Page:Vedic Grammar.djvu/35

 I. PHONOLOGY. THE FIVE CLASSES OF MUTES. GUTTURALS. PALATALS. 25 of cognate languages shows, derived from IE. velars. Gutturals are found interchanging to some extent with sounds of the four other classes. s-s 1. Under certain conditions they interchange with the new palatals (cj h) which are derived from them²; with the old palatal s³ (also old j and h) only when followed by s (which then becomes ş) 4. Between this ks and kṣ=k-s it is possible to distinguish by the aid of Iranian, where the two are represented by different sounds 5; and the original value of the k can thus be determined even in words in which no form without the sibilant occurs. This evidence shows that, in the following words, ks represents. a. s-s: ákşi-‘eye'; ṛkṣa- 'bear'; kákṣa- 'armpit'; kuksi- 'belly'; kṣi- ‘dwell’; ksidh- 'hunger'; caks- see'; taks- 'fashion'; dákṣina- 'right'; pákṣman- (VS.) 'eyelash'; makşú 'quickly'; raks- (AV.) 'injure'; ráks-as- 'injury'; b. k-s: ksatrá- 'dominion'; kşáp- 'night'; ksi- 'rule'; ksip- 'throw'; kṣīrá- 'milk'; ksud 'shake', kşód-as- 'rush of water', kṣudrá- 'small' (VS.), n. ‘minute particle'; kşübh- 'swift motion'; tvaks- 'be strong'; vyksá- 'tree'. 2. In a few instances k stands for a medial t: in vrkkáu (AV.) ‘kidneys', for *vrtkáu; prkṣú (SV.) prtsú in battles'. In these two forms the sub- stitution is due to Prakritic influence; this is probably also the case in skambh- beside stambh- 'prop". The guttural only seems to stand for a dental in ásiknī- beside ásita- 'black', pálikni- beside palitá- 'grey', and háriknikā- (AV.) beside hárita- 'yellow', as there is no etymological connexion between -knī- and -ta-⁹. L IO 3. In a few words a guttural interchanges with a labial medially: kakárdu- beside kaparda- 'braid of hair'; kulikā (VS.): pulikā (MS.) a kind of bird; kulikáya- (TS.): pulīkáya- (MS.), kulīpáya- (VS.): purīkáya ¹ (AV.) a kind of aquatic animal; nicunkuná- (TS.) nicumpuná- 'flood'; and in the TS. (B.) tristigbhis and anuştigbhyas occur beside tristúb-bhis and anuştúbbhyas**. = 4. In a few verbal forms from three roots k stands for s before suffixal s¹², though this never made its way into the loc. pl. (where only -s-su- or -h-sṣu, -t-su occur). The only example in the RV. is pinak (for pinak-s) 2. sing. impf. of pinas-ti (pis- 'crush'). In the AV. occur dvik-s-at, dvik-s-ata, aor. of dviș- ‘hate'; šišlikṣate, -śiśliksu-, desid. of śliş- 'embrace'. Other possible examples from the RV. are rksará- 'thorn' (if from rs- 'prick'); ririkṣa-ti and ririkşú-, desid. (if from ris- 'injure'); vivekşi (if from vis-'work') ¹3. 35. The palatals. These are pronounced in India at the present day as a close combination of a t-sound followed by a palatal spirant s The evidence of the Greek reproduction of Indian words 4 points in the same That is, the q-sounds; some, however, are derived from IE. labio-velars or qu-sounds; see BRUGMANN, KG. 1, 244 and 254; WACKERNAGEL I, 115. 2 See BRUGMANN, op. cit. 244- 3 Op. cit. 233. 4 See below 56. 5 That is, ss by š and k-s by hš; thus vakşi, from vas vaši; vaksya-mi, from vak- (for vac-) = vahšyā. = 6 The two components of ks cannot yet have coalesced when s dropped out between two mutes in abhakta, for abhak-s-ta (aor. of bhaj), and ataşta for alak-s-ta- from taks- (Av. taš) 'fashion'; otherwise the two differ- ent original sounds could not have been kept apart in these two forms. See above 30, note 4. 8 Cp. WACKERNAGEL I, p. 136, note (top). 9 Cp. J. SCHMIDT, Pluralbildung 398. 10 See ZDMG. 33, 193. ¹1 See WEBER, IS. 8, 40. 54; 13, 109. 12 This probably started from the paralle- lism of the 3. sing. of roots in s and s: thus dves-ti from Udvis-, and vaș-ți from V/vaš-; then the 2. sing. dvek-şi for dueş-şi, followed vak-şi. 13 The relation of the k in dadhŕk 'firmly', to dadhrsá-, dadhrs-váni- bold', is uncertain. Cp. WACKERNAGEL I, 118, note (end). 14 Thus Távdavov candana- sandal- wood'; Tiaorávns = castana, N.; Ilatáhai = pañcala-, N. of a people; LaudpóKUNTOS candragupta-, N.; 'Õ¢ýv=ujjayinī- (Prakrit = -