Abbildungen der Seite

Lith. vidus, Lett. widus 'inner part' may be identical with Ved. vidhú'lonely.' Lith.lengvùs (lengvas) 'light,'tho in imperfect coordination with I.E. Inghú- (Ved. raghú), Gr. Éhaxús, (cf. OBulg. ligůků) reflects the old ú-stem. The same seems to be true of Lith. grabùs 'skilled with fingers,' Ved. gr(b)hú- 'beggar'; Lith. drąsùs in its relation to the I.E. stem dhysú. 'bold' (see above p. 88); and Lith. kartùs, OBulg. kratůků 'short.' Perhaps also Lith. lipùs 'sticky': Ved. ripú- 'deceptive', and Lith. at-stùs 'being far': Ved. su-şthú- being well' are of old stock.

The list of about two dozen Lith. u-adjectives with u in the root 15 displays only the fact that succession of two u's has become a favorite fonism in that language: dubùs 'deep,' is invalidated by discrepant OBulg. duplŭ 'hollow,' and Goth. diup(a)s 'deep'; rupùs 'rough' is discredited by rúpas; skubùs 'hasty' has invited, very properly, comparison with Goth. af-skiuban 'shove off,' OHG. scioban 'schieben,' Skt. root kşubh 'shake, but is not supported by any u-stem in the related languages. There is nothing to disprove that the succession u-u is part of the wide but secondary movement of the language in favor of u-adjectives. Predilection for the succession u-u appears also in the confluence of -10- stems, with -ru- stems, which cannot be differentiated historically.16 Many of these show the succession u-u; budrus 'watchful (see above, p. 91), kutrùs diligent,' gudrus 'wise,' skubrùs 'hasty,' skudrus 'sharp,' sudrùs 'abundant,' sukrùs 'adroit' (Lett. sukrs), tukrùs 'fit to fatten,' mustrùs 'cautious,' mudràs and mundrus 'joyous” (also muñdras). It is barely possible, though not provable, that these forms reflect a blend of -2- and -ro- stems, based upon their historic suppletive character.

Suppletion of -2- and -10- stems is indicated by occasional formations with both suffixes. From I.E. times come è appós, OHG. lungar 'light,' by the side of laxus 'light' and its correspondents. So also Lat. plērus, Gr. Tanpów 'fill,' ampó-ons 'fulness,' by the side of I.E. pfiú- (Gr. Tolús); cf. Lat. plēnus: I.E. płno- in Skt. pārna-, etc. Ved. rjrá 'straight in rjrāgua - Avest. ərazrāspa- 'whose horses go straight': Ved. rjú- Avest. ərazu- 'straight'l7; ripú- 'deceitful' (cf. Lith. lipus "sticky'): ripra-m, neuter, 'impurity'; Ved. mandú- 'joyous’: mandra- 'agreeable' (cf. Lith. mándras, mandràs, OBulg. mądrů 'cheerful'); grdhnú'eager' (borrowing, perhaps, the n of dhrşņú- 'bold'): grdhra- 'eager,'

u Leskien, l. c., 244.
15 Leskien, I. c. 257, 258.
16 Leskien, ibid. 440ff.
17 Cf. Geldner, Ved. Stud. ii. 165 ff.; Brugmann, IF 17.361.

mascul. 'vulture. Ved. du- in pdūdára, dupa, rdüofdh seems to mean ‘moist, comparable with ārdra- ‘moist.' The Greek grammarians report ulkus as the equivalent of both ulkkos (Doric) and mixpós 'small'; also the diminutives μικυθος and μικύθινος (Hesych. μικρών και νήπιον). Cf. also the apparent blend of 2- and ro-suffixes in Gr. Acyus : Acyupós ‘shrill’; yAukús: Aukepós 'sweet'; Ved. madgús diving bird': Skt. madguras 'diver'; cf. rudhi-krá 'smeared with red blood' (designation of a demon): rudhiras for *rudhrós 'red,' Wackernagel, 2.1. 61.

The Germanic languages have two words from different roots for 'thick'; one an ú-stem, the other a ró-stem. The ú-stem in ON. þykkr, accusative þykkuan, OHG. dicchi, Erse tiug; the ró-stem in Goth. *digrs in digrei 'thickness,' ON. digr 'thick' MHG. adverb tiger 'complete,' Goth. deigan 'knead' (I.E. root dheigh). Class affinity between the two suffixes appears also in opposites: the Germanic words for 'sweet' based upon *svotus (OHG. suozi, etc.), though displaced in Gothic by suts (cf. Vedic súdas ‘sweetness'), is contrasted everywhere with the -stem, ON. bitr; OHG. bittar. Goth. baitrs 'bitter. Similarly Gr. TIKpós “bitter': adús and yukús; cf. the Slavic opposites sladůků 'sweet,' and bridůkŭ 'bitter,' assimilated to identical stems.

Caland in KZ 31. 267 (1891), 32. 592 (1893), has pointed out that Indo-Iranian adjectives in -rá- are in suppletion with i-stems when these latter are first members of compounds. Since then this relation has been carried, to be sure on the basis of few examples, into the I.E. time. The subject is summarized in Wackernagel, Altindische Grammatik, 2. 1. 59ff. (824), where the literature is given; and later by Brugmann, Vergleichende Grammatik, 2. 1.2 78; cf. also, last, Jokl in Streitberg Festgabe 176 ff. Examples are: Ved. çoitrá- 'bright:' gvity-á ñc 'clear,' Avestan spiti-gaona ‘of white color'; Ved. krūrä-, Avest. xrūra- 'bloody:' Avest. Xrvi-yni 'slaying cruelly'; Gr. Kūdpós 'glorious:' KŪdc-ávelpa 'having distinguished men.' In one case at least such a pair is joined by an ú-stem: Ved. rjú-, Avest. ərəzu- 'straight': Ved. rjrágua-, Avest. ərazrāspa- 'he whose horses run straight': Ved. rjipya-, Avest. ərəzifya'going straight,' 'eagle,' or 'hawk' (cf. also Ved. fjiti- 'going straight' tjipin- 'hasting,' rjtçuan, nom. pr., and perhaps rjīşin-, obscure, perhaps liturgic epithet of Indra).

Anent the triad rjú-, rji- and rjrá-, Brugmann, IF 17. 361 (1905), has made the remarkable statement that rji- is the compositional form of rjú-. But he has, apparently, not repeated this suggestion since. There are, perhaps, three or four more cases of this apparent suppletion of an u-stem with an i-stem in composition: Avestan gərəđi- "eager," in

gərədy-aoxsa- 'howling eagerly':{Ved. grdhnu-' (for *grdhú, above p. 93) and grdhra- 'eager'; darşi- in darşi-dru 'with bold weapon': Ved. dhrşņú- (above, p. 88); āçi-vişá 'quick-poison,' AV. 12.5.34 (designation of a serpent): äçú- 'quick’; and doubtfully mərazi- in the patronymic mərazīşanya- which might be compared with Bpaxus, if that word be radically related to Goth. ga-maúrgjan 'shorten,' rather than to Lat. brevis. Perhaps also Gr. åpyu- in špyupos 'silver,' áprupos 'of silvery sheen”: ápyr- in åpyuxépavvos 'with shining lightning'; Thracian ápyidos d Mûs18: åpyós 'shining,' for åpypós =Ved. rjrál9 'reddish.' This is all too poor to establish a direct relation between the ú and i-stems, and does not, therefore, solve the riddle of the suppletion between the latter and the ró-stems. As the ró-stems are in suppletion with both ú-stems and i-stems the occasional correspondence of both ú-stems and i-stems with -stems is probably hap-hazard.

18 See Jokl, I. c. 173 ff.

10 This rjr&- is to be differentiated from rjrá- 'straight'; see last, Neisser, Zum Wörterbuch des Rigveda 188.

[ocr errors]




BERKELEY, CAL. The state of confusion of our knowledge of the linguistics of Central America needs no better comment than the hodge-podge in that section of Les Langues du Monde, recently edited by Meillet. Oaxaca is a veritable tangle-knot of linguistic stocks. One is reminded of the Ca'ifornia of early days with its twenty-one unconnected families. Oaxaca is a comparatively small state, but there are fourteen tongues spoken within its borders. With the exception of Aztec not one of them has been seriously studied until now. As for the classifications, they vary all the way from Belmar, who reduces them all to three families, to Meillet, who has eight different stocks represented in the state, four of which would be made up of one single isolated language. I hope that the present paper will help to clear the situation.

The material on which it is based was collected in the field, when I went to Oaxaca in 1922, at the behest of Manuel Gamio, the Director of Anthropology of Mexico, to collaborate as linguist in the Anthropological Survey of the Mexican Republic which he has planned. The monumental work on Teotihuacan, as typical of the Aztec culture, had just been completed and the field of investigation was now shifted to the Zapotecan region.

Although the primary object of my work was to be a study of the Valley Zapotec proper I had made up my mind to have a look at the other languages if possible. Accordingly I first went to live at Teotitlan del Valle, close to the famous ruins of Mitla, and commenced my study of the language. It was not long before I realized that I was dealing with a type of language unfamiliar to me: a very sparse form; practically no treatment of the verb, that is to say, the verb reduced to a simple expression of the infinitive preceded by a semi-prefixed particle of temporal aspect and followed by a half-agglutinated, half-fused pronominal ending; the same pronominal ending added to any form of noun to express personal relation or possession; the burden of the expression of relation laid on word-order, except for a division of the world of objects into the

three classes of animals, men, and things, expressed in a somewhat confused manner by the pronominal endings, or rather mixed up with these; the pronominal endings a mere garbled repetition of the independent pronouns or demonstratives, or even of the very words for "man," "animal," or "thing," all of it based perhaps more strongly on tone pattern than on vocalic or consonantal symbolism, pitch tone forming a very strong element in the grammatical processes. In short, the whole thing gave me the impression of a language caught in the midst of rapid evolution. But whither and whence? I was puzzled.

I returned to the town of Oaxaca and there I found a marvelous opportunity to obtain informants for the several different tongues spoken in the state; for Indians from all parts sooner or later make the pilgrimage to the market-place at Oaxaca. There each Saturday you may recognize, after you have learned their characteristics, mountaineer Zapotecs who come from as far as the Rincon to sell their fine rope-work and hamacas, the wild looking and suspicious Mixes with enormous loads of chile, Chinantecos with slanting eyes and high cheek-bones conversing in their extraordinary nasa sing-song, Chatinos from the south coast with cargoes of pineapples, cargadores from all parts of the Mixteca, Chocho hat weavers, sometimes Cuicatecos and Mazatecos, even an occasional Chontal. The Triques alone never come down from their five villages way up in the mountains. The Huaves of Tehuantepec are too far away to come except very rarely on the occasion of some big fiesta. I never obtained any information on those two languages, nor on Amuzgo. This last remained a mystery to me. Several informants who claimed to be Amuzgos spoke pure and simple Mixtec! Thus I worked

Chinanteco, Mixteco, Chocho, Chatino, Mazateco, Cuicateco, Mixe and Chontal, all at the same time with different informants. Later I made two roundabout tours of the state with a pack burro and visited most of those regions to confirm my data and collect dialectical variations.

I will dispose at once of Chontal and Mixe since they have no connection whatever with the rest. In a study of Chontal which will appear soon I confirm Kroeber's classification of this tongue with the Hokan family. In another paper I present my evidence for classifying Mixe with the Penutian family.

Now to return to the other group. I had no sooner commenced taking down vocabularies than I was astonished at the utter lack of semantic similarity. I knew that Thomas and Swanton in their excellent map had lumped them all, with the exception of Chinanteco, in one group on the

« ZurückWeiter »