Nyelvtudományi Közlemények 91. kötet (1990)

Tanulmányok - Larsson, Lars-Gunnar: The Origin of Mordvin M šaba, E žaba ’child’ and Cheremis šußo 147

148 LARS-GUNNAR LARSSON to the Volgaic languages by the Mishar Tatars. Not even Paasonen himself seems, however, to be convinced by the etymology, although he emphasizes that the Mordvin word gives the impression of being foreign due to the non-initial -6-, a sound that, according to him, does not exist in indigenous Mordvin words (cf. also Paasonen 1893. 15f.). I have not been able to find any such Mishar Tatar word; it is not attested in Budagov (1869), nor is there any word *zaba meaning 'dwarf in Radioff's dictionary. Even if such a word were to exist, the semantic development 'dwarf —• 'child' would seem dubious. Being a child is, after all, a more normal state than that of being a dwarf. Cygankin-Mosin (1977. 20), however, seem to accept the etymology suggested by Paasonen and report mdM saba, E zaba as being of Turkic origin. Genetz (1897.39) compares Mordvin saba, zaba with Cheremis "sa6a heikko" (cf. sava 'weak' in MRS) and combines them with Finnish hievukka 'weak, thin' and Finnish sievonen 'nice, clean'. Such a comparison should, however, be discarded since it would not represent any regulär sound development (cf. Finnish liemi 'broth, soup'/md. /em/cher. lem; FUV). Instead of the Mishar Tatar origin of mdM saba, E zaba and Cher. sußo suggested by Paasonen (1897.51) and accepted by Cygankin-Mosin (1977. 20), a word in Sanskrit seems better suited as origin. Skr. säva-, also attested in a diminutive form sävaka-, means 'the young of any animal' (Monier-Williams). Both the phonetic shape of the word and its meaning suggest a comparison with mdE zaba, M saba and Cher. sußo. It is well-known that there are Indo-Iranian loanwords in the Volgaic languages, both such as are usually assumed to have entered Proto-Finno-Ugric, e.g. mdE sado, M sadä/Cher. sű6ö '100', and such as have come into use by later contacts, e.g. mdE ksni, ksiie/Cher. kartni, kürtni 'iron'. The latter word, to which there are corresponding forms also in Permian languages and in Vogul, is regarded by Korenchy (1972. 58f.) and Joki (1973. 273) as being an example of later separate contacts between Iranian and Volgaic peoples. To support the idea of mdM saba, E zaba and Cher, sußo 'child' as an Indo-Iranian loanword, words correponding to Skr. säva- should be searched for. This, however, is a rather complicated matter. According to Walde-Pokorny (1959) Skr. säva- 'das Junge eines Tieres' dérives from an Indo-European stem *k'eu- 'schwellen, Schwelbung, Wölbung; Höhlung, hohl'. They do not supply any direct etymologically corresponding form from any other Indo-European language, although the stem is présent in several other words, all of which Walde-Pokorny (1959) group under the basic meaning 'Wölbung nach aussen oder innen'. According to Peterson (1916.139), however, Skr. säva- 'the young of any animal' could be combined with a Nyelvtudományi Közlemények 91. 1990.