Compound agent nouns in Russian: A comparison of rival word-formation constructions

The aim of this dissertation is to analyze and compare rival word-formation constructions giving rise to compound agent nouns in modern Russian. Given the almost total lack of investigations on compounds in Russian, this research aims to partially fill this gap in the literature by focusing on one of the most productive group of compound constructions in Russian, i.e. synthetic agentive compounds (e.g. basn-o-pis-ec ‘fable writer’). Apart from the descriptive aim consisting in a thorough analysis of the formal and semantic features of such compounds, this study also exploits quantitative corpus-based methods to investigate the distribution and productivity of rival word-formation constructions giving rise to synthetic agentive compounds, and thus contributes to demonstrating the importance of quantitative investigations in studies on word-formation. The first part of this dissertation is mainly concerned with theoretical questions regarding the phenomenon of compounding in Russian (Chapter 1), the constructionist approach to compounding (Chapter 2), and the polysemy of agentive constructions (Chapter 3). The second part is devoted instead to qualitative and quantitative investigations of the semantics, distribution, and productivity of the rival agentive constructions selected for the analysis (Chapters 4 and 5), i.e. the compound constructions formed with the suffixes -ec, -lec, -tel’, -nik, -ščik/čik, -l’ščik, -ka, -lka, and the suffixless construction (-ø). The analysis highlights some major differences among the rival constructions investigated. The low-frequency constructions in -lec, -l’ščik and -lka show no variation at all and their productivity is restricted to specific constructions based on a limited number or base verbs. By contrast, the high-frequency constructions show greater variation (especially ec, -tel’ and -ø; to a lesser extent, -nik, -ščik/čik and -ka), but the degree of variation depends on the parameter considered. Among all the parameters employed for the analysis, those that better discriminate the distribution of rival constructions are the aspect of the compound’s verbal element (which brings out the different behavior of -tel’ with respect to the other constructions) and the semantics of the compound, (which allows identifying diverging tendencies especially among the high-frequency constructions in -ec, -tel’ and -ø). Rival constructions show diverging tendencies also in terms of diachronic and stylistic distribution, and in terms of productivity. Although determining the synchronic productivity of these constructions appears as a complicated task due to the nature of the data at my disposal, I assess the productivity of rival constructions based on their diachronic developments and on possible restrictions depending on the nature of the compounds’ constituents (i.e. the availability of autonomous deverbals outside compounding and the variability of the verbal bases).