The Vedic Word for "King"

Abstract

OUR INDO-EUROPEAN REFERENCE WORKS list the standard equation Vedic rdjking, Latin rix king and Old Irish rY (gen. rig), going back to an IE reking. rdjan king, much more common in Vedic and classical Sanskrit, regarded as an Indian innovation that replaced archaic raj-.' The seemingly obvious connection with the Latin verb rego order, direct was rendered doubtful through the observation that the verb means in older Latin only send/go/arrange in a straight line that still evident in classical Latin r&,tus straight.2 J. Gonda3 and J. Santucci4 have tried to save this connection with the assumption that the root reg originally meant ausbreiten, (sich aus)strecken. Gonda explained the title *rjg as (der M ittler), der sich oder seine Arme (schUtzend, helfend, gebend, segnend) ausstreckt und seine Macht (Wtrde, Reichtum) uber Land und Volk ausbreitet.5 E. Benveniste6 connected rex with rego as celui qui trace la ligne, la voie a suivre. A. Sihler7 has pointed out problems of meaning and of form with Gonda's explanation, agreeing at the same time with Gonda's critique of earlier attempts. Gonda's 'original' meaning is not actually to seen in any of the reflexes of *reg-, and the 'reach out' meanings are visible only in Latin in verbs like porrigo, there the meaning presumably the result of the preverb.8 On the formal side, it striking that the lengthened form with /e/ found in all existing derived forms. If rjgwere a root noun with vrddhi, we would expect to see derived forms with the shorter vowel form-but there are none. The lengthened vowel supposedly was generalized in vom Nom. Akk. Sing. aus durchgefUhrter Dehnstufe.9 Sihler proposes to see in all these words for king derivatives of a root *reHig be efficacious, have mana; 10 a root noun in full grade would appear in Sanskrit as rdj(nom. sing. rat), in Latin as rig(nom. sing. rex). The weak grade of the root, *rHij, would appear in Sanskrit as urj-, which Sihler recognizes in the feminine noun u-rjstrength, nourishment. Since the paradigm of such a root noun should show ablaut comparable to that of nom. sg. anadvdn ox, gen. sg. anacluhas, Sihler posits a hypothetical pair $rat: $iirids, where the symbol $ means synchronically incorrect.'1 Sihler failed to complement his elegant analysis with an investigation of the Vedic texts. The noun ria occurs only three times in the RV-each time in the nom. sg., and at the end of a second pada of a tristuhh-a limitation that indicates archaic usage.