________________
INDO-EUROPEAN SOUND SYNTHESIS
3
or external sandhi also developed in the proto IndoEuropean speech, but it was not com pulsory. The earliest evidencos in IE historical lao guages show that sandhi in compounds was also originally optional; cpvd yuktaaśvaḥ, Ay yuxta-aspo, 'yoked-horse OCS dobro-oků *beautiful-eyed, etc. (vide Brugmann : CGIGL I p. 454) boside Skt yuktas vaḥ, Av yuxtaspo etc.
Sentence sandhi was more optional in IE and it was independently operative in the historical languages, for a pretty long time.
In a very early stage of IE even internal sandhi was optional. This is to be assumed because of the two different types of dialectal treatment of IE voiced aspirate
+-t & s ( vide below 14, 15, 16), which have conti nued for some time before it could give rise to the twodifferent types of forms like bhit-tos (14) & bhidd hos ( 15 ). But in course of time, towards a later stage of IE, internal sandhi was compulsory, although at that stage, sandhi in compounds was not compulsory. Its optional character continued upto to an older stage of the individual historical languages.
In internal sandhi also combination of a root and a suffix to form a stem and combination of a stem with a case-ending should preferably be treated as two different categories of sandhi, from the point of view of history and development of sandhi. In one stage of proto IndoEuropean, sandhi of a root and a suffix, to form a stem, was more rigid than sandhi of a stem and a case-ending Vedic and Avestan, which represent two very old IE historical languages, sometimes show forms, where caseendings are used without sandhi; cp early RV parastaat (VI 54.90) for late RV (and cl) parastāt (vide RV X.
Jain Education International
For Private & Personal Use Only
www.jainelibrary.org
