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

LÁSZLÓ DEZSŐ to Finno-Ugric, but it requires further examination (Sammalahti 1988: 482, 486). The loss of laryngeals, however, is not the only source of long vowels. At présent, laryngeals are not reconstructed for the Altaic group and their présence is also problematic in Uralic. The position of accent differentiated the languages of West Eurasia: in Uralic and the Altaic group accent was fixed on the first syllable; in Indo-European it became variable. It is relatively easy to account for the conséquences of a fixed accent. If it is variable, a complex System of changes will arise. If one considers the situation in Modern Slavic languages the situation in the area will become more compréhensible. Slavic languages with fixed accent on the first syllable without a change in vowel quality are opposed to Russian with variable accent and with degrees of réduction combined with change of quality. 1.3. Roots in Indo-European Benveniste's model is abstract, useful also for cross-linguistic comparison and specific enough to présent the phenomena of Indo-European. Its starting point is the CeC structure, which is the only two-consonant scheme possible in Pre-Indo-European, then two further chains are reached by the extension by a formative and the apophonic réduction of the unstressed vowel: CeC-C and CC-eC (where the symbol e usually stands for e, but can represent other vowels). The chain can be extended further, but we shall not follow it (cf. Lehmann 1993: 94). The extended variants can be interpreted as new roots CCVC or CVCC with their rules of structuring, which mostly concern the consonants and will be analyzed later. The apophonic system was not complète in Pre-IE. In Proto-Uralic the CVCV structure was prévalent, which differs from the Indo-European two-consonant root because of the présence of a final root vowel. As we have seen, the Uralic CVCCV structures could arise from the conjunction of a root dement CVCV and a formative CV; they gave rise to CVCCV structure after the loss of the root final vowel of the first. We can rewrite the Uralic structures according to Benveniste's model: CeCe and CeCe-Ce > CeC-Ce; we have the loss of unstressed vowel here also, but the fixed accent does not permit the variant CCeC which is the condition for a complète ablaut system in Indo-European. Gamkrelidze and Ivanov (1995: 185-9) présent seven structural types of morphèmes of roots and suffixes of late Indo-European. The types are presented according to canonic forais, in which C° dénotes any consonant, C = obstruent, R = sonant, H = laryngeal. I shall give each type with restrictions established from the possible realization of the type; then I shall make some conclusions relevant to my treatment of the topic. No concrète examples will be given here. These types are more concrète than the root classes of Uralic but the classification is not accompanied by statistics.