to spin, and with KART, to join, because in the earliest time spinning and platting were frequently combined (cf. crâtes, mat, kάρT-aλos, basket), and platting consisted in cutting and slicing reeds together. Again, DÂ, to cut, would have to be identified with DÂ, to bind, because in many cases binding is performed by cutting small threads and slicing them together, just as to sort, for instance, means to separate, but also to join. Fortunately, however, we are not driven to these forced explanations by any inherent necessity, and we ought not to resort to them, except in cases where we have real analogies to guide us. It is as yet a purely gratuitous assumption that one sound can at first express one concept only. It rests on another assumption, that there is some mysterious bond between the sound and meaning of a root. But even if there were such a bond, that would not exclude the perfect freedom with which at different times and in different places similar or even identical sounds may accompany different acts which they afterwards signified. There is, for instance, a root R, which is explained by such general terms as to go, to send. Root . But the question is whether AR, AR, to go, red. perf. âra, is really the same root which we find in ri-no-ti, op-vu-μ, to excite, red. perf. op-wp-a, I have risen, or-ior, or-igo, and again in ar-us, wound, and ar-is, enemy. Curtius holds that the Greek and Latin languages split the root into three, namely AR, in ap-ap-ioкw, ar-tus, ap-ów; ER, in èp-éσow, rêmus; OR, in op-vu-μ, or-ior, etc. The same distinction, however, though not expressed phonetically, at least in writing, may have existed Ab-olesco, to decay, to vanish; as nomen vetustate abolevit. Obs-olesco1, êvi, êtum, to decay, to wear out, as in vestis obsolêta. (2) In the sense of making grow, we find the same root in Ålo, alui, altum, to nourish, as in Plaut. Rud. 3, 4, 36, Athenis natus altusque. Altus, high, like great from to grow. Alumnus, foster-son, nursling, pupil. Al-mus, kind, genial, particularly of Ceres, Venus, Terra. Ab-oleo, evi, -olitum, -êre, to destroy, to re move. Ad-oleo, -evi, -ultum, to increase the gods, to worship them. Non. 58, 21; Virg. Aen. i, 704, flammis adolere Penates. Lastly elementum, lit. what makes grow, for olementum, the o changed into e as in bene, velle, etc. It cannot be denied that to go, as the original concept of Æ, might have been developed so as to express both the peaceful going together and the aggressive going against a person. Some scholars might even maintain that the idea of striking the earth or the water in ploughing or rowing was derived from that of attacking, nor would it be impossible to find some analogies for a transition of meaning from going to growing, and making grow. But all this could only be approved of or tolerated if it had been established once for all that identity of Bopp, Comp. Grammar, § 96, remarked, that we divide with Vossius ob-solesco, not with Schneider, obs-olesco, requires no excuse.' Yet Schneider was a safer guide. sound in roots proves always identity of origin, a principle which, as far as I can judge, receives no support either from theory or from facts. Professor Fick, for instance, in his Wörterbuch, often keeps homophonous roots apart, but quite as often he assigns the most incongruous meanings to one and the same root. He distinguishes, for instance, between SKARD, to shine, SKARD, to break, SKARD, to vomit, and SKARD, to jump, where the meaning of shining (éclater), breaking, and vomiting (brechen) might easily have been reconciled. But he treats SKAR, to divide (cer-no), KAR, to feed (kópos), and KAL, to conceal (oc-culere), as mere varieties of the root SKAR, to pour out, to scatter, which seems to require a very great effort of combinatory imagination. Many of the questions connected with the original meaning and the later development of roots can at present be answered provisionally only. SVAR, to shine and to sound. Should we separate, for instance, the root SVAR, to sound, from SVAR, to be bright, to glisten? It is true that the second root has yielded no verbal forms, but svàr, sky, súra and surya, sun, require the admission of such a root. This being the case, it seems possible, even for us, to comprehend light and sound under the same concept of clearness or brightness, and one feels all the more inclined to admit the same process for the poetical language of our earliest ancestors. shine and Again, may we admit a relationship between BHÂ and BHÂS, to shine, and BHÂSH, to BHÂ, to speak? The analogy of paive and oŋuí is tempting, but hardly convincing. Curtius, according to his system, admits a primitive root BHÅ, which became specialised as BHAN, BHÅS, BHAV, D d to sound. and BHAK. The simple root he sees in pn-ui, pá-Tis, pw-vý, Sk. bhâ-nus, light, sun, Lat. fâ-ri, fâ-ma, fâ-s. The root BHAN appears, according to him, in φαίνω, φανερός, φά-σις, Sk. bhan-âmi, I sound, bhan-âmi, I speak. The third form BHÂS occurs in Sk. in the sense of shining, and as BHÂSH, speaking. The fourth form BHAV is traced by him in paû-os1, pas, light, and -páwv; the fifth in fac-ies. From my own point of view I should look upon these five roots as parallel varieties, but even then certain phonetic difficulties remain. Curtius could not explain the transition of s into sh, as in BHÂS and BHASH (not Vedic), nor of BHAN into BHAN, for the change of a dental into a lingual always requires a motive. It may be that the pronunciation of the indigenous races produced BHASH and BHAN with linguals, instead of BHÂS and BHAN with dentals, but in that case we should expect BHASH to have the same meaning as BHÅS, which it has not. I am not convinced therefore that the concept of shining was raised into that of speaking, and I think we must admit the existence of two roots, one being BHA, to shine, the other BHÂ, to speak, each with its varieties; BHÂ, to shine, with BHAN, BHÂS, BHAV, and BHAK; BHÂ, to speak, with BHAN, BHASH2. Curtius points out that in Greek paivw is often applied to the showing forth of speech, as in Soph. Antig. 621, κλεινὸν ἔπος πέφανται. He might have quoted a still stronger instance from Aristotle, Categ. c. 6, ἀπόφασις δέ ἐστιν ἀπόφανσίς τινος ἀπό τινος. But even that would hardly be sufficient evidence to show 1 Cf. Saussure, Système Primitif des Voyelles, p. 54. that to shine by itself was ever conceived as to speak. The nearest analogy is that of KÂS1, to shine, and KSÂ, pronounced KHYÂ, to tell: all other supposed parallel cases seem to me to break down on closer inspection. After these preliminary remarks I now give the first attempt at a classification of San Classification Roots. skrit roots according to their meaning. of Sanskrit From what I have said it will be clear that I myself look upon such a classification as purely tentative. I have tried to ascertain what is most likely to have been the original meaning of every root, but having to select one meaning only out of a great variety, the selection has often been a choice between two or even many evils. In a first attempt of this kind the chances of error must be very great, still I hope that those who will carefully examine the results at which I have arrived, will admit that they prove by overwhelming evidence that the meanings of roots are really what we expected them to be, and that they express the primitive social acts of primitive social men, and the states more or less closely related to such acts. The order in which the concepts succeed each other is not very systematic. I have Order of Roots. tried, as much as possible, to keep the special acts, such as to dig, the general acts, such as to find, the special states, such as to cough, and the general states, such as to stand, together. But it was impossible to adhere strictly to such a plan, because there are roots which express both acts and states, while in many cases it is difficult to determine whether the special or general meaning pre 1 On KÂS see Hübschmann, Indog. Vokalismus, pp. 56, 63. |