Nyelvtudományi Közlemények 96. kötet (1998-1999)

Tanulmányok - Dezső László: Typological Comparison of Root Structuring in Uralic and Early Indo-European. [Az uráli és korai indoeurópai tőstruktúrák tipológiai összevetése] 3

28 LASZLO DEZSŐ unstressed position according to Rédei's version. When one compares the System of normal vowels: i, e, ä and u, o, e, a to that of reduced ones e, ä, a, then i, u, o lack in the latter and only the e can „stand for them" being a middle sound. In Sammalahti's version the normal vowels are front i, ü, e, ä and backy, u, o, a (à) opposed to reduced ones: i, y, ä, a (à), here instead of e one fínds i and y which admits a complète front-back vowel harmony. Two generál conclusions can be made from the présentation of Uralic. First, the wide phonetic range of e, the variants of which may or may not become phonèmes. Second, the uncertainty of phonological analysis within the range of e sounds; it is particularly clear in the case of the unstressed e within the unaccented vowels and in the entire system of vowels according to Rédei's reconstruction on which the UEW is based. It seems likely that UEW followed truly the reconstruction of phonetic realization of etymons; in Sammalahti's version the phonological reasoning seems to dominate, not without phonetic foundation however. In Indo-European behind the rather unusual frequency of e as V, there are manifold realizations similar to those of Uralic e sounds. Among the other possible realizations of V there was also an a and an o sound which appear under certain phonetic conditions. Another characteristic of the earliest reconstructable stage of Indo-European is the lack of a clear phonemic system. If one considers that in accented position most vowel phonèmes of Uralic are clearly distinguishable, but there are considérable différences in the interprétation of unaccented vowels, which, however, effects the reconstruction of the whole vowel System, then the phonemic évaluation of Indo-European will seem less stränge. The scarcity of early clear vowels without apophony, the dominance of apophonic variants with unstressed, reduced root vowels, the strong influence of neighbouring consonants on the vowels of accented and, especially, of unaccented syllables, the frequency of an unclear e explain the uncertainty of phonetic reconstruction and its phonemic interprétation. Lehmann (1993: 138-139) does not assume phonèmes for the earliest stage, even the phonetic shape of vowels appears unspecifíed when generalized in the évaluation of vowel and accent system. For the earliest stage A he reconstructs three sets of consonants (obstruents, résonants, laryngeals) and nonsegmental syllabicity. In stage B under maximum stress a vowel [e] appears, under minimum stress the vocalic allophones of résonants, to them an [e:] is added if an [e] stands before a syllable which has lost its syllabicity. Phonèmes appear only in stage C: [e e: ë], of which [e] has allophones in the neighbourhood of vowel-coloring laryngeals [e a o]. We can treat Pre-Indo-European as a reconstructed System which will never represent a concrète functioning language; however, one should not forget that our objective is to know as much as possible about its functioning even in pho-