Suomen nominien taivutuksesta: rytmi-, sivupaino- ja agglutinaatiohypoteesien testaamista

• Alexandre Nikolaev
• Jussi Niemi

Avainsanat: korpuslingvistiikka, nominit, optimaalisuusteoria, taivutus

Abstrakti

Inflection of Finnish nominals: Testing of rhythm, secondary stress and agglutination hypotheses (englanti)

4/2005 (109)

INFLECTION OF FINNISH NOMINALS: TESTING OF RHYTHM, SECONDARY STRESS AND AGGLUTINATION HYPOTHESES

The article analyses the free variation claimed to be typical in plural genitives and partitives of Finnish trisyllabic nominals. On the basis of a large text corpus (approx. 109 million words), the writers calculated the surface frequencies of lexemes in certain trisyllabic paradigms classified by the Suomen kielen perussanakirja (A basic dictionary of Finnish) as containing highly variable competing forms. These were paradigm 2 (example: palvelu service), paradigm 4 (laatikko box/drawer), paradigm 13 (katiska weir/(fish) trap), paradigm 14 (solakka slim), paradigm 6 (paperi paper) and paradigm 11 (omena apple). Each paradigm was then divided into sub-groups on the basis of the mora values of the syllables, since the writers assumed that what appeared to be free variation was not entirely free but guided by a number of principles that are reflected in, for example, syllable weight.

The variation was analysed with the aid of three hypotheses covering rhythm, secondary stress and agglutination. Each of these factors presents its own constraints on the eventual surface forms of the paradigms, and violations of these constraints are manifested as a lowered degree of productivity. This, in turn, is reflected in the preferential use of different forms by native speakers and through the relative frequency of use of the competing forms.

The rhythm hypothesis may be seen as an inherent principle of spoken Finnish, and it turned out to be the primary explanation in a majority of the instances studied. The secondary stress hypothesis was in some cases the primary and in others the secondary explanation of the variation at hand, while the agglutination hypothesis was applicable only to the inflectional paradigms of laatikko and paperi, where, in any case, its role remained secondary. By applying these hypotheses to an optimality theory analysis and comparing the outcome of this with the frequencies determined from the corpus analysis, the writers also investigate other structural factors affecting the rank order of the paradigmatic variants.

Alexandre Nikolaev & Jussi Niemi