|Title||Revisiting ‘agglutinative’ infection in Cappadocian Greek|
In this paper, I revisit the synchronic status and the diachronic development of ‘agglutinative’ inflection in Cappadocian Greek. The term refers to the use of the endings -ιου and -ια that are characteristic of the ι-neuter inflectional class, which prototypically contains inanimate nouns such as σπιτ ‘house’ or φτι ‘ear’, to form the genitive singular/plural and the nominative/accusative plural of nouns that do not historically belong to the ι-neuter class. Compare, for example, the inflection of ζωνάρ ‘belt’, a historical ι-neuter noun, with that of γύπνος ‘sleep’ in Axó Cappadocian (1a,b) (data from Mavrochalyvidis & Kesisoglou 1960). Note that in other Modern Greek dialects the cognates of γύπνος belong to the ος-masculine inflectional class.
b. Axó Cappadocia
Cappadocian ‘agglutinative’ inflection is most commonly treated as the outcome of heavy structural borrowing that resulted from language contact with Turkish (Matras 2009; Thomason & Kaufman 1988; Winford 2005). -ιου and -ια in (1b) are analysed as single exponents expressing solely genitive case and plural number respectively in a fashion similar to the equivalent Turkish endings -nun and -lar in the inflection of uyku ‘sleep’ (1c). The unit γυπνος to which -ιου and -ια attach in inflection is further taken to be a free base, having the same structural properties as Turkish uyku (Janse 2009; Ralli 2009). This approach is, however, problematic, in that it relies too heavily on the superficial structural similarity and linear intermorphemic correspondence between inflected forms in Cappadocian and Turkish, which are employed as evidence to establish language contact with the latter as the single cause for the development of ‘agglutinative’ inflection in the former.
In this paper, I take issue with this approach. From a synchronic point of view, I argue that noun paradigms such as that of γύπνος are not actually agglutinative. -ιου and -ια are cumulative exponents expressing a bundle of morphosyntactic features (case, number, inflectional class) whereas the unit serving as the basis of inflection is best analysed as a bound stem when examined in the context of the system defining properties of noun inflection in Cappadocian. Considering that the case/number combinations making up the nominal paradigm are expressed by the same set of endings in the inflection of both ζωνάρ and γύπνος and that the inflected forms of both nouns are structured upon the same type of stem allomorph, I take the two nouns as belonging to the same inflectional class, namely the ι-neuter one. By extension, I make the claim that all nouns that inflect according to the ι-neuter inflectional class belong to it, irrespective of their former inflectional class membership. From a diachronic point of view, I interpret the paradigms of nouns such as γύπνος as evidence of historical shifts to the ι-neuter inflectional class. Based on data from Cappadocian varieties in which ‘agglutinative’ forms have the most limited distribution, I propose that the early shifts to the ι-neuter class were the result of the combined effect of gender class prototypicality and of the wider tendency for inanimate nouns in Modern Greek to belong to the neuter gender and the ι-neuter inflectional class in particular. I therefore suggest that shifts were triggered in order to repair prototypicality deviations within the non-neuter inflectional classes by assigning inanimate nouns to the semantically appropriate and morphologically productive ι-neuter class.
|Conference||10th International Conference on Greek Linguistics|
|Accepted author manuscript||karatsareas_icgl10_slides.key.pdf|