What We Learn From Old Aryan Words

THE discovery of the Aryan family of languages, as elucidated in the paper printed in The Atlantic for February, was the first and most conspicuous consequence of the zeal for Sanskrit studies which ensued upon the English conquest of India. Surely, this in itself was no small thing. It was in every way stimulating and suggestive to have detected a specific bond of relationship, in speech and in culture, between such different peoples as the English and the Hindus, who had not previously been suspected of possessing anything in common save their common humanity. It had indeed been long ago maintained that languages the most diverse in superficial aspect were descended from a common source, but such views were based merely on a languid assent to an ill-understood tradition, and no one had the least conception of the proper method of tracing linguistic affinity. Down to the beginning of the present century the labors of etymologists had all the crudeness of astrological speculations, or of barbarian theories of the universe. And no wonder, since attention had been chiefly directed toward Hebrew, a language entirely unrelated to those of Europe, so that any attempt to explain the latter by a reference to the former could end only in mental confusion. It was a very striking discovery that was made when it was proved that though no likeness whatever exists between the European tongues and Hebrew, yet the closest similarity is manifest between these tongues and a much more remote Asiatic language. The completion of this discovery was no less striking when it was shown that while linguistic relationship can be clearly traced, according to fixed rules of inference, among all the various members of the Indo-European group, yet the moment we step outside of this group we can neither detect relationships nor establish rules of inference, but have before us a new set of facts, quite incongruous with the old ones. Such a contrast was just what was needed in order to indicate what the true signs of linguistic relationship are, and thus our whole mental horizon was shifted, as far as concerns the study of language. In the act of establishing the existence of our own great family of speech, scientific methods of comparison were gradually worked out, and the results of this have been far-reaching enough.

In the present and following papers I propose briefly to notice three departments of study which have been actually created by the comparative investigation of Aryan languages. Under the first head I shall call attention to some characteristics of scientific etymology; under the second, we shall get a glimpse of the prehistoric culture of the Aryans ; under the third, we shall make a beginning in the study of myths.

First, as regards etymology, we need consider only a few facts which show how systematic and orderly inference has been substituted for what once was mere random guess-work. In comparing different languages, similarity and dissimilarity are still, as formerly, the principal tests of relationship ; but in applying these tests we are strictly limited by rules which formerly were ignored. Once a vague resemblance between the vocabularies of two languages was considered sufficient ground for assigning them to the same class; now even a close and sustained likeness in vocabulary is not enough, unless it be accompanied by likeness in grammatical forms. Thus, the possession of innumerable Latin words, such as opinion, reflect, admire, umbrella, honor, color, contemplate, criminal, etc., does not make English a language of the Italic class, nor does it even show any original kinship between English and Latin. Such words have simply been adopted from Latin, just as ennui and naïveté have been adopted from modern French, and such borrowing and lending as this can go on between any two languages. It is just as easy for us to use Arabic words like alcohol and cipher as if Arabic were a kindred language. Nearly half the vocabulary of modern Persian has in this way come to be made up of Arabic words, yet there is no kinship whatever between Persian and Arabic. But while mere vocabulary does not determine the place of a language, the peculiar style of making sentences does determine it. Though more than half the words we use are Latin, English is not an Italic language, because we cannot make a single sentence out of Latin materials alone. English, on the other hand, is a Teutonic language, because we cannot make a single sentence without introducing some Teutonic shibboleth. Suppose we say, “ Atheism desecrates deity: ” here we seem to have simply one Greek word followed by two Latin words; but the Teutonic shibboleth comes out in the terminal s of “ desecrates,” which is the peculiar shape in which English has retained the old Teutonic verb-ending th, as it would appear in “ desecrateth.” Again, if I say, “ I can go to Boston,” my phraseology is purely Teutonic; but if, like Dr. Johnson, I have a weakness for big words, and say, “It is possible for this individual to traverse the vast area intervening between this locality and Boston,” I have not yet escaped the boundary of Teutonic speech: for although I have introduced seven Latin words of secondary importance, yet the little words which enable me to knit the sentence together are still Teutonic, as before. So when we say, “ I have, thou havest or hast, he haveth, hath, or has,” the Teutonic shibboleth comes out in this style of inflection. In short, it is easy enough for us to acquire new words, but we cannot abandon our habits of sentence-making without giving up our language altogether. Now the demonstrated community of the Aryan languages rests not merely on their possession of a common vocabulary, but on their retention, in various degrees, of grammatical forms originally common to all. We can hardly find a better instance than in the conjugation of the verb just alluded to :1


Gothic, haba, habai-s, habai-th; haba-m, habai-th, haba-nd.

Pers. -m, -d; -m. -d, -nd.

Kelt. -m, -d; -m, -d, -t.

Lith. -mi, -si, -ti; -me, -m, -te, -d, -ti.

Slav. -mi, -si, -ti; -mu, -te, -nti.

Lat. habeo habe-s habe-t; habe-mus, habe-tis, habe-nt

Gr. -mi, -si, -ti; -mes, -te, -nti.

Skr. -si, -ti; -masi, -tha, -nti.

Community of vocabulary is, however, a very important matter, when rightly considered. It is true that any language may borrow a large proportion of its words from an entirely alien source, as Persian has borrowed from Arabic. But in comparing the various forms of Aryan speech we have found a criterion which enables us to distinguish

between words that are alike in two languages because one has borrowed them from the other, and words that are alike because they are simply modified forms of the same aboriginal word. This supremely important point can be here treated but roughly ; yet I hope that, with a few illustrations, it may be rendered intelligible. vergence of a language, originally uniform, into two or more distinct dialects is to be found in those differences of pronunciation which arise, one hardly knows how, in different localities. The most curious feature of these differences is that they are often so extremely systematic. Every one has heard of the Englishman who inquired, “ If a haitch and a ho and a har and a hess and a he don’t spell ’orse, what the deuce does it spell, you know ? ” The infallible accuracy with which the cockney omitted his h where it belonged, and supplied it where it did not belong, used to excite my wondering admiration when I was in London. Had there been any caprice in the usage, it would have seemed less marvelous. But so unerring is the instinct that when a friend of mine once purposely spelled his name out as U-t-t-o-n he was correctly announced by the waiter as Mr. HUTTON ! Is not this what our High German friends, with equal felicity, and in illustration of the same point, would call a very eggsdraortinary zirgumsdance ? Yet after all, so far from being extraordinary, such phenomena occur so regularly in a comparison of the Aryan languages that they have been reduced to a systematic form of expression in what is known as “ Grimm’s law.” Take, for example, the word " father.” This is the same in all the Aryan languages, save for the differences in pronunciation which make the German say rater, while in Latin, Greek and Sanskrit, we have pater. On the other hand, brother, in German bruder, appears in Latin and Sanskrit as frater or bhratar, in Greek as øραης, the member of a brotherhood or fraternity. That is, where we pronounce an f the Greeks, Romans, and Hindus pronounced a p, but where we pronounce a b they pronounced an f, or something like it. Similarly, where we say gard-en the Greek said χоρτς and the Latin hort-us ; and our goose, which appears more fully in the German gans,

One of the chief reasons for the di-

is found in Greek as χηυ, in Sanskrit as hansa, in Bohemian as hus, the name of the celebrated martyr. But conversely, where we say heart the Greek said καρϐ and the old Roman cord, and where the German says haupt the Roman said caput. That is, a Teutonic g answers to a Greek, Latin, Sanskrit, or Slavonic h, but a Teutonic h answers to a K in the latter languages. Now this group of facts is not precisely analogous to the cockney’s transposition of his aspirates, but it is certainly very similar, and it is equally mysterious. Why this curious alteration of sounds should have occurred so systematically, and on so great a scale, no one has ever succeeded in explaining. It is none the less to the purpose, however, that it has occurred. Although an empirical rule, Grimm’s law is nevertheless a well-established rule, and in the study of Aryan etymology it has to be taken into account at every step. It is easy to see what a revolution the establishment of this law has worked in our methods of comparing words. Formerly the etymologist looked, though in a vague, intermediate way, for mere resemblances; and this was natural enough. But now a too strict resemblance sometimes becomes a suspicious circumstance. The Greek word for “ whole ” is ολος, and what could he more plausible than to suppose it identical with the English word ? But here Grimm’s law makes us suspicious. We ought not to expect a Greek to pronounce “ whole ” like an Englishman, any more than we ought to expect to hear a cockney say “ horse.” What the cockney says is " orse,” and what the Greek would naturally say is not ολος, but κóλος; and in point of fact it has been otherwise proved that our suspicion is here well grounded, — the resemblance between the English and Greek words is purely accidental. Mere resemblance is thus a very treacherous guide in etymology. In French we have louer, “ to hire,” and louer, “ to praise.” Some philological dreamer tried to show that these words might be connected, because you praise your lodgings or horses when you wish to induce some one to hire them! In fact, the one word has been clipped down from Latin locare, “ to hire,” and the other from Latin laudare, “ to praise.” In striking contrast to this, let us observe how two English words, pen and feather, are closely connected in origin, in spite of their entire dissimilarity. There was an Old Aryan verb pat, “to fly,” which still appears in the Greek πϵτομαι. There were also such suffixes as tra and na, denoting the instrument with which an act is accomplished. Pat-tra thus meant " a wing,” and a Hindu might perhaps thus understand it; but in Gothic we find fath-thra, and in English feather, just as Grimm’s law has taught us to expect. Put-na had the same meaning, and passed into old Latin as pes-na, which later Latin clipped down to penna, a wing or feather, and finally the quillfeather with which you write. In these days we have applied the word to little implements of gold or steel which have nothing to do with flying, unless the soaring of Pegasus he supposed to keep up the association of ideas.

This example of pen and feather is a trite one, but I have cited it because it further illustrates a very important point, toward which my argument has been for some time tending. Looking at these two words, with reference to the whole extant Aryan vocabulary, we find that their very forms disclose their past, history. We see that the word feather, which has undergone the change of pronunciation indicated in Grimm’s law, in common with Teutonic words in general, is a genuine Teutonic word, and appears in the English language today because it has always belonged to English speech. But the word pen, which has not undergone this change, shows thus on its very face that it has not grown up in company with Teutonic

words, but has been adopted at a recent date from another branch of the Aryan family. The changes formulated in Grimm’s law took place in early times, long before people had begun to think critically about their pronunciation or their diction. When we adopt Latin words in modern times, we do not refashion them in accordance with the twisted pronunciation of our barbaric ancestors, but we take them as they are. From pater we take paternal, without trying to make it sound like its equivalent, fatherly. Thus we arrive at a safe criterion for distinguishing between words which have been passed about from one Aryan language to another, in the course of recent intercommunication of culture, and words which have descended, with divers modifications, from a common original. Words of the latter sort, where they exist in different classes of Aryan speech, have obviously been handed down from primeval times; they must have formed part of the vocabulary employed in Aryana Vaëjo, and the most convincing proof of their genuineness is to be found in the peculiar nature of the wear and tear they have undergone. To recur to an example previously cited, the existence of such English words as color, opinion, admire, etc., not only fails to prove kinship between English and Latin, but it does not even prove that English is an Aryan, language, since these words are manifest importations, and the case of Persian and Arabic shows that nothing is easier than for one language to adopt half its current words from another that has no relationship with it. But on the other hand, when we compare such words as corn with Lat. granum ; horn with Lat. cornu; who and what with Lat. quis and quid, Skr. kas and kad; queen with Gr. yovy; beech with Lat. fagus ; doom with Gr. ϐεις; tear with Skr. dar ; bear with Skr. bhar, Gr. and Lat. fero; tooth (Goth, tunthus) with Zend and Skr. dant, Lat. dens, — when we find a thousand such cases of systematic divergence, we get clear proof of the original identity of the English vocabulary with the others brought into the comparison. For the divergences in themselves are incompatible with any theory of modern borrowing and lending, while the extreme regularity of their recurrence is explicable only as the result of common processes operating on common materials.

The symmetry of consonant-changes throughout the Aryan languages is at first sight a wonderful phenomenon, and the tracing of correlated words in accordance with such laws as Grimm’s never ceases to be a fascinating study. The laws of vowel-change — whereby, for example, the Skr. matar corresponds to Lat.mater, Gr. μητηρ Gaelic math air, Germ, mutter, and Eng. mother — are hardly less interesting. But to do justice to such a subject as etymology would require much more time than we have at our disposal. In the present paper I have not attempted to make anything like a full statement even of Grimm’s law, but have given only such scanty illustrations as may serve to render the outline of my argument intelligible. That other sound generalizations have been made concerning phonetic change I must ask the reader to admit, while I go on to point out one of the largest of the results that have come from this minute study of consonants and vowels. From this minute study the laws of the permutation of words have been wrought into such a complete and harmonious system that it has become possible to reconstruct large portions of the common Aryan mothertongue by comparing together the curiously modified forms of its modern descendants. The problem is quite similar to what it would be if classical Latin were extinct, and we were required to reproduce as much as possible of it from an elaborate comparison of the vocabularies and grammatical forms of French, Spanish, Italian, and their allied modern dialects. Such a task would no doubt be delicate and difficult; but there is also no doubt that a great deæ of good Latin could be reconstructed in this way. The restoration of the Aryan mother-tongue seems at first sight a still more formidable task ; but it is a task for which we have also more abundant materials in the wider variation among Aryan words as compared with Romanic words. Thus by a comparison of French mois with Span, mes and Ital. mese, knowing besides the general habits of the Romanic languages, we might probably infer the Lat. mensis as the common original of the three ; but on looking over the whole Aryan field, and comparing Lat. mensis with English month, Gr. μηυ. Lith. menesis, O. H. G. manat., and Skr. masa, we arrive with even stronger probability at the Old Aryan mansa as the only form which could have given rise to all these. This work of reconstructing the Old Aryan has gone on so vigorously of late years that we have now a dictionary of it in three stout octavos ; 2 and one great philologist, August Schleicher, has gone so far as to write a short story in this prehistoric language. Very likely a primitive Aryan, if we could resuscitate him. would have some severe criticisms to make upon this attempt. In so bold an undertaking, frequent and extensive revision will no doubt be called for. But the mere possibility of making such an attempt shows forcibly how thoroughly scientific the study of etymology has become.

In no respect do these philological inquiries appear more interesting than in the light which they throw upon the prehistoric civilization of our Aryanspeaking forefathers. No historic record, not even a vague tradition, is preserved of the time when the ancestors of Kelt, Greek, Roman, and Teuton cdwelt in a single community with the ancestors of Persian and Hindu. We have no clue even to the date of this epoch of common Aryanism, though we may very fairly allow for it at least three or four thousand years before the Christian era. Even the oldest Aryan legends, as those of the Vendidad, preserve only a dim reference to a time when the Indo-Persian branch of the family had not yet become divided. Yet concerning the degree of culture reached in those remote times, so far antedating all conscious historic tradition, the unconscious record of language has given us some trustworthy information. From the seemingly dry study of consonants and vowels an easy process of inference opens up to us, as with a magician’s wand, a fascinating picture of the life and pursuits and habits of thought of the people from whose longperished form of speech our vowels and consonants are derived.

Wonderful as this may seem, what is simpler, when we have once ascertained that a certain word belonged to the Old Aryan language, than the inference that the word was used to describe some object or express some thought? And where the meaning of the word has remained uniform throughout all the vicissitudes of pronunciation and inflection to which it has been subjected, what better guarantee do we need that the word was used with the same meaning in the mother-tongue ? It requires no extraordinary insight, when one has mastered the rules of comparative grammar, to see that the primitive Aryan called his nearest relatives by the names patar, matar, bhratar, svasar, sunu, and dhugatar ; or that when he learned to count up to ten he said something like aina, dva, tri, katvar, paukan, ksvaks, saptan, aktan, navan, dakan.

Proceeding in this way, we find abundant evidence that the early Aryans had outgrown the nomad stage of civiliza-

tion and acquired settled habitations, not merely in villages, but even in fortified towns. The Lat. domus reappears, with hardly any change, in Gr. ϐóμος, Skr. dama, Armen. dohm, Irish daimh, and Russ. domu, always with the meaning of house.” In the Teutonic class we do not find this word in precisely the same sense ; but we have the Germ, zimmer, “a room,” connected with Goth. timrjan, “ to build,” and Eng. timber, or building material; ” and these words, compared with Gr. ϐεμϵιυ, carry us back to Old Aryan dam, “ to build,” so that the domus of our forefathers was not a mere hole in the rocks, but a dwellingplace put together by the arts of the carpenter.3 In Greek the more common word for house is οικος, originally Ϝοικος, “a place that one goes into.” This word runs through all the Aryan languages, but the original sense of “ entering” is forgotten, and it only means “ a place where one lives,” — sometimes a house, but more generally a village. Thus we have Skr. veça, Zend viç, Russ, vest and Polish ivies, Lat. vicus (whence the diminutive vicula, villa, village), Irish felt, Kymric gwic, Goth. weihs, Eng. wick. The Old Norse language shows a curious deviation4* from this general agreement in meaning ; for whereas the word generally describes an abode on the land, to the sea-roving Norseman a wick was a creek or sheltered bay serving as a station for ships, and hence their famous name of Vikings or “ men of the fjord.” So, ending the ending wick or wick is very common in Old English names of inland towns, it occurs frequently also on the British coasts in the Norse sense, as inSandwch and Berwick, favorite stations for pirates, But with this characteristic divergence, the generally uniform significance of the word, in languages so widely scattered, points clearly to the existence of village communities among the prehistoric Aryans. The of the English word town are equally instructive, though not quite so numerous. The Old English form tun has its counterpart in Old German zun, “ an inclosed or fortified place,” with which the modern German zaun, “ a hedge,” is connected. Now, in accordance with Grimm’s law, we find Armenian dun, “ a house,” Kymric din, “ a fortress,” Irish dun, a “ fortress ” or “ camp ” or “ walled town.” This Keltic form appears in many geographical names, such as Thun, in Switzerland; Lug-dun-um on the Rhone, now Lyons; Lug-dunum in Holland, Leyden ; Dun-Keld, the “fort of the Kelts;” Durn-barton, the “fort of the Britons ; ” Dundee, London, Clarendon, etc. In the remote Himalayas the same word reoccurs in the names of hill fortresses, such as Kjarda Dhun, Delira Dhun, etc. ; and again it is a fair inference that where a word turns up in so many parts of the Aryan domain with the very same determinations of meaning, it must have belonged to the primitive vocabulary of the race. So that our forefathers would appear to have been acquainted not only with houses and villages, but also with some kind of walled towns.

The name of the rampart with which such fortified inclosures were surrounded was also contained in the Old Aryan vocabulary. From the old root val or car, to “ protect ” or “ surround,” we have Skr. varana, Old Germ, wari, Pol. warownia, Lat. vallum, Lith. wolas, Irish fal, Kymric gwal, Eng. wall. The partition wall of a house, on the other hand, is more properly described by a root which in Sanskrit seems to be applied to wicker-work, but which in the European tongues appears, with hardly any variation either in sound or sense, as Lat. mums, Lith. muras, Old Germ. mura, Modern Germ, mauer, Irish, Kymric, Anglo-Sax., and Pol. mur. The name for “roof ” is similarly ubiquitous : in Skr. we have sthag, “ to cover,” in Lith. stogas, “ a roof,” in Gr. στϵΥυς, a “ roof ” or “ house,” and στϵΥ, “ to cover ; ” but the word appears about as often in Greek as τϵΥος, with the initial letter dropped; and so in Irish we find teg, “ a house,” in Lat. tego and tectum, in Old Eng. thccan, in Eng. deck and thatch. In door there has been even less variation than this: Skr. has dear, and also dur in the Vedas ; Zend dvara, Pers. dar, Gr. Ovpa, O. H. G. tura, Goth. daur, Old Eng. duru, Irish and Welsh dor ; the Lithuanian has lost tlxe singular, but retains the plural durrys for folding-doors. The word meant originally “that which obstructs or keeps out.” Another old name for the door, which appears in Skr. as arara, has been preserved in Europe only in the Irish or air, a “porch” or “vestibule,” and Welsh oriel. This latter is one of the very few Keltic words to be found in English, where it has become the name of a kind of bay-window.

Among the Aryan words for “window ” there is no such identity, though there is a most curious similarity in the metaphors by which they have been constructed. In Sanskrit the window is grhaksha, or “ the eye of the house,” and a big round window is called gavaksha, a compound of gau, “ cow,” and aksha, “eye,” which is about equivalent to our expression “ bull’s-eye.” The Slavonic languages have ohm, from oko, “an eye,” while Gothic lias augadauro and O. H. G. augatora, or “ eye-door.” The meaning of our English word is not so immediately apparent, but in one of our nearest relatives, the Danish, it occurs as I'indue, and in Old Norse this was vindauga, that is, “an eye or hole for the wind to blow through.” These coincidences are interesting as showing how easily and naturally the same association of ideas may occur to different people, for these words have been independently formed. Whether we are entitled to infer from this that the Aryan mother-tongue had no word for window, and that therefore the people who spoke it lighted and aired their houses only through the door-way, it is not easy to decide. Such an inference might seem probable ; but here, as elsewhere, it is very unsafe to rest a conclusion upon negative evidence. The Old Aryans certainly might have had a name for window which among various tribes came to be supplanted by various other expressions, and this possibility must prevent our holding a positive opinion. We can only say that, while we are perfectly sure that they had doors, it is quite uncertain, so far as philology goes, whether they had windows or not. And in general, while the occurrence of the same indigenous name for any object, throughout the different classes of IndoEuropean speech, is sufficient proof that the primitive Aryans knew and named the object, on the other hand, the nonexistence of such a common name raises only a negative presumption, which we have seldom any further means for testing.

The ancient Aryan gained a livelihood chiefly from rearing cattle and tilling the ground. The names of our principal domestic animals are found in all parts of the Indo-European territory. The various Teutonic forms cow, ku, chuo, reappear with the proper change of guttural in Lettish gôws, Pers. gâw, Armen. gov, Zend gao and gava, Skr. gaus, gava, and gu. A peculiar twist, by which a labial was pronounced instead of an original guttural, may be observed quite frequently in the GnecoRoman and Keltic languages, and here we have Gr. βους, Lat. bos, Irish bo, and Welsh bu. The meaning of the word has been variously explained, but, as we have beside it the Skr. gu, Gr. yodio and βοáω, Lat. boan, to “ bellow,” it is most likely an imitative sound, like our moo and mooley. In the dialect of the Vedas a bull is called υáksha, in later Skr. and Zend uksha; in Gothic this appears as auhsa, and in Anglo-Sax. as o.ra, whence Dur ox. Sthira, again, is a Skr. name

for bull, meaning the “ powerful ” animal. In Zend çtaora means a strong beast of burden : in English we have kept the full word steer, but the initial s has generally been dropped, so that we have Dan. tyr, Gr. and Lat. taunts, Russ, turn, Irish tor. The word bull itself is descriptive of the strength of the animal, and appears in Skr. balin, Irish bulan, Lith. bullus, and in many other languages. There are a great, many other Aryan names for these animals, but without spending time on them we may note that several of the words just cited have been borrowed by nonAryan languages, such as those of the Finno-Tataric class, and even the Japanese and Chinese; from which it would seem probable either that the primitive Aryans were the first to domesticate cattle, or at least that they were very preëminent as a pastoral race, and furnished to their neighbors great numbers of these most useful animals. The prominence of the cow in early Aryan thought is shown both by the multitude of synonyms for the creature, and by the frequency of similes, metaphors, and myths in the Vedic hymns in which the cow plays a part. In those days, moreover, which were before the days of “soft” or “ hard ” money, wealth was reckoned in cows, and cows were the circulating medium, with sheep and pigs for small change. Every one knows that Lat .pecunia is derived from pecits, “a herd ; ” the same is true of peculium, “a man’s private property,”from which we have obtained peculiarity, or “ that which especially pertains to an individual.” Pecus, Lith. pekus, Skr. and Zend payu, “ the animal that is tied or penned up,” reappears with the regular change in Goth. faihu, Old Eng. feoh, modern Germ. Vieh; in modern English the word has become fee, a “ pecuniary reward.” In Irish we have bosluaiged, “ riches,” from bosluag, “ a herd of cows.” When you go to a tavern to dine you pay your shot or scot before leaving ; or you sometimes, perhaps, get into a very ticklish situation, and still escape scot-free. In Old Eng. sceat was “ money,” and the Old Norse skattr and Goth. skatts had the same meaning; but the Irish scath means “a herd,” and Old Bulgarian skotu was one of the many Aryan words for cow. Another of these words, in Skr., is rupa, whence are derived rûpya, " money,” and the modern rupee of Bengal.

More than a hundred different names for the horse have been counted in Sanskrit, but most of these are comparatively modern in origin. The only one we need notice is açva, from an Old Aryan akva, meaning “ the swift.” In Lith. the same word aszwa is the name of the mare only, but the Lat. equus preserves the old meaning. The classic Greek ιππος does not sound so much like equus as one might expect, but we find the requisite transitions in the Aiolic ικκος and Old Aiolic ικϜος. In Irish nothing is left but the first syllable, ech. In Gothic the word reappears quite regularly as aihva, and in Old Eng. this is clipped down into eoh. Modern English, however, and the other modern Teutonic languages have lost this word and replaced it by another, which goes back to the times of Teutonic unity, but does not seem to have been known to the primitive Aryans. The Old High Germans and the Norsemen pronounced this word hross, but the Anglo-Saxons called it horse, and the modern Dutch, like the cockneys, call it ors. It appears to he a mimetic word, meaning the “ animal which neighs.” Modern High German, in turn, though it has not lost the word ross, has adopted a new name, pferd, which is in more frequent use, and the history of which is extremely curious.

One of the few Keltic words which the Roman conquerors adopted from their Gaulish subjects was the word rheda, used to describe a light fourwheeled carriage. Such carriages were used for posting, and the light, swift animal which drew them received a special name, made by compounding the root of veho, to “ draw ” or " carry,” with the name of this kind of carriage. Thus arose the word veredus, the “ drawer of the rheda,” the post-horse, or courier’s horse ; and so veredarius was a postclassic Latin word for “ courier ; ” but the name veredus was not long in becoming generalized, for in Martial wo find it used for a light, fleet hunting horse. At the same time there came into general use the curiously hybrid word paraveredus, made by prefixing the Greek preposition mpa, meaning “ beyond,” to veredus, to denote an extra post-horse for extraordinary occasions. This mongrel word paraveredus, thus oddly made up out of Greek, Latin, and Keltic elements, seems to have been a favorite name for the horse in the Middle Ages. In Ducange’s great dictionary of mediæval Latin we find parvaredus, parafredus, and palafredus, along with many other forms. From palafredus came the French palefroi and the English palfrey ; while the simple contraction and abbreviation of the older paraveredus resulted in the form pferd adopted by the modern German.

As the Teutonic languages have thus adopted new words to designate the horse, so the modern Romanic languages have generally forgotten equus and substituted for it the name which appears in French as cheval and in Italian as caballo, and from which we have obtained such words as cavalry, chevalier, and chivalry. Ancient Greek and Latin both had this word caballus, which, as kobyla, is the common name for a horse in the Slavonic languages, and appears also in Irish as capall and in Welsh as eeffyl. We do not find any such name in Sanskrit, but in the Kawi of the island of Java, which is a non-Aryan Malay language, as full of Sanskrit words as English is of Latin words, we find the horse called capala, and side by side with this we have in Sanskrit the adjective çapala, “swift.” The Sanskrit quite generally corrupted Old Aryan sounds in this way, as we corrupt Latin sounds in English when we say serebrum and Sisero instead of Keerebrum and Kiker o ; and I have no doubt that in this word for “ swift ” we have the explanation of caballus. Curiously enough, the modern Greek has also dropped the classical name for the fleet-footed beast, and substituted αλοΥου, which means “unreasoning,” and in former times was applied to brutes in general. It is quite remarkable that there should have been such vicissitudes in the career of the words which describe so familiar an animal, and we need no better illustration to convince us of the danger, above pointed out, of relying too confidently upon negative evidence in such inquiries as we are here making. Looking at the contemporary names only, we find the principal Low German language saying horse, “ the neigher,” while High German, French, and Greek say pferd, the extra drawer of a post-carriage,” cheval, “ the swift creature,” and dAoyov, “ the brute,” — names quite distinct both in sound and in meaning. If all the other forms had been lost and replaced by new words, — as might easily be the case where there are so many synonyms for the same object, —we might perhaps have inferred that there was no common Aryan name for the horse, and that hence the animal was not known until after the separation of Aryan tribes had begun ; but this would have been very plainly a mistake.

Besides the horse and cow, the primitive Aryans had domesticated sheep, goats, and pigs, as well as dogs. With regard to the cat, the case is less clear. That wild species of the cat family were known seems probable, and the word puss has some claim to an Old Aryan pedigree, lor we find pushak in modern Persian, puizé in Lithuanian, pusag and puss in Irish, whence we have adopted the word; but whether the primitive form of these names was applied to a wild or to a domesticated cat is uncertain. With this exception, the IndoEuropean names are all different. In Latin we have felis, in Greek αιλουζος; but we know otherwise that the Greeks and Romans had no domestic cats, but kept a kind of weasel to destroy their rats and mice. In our own and most other modern European languages the principal name of the animal is borrowed from Latin ; but the Latin cattus is itself an imported word from a nonAryan source. It is the Syriac Kato, Arabic Kitt, indicating that the cat was introduced into Europe from the Levant, at a comparatively recent period.

But whether the Old Aryans had domestic cats or not, they certainly needed them, for the word mouse occurs, with hardly any variation, in nearly all the Indo-European languages. In Latin, Greek, Old Norse, Old German, and Old English it is mus ; in Russian we have myshi, in Bohemian mysh, in Persian mush, in Sanskrit musha, the “ pilfering creature,” the “ little thief.”

Flies are also to be numbered among the household pests of Aryana Vaëjo; the old name was makshi, the “ buzzing creature,” and is preserved in Zend and the modern Indian languages. In Europe we have Lith. musse, Bohem. mussha, Lat.musca, O. H. G. muccha, Swed. and Old Eng. mygge, Eng. midge, of which the diminutive midget, or “ little fly,” has been applied as a caressing epithet to children. The meaning of the more common Teutonic name “ fly ” is too obvious to require mention.

The ordinary Aryan name for “ bee ” — Skr. bha, O. H. G. pia, Old Eng. beo, Eng. bee — refers to the bright color of the insect, but the Lat. apis is the “ thrifty creature ” and the Greek μϵλισσa is the “ maker of honey.” The Old Aryans not only kept bees for their honey, but out of the honey they made an intoxicating drink called madhu, from which we have the Zend madhu and Greek μϵϐυ, “ wine,” Russ, medic, Irish meadh, Old Eng. medn, Eng. mead. Wine and must are Old Aryan words, and the same is probably true of ale : but in this latter instance we cannot safely infer that what we call ale was brewed, for the meaning of the word has varied considerably. Lith, alus, Old Norse öl, Old Eng. eala, mean “ beer,” but the Skr. ali means a spirituous liquor, and the Irish ol is applied to any kind of drink. As for the word beer itself, it is doubtful if it can be traced outside of the Teutonic languages ; for although it occurs in Irish, Welsh, and modern Persian, it does not conform to Grimm’s law, and has thus most likely been borrowed from English or some other Teutonic source.

Whether our Aryan forefathers brewed ale or not, they certainly cultivated barley arid probably wheat, and ground them into meal in mills. They were familiar with the plow, the yoke, and the spade. Their harvests were reaped with the sickle, and the grain was duly threshed and winnowed, and carried to mill in wagons fitted with wheels and axle-trees. The blacksmith’s work with hammer and anvil, forge and bellows, was also carried on. Sewing and spinning were feminine occupations, and garments were woven out of sheep’s wool. The art of tanning was also practiced, and leather shoes were worn. The entire career of the Aryans has been that of a warlike people. In the primitive times of which we are treating their principal weapons were the lance, the bow and arrow, the sword and dagger and mace, with helmet and buckler for defense.

That the early Aryans were acquainted with the sea seems unquestionable, for the name occurs, with very little change in sound and hardly any in meaning, in nearly all the Indo-European languages. The Lat. mare, whence our adjective marine, appears in Skr. mira, Russ. moru, Lith. mares, Irish muir, Welsh mor, Goth, marei, O. H. G.mari, Old Norse mar, Old Eng. mere. In English meer is an archaic word, still used in poetry in the sense of “ lake,” and it appears in many well-known names of English lakes, as Grasmere and Windermere. The original sense of the word has something poetic in it, for it means the barren, desolate waste, just as we find it commonly described in Homer. The Teutonic languages, however, have generally adopted another name. In Skr. sava means simply “ water,” but the more specific sense appears in Goth, saivs, O. H. G. seu, Old Eng. sewe, Eng. sea. It is noticeable that while modern English applies this name to great bodies of water, and keeps meer only in the sense of lake, in modern German the case is just the reverse, — in German meer is the sea, but see is a lake. The only other conspicuous deviation from the general Aryan usage is a very characteristic one. The Greeks, who were the most maritime of all peoples that have existed, save the English, had three names for the sea, of which the later ϐáλaσσa and πϵλaος referred to the boisterous, white-crested waves, but the earlier πουτος meant a “ pathway for travel.” What large bodies of water the primitive Aryans could have known is perhaps not fully ascertained, but they were most likely the Caspian and the Sea of Aral. On these inland seas, or along the great rivers which flowed through their country, the Aryans would seem to have plied in boats rowed with oars ; but whether they had advanced farther than this is uncertain. At all events, there is a singular lack of agreement among all the common words indicative of a higher acquaintance with the art of navigation.

With these illustrations we must bring our exposition too abruptly to a close. By the course of inquiry we have followed, something might be brought out concerning the political organization of the primitive Aryans, which appears to have been extremely simple. “ The people uras doubtless a congeries of petty tribes, under chiefs and leaders rather than kings, and with institutions of a patriarchal cast, among which the reduction to servitude of prisoners taken in war appears not to have been wanting.” 5 This inquiry, however, would take us far beyond our limits, and might be more advantageously conducted in another connection, where we might avail ourselves of the harmonious results which Sir Henry Maine, Mr. Freeman, and others have elicited from a comparative survey of Indo-European politics and jurisprudence. But this most interesting and profitable study must be postponed to another occasion. In the present paper, confining myself chiefly to the material circumstances of the primitive Aryans, I have endeavored only to give some idea of the method by which sound conclusions are reached, through the study of words, concerning the civilization of an age of which the historic tradition has been utterly lost. More than this could not well be attempted in so brief an exposition. My examples have necessarily been scanty, and from the nature of the subject I fear they may have seemed rather dry. It is not in a moment that one can become fully possessed with the rare fascination which surrounds the study of the historic lessons conveyed in words. Yet possibly to some reader it may have come as a novel and striking thought that out of mere grammars and dictionaries a trustworthy picture of the long - forgotten past maybe reconstructed. Inadequate as our illustrations have been, none can fail to perceive the historic interest and value of the information which has been gained in this way. Inquiries of this sort need, no doubt, much caution and sagacity to be conducted successfully; but when properly sifted there is no more unimpeachable testimony to the past than that which the aspect of words gives us. For the changes of vowel and consonant proceed according to general laws which observation may detect, but with which no individual will is able to tamper. And thus it is that in the winged word which seems to perish in its flight through the air we have nevertheless the most abiding record, though unwittingly preserved, of the knowledge and achievements of mankind.

John Fiske.

  1. Whitney. Study of Language, page 199.
  2. Fick, Vergl. Woerherb. d. Indog. Sprachen. Third edition, Goettingen. 1874-1876.
  3. See, however, Pictet and Fick.
  4. Taylor, Words and Places, page 61.
  5. Whitney, Study of Language, page 207.