 
 colour words, simply because it happened to be 
 common to two original colour adjectives, blue and 
 yellow, Latin fldvos and helvos, whence it spread. 
 Pairs like health and weal, male and femel, hither and 
 tJiatlier, cither and notJier, assimilated one another and 
 produced forms which could only reduce phonetics to 
 chaos if the ever-present influence of analogy were 
 not recognised. It is hardly going too far to say that 
 whenever a single word shows serious difficulties in its 
 formation, the first instinct of a modern philologist is 
 to search for some twin word which could work upon 
 it by analogy. Now in all this the question will 
 natural!}' occur to you, " How did Language choose 
 between altering A to match B and altering B to 
 
 o o 
 
 match A ? Granted that we should not be likely to 
 say / sat but we sit, I sang but we sung-, how is it that
 
 28 The Science 
 
 we did not come to say in the past tense / sit, I sungt 
 Why not mell and femcl, other and notiier, heal and 
 luealT' To this question \ve can hardly give a general 
 answer, any more than we can to the allied question how 
 taught a.\\d. feet resisted the processes which destroyed 
 rang/it and beck, why drownded is vulgar, but sounded 
 correct, and so on. In many cases we can see a 
 reason. The form that proved the survivor shows 
 itself to have been in one way or another the fittest to 
 survive. It occurred in the larger number of forms, 
 or in those which were most in use it avoided 
 an ambiguity it was easier to pronounce all 
 manner of such forces turned the scale in individual 
 instances. Sometimes the scale is only turned after 
 oscillations to and fro. Milton says / sung in the 
 seventeenth century, and Charles Wesley he begun in 
 the eighteenth ; and our existing sang and began may 
 have to yield their place, for all we know, before this 
 century is out. The systematisation of these mental 
 processes, by which Language has so much simplified 
 itself, will, I believe, afford plenty of work to the 
 student of Indogcrmanic philology for years to come, 
 the sphere of purely phonetic development leaving 
 us now, it would seem, little beyond details to work 
 upon. 
 
 This morning's sketch and I need not point out 
 to you how absolutely sketchy a single lecture upon 
 so vast a subject must be has indicated some of the 
 lines on which the modern languages have developed 
 throughout their long history. We have seen in
 
 of Language 29 
 
 closing how they have been enabled to rid themselves 
 of accumulated lumber, and evolve the combination 
 of simplicity in form with wealth in resource which 
 can fit them for the complex needs of modern life. 
 Obviously the language which has most successfully 
 shaken off what is useless, while keeping all that 
 makes it expressive and concise, is likely to outrun 
 its rivals in the struggle to become the leading lan- 
 guage of the world. Heaven forbid that, with guests 
 present from so many other nations than our own, I 
 should attempt for one moment to argue which that 
 victorious language is likely to be ! 
 
 The principles which have this morning led us to 
 the latest developments of speech will lead us also 
 into its first beginnings. It will be my duty to- 
 morrow, not indeed to venture on the thorny but 
 fascinating problem of the Origin of Language, but to 
 enquire what Language can teach us of our ancestors' 
 lives in the distant past which, though recent when 
 compared with man's earliest appearance on this 
 planet, lies far behind all literature and all history. 
 The investigation, even though it fail to open wide 
 the door and reveal to us in clear and brilliant light 
 the long vista of the past, will at least tell us some- 
 thing more of what Language is and what Language 
 can do.
 
 II. LANGUAGE AND PRIMITIVE 
 HISTORY. 
 
 " Linguistic Palaeontology," as the method of re- 
 search which I am to describe this morning is usually 
 called, is one among many tools which we may use to 
 excavate the prehistoric past. Archaeology studies 
 its material relics. Geology offers to tell us under 
 certain conditions the dates to which they belong. 
 Botany and Zoology come in occasionally to pro- 
 nounce upon arguments drawn from trees or animals 
 which are brought into association with primitive 
 man. Craniology measures the skulls of those who 
 were considerate enough to leave them behind, and 
 Ethnology pursues other methods of classifying their 
 racial characteristics. Anthropology and Folklore 
 investigate primitive man by studying his equally 
 primitive representatives among the savages of to- 
 day, and by following out through modern survivals 
 the history of customs and institutions, superstitions 
 and magic. The Science of Language, as we have 
 seen, can do something towards reconstructing the
 
 Language and Primitive History 31 
 
 speech and vocabulary of the parent Indogerman 
 peoples, who lived ages before the dawn of history, 
 and it is clear that this reconstructed vocabulary, if 
 properly used, can tell us many facts about the life of 
 these interesting forbears of ours. The methods I 
 have described are of course allies, to be regarded as 
 necessary to one another's efficiency, and on no account 
 to be used exclusively or with exaggerated belief in 
 their powers when standing alone. 
 
 There are scholars who seem to regard Language 
 as almost entirely useless for this purpose. They 
 press the weaknesses of Language as evidence till 
 they persuade themselves that it is sheer waste of 
 time to study Linguistic Palaeontology at all. As 
 usual, the truth would seem to lie between two ex- 
 tremes. To reject the mass of evidence, the nature 
 of which I am to describe, is surely scepticism gone 
 mad. To imagine Language capable of proving what 
 we ask of her, without help from sister sciences, is 
 an equally foolish presumption. The truly scientific 
 method is to examine most carefully the cautions 
 with which the argument from language must be 
 applied, and then to test every conclusion by evi- 
 dence derived from other sources. Proceeding in 
 this way we can hardly fail to get some trustworthy 
 results. 
 
 Let us then note firstly the cautions to be observed 
 in making our vocabulary of primitive speech, and 
 secondly those which come in when we seek to use 
 what we have made. Under the first head we note
 
 32 Language 
 
 that the presence of a word in two or more branches 
 of the Indogerman family is not conclusive for its 
 right to a place in the primitive vocabulary. It is 
 impossible to distinguish decisively between a case of 
 inheritance and a case of early borrowing in " culture 
 words," for the name naturally spreads with the thing 
 when a new addition is made to the possessions of a 
 people. For instance, the \vord jis/i, which is common 
 to the Italians, Kelts, and Germans, but not extant else- 
 where, may have arisen in one of these contiguous 
 peoples and spread to the other two : Language 
 perhaps does not deny, but it assuredly does not 
 encourage the idea that a prehistoric Izaak Walton 
 taught his art to the undivided Indogermans. We 
 can only feel confident when a word is found in 
 branches widely separated in geographical position, 
 and of course the more there are of them the safer 
 we feel. On the other hand, the absence of a word 
 from many branches is not conclusive evidence 
 against it, for the loss of old words is a perpetual 
 phenomenon in all languages. The primitive words 
 so)i l and daughter have disappeared in Latin, father 
 and mother in Gothic, and sister has so far disappeared 
 from Greek that only an old lexicon gives evidence 
 for its former existence. We can ourselves watch the 
 superseding of hound by the foreign importation 
 " dog," and many similar cases may be observed.
 
 and Primitive History 33 
 
 Moreover the absence of a common name does not 
 prove the absence of the tiling. The Indogermans 
 had no common word for "one," though they had 
 numerals from two up to hundred, no word for 
 "hand," though foot is found everywhere. We can 
 hardly infer that their physical and mental equip- 
 ment was so deficient as the argument from silence 
 would suggest in these cases. Besides this we have 
 to remember that care is needed before we set down 
 a word as absent from any particular language. 
 Gothic comes down to us only in the Biblical version 
 of Wulfila, the important Umbrian dialect of Italy 
 only in some scanty ritual tablets. Clearly we can 
 only argue absence in such cases when a missing old 
 word is very frequently replaced by another, as we 
 saw just now happened to father and mother in 
 Gothic. Even in Greek, with its enormous extant 
 literature, we find the dictionary enlarged whenever 
 a new work is unearthed, or a new batch of inscrip- 
 tions or papyri. 
 
 Then in using our vocabulary we must take note 
 of the warnings received from " Semantics," the study 
 of the meanings of words. The Latin cognate of 
 beech agrees with the English, but the Greek means 
 " oak " ; the Greek answering to tree also means 
 "oak," while in some languages it means "fir" or 
 " pine." Nor must we be hasty in drawing con- 
 clusions from the existence of words with ascertained 
 meanings and indubitable authority. Early philo- 
 logists drew very rosy pictures of home life in the
 
 34 Language 
 
 primeval age on the strength of their etymologies 
 father meant "protector" and dangJitcr "milkmaid." 
 In these days we prefer not to dogmatise about the 
 etymology of words which come down to us entire 
 from the earliest period. Then, to take another kind 
 of example, the cognates of the Latin cquos, which 
 are found in almost every main language, and always 
 with the meaning " horse," have been regarded as 
 proof that the Indogermans tamed and used the horse 
 for agriculture or war. The least thought will show 
 
 o o 
 
 that the mere word proves nothing but their familiarity 
 with it and its being sufficiently important to them 
 to be \vorth naming. A prehistoric cowboy on horse- 
 back cannot be deduced from language alone, and for 
 all the linguist can tell to the contrary, the interest of 
 the Indogerman in the quadruped may have been 
 purely culinary. 
 
 In the former lecture I gave some account of the 
 main branches of the Indogermanic family, as ar- 
 ranged according to language. It is remarkable that 
 from the earliest dawn of history the six principal 
 languages belong to races arranged relatively as we 
 see them now. Indian and Iranian, Slavonic and 
 Lithuanian, the two main branches of the east 
 Indogermanic languages, are found still in their 
 relative position, with Germanic, Keltic, Italic, and 
 Greek following in this order down the western side 
 of the map. If we squeeze them all together, leaving 
 the Lithuanians and Germans near the shores of the 
 Baltic, and the rest on lines running to the south-east
 
 and Primitive Plistory 35 
 
 and south-west or south respectively, only shorter 
 than they were before, we shall obtain a very pro- 
 bable position for the races speaking those languages 
 at the dawn of history. 
 
 At this point the researches of archaeologists and 
 historians come in to show us that the great north 
 European race, associated especially with the Keltic 
 and Germanic languages, in the prehistoric period 
 sent out successive swarms of sturdy invaders who 
 established themselves as a conquering caste in 
 various southern lands. Their tall stature, yellow 
 or sandy hair and blue eyes contrasted strongly 
 with the features of the short, dark-haired and 
 swarthy races which inhabited the shores of the 
 Mediterranean. In his brilliant book on The Early 
 Age of Greece Professor Ridgeway has made it highly 
 probable that Homer's Achaeans belonged to this 
 conquering northern race, before whom the indige- 
 nous " Pelasgians," represented most faithfully by the 
 Athenians and lonians generally, were forced to 
 bow. Traces of northern origin remain in Greek 
 traditions of places of ceaseless sunshine, and places of 
 perpetual darkness where the air was full of wool of 
 floating rocks (icebergs ?) that crashed together over 
 the sea of the quest of the "golden-horned hind," 
 which can only be the reindeer. Baltic amber has 
 been found in Greek tombs. The study of manners 
 and customs and beliefs tells the same tale. It is 
 remarkable also that some conspicuous heroes of 
 the Iliad and Odyssey have names which cannot be
 
 36 Language 
 
 accounted for as Greek, whereas native Greek names 
 are almost always transparent in their etymology. 
 It is interesting to add to Professor Ridgeway's 
 point an unconscious confirmation from the great 
 philologist August Pick, who found a congener 
 for Achilles in the old German name Agilnlfs. 
 Dr Ridgeway believes that the same origin can be 
 asserted for the sturdy Sabine race, who formed the 
 " patrician " nucleus of the early population of Rome, 
 and whose kindred in later ages made so brave a 
 struggle for independence against the growing power 
 of the great city. We can probably trace the 
 same strain further east. Cyrus and his victorious 
 Persians present many of the characteristics which 
 we note in the patricians of Rome and the Achacans 
 of Greece, and in all these we have the same sub- 
 sequent history: a comparatively small host of 
 warriors, greatly superior in physique and equipment, 
 easily conquers a weaker population, and in a few gene- 
 rations loses most of its distinctive features by inter- 
 marriage with the indigenous race. There are traces 
 still remaining of these kinsfolk of our own who in dis- 
 tant ages overran the rich lands of southern Europe 
 and western Asia, inhabited by races physical!}- weaker 
 but more intellectual than themselves. In India the 
 rigid preservation of caste barriers enables observers 
 still to note the decidedly northern physiognomy of 
 some representatives of ancient royal houses. And 
 among the Ossetes, an Iranian-speaking folk in the 
 centre of the Caucasus, modern travellers have been
 
 and Primitive History 37 
 
 curiously agreed in noting customs strongly remind- 
 ing them of Germany. 
 
 Between these two 1 widely different races we shall 
 find immense differences of customs and culture. 
 The question will of course be asked, which of the 
 two races represents the primitive Indogermanic 
 people? It may be doubted whether we shall ever 
 be able to answer such a question. If we must 
 choose, we can hardly doubt that the northern race 
 has the prior claim, and the affinities between Indo- 
 germanic and Finnish speech, accepted as proved by 
 the high authority of Dr Sweet, undeniably make in 
 this direction. On the same side stands Dr Ridge- 
 way's proof that the Achaeans brought with them 
 the worship of Zeus, an unquestionably Indogerman 
 deity, while the indigenous Poseidon bears a name 
 which has so far defied analysis. But, on the other 
 hand, the non-Achaean Greek and the non-Sabine 
 Latin only differ dialectically from the language of 
 the invaders, so far as we can disentangle them. The 
 same conditions seem to prevail in Media, where the 
 non-Aryan' 2 population cannot be shown to have 
 spoken a language radically different from that of 
 
 1 The blond northern race is here spoken of for convenience as 
 one. Hut the conquerors of southern Europe included not only long- 
 headed but also short-headed men, who lived in great numbers in 
 south Germany ; so that there were really two different stocks among 
 them. So Professor Kidgeway informs me. 
 
 - I am of course using the term Aryan in its strict sense, as the self- 
 chosen name of the family speaking the closely-related Indian and 
 Iranian languages.
 
 38 Language 
 
 their Aryan conquerors. It must not be forgotten 
 that race and language cannot be treated as neces- 
 sarily going together. Two totally distinct races 
 may very well speak closely related languages, and 
 the difference of speech may be the result of the 
 fact that one race learnt the language as a foreign 
 idiom. But if the Mediterranean race did learn their 
 language from the peoples of the North, it must have 
 been in a prehistoric period at which it is absolutely 
 impossible for us to arrive \vith our present methods. 
 The period of this assumed process antedates the 
 dialectic division between the eastern and western 
 Indogermanic languages, a division which goes back 
 as far as our knowledge can take us. The typically 
 northern Keltic and Germanic lie in this respect on 
 the same side as the typically Mediterranean Latin 
 and " Pelasgian " Greek. It is better therefore to 
 acquiesce in our ignorance, and say that both races 
 spoke Indogermanic at the very earliest period to 
 which our science can approach. 
 
 And where did these primitive people live ? 
 "Somewhere in Asia" was the answer universally 
 given till comparatively lately. Whether the Book 
 of Genesis was supposed to demand this, or whether 
 it resulted from a general idea of the fitness of things, 
 is rather hard to say. In 1851 R. G. Latham of 
 whom as in private duty bound I must chronicle that 
 lie was a Fellow of King's College, Cambridge pro- 
 pounded the revolutionary view that the home was 
 to be found in Europe. His reasons centred on the
 
 and Primitive History 39 
 
 antecedent probability that the dog would wag the 
 tail and not the tail wag the dog. He was scoffed at 
 as a " crack-brained Englishman " by the superior 
 Germans. But the whirligig of time brings its re- 
 venges, and there is hardly a German now with a 
 reputation to lose who does not hold with Latham. 
 There is one very interesting exception, that great 
 philologist Johannes Schmidt, recently deceased, who 
 invented an argument for the Asiatic hypothesis, in- 
 sufficient indeed to bear the weight he laid upon it, 
 but well worth respectful consideration. He pointed 
 out that there were in European languages consider- 
 able traces of a sexagesimal system of numeration 
 crossing the decimal the peculiar formation of our 
 own eleven and twelve, and the Germanic " long 
 hundred," for 120, will serve as examples, and he 
 argued that this must be due to a very early contact 
 with the Babylonians 1 , with whom 60 was the pivot 
 of the numeral system. It seems clear that we should 
 call our European peculiarity duodecimal rather than 
 sexagesimal, and that other causes must be accepted 
 
 1 Schmidt throws in other supposed borrowings from Babylonian. 
 The most plausible of these is the Greek TrAe/cus "axe," Sanskrit 
 parafus, compared with Assyrian pilaqqu. I cannot feel satisfied with 
 his account of the relation between the Assyrian q and the Sanskrit 
 sh sound, and I think this is to be added to the remarkably large 
 category of purely accidental similarities. TreXe/cus and parafiis will 
 suit the root pclck (Latin flecto, our_/?y), and the original meaning may 
 have been a knife for cutting hides. Schmidt's doctrine has received 
 a serious blow from the other side in l)r Ridgeway's proof that the 
 sexagesimal system was of quite recent date in Babylonia. 
 
 M.
 
 4O Language 
 
 for its appearance than that which Schmidt de- 
 manded, but we may still admire the ingenuity of 
 the only serious argument which recent years have 
 brought to the support of a theory once taken for 
 granted as an axiom. 
 
 We have no time to indicate the various lines of 
 evidence which converge on northern Europe as the 
 most probable centre of radiation for the emigrations 
 of prehistoric ages. It must be carefully noted that 
 we only seek to determine the home of the Indo- 
 germans at the latest period at which they were 
 speaking mutually intelligible language. Of their 
 earlier home and culture Language can naturally 
 tell us nothing whatever ; and if other methods of 
 research decide that our forefathers ultimately started 
 from Asia or from the North Pole, our science cannot 
 say no, For the period to which Language leads us, 
 a northern origin seems at least favoured by the 
 prominence of winter and cold. Snow is certainly 
 Indogermanic, and possibly ice 1 as well, while winter 
 was known by the name found in the Latin Jdcmps 
 and the Sanskrit Himalaya. (" abode of snow 1 '). Spring 
 also (Latin iw) was a very well-marked season. 
 An argument has been found in the names of trees 
 which are shown by language to have been known 
 to the Indogermans. The oak and the pine, and less 
 
 1 The Zend isi, if we might slightly alter the sibilant in the one 
 place where it occurs, suits the (lermanic ice very well, and an attesta- 
 tion from Aryan and Germanic means more than one from any other 
 two Indogermanic languages, as these are the farthest apart.
 
 and Primitive History 41 
 
 conspicuously the willow and the birch, are decisively 
 proved to have existed in the primitive period, and these 
 four are said not to be found together outside Europe 1 . 
 The argument is complicated by the extraordinary 
 propensity of the tree-names to change their meaning. 
 When people migrated into a region where a par- 
 ticular tree was rare or absent, they would use its 
 name for " tree " in general, or for some other tree. 
 Thus in Italy fraxinus, the cognate of our birch, is 
 used to denote the ash. I may mention two ex- 
 amples of the methods by which we may fix the 
 original meaning of these words. Tree means "oak" 
 in Greek and Keltic, " wood " or " tree " in Germanic 
 and Albanian, "larch" in Latin. The cognate tar, 
 with one of similar meaning in Lithuanian, decides us 
 for a resinous tree, presumably the pine, as the 
 original. Fir means " oak " in Latin (guercus), a 
 kind of fig-tree in Sanskrit, and it has indeterminate 
 derivatives in Keltic, Gothic, Lithuanian, etc. But a 
 word for the Thunder-god derived from it argues for 
 " oak " as the earliest meaning, for the oak-tree and 
 the thunder were old associates, as we shall see later. 
 Pursuing our enquiry, we find that the Indo- 
 germans were familiar with the sea (mere, Latin 
 mare). The older investigators were most unwarrant- 
 ably sceptical on this point. Holding as they did an 
 exaggerated view of the importance of the Aryan 
 
 1 So Ilirt, an able philologist. But an excellent botanist tells me 
 he is wrong. How necessary it is to "verify our references " ! 
 
 42
 
 42 Language 
 
 branch in all such enquiries, they denied Indogerman 
 antiquity to words which were not to be found in 
 Indian or Iranian languages. Since Aryan is now 
 found to be closely united to its neighbours, the 
 Slavonic on one side and the Armenian on the other, 
 we cannot lay the same stress on the absence of 
 Aryan attestation. The word mere has in its form 
 strong witness to its high antiquity ; and, the Aryan 
 migration being entirely inland, they may easily have 
 lost the word before they reached the Caspian or other 
 great sheet of water. I cannot stay to argue the case 
 for the Baltic as the sea of the Indogermans. I need 
 only point out that it meets all the conditions, so 
 long as we do not insist on the immediate contact 
 of the whole people with the seaboard, but conceive 
 of them as spreading inland to the south-east and 
 south-west in two streams as described already. That 
 the converging point on the Baltic is still the home 
 of the Lithuanians is itself no mean argument, for this 
 people retains to the present day a most dispropor- 
 tionate amount of primitive features in its language, 
 which is at once accounted for if we take them to be 
 descendants of the original folk occupying still the 
 original home. The region of our hypothesis still 
 harbours the wolf and the bear, beasts thoroughly 
 familiar to the primitive people, who knew nothing 
 (so far as language can tell us) of the Asiatic elephant 
 or tiger. There are bees to supply honey for 
 making mead, on which I fear the primitive Indo- 
 german had learnt to make himself drunk to an
 
 and Primitive History 43 
 
 extent quite worthy of his noble descendants, the 
 heirs of all the intervening ages. Gigantic trees 
 were there to be felled and hollowed by fire to 
 make the "dug-outs" in which they navigated the 
 streams ; and in the vast and awesome woods we 
 can picture them worshipping Dicus Perqiinos, Zeus 
 of the oak, with rites not greatly differing from 
 those of the Druids in historical times. 
 
 From this centre we may conceive the Indo- 
 germans spreading as their numbers grew, and their 
 dialects gradually developing in the way I tried to 
 describe in the former lecture. It would naturally 
 happen that some of the tribes thus formed would 
 remain in contact with one another long after they 
 had ceased to be in touch with their other kin, so 
 that new culture characteristics and new terms to 
 describe them spread throughout a limited area 
 without affecting those who had detached themselves 
 before these new departures took place. A good 
 example may be found in the evidence mentioned 
 earlier in this lecture, connecting the common culture 
 of the Italians, Kelts, and Germans with the period 
 of the Swiss Lake-dwellers, who differ from the 
 earlier Indogermans by the prominence of fish in 
 their diet 1 . 
 
 1 I should remark here that Professor Ridgeway, speaking as an 
 archaeologist, pronounces primitive man, wherever found, largely a fish- 
 eater. I am dealing only with the evidence of language, and if the 
 archaeologists press this point we must of course assume either (i) that 
 fish (Latin pistil,, Old Irish iasc) was lost in the other five main
 
 44 Language 
 
 It remains for us to ask what light our science 
 can cast on the arts and accomplishments, the family 
 life and the religion of this primitive people. It is 
 usual to gauge the progress of nations in early stages 
 of development by the extent of their acquaintance 
 with metals. The test does not pretend to be com- 
 plete, for it is obvious that a people still in the Stone 
 Age may be more advanced and more civilised than 
 one which has learnt something of the use of metals. 
 But practically the test is found quite good enough 
 for its purpose, and it is therefore the best course for 
 us to ask first what was the condition of the Indo- 
 germans in this respect. We find that there is fairly 
 decisive evidence that they were acquainted with 
 copper, which they knew by two names, one the 
 ultimate origin of our word iron, the other possibly 
 a derivative from the adjective red. The only other 
 metal they could have known is gold, which like 
 copper is often found on the surface pure. The 
 Germans, Letts and Slavs -contiguous peoples, be 
 it noted must have learnt its use very early, and 
 agreed to name it by its present name, derived from 
 the root of yellow. Some rather tempting Sanskrit 
 evidence for this word has been alleged, but it is 
 
 Indogerman dialects, especially by those whose migrations took them 
 across the steppes, or to seas where the tish were unappetising, or 
 (2) that the Indogermans, as in the ease of " one " and " hand " 
 (p. 33), possessed the tiling but not the name in common. It seems 
 to me, however, at least a striking coincidence \\\z\. fish belongs exclu- 
 sively to these three contiguous peoples, who may so well have 
 radiated from the Lake-dwellers' area.
 
 and Primitive History 45 
 
 hardly to be relied upon ; and, if it were, it only 
 involves extending the chain of contiguous tribes 
 possessing the name one more link to the east. At 
 first sight, the case for silver appears stronger, inas- 
 much as Aryans, Armenians, Italians and Kelts use 
 the same word, while Greek has the same root with 
 the suffix changed. But on further examination we 
 find that it only involves the coincident use of a word 
 for "white," to which not very recondite piece of 
 nomenclature the great wits of various tribes may 
 well have jumped independently. For other metals 
 no semblance of a case can be made out. It will be 
 obvious that the absence of a common name for the 
 various metals is in this case a very serious argument, 
 for though a word for " hand " may as we have seen 
 be missing without our drawing the inference we 
 draw here, we have the best possible evidence from 
 archaeology that the Inclogermans did possess hands, 
 evidence which is conspicuous by its absence in the 
 case of the metals. We conclude that the Indo- 
 germans were in the Neolithic stage of culture, 
 slightly modified by their use of copper. This fact 
 incidentally supports our decision against the Asiatic 
 home ; for the Semites, while still one undivided 
 people, knew gold, silver and copper, and it is very 
 hard to believe that the Inclogermans would have 
 remained ignorant of all but copper, had they started 
 from a country comparatively near the Semite home, 
 especially if, as Schmidt's theory demands, they had 
 actual contact with the Semites in a very early period.
 
 46 Language 
 
 We soon find that dulce domum was an idea 
 familiar to the earliest Indogermans, for domus (our 
 timber, German Zimmer) is a word found everywhere. 
 It seems moreover to connote "home" as well as 
 " house," for it is used to describe the household as 
 well as the dwelling, and the "house-master" and 
 " house-mistress " (cf. the Greek word from which we 
 get our despot) took their names from it. We must 
 leave the archaeologists the last word as to the 
 character of the house. Language supports the as- 
 sumption that there was one room specially named 
 from the central hearth, with a large opening in the 
 roof above to let the smoke escape. Oven and cook 
 are both primitive words, but we must not let our 
 ideas approach too near the processes of a modern 
 kitchener ; an earthen pot hung from three sticks over 
 the fire is perhaps as likely as anything. The fact 
 that the wagon was known, with axle and wheels 
 presumably in one piece, has suggested that the Indo- 
 germans were gipsies, differing from their modern 
 descendants only in the inferior finish of their caravan. 
 But there is no necessary connexion, and wagons 
 originally used for living in may well have been 
 retained for agricultural use. That agriculture was 
 practised, probably side by side with the keeping of 
 flocks and herds, for which con.', steer and civc are 
 primitive words, seems certainly established. Corn, 
 and the German Gcrste, are well-attested primitive 
 words, and there are too many others of the kind to 
 make it probable that the Indogermans only collected
 
 and Primitive History 47 
 
 wild-grass seeds. The words qncr)i and mill are 
 equally old, and so are the verbs car and sow, and 
 probably the name of the finished article, loaf. The 
 names for " plough " often tell their own tale by their 
 connexion with words meaning " branch." The older 
 scholars' hesitation to accept agriculture as Indo- 
 germanic arose largely from the fact that most of 
 these words are either absent or have become colour- 
 less ("move," "go," and the like) in the Aryan 
 languages. But there is no reason why we should 
 allow Aryan a decisive \vcight which we should never 
 think of granting to any other single branch of the 
 family ; and when we have recognised that the 
 Aryans' migration to their present homes took them 
 across the steppes we have at once the sufficient 
 reason for their losing words denoting processes of 
 agriculture, which could be rarely applied. 
 
 It is naturally impossible here to attempt a com- 
 plete sketch of the life of this prehistoric people. 
 I might mention that they had discovered arts to 
 which the words i^cai'c and seiu attached themselves : 
 of course they would be in the most elementary stage 
 of development. What gives us a yet higher idea of 
 their progress is the apparent fact that they achieved 
 a very near approximation to the length of the solar 
 year. ( Year, like inontJi, is primitive, but probably 
 meant " spring," as does sometimes its Greek cognate 
 from which comes the word " hour.") Twelve lunar 
 months, 354 days, seem to have been the first approxi- 
 mation, which would soon be found to be too short.
 
 48 Language 
 
 A " little year " of twelve days brought the year to 
 what is practically correct. We find these twelve days 
 kept with special ceremonial at the two extremes of 
 the Indogermanic area ; and each day up to "Twelfth 
 Night " was supposed to forecast the weather of a 
 corresponding month in the year following. 
 
 The family conditions of the Indogermans claim 
 our attention before we begin to describe their religion, 
 the last topic on which we shall have time to dwell. 
 The very large number of relationship words strikes 
 every observer. Father, mother, son, daughter, brother, 
 sister, are our share in the primitive inheritance ; and 
 to these may be added words for grandson and grand- 
 daughter, son's wife, husband's father and mother, 
 husband's brother and sister (?), and perhaps grand- 
 father, stepmother, uncle, son-in-law, brothers' wives, 
 sisters' husbands. The relationships seem mostly to 
 be on the male side, but there are possible survivals 
 of a " matriarchate," and in any case the position of 
 women seems to have been passably good for so early 
 a stage of civilisation. There was a great difference 
 between the North European and the Mediterranean 
 races in this respect. Everyone notices the extra- 
 ordinary contrast between the position of women at 
 Athens, in the age of her greatest fame, and that of 
 the Homeric women, some six centuries earlier. 
 The apparent degeneration is explained at once by 
 Professor Ridgeway's theory, for the Achaean women 
 are in just the same social condition as those of 
 ancient Germanv, as described by Tacitus, which is
 
 and Primitive History 49 
 
 also reflected in the life of the patrician matrons at 
 Rome. It was not from the northern invaders that 
 the slavery of women in India came. At the same 
 time it is suggestive that widow is a primitive word, 
 while there is none to describe the widower: we may 
 probably infer that the widow was at any rate not 
 re-married. 
 
 The clan was apparently the highest political unit. 
 The wife would be brought in by purchase from 
 another clan : the word wed is specially associated 
 with this purchase-money. The slaying of a member 
 of the clan originally denoted in Italy by the word 
 parricide (which only popular etymology connected 
 with pater-'] imposed a special obligation on the 
 survivors. The northern race were content to exact 
 a wergild from the homicide or his clan ; but the 
 Mediterranean people, believing as they did that 
 the shed blood cried from the ground, and that the 
 unsatisfied spirit was always near the living and 
 powerful to harm, insisted on blood revenge, nor has 
 the practice died out in our own day. 
 
 We have come to a point at which it is natural to 
 go on to the investigation of Indogennanic religion. 
 The wide differences of view which have been held 
 on this subject suggest that a solution is most likely 
 to be found on the lines of Professor Ridgeway's 
 fruitful theory. There arc plentiful traces of ancestor- 
 worship. The Romans h;id their Manes, and the 
 Greeks never lost their deep-seated conviction that 
 the souls of the dead dwelt in or near their tombs,
 
 50 Language 
 
 mighty to help or hurt those whose most sacred 
 duty was to care for their ancestors' graves, and give 
 the shades a momentary taste of renewed life by 
 drink-offerings of blood poured through a little 
 opening into the tomb. So indestructible was this 
 belief, that in the island of Thera we find a number 
 of Christian tombs with the Greek inscription " The 
 angel of so-and-so": the old pagan ancestor-spirit 
 had a new lease of life when christened as a guardian 
 angel. On the other hand we find that the Achaeans 
 burnt their dead, and believed that the spirit after the 
 burning flew away to the Isles of the Blest, never to 
 return. The mixture of these two contrary ideas 
 produced inextricable confusion in the eschatology 
 of the later Greeks ; but among the kindred of the 
 Achaeans in Northern Europe we find their ideas in 
 their original purity. There is moreover from all 
 quarters evidence of nature-worship as the creed 
 of original speakers of Indogermanic languages. 
 Classical writers portray for us the religion of the 
 ancient Germans and Gauls and Persians, and the 
 portraits agree in the prominence assigned to the sun 
 and moon, and to the associated worship of heaven 
 and earth, which latter were regarded as father and 
 mother of all. These ideas are reflected in the one 
 divine name which is common to many branches of 
 the Indogermanic family. The Greek Zeus, the 
 Italian Jove, the Germanic deity whose name we 
 preserve in Tnes-day, answer to the old Vedic 
 Dyaus, whose worship evidently failed to take
 
 and Primitive History 5 i 
 
 root among the peoples of India conquered by the 
 northern warriors, for even in the earliest Veclas he 
 is a name and little more. I am myself tempted 
 to conjecture that when Herodotus tells us of the 
 Persian popular faith, and observes that they " called 
 the whole vault of heaven Zeus," he was not merely 
 using the familiar name of the Greek supreme deity, 
 to describe the supreme deity of another people, but 
 reproducing the very word he had heard in Persia 1 . 
 Closely akin to this name is the general word for 
 God, whence the Latin dens and divinus descend 2 . 
 The name presumably means " shining," and fits the 
 nature deity of the northern race, as completely as 
 the Greek $609, if its congener is really to be sought 
 in Lithuanian and German words for " ghost " or 
 " spirit," fits the ancestor-worship of the Mediterranean 
 peoples. It is impossible for us to enter further here 
 upon the features which language and archaeology 
 combine to authenticate for early times. One in- 
 teresting point may be noted with reference to the 
 
 1 Ilcrod. I. 131 TOV KIIK\OI> TTO.VTO. Tou ovpavov Aid Ka\eovres. The 
 old word for " heaven " probably survives in one passage of the Avesta; 
 and, as the present writer has tried to show, in a forthcoming article 
 on the religion of Persia, the common people probably kept up the 
 old nature-worship long after the court had adopted the reform con- 
 nected with the great name of Zoroaster. If the name Dyaus still 
 survived, a Greek could only suppose it Zeus, especially if he heard 
 it in the accusative case. 
 
 2 And, according to one excellent authority, Tues(day), which 
 would thus become in its origin a common and not a proper noun. But 
 the older view cannot be regarded as conclusively ruled out by the 
 phonetic objection raised against it.
 
 52 Language 
 
 god who seems clearly marked out as supreme, at 
 any rate among the northern folk. It may be re- 
 garded as fairly certain that Dieus had not only the 
 title " Father " but the further appellative Perqimos, 
 " of the Oak." The latter name has become a title 
 of a rain or thunder god among the Vedic Indians 
 (probably), the Norwegians who applied it to the 
 mother of Thor and the Lithuanians, while among 
 the Albanians it denotes " God " or <( heaven." We 
 have from Homer evidence of a sacred oak in 
 Dodona, where special priests of Zeus ministered, 
 "with unwashen feet and making their bed on the 
 ground " instances of the holiness of dirt which may 
 be abundantly paralleled in the Fakirs of Hinduism 
 and other religions. The northern cult of Zeus was, 
 on Dr Ridgeway's theory, brought with them by the 
 invading Achaeans through Epirus to Greece proper, 
 and the shrine of Dodona preserves this worship in 
 its last stage before it was fused and harmonised with 
 the cult of Poseidon and other deities which existed 
 in Greece before the invasion. The combination of 
 Sky and Oak is not an obvious one to our minds, 
 but to the primitive man the connexion was easy. 
 The Sky-god sent the lightning which blasted the 
 Oak, the chief of trees, and presumably the most 
 abundant in the region where the cult arose. A 
 peculiar sanctity was always, even in much later 
 ages, attached to objects and persons struck by 
 lightning, and the blasted oak was considered to be 
 the special abode of the Sky-god who had thus come
 
 and Primitive Plistory 53 
 
 to his own. Apart from the sanctity of the oak, the 
 fact that words for " acorn " are widely attested, and 
 have primitive appearance, would seem to suggest 
 that the tree provided food for man as well as beast. 
 It is interesting to note how the " rain from heaven 
 and fruitful seasons," declared by St Paul to be God's 
 witness of Himself to the heathen world, produced 
 among the remote ancestors of those to whom he 
 spoke a cult of the sky that gave rain and the tree 
 that bore fruit. And "the times of that ignorance 
 God overlooked," well pleased we are encouraged to 
 believe that even by the imperfect nature-worship 
 of the "Sky Father" He had led His children to 
 know that they were " His offspring." 
 
 The other strain of primitive religion, the ancestor- 
 worship, is hardly supported at all by evidence from 
 language. The nearest approach to a common word 
 for " spirit " is seen in the cognates of the Greek 
 tJicos, already noticed 1 . The evidence is decidedly less 
 satisfactory than that which proves nature-worship. 
 Nearly all the Indogermanic peoples, however, seem 
 to have practised this cult, even though they may not 
 have used common words for it. The fusion of the 
 two conceptions of religion may well be as old as 
 the spread of the common language over two very 
 distinct races, and in this case the science of language 
 cannot come to the aid of history. It will remain 
 only a theory, made plausible rather by archaeology 
 and history than by linguistic evidence, that ancestor- 
 
 1 See Schrader, Reallexikon, p. 28.
 
 54 Language and Primitive History 
 
 worship was originally the religion of the Mediter- 
 ranean people, and filtered northwards just as the 
 worship of Dieus Perqunos spread over the south. 
 The two strains met conspicuously in the Greek 
 religion, and very evidently played their part in 
 preparing the Greek mind for its ultimate work in 
 the propagation of a world religion. 
 
 So must end this brief and fragmentary intro- 
 duction to a fascinating study. That many phases 
 of primitive Indogermanic life, on which the evidence 
 of speech is attainable, have been entirely passed 
 over, while those that are not passed over have only a 
 few leading features indicated, is obvious enough, and 
 belongs to the conditions under which the present 
 task is undertaken. But I hope it will be found that 
 the selections given will whet the appetite for more, 
 and that even this meagre sketch will do something 
 to show how wide and how suggestive a study lies 
 behind the words we unthinkingly use in daily life.
 
 55 
 
 BIBLIOGRAPHY. 
 
 A short list of books on the subjects sketched in these 
 lectures, available for English readers, may be found useful. 
 Comparatively few of them will appeal to the general reader, 
 but the student will find them within his grasp if he has 
 a fair acquaintance with the structure of Latin and Greek 
 and English. The books are classified, the easiest being 
 put first in each section, with a few notes on each. 
 
 A. 'I'm-: STUDY OF LANGUAGE (GENERAL). 
 
 Sweet, II. .77/1? History of Language. (Temple Primers.) 
 Dent, 1899. 
 
 Though only a primer, this little book contains a large 
 amount of original matter, especially on the relations of the 
 Finnish and other languages to the Indogermanic family. 
 
 Strong, Logeman and Wheeler. Introduction to the Study 
 of the History of Language. Longman, 1891. 
 
 An adaptation of the difficult, but most important, work 
 by H. Paul, on Principles of Language-. (The English trans- 
 lation of this latter work, by Strong, is not recommended.) 
 M. 5
 
 5 6 Bibliography 
 
 Whitney, W. D. Life and Growth of Language. (Internat. 
 Science Series.) King, 1875. 
 
 Due allowance must of course be made for the immense 
 development of the subject since the publication of this 
 work ; but the great American scholar's teaching in many 
 ways cannot go out of date. 
 
 Oertel, H. Lectures on the Study of Language. Arnold, 
 1901. 
 
 By far the best general account of the subject, acute, 
 correct, and original. Like the last, this book is due to an 
 American scholar. 
 
 Giles, P. " Philology" (in new Supplement to Encyclopaedia 
 Britannica]. See under C. 
 
 B. ETYMOLOGY. 
 
 Skeat, W. W. Primer of English Etymology. Clarendon 
 Press, 1892. 
 
 ,, ,, Principles of English Etymology. 2 vols. 
 Clarendon Press, 1891, 1892. 
 
 ,, ,, Etymological Dictionary of the English 
 Language. 3rd ed. Clarendon Press, 1898. 
 
 ,, ,, Concise do. Latest edition (rewritten), 1901. 
 
 For practical use this last supersedes everything else, con- 
 taining as it does the matured results of many years' work 
 by the leading English authority on Middle English.
 
 Bibliography 5 7 
 
 Kluge, F., and Lutz, F. English Etymology; a select 
 
 glossary. Blackie and Son, 1899. 
 
 A much smaller book than Skeat's Concise Dictionary, 
 but exceedingly sound and reliable. 
 
 Taylor, I. Names and their Histories. Rivingtons, 1896. 
 
 A glossary of place-names, selected from various countries, 
 with an introduction and appendices. 
 
 C. M KAN INGS OF WORDS. 
 
 BRKAL, M. Semantics. Heinemann, 1900. 
 
 Practically a first attempt to deal with the science of 
 meanings systematically. Dr J. P. Postgate adds some 
 valuable discussions to the English edition. 
 
 Murray and Bradley. The New English Dictionary. 
 Clarendon Press (in progress). 
 
 This colossal work traces the history of every English 
 word by (potations reaching back to its earliest appearance: 
 the etymology is also examined. It is of course the final 
 court of appeal in all matters concerning English words, no 
 other dictionary pretending to cover more than a fraction of 
 the ground here occupied. 
 
 Giles, P. " Philology'' (in new Supplement to Encyclopaedia 
 Britannica], 
 
 This article is mentioned here specially because there is 
 so little literature as yet dealing with the general principles 
 
 52
 
 58 Bibliography 
 
 of Semantics: it is however to be consulted on most of the 
 subjects referred to in this little book. See also Mr Giles's 
 article, Writing, in the same Supplement. 
 
 D. INDOGERMANIC LANGUAGES. 
 i . General. 
 
 Brugmann, K. Elements of the Comparative Grammar of 
 the Indogermanic Languages. 4 vols. and Index 
 (a 5th vol.). Triibner, 1886-1897. 
 
 This encyclopaedic work, translated by Wright, Convvay 
 and Rouse, is the thesaurus of the results of the new philo- 
 logy, covering the phenomena of the eight main branches 
 of the Indogermanic family in their older developments. 
 The first volume, dealing with phonology, is superseded by 
 the new German edition (1897); the second part, by 
 Prof. Delbriick, which treats the Comparative Syntax on 
 similar lines, is unfortunately not translated. Prof. Brug- 
 mann has been described above (p. 13) as the great pioneer 
 of the new school, and the learning, clearness, and penetra- 
 tion visible throughout his work are extraordinary. He is 
 now publishing (in German) an abridgement of it. 
 
 2. Classic Y7/. 
 
 Giles, P.' Manual of Comparative Philology for Classical 
 Students. 2nd ed. Macmillan, 1901. 
 
 This admirable work, which has been translated into 
 German, entirely supersedes all previous summaries of the 
 history of the Greek and Latin languages. It includes also
 
 Bibliography 5 9 
 
 an excellent account of the principles of the science, with 
 an application of these principles to syntax, a field which is 
 only beginning to be worked on the lines of the newer 
 philology. Considerable attention is given to English. 
 
 ('lark, J. Manual of Linguistics, Thin, 1893. 
 
 This book is much slighter than Giles's, but has some 
 very good material, and deals with English more fully in 
 proportion. 
 
 3. Germanic. 
 
 Henry, V. Comparative Grammar of English and German. 
 
 Sonnenschein, 1894. 
 
 Those whose knowledge is mostly confined to modern 
 languages can study the principles of linguistics to excellent 
 purpose within the narrow field of Germanic, as is shown 
 in these lectures above. Prof. Henry, of Paris, is a 
 thoroughly trustworthy writer. 
 
 West, A. S. JUements of English Grammar. Cambridge 
 
 University Press. 8th edition, 1902. 
 An excellent sketch of historical .English. 
 
 Toller, T. X. Outlines of the History of the English 
 
 Language. Cambridge University Press. 1900. 
 A very sound and readable summary. 
 
 Mayhew, A. L. Synopsis of Old English Phonology. Claren- 
 don Press, 1891. 
 
 An exceedingly compact account of the relations be- 
 tween English sounds and those corresponding to them in 
 cognate languages.
 
 60 Bibliography 
 
 Behaghel, O. Short Historical Grammar of the German 
 Language. Translated and adapted by Dr E. 
 Trechmann. Macmillan, 1891. 
 
 Wright, J. Primer of the Gothic Language. Clarendon 
 Press, 2nd edition, 1899. 
 
 The study of these cognate Germanic dialects, the 
 second-cousin and the great-aunt, respectively, of modern 
 English, will of course greatly enlarge the understanding of 
 our own language. 
 
 E. LANGUAGE AND PRIMITIVE HISTORY. 
 
 i . General 
 
 Kendall, G. H. The Cradle of the. Aryans. Macmillan, 
 1889. 
 
 A sound little resume for its date, but very much has 
 been done in the past fourteen years. 
 
 Taylor, I. The Origin of the Aryans. Scott, 1889. 
 
 On a larger scale than Kendall's essay, this book is 
 necessarily more out of date, especially on the linguistic 
 side. I am not competent to criticise its craniology and 
 archaeology. Canon Taylor did excellent service by 
 popularising newer views, and the book may still be used 
 with profit if carefully checked by later work.
 
 Bibliography 6 1 
 
 Schrader, O. (tr. I)r F. H. Jevons). -Prehistoric Antiquities 
 of the Aryan Peoples. 1890. 
 
 This important work has been superseded by DrSchrader's 
 recent Rca/lexikan, a dictionary of words bearing on the 
 culture of primeval times. The author is decidedly the 
 leading specialist on linguistic palaeontology. 
 
 There is also an excellent account of the controversy in 
 French by S. Reinach: II Origine des Arycns (Leroux, 1892). 
 
 2. Religions, 
 
 Stetiding, H. Greek and Roman Mythology and Heroic 
 
 Legend. (Temple Primers.) Dent, 1901. 
 An excellent little manual. 
 
 Hopkins, K. W. The Religions of India. Arnold, 1896. 
 I)e la Saussaye, P. D. C. The Religion of the Teutons. 
 
 Ginn, 1902. 
 
 These two works, in Prof. Jastrow's "Handbooks on the 
 History of Religions," give admirable accounts of religions 
 which, with the Greek, the Roman, and the Persian, supply 
 our main evidence for deducing the religion of the primeval 
 Indogerman. 
 
 Ridge way, W. The. Early Age of Greece. Cambridge 
 University Press, 1901 (vol. i.), 1903 (vol. ii. in 
 the press). 
 This brilliant work, frequently referred to above, contains 
 
 very important discussions on the history of Greek and 
 
 Roman religion, as well as on the origin of their culture. 
 
 I abstain of course from pronouncing on the archaeologists' 
 
 debate of which this book is the centre.
 
 62 Bibliography 
 
 Chadwick, H. M. The Oak and the Thunder-god. Harri- 
 son, 1900. 
 
 A paper presented to the Anthropological Institute, 
 giving an excellent account of the evidence for the cult 
 described on p. 52 above. My lecture was in print before 
 I saw Mr Chadwick's paper, from which I venture to 
 dissent only in the account he gives of the rationale of the 
 link between the oak and the thunder. 
 
 3 . An th r apology. 
 Tylor, E. B. Anthropology. Macmillan, 1895. 
 
 Frazer, J. G. The Golden Bough. 2nd ed., 3 vols., Mac- 
 millan, 1901. 
 
 Prof. Tylor' s manual has chapters on Language, and on 
 the various elements of culture which are taken up from the 
 other side by linguistic palaeontology. l)r Frazer's famous 
 book illustrates phenomena of language very freely, and its 
 study of comparative institutions must be taken into account 
 whenever we examine what the evidence from language 
 proves.
 
 INDEX I. 
 
 Names of modem writers are in italics. For convenience, the name.* 
 :if foreign scholars are given with initials only. 
 
 Ablaut see Vowel-gradation. 
 
 Accent 13 f, 21, 25 
 
 Accidental similarities 16, 39 
 
 Achaeans 35, 37, 48, 50, 52 
 
 Acorns 53 
 
 Agriculture 46 f 
 
 Albanian 6, 52 
 
 Analogy 23 tf 
 
 Ancestor-worship 49 ff 
 
 Angel, for ancestor spirit 50 
 
 Animals, Indogermanic 42 
 
 Anthropology 30, 62 
 
 Archaeology 30, 35, 45 f, 53 
 
 Armenian 6, 42, 45 
 
 Articulation and inherited speech- 
 organs 21 
 
 Aryan set; Indogermanic. 
 
 Aryan ( = Indo-Iranian) 6 f, 37 f, 
 40 IT, 45, 47 
 
 Asia as Indogerman home 38, 
 
 45 
 
 Assyrian 39 
 Athenians 35, 48 
 Athens, St Paul at 53 
 Avesta 51 
 
 Babylonians 39 
 Baltic 34 f, 42 
 Behag/iel, O. 60 
 Blood-offerings 50 
 
 Blood revenge 49 
 Bofp, F. 5 
 
 Borrowing of culture words 32 
 Botany 30, 41 
 Bradley, //. 57 
 Brcal, M. 57 
 Brugmann, Karl I3f, 58 
 Burning and burying 50 
 
 Cambridgeshire 19 
 
 Celt see Keltic. 
 
 Cluuhvick, H. M. 62 
 
 Chaucer 25 
 
 Clan 49 
 
 Clark, J. 59 
 
 Classical philology 5, 58 f 
 
 Colour words 27 
 
 Comparative mythology 1 1 f 
 
 ,, syntax 58 f 
 
 Conquest, effects on language 23, 
 
 3 r >- .? 
 
 Conscious phonetic change 10 
 Conway, Prof. 20, 58 
 Copper 44 f 
 CinveU, Prof, ix f 
 Craniology 30 
 Curt ins, Georg I i f 
 Cyrils 36 
 
 Dawn-goddess 1 2 
 Decimal system 39 
 Delbriick, B. 58 
 Dialect mixture 20 
 Dialects, growth of 22
 
 6 4 
 
 Index I 
 
 Dialects of English 19 
 Dieus Perqunos 43, 52, 54 
 Digamma 1 2 
 Dirt and religion 52 
 Dodona 52 
 Druids 43 
 "Dug-outs" 43 
 Duodecimal system 39 
 Dutch 8 
 
 Eastern and Western Indogennans 
 
 6, 14. i?. 3 8 
 
 Educational value of Language 2 
 English Language 8, i8ff, 25, 27, 
 
 56 f, 59 f 
 Epirus 52 
 Eschatology 50 
 Ethnology 30 
 Etymology i f, 9, 34. 56 f 
 Europe as Indogerman home 38 ff 
 European races 34 f, 37, 53 
 Exceptions to phonetic law 15 f 
 
 Fakirs 52 
 
 " False" Analogy 23 
 
 Family life 48 f 
 
 Fatherhood, divine 50, 52 f 
 
 Pick, A. 36 
 
 Finnish 37, 55 
 
 Fishing 32, 43 f 
 
 P'ive, word for 14 
 
 Formal Analogy 24 
 
 Frazer, Dr J. G. 62 
 
 Gauls 50 
 
 Geology 30 
 
 German 8, 10 
 
 Germanic languages and peoples 
 
 7. 9> J3, l8 > 2 7> 34 ^ 3 ff > 43 f > 
 48, 50, 59 f 
 
 Germans, in Linguistics and Eng- 
 lish studies 4 f, 58 
 Giles, P. ix, 56 ff 
 Gipsies, Indogermans as 46 
 Gold 44 
 
 Gothic 4, 32 f, 60 
 Greek 2 f, 6, 8, 12, 16, 25 f, 32 f, 
 
 3 8 ' 45: 49 
 
 Greek religion 37, 49, 52 ff, 61 
 Grimm, Jacob 7 
 Grimm's Law 7 ff , 13, i8f 
 Gutturals 14 
 
 Hand, name for 33, 45 
 
 Hearth 46 
 
 Heaven and earth 50 f 
 
 Hebrew 16 
 
 Henry, V. 59 
 
 Herodotus 5 i 
 
 High Germans 10 
 
 Ilirt, H. 41 
 
 Home life 46 
 
 Homer 35, 48, 52 
 
 Homicide 49 
 
 Hopkins, E. IV. 61 
 
 Horace 20 
 
 Horse 14, 34 
 
 House-master and -mistress 46 
 
 India, northern immigrants into 
 
 36, 49 
 
 Indian languages 6, 34 
 Indogermanic, the name 7 
 
 ,, peoples, their home 
 
 34 ff 
 
 ,, , eastern and west- 
 
 ern 6, 14, i 7, 38 
 ,, , sounds of 9. 13 
 
 ,, . cautions in re- 
 
 constructing i4f
 
 Index I 
 
 Indogermanic religion 49 ff 
 Invariability of phonetic law 15 ff 
 lonians 35 
 Iranian 6, 34, 36 
 Isles of the Blest 50 
 Italic 6, 20, 34, 43, 45 
 
 Jastrow, M. f\ i 
 /c-rww, Dr F. />'. 6 i 
 Jones, Sir ll r . 3 ff 
 
 Lake-dwellers 43 f 
 Language, as a Science i 
 
 ,, its educational value 2 
 
 ,, and Race 38 
 Latham, Dr R. G. 38 f 
 Latin 3 f. 6, 8, i(i, 19, 27, 32, 38 
 Law in Language 12, 15 
 Letts 44 
 Lcskicn, A. 15 
 Lightning 52 
 
 Linguistic Palaeontology 30 ff 
 Lithuanian 6 f, 34, 42, 52 
 ' Little Year " 48 
 Logical Analog}' 24 
 " Long hundred '' 39 
 Loss of old words 32, 43 ff 
 Lutz, F. 57 
 Lunar year 47 
 Lystra, St Paul at 53 
 
 MacAdam 26 
 Manes (Lat.) 49 
 Marriage 49 
 Material Analogy 24 
 Matriarchate 48 
 May hew, A. L. 59 
 
 Media 37 
 
 Mediterranean race 35, 38, 48 ff 
 
 Metals 44 f 
 
 Milton 28 
 
 Mitllc-r, Prof. Max 9, r i 
 
 Murray, Dr J. A. II. 57 
 
 Nature-worship 50 
 
 Neogrammarians viii, 15, 58 
 
 Neolithic culture 45 
 
 Nomadic life 46 
 
 Norse 8, 52 
 
 North European Race 35, 37 f, 48 ff 
 
 Numerals 33, 39 
 
 Oak-tree and thunder 41, 52, 62 
 Ocrt.'l, If. 56 
 One, word for 33 
 Ossetes 36 
 
 Pastoral life 46 
 
 Patricians at Rome 36, 49 
 
 Paul, H. 55 
 
 Paul, St 53 
 
 Pcile, Dr J. v, viii f 
 
 Pelasgians 35, 38 
 
 Persians 36, 50 ff, 6r 
 
 Phonetic change, unconscious 10 f 
 
 Plough 47 
 
 Poseidon 37, 52 
 
 I\>st!fate, DrJ. P. 57 
 
 Primitive Indogerman home ^4, 
 
 38 ff 
 
 ,, ,, religion 37, 
 
 49 ff, rt,f 
 
 ,, ,, animals 42 
 
 ,, ,, metals 44 f 
 
 Psychology in Language 2, 28 
 
 Purchase of wife 49 
 
 Race and Language 38
 
 66 
 
 Index I 
 
 Races of Europe 34 f, 37, 53 
 Rapidity of utterance 20 f 
 Reinach, S. 61 
 Relationship words 48 
 Religion of Indogermans 49 ff 
 Kendall, Dr G. PL 60 
 Ridgeivay, Prof, vii, 35 ff, 39, 43, 
 
 48 f, 52, 61 
 
 Roman ancestor-worship 49 
 Romance languages 6 
 Rouse, Dr W. H. D. 58 
 
 Sabine 20, 36 
 
 Sanskrit 3, 6, 17, 40, 44 
 
 Saussaye, P. D. C. de la 61 
 
 Sihlciiher, August I3ff 
 
 Schmidt, Johannes 39, 45 
 
 Schrader, 0. 53. 61 
 
 Sea 41 
 
 Semantics 33, 41, 57 f 
 
 Semites 45 
 
 Sexagesimal system 39 
 
 Silver 45 
 
 Simplification of language 28 
 
 Sixty 39 
 
 Skeat, Prof. 19, 56 
 
 Sky and Oak 52, 62 
 
 Slavonic 6, 34, 42, 44 
 
 Sonant liquids and nasals 13, 21 
 
 Sound-shifting 10, 18 
 
 Speech-mixture 23 
 
 Spirits 49, 51, 53 
 
 "Sporadic" change 12 
 
 Spring 40. 47 
 
 Steppes, migration over 44, 47 
 
 ^tending, H. 6 1 
 
 Stone Age 44 
 
 Strong, Logeman and Wheeler 55 
 
 Strong perfect, English and Greek 
 
 25 
 Sweet, Dr H. 37, 55 
 
 Syntax 58 f 
 
 Tacitus 48 
 
 Taylor, Dr I. 57, 60 
 Thera 50 
 
 Thunder-god 41, 52, 62 
 Toller, Prof. 59 
 Trechmann, Dr 60 
 Trees, argument from 40 f 
 Tree-names 27, 33, 40 f 
 Twelfth Night 48 
 Tylor, Prof. 62 
 
 Ulfilas see Wulfila. 
 Umbrian 33 
 
 Yedic Sanskrit 6, 13, 51 f 
 
 Yerner's Law 13 
 
 Vocabulary of primitive speech 
 
 3i ff 
 
 Vowel-gradation 24 f 
 Vowels, Indogermanic 13 
 
 Wergild 49 
 
 Wesley 28 
 
 West, A. S. 59 
 
 Whitney, W. D. 56 
 
 Winter 40 
 
 Women, condition of 48 
 
 Wright, Prof. 58, 60 
 
 Writing 58 
 
 Wulfila 33 
 
 Year, length of 47 
 
 Zend 40 
 Zeus 37, 50 ff 
 
 of the Oak 43, 54 
 Zoology 30 
 Zoroaster 5 t 
 Zulus 2i
 
 hNDEX II. 
 
 In this word-index the words in italics are those quoted as reprt 
 senting Tndogermanic originals. 
 
 accent 21 daughter 32, 34, 48 
 
 Achilles (dr.) 36 deism 16 
 
 Agilulfs (old Ger. ) 36 despot 46 
 
 ai (Fr.) 25 dcits (Lat.) 16, 51 
 
 aime (Fr.) 25 Digentia (Lat.) 20 
 
 alnus (I. at.) 27 ttirimts (Lat.) 51 
 
 amo (Lat.) 25 dog 32 
 
 Aurora (Lat.) 3 downs (Lat.) 46 
 
 avons (Fr.) 25 dough 18 
 
 drownded 28 
 Dvaus (Skt.) 50 f 
 
 ear 47 
 Faster 3 
 either 27 
 eleven 39 
 t y?<w (Lat.) 34 
 Frin (Ir.) 7 
 -etum (Lat.) 27 
 
 feet 25, 28 
 female 27 
 
 fir 4i 
 
 call 8 fish 32, 43 
 
 cook 46 fire 14, 27 
 
 corn 46 flavos (Lat.) 27 
 
 cover i 6 Hay 39 
 
 cow 46 ><?/ 25, 33
 
 68 Index II 
 
 four 27 macadamise 16 
 
 frater (Lat.) 8 male 27 
 
 fraxinus (Lat.) 41 mare (Lat.) 41 
 
 mead 42 
 
 Cast (Ger.) 17 Medise 26 
 
 Gerste (Ger.) 46 /- 41 f 
 
 7^.r/ 17 w/7/ 47 
 
 month 47 
 
 health 27 TW//W 32 f, 4 S 
 helvos (Lat.) 27 
 
 hiemf>s (Lat.) 40 neither 27 
 
 ///wtf(laya) (Skt.) 40 nescius (Lat.) 3 
 
 hither 27 nice 3 
 
 hard (Gr.) 47 niclit (Ger.) 14 
 
 hound 32 north, northern 27 
 hundred 33 
 
 odor (Lat.) 19 
 
 ia.sc (Ir.) 43 oI5a, o'iSafJ.fv (Gr.) 25 
 
 zV* 40 olere (Lat.) 19 f 
 
 ingens (Lat.) 3 onto(logy) 3 
 
 iron 44 oven 46 
 
 -ise 26 
 
 . . , n paraholo (Lat.) 2=; 
 
 isi (Zend) 40 
 
 ir- \ paracus (Skt. ) 39 
 
 tff/j.ei> ((^r.) 25 , ,1- \ 
 
 . parole, parlons (Fr.) 25 
 -ize, see -ise 
 
 parncMti 49 
 
 7^ (Lat.) 50 A^'' 49 
 
 TrAe/c^s (Gr.) 39 
 
 KaX (Gr.) 9 perhaps, p'raps 20 
 
 kapliar (Ilcb.) 16 Perqunos (In<lo R .) 52 
 
 knave 3 pfund ( Ger -) '9 
 
 (frparrip (Gr.) 8 
 
 larch 19 pilaqqu (Assyr.) 39 
 
 larix (Lat.) 19 pinus (Lat.) 27 
 
 lender (M. L.) 25 jiiscis (Lat.) 43 
 
 Licenza (Ital.) 20 j>lecto (Lat.) 39 
 
 lingua 19 f plough 18 
 
 loaf 47 pondus (Lat.) 19 
 
 loch 14 praedico (Lat.) 19 
 
 long 25 preach 19, 24 
 
 Lough 1 8 predige (Ger.) 19
 
 Index II 69 
 
 (pre)scnt 3 taught 28 
 
 teached 24 
 
 (jnercus (Lat.) 41 tear 26 
 
 <jnei-n 47 theism 16 
 
 0eos (Or.) 10", 51, 53 
 
 raught 24, 28 thither 27 
 
 reach 24 timber 46 
 
 m/ 44 tongue 19 
 
 right 20 tore 26 
 
 righteous 20 /ra; 20, 33, 41 
 
 right-you-are 20 T"ues(day) 50 f 
 
 twelve 39 
 
 sang 25, 27 f two 33 
 
 scrope 26 uncouth 3 
 
 sew 47 
 
 silly 3 venous (Fr.) 25 
 
 sister 32 48 7'er (Lat.) 40 
 
 sit 25, 27 viens (Fr. ) 25 
 
 stiinu 40 
 
 -snus (Lat.) 27 wagon 46 
 
 .sv 32, 48 -warm 17 
 
 sooth 3 weal, wealth 27 
 
 sore 24 wear 26 
 
 sorrow 24 weave 47 
 
 sorry : 4 uvd 49 
 
 sounded 28 widow 49 
 
 south, southern 27 wit, witen 25 
 
 sow 47 wore 26 
 
 .r/tvr 46 -wos (Lat. and Prim. Ger.) 27 
 
 Strasse (Ger.) 19 wot 25 
 
 strata (Lat.) 18 writ, wrote 25 
 
 street 19 
 
 sung 25, 27 f _ymr 47 
 
 suppose, s'pose 20 yellow 27, 44 
 
 suttee (Skt.) 3 
 
 Zeus (Gr.) 50 f 
 
 // 41 '/.immer (Ger.) 46 
 
 CA.MBKIlJC.I-:: I'KINTKI) BY J. AND C. F. CLAY, AT THE UNIVERSITY i'KESS
 
 A 001 428153 9