ON NUMERALS AS SIGNS OF

PRIMEVAL UNITY.

SIMILARITY in the names of numerals, and especially of the numerals from 'one' up to ten,' is commonly and justly regarded as strong evidence in favour of an original connection between any languages in which such similarity is observed to exist; and, indeed, where neither the supposition of fortuitous resemblance, nor yet that of borrowing by ⚫ one independent language from another, can adequately explain how coincidences of this nature arose, then those coincidences may be said to prove a common origin for the words in question, and thus to imply, at least probably, though not certainly, a common origin for the nations which employ them. Thus the relationship which unites what are called the Semitic languages and the Semitic nations is plainly exhibited in their numerals; and the Aryan numerals form one most important part of the mass of evidence by which all the members of the Aryan race, from the Celts to the Hindoos, have been traced up to a single clan, if not a single household, once dwelling in Mount Imaus.1 But it is not impossible that what the science of language has been made to do for the Aryans, it may be made to do in some measure for the human race as a whole. At any rate, it may be worth while inquiring if the belief that all mankind sprang from one family in Western Asia, or nearly in the centre of the Old World, is borne out by anything in human speech at the present day. 1 Max Müller, Science of Language, Lecture 5.

56

B

Should such linguistic signs of primeval affinity still survive the changes of several thousand years, it is in numerals that they would very likely, if not most likely, be detected. For the names of numerals commonly carry in themselves the proofs of their own great antiquity, as their mode of formation indicates anything but an advanced state of culture. Many uncivilised nations still exist upon earth, and some among them who may have fallen below the primitive state of their remote ancestors, while highly civilised nations would, on the other hand, have risen far above such a condition. But inquiry shews that it makes no great difference in the derivation and composition of a nation's numerals, whether that nation be now civilised or uncivilised; for civilised nations take here after the uncivilised. And the manner in which uncivilised nations habitually compute is this. They employ as numerals, either singly or in combination, various words for hand' and 'finger', or sometimes 'foot' and 'toe'; a mode of numeration which would hardly be consciously adopted or employed by a nation which had made much progress in civilisation. "Alle zahlwörter", says Grimm, "gehn aus von den fingern der hände." What, then, would be the inference, when it is discovered that the Aryan numerals have been formed after this manner, and that the Aryan decade contains two words for 'hand', and one for 'fingers' or 'toes'? That inference would not be, that the original Aryan family or clan in Mount Imaus was quite uncivilised, for the common Aryan vocabulary would imply that it was not so; and language must, besides, have existed for ages before it could have taken an inflecting form like the Aryan. The truth is, that the Aryan numerals, or the elements of the Aryan numerals, carry us back to a time when, properly speaking, there were no Aryans at all; when the distinction between Aryan and Turanian, and perhaps other races, was not yet established; and when the common ancestors of all counted upon their hands and

fingers, employing as numerals the names of those members. The original Aryan family in Mount Imaus would not have invented a new language for itself, but would have selected and combined in a manner which became characteristic some portion of the words used at a particular epoch in the region of the world where those mountains lie. This, at least, is a natural inference, and the evidence supplied by the Aryan decade will be found in favour of such a supposition. For Aryan 'fives' and 'tens' are not merely similar to several non-Aryan 'fives' and 'tens', but likewise to words for 'hand' or 'foot', and 'finger' or 'toe', which are quite as much non-Aryan as Aryan, and would have been employed numerically by both races. Thus, to take examples affecting Aryan tens:-we should at first, it is probable, be inclined to regard as no more than accidental the resemblance which the English twen-ty and the equivalent Old Norse tu(t)-tugu bear to such Yeniseian forms in Siberia as the Kamacintzi tonga-tu, 'thirty' (tonga, 'three'), and hkelina-tugu, 'seventy' (hkelina, seven'), with the corresponding -tukn and -taga, '-ty', in the Assan dialect of the Yeniseian. But the cradle assigned to the Aryans is not so very far from the Yenisei; and we find, moreover, in other Yeniseian dialects, the words tok and tokan for 'finger', and toigen for 'foot', in addition to the previously cited Yeniseian forms, -tukn, -tugu, -taga, -tu, '-ty', i. e. 'ten', while we know that, in like manner, zehn has been connected with zehe, decem with dig-itus, and déka with dákTUXOS. Nor do such resemblances occur in two classes of languages only, the Aryan and the Yeniseian, for in Africa we meet with tuko, 'toe', tukui, 'hand', and toko, 'arm', and also with tek, teku, toko, 'one', and atuk, 'ten'; as well as, in North America, with atoken, 'one', and atek and aduk, 'ten', the transition to which from the Yeniseian tok and tokan, 'finger', and -tu, -tugu, -taga, -tukn, '-ty', i. e. ‘ten', is facilitated by the Kurile dek and tegi, 'hand', terms

which are, again, nearly identical with dak, tekha, and takha, three words used for 'hand' in the mountains between Assam and Burmah, where also dug-, dugu-, and duku- are employed as prefixes in the numerals of one decade, as if they had once meant 'finger'. All this may, no doubt, be chance; but it may, on the other hand, be more than chance, and the alternative seems worthy of consideration.

A great number of coincidences of this double kind, affecting not only numerals, but also the names of the members of the body from which those numerals are derived, may be detected in languages far removed from each other in position, and will be found in the following pages arranged in groups. In such languages as are little known, the materials have been chiefly derived from Dr. Latham's Elements of Comparative Philology, from the Polyglotta Africana, from Dr. Hunter's Non-Aryan Languages of India and High Asia, and from Professor Pott's Zählmethode. Each group of coincidences presents a certain body of facts, of which the right explanation is to be sought, and which give rise to three questions. Are such coincidences the result of chance? Or are they the result of borrowing between nations originally unconnected in blood and speech -an hypothesis which would imply the existence of early intercourse, either direct or indirect? Or, finally, are they the result of primeval affinity—indications of unity of origin in human speech and probably in the human race? Each of the three suppositions is possible: which of them is most in accordance with the evidence? To that evidence we now proceed.

The first group of coincidences to be noticed is the most important of all, both on account of its significance, and of the races that it affects, which comprise the most remarkable peoples that have appeared in the world's history. North America presents us with the following words, of which different names for finger' supply the elements:

By comparing the Pawnee with the kindred Caddo, which stands below it in the previous table, we may see that -koo, in the Pawnee as-koo, one', and peet-koo, 'two', is some suffix, which may be here left out of consideration, and will appear eventually to be probably a generic term for 'finger' or limb', while as- and peet- may be rather the proper or 1 Here -t is replaced by the Mexican -tl, to pass below into l and l. In some Aryan languages, t is frequently weakened into 7.