I. PHONOLOGY. SECONDARY SHORTENING OF iur. CONSONANTS. 21 'kindle'; from dhi- 'think', di-dhi-ma, 1. pl. perf., di-dhi-ti- 'devotion', beside dhi-ti- 'thought'; from kr- 'commemorate', cár-kr-se, intv., car-kr-ti- 'praise', beside kīr-ti- 'praise'; from pr- fill', pi-pr-tám, 3. du., beside pur-ná- and pur-tá-. Such shortening often occurs in red. aor., e. g. bī-bhis-a-thas, 2. sing. mid., beside vi-bhis-ana- 'terrifying'. It is also found in a few nouns; e. g. si-si-ra- (AV.) 'coolness', beside si-tá- 'cold'; tū-tu-má- beside tú-ya- 'strong'*. The Consonants. 30. Doubling of consonants.-All consonants, except r hl, Anusvāra, and Visarjanīya, can be doubled, and the distinction between double and single consonants is known to the Pratiśākhyas as well as to Pāṇini. Aspirates are, however, nearly always written double by giving the first in the unaspirated form. A double consonant is pronounced by the organs of speech dwelling longer on it than on the single sound. Within words 3 a double consonant appears: 2 I. as the result of the contact of the same consonants or the assimilation + of different ones; e. g. cit-tá- 'perceived' (= cit-ta-); uc-cá- 'high' (= ud-ca-); bhet-ty- 'breaker' (= bhed-tr-); án-na- 'food' ( ad-na-). - 2. in a few onomatopoetic words: akhkhali-kŕtya 'shouting'; cicciká- a kind of bird; kukkuțá- (VS.) 'cock'; tittíri- (VS.) and tittiri- (TS. B.) ‘quail'; pippaka- (VS.) a kind of bird. 3. in the case of the palatal aspirate, which regularly appears as cch between vowels (though often written as ch in the Mss.), for it always makes the preceding vowel long by position and is derived from an original conjunct consonants. Some forms of khid- 'press down', are doubled after a vowel in the TS. (akkhidat, á-kkhidra-; ā-kkhidaté, pari-kkhidaté). In the TS.6 bh appears doubled in pári bbhuja. In a school of the White Yajurveda initial was regularly doubled 7. 4. when final n is doubled after a short vowel if followed by any vowel sound 8. a. In the Mss., when double consonants are preceded or followed by another con- sonant9, one of them is frequently dropped, because in such consonantal groups there was no difference in pronunciation between single and double consonants. Hence the VPr. (VI. 27) prescribes a single t in ksattrá- dominion' (= kşad-tra-), and in sattrá- 'sacrificial session' (= sad-tra-). Such shortening is further presupposed by the analysis of the Pada texts in hr(d)-dyotáḥ (AV. I. 22¹) and hrd)-dyotanaḥ (AV. v. 2012) as hr-dyotáh and hr-dyotanah; in tá(d)dyám (AV. Iv. 196) as tát yam (instead of tát dyam); and in upástha-, which appears in the RV. Pada as upá-stha- instead of upás-stha-, if GRASSMANN'S suggestion is right 1o. In some instances this reduction is IE., as in satrá- (IE. setlo-)11. ¹ On variations in cognate forms between | as in dvikṣat (AV.) — *dviş-şat, aor. of dviş- i u and i u 7 in some other words see 'hate', is also an Indian innovation. WACKERNAGEL I, 86. 3 On double consonants in Sandhi, see below 77. 4 The evidence of the Avestan form varadka- shows that the double consonant in vṛkká- (AV. VS.) ‘kidney' is due to assimilation. 5 See below 40; WACKERNAGEL 1, 133. 6 See TPr. XIV. 8. 2 Sometimes a single s represents the double sound, as in ási 'thou art' (= as-si); apásu (RV. VIII. 4¹4), loc. pl. of apás- 'active'; ámhasu (AV.), loc. pl. of ámhas- 'distress'; jo-si 'thou shalt taste' (juş-); probably also in gho-si (from ghus- 'sound'), in uş-ás, gen. sing., acc. pl. of us, weak stem of uş-ás- 'dawn' (for uss-as), possibly in usr- 'dawn' (for
- uss-r-). As the singles in such forms is
shown by cognate languages also, it seems here to be pre-Vedic, and the double ss in forms like rájas-su, loc. pl., is probably an Indian innovation. The change of ss to ks, A 7 See WEBER, Abh. d. Berliner Ak. d. Wiss. 1871, p. 83 f. 8 See below 46; cp. WACKERNAGEL I, 279 a. 9 Cp. ROTH in ZDMG. 48, 102 f. 10 See his Lexicon, s. v. upástha-. 11 See WACKERNAGEL I, 98 b, note.
