I. PHONOLOGY. CEREBRALS. 33 of the) head. They are described by the Prātiśākhyas as pronounced by turning the tip of the tongue up to the roof of the mouth and bending it backwards. They were therefore pronounced farther back in the mouth than the palatals. This is also their pronunciation at the present day in India. An indication that it was such even in Vedic times is the fact that d is sometimes found in the later Samhitas interchanging, between vowels, with 7 (which itself interchanges with ), and that in the RV. itself d dh become !h between vowels. It is also to be noted that the Greeks reproduced d not only with d, but also with p². The cerebrals, however, were a specifically Indian product, being unknown in the Indo-Iranian period ³. They are still rare in the RV., where they occur medially and finally only. According to most scholars, they are due to aboriginal, especially Dravidian, influence*. As a rule, they have arisen immediately after y or an sound from dentals. But before consonants and finally they may represent the old palatals i sh. a. The voiceless cerebrals t țh take the place of the dentals t th after s s, or j); e. g. vṛṣ-ți- 'rain' (suffix -ti); dus-tára- ‘invincible' (— dus-tára-); nákiş te (= nákis te); vás-ți 'wishes' (— váś-ti); mṛṣ-tá- 'cleansed' (= mrj-tá-)5. Similarly the voiced cerebrals d dh take the place of the dentals d dh after z (= s or old palatal j, h), which has disappeared6; e. g. nīḍá- nesť (= IE. nizdó-); du-dhi- 'ill-disposed' (= dus-dhí-); id-é 'I worship' (*iz-d-— ij-d- for yaj-d-); dṛḍhá- ‘firm' (— drh-tá-). The preceding voiced sibilant * (= Ś and ş) has (instead of disappearing) itself become d in didid-dhi (from dis- 'show') and vivid-dhi (from vis- 'be active')¹. a. When the dental here was immediately followed by an » sound, the cerebralization seems originally to have been stopped. Hence drdhrá- (= drh-tra-), beside drahá- (=drh-ta-) 'firm'; and though str occurs several times in the RV. 8, the seems to have been dropped in pronunciation, as the only stems ending in -stra- which show a case- form with n, do not cerebralize it: ústrānām and rāṣṭrānám, as if no preceded. In TS. 1. 2. 5², is actually dropped after șt in tvășțīmatī- ‘accompanied by Tvaṣṭrī”. b. In several instances a cerebral appears by an evident Prakritism, in place of a dental originally preceded by an ♂ (or 7) sound; thus vi-kata- 'monstrous', beside kr-tá- ‘made'; kāṭá- 'depth', beside kartá-9 'pit'; avațá- (SV. VS.) 'pit', beside avár¹0 'down'"; and as shown by comparative evidence, kátuka- 'sharp'; kūḍayati 'singes'; kévata- 'pit'; jáḍhu- 'dull'; kūtá- (AV.TS.) 'horn- less'. In the following words, though cognate languages show 12, the cerebral is similarly based on Indian r or ¹3+ dental: kúta- 'frontal bone'; jathára- 'belly'; tadit-'contiguous', táḍa- (AV.) 'blow'; pinda- 'lump'; kanda- (AV.) 'piece'¹4. c. Cerebrals have in some instances supplanted dentals owing to the ¹ Cp. VPr. IV. 143; 28, 298. 2 Şee WACKERNAGEL I, 143, note. 3 Op. cit. I, 144. 4 Ibid., note. V. BRADKE, KZ. if from krt- 'cut'; but see BARTHOLOMAE, IF. 3, 180 f. 12 IE. / by rhotacism became » in IIr. 13 MS. II. 47 has the reading jinva rāvát for that of TS. II. 4. 7¹ jinvár ävýt and K. XI. 9 jinva ravat. Cp. below p. 70, note 4. ¹4 The cerebral could be similarly accounted for in kúta- (RV¹.) "house'(?), if it is related to kula-(pa-) family', and kulaya- (AV.) 'nest'. In dandá- 'staff' if identical with Gk. dévòpov (J. SCHMIDT, KZ. 25, 52, note 1) we seem to have an instance of a cerebral for a dental + following , but such a change seems not to be in accordance with the phonetic laws of either Vedic (cp. drdhrá-, above a, a) or Prakrit. On two other 10 Cp. BARTHOLOMAE, IF. 3, 179.
- Perhaps also renúka-kata- 'stirring dust', examples of this supposed change, andá-
Indo-arische Philologie. I. 4. 3 5 Cp. above 41, I. 6 Cp. above 17, 5. 7 See WACKERNAGEL 1, 145 a, note (end). Cp. 42 d (p. 34) end. 8 In rastrá- dominion', ústra- buffalo', destri Directress'; dámstra- 'tooth'; - in á-ni-strta not shaken off', tvástr-mant 'accompanied by Tvastr; ndr in kundr- naci- 'house-lizard' (?). 9 WACKERNAGEL 1, 146 a.
