Discussant : 

Nina Sumbatova, Université Russe d'Etat des Humanités, Moscou

Abstract :

Grammatical gender is a pervasive category in the Nakh-Daghestanian languages (Comrie, 2013); its members possess between two and five gender classes, which can be marked on certain verbs (agreeing with the nominative argument), as well as on some adjectives, adverbs and numerals (agreeing with their head). Gender assignment principles are specific for each language, but are known to be primarily based on semantic factors, while phonology may be a secondary influence. A typical four-gender Nakh-Daghestanian language features a gender for male humans (M), one for female humans (F, see (1a)), and two for animals and inanimates (see (1b)), here named for their agreement markers: B, D, J. The latter two classes display more fine-grained sub-divisions, such as animals in the third class and liquids in the fourth (Klimov 1978: 67; Corbett 1991: 25ff).

(1a)    Sanzhi Dargwa (Forker, 2020)

r-uqna aba               le-r=de                di-la

F-old mother(F)        exist-F=PST        1SG-GEN

‘My old mother was alive.’

(1b)    Tsova-Tush (Hauk & Harris, forthc.)

tišiⁿ          c’a                         daħ           d-ox-d-Ø-o-t                   ve

old          house(D)                 PV            D-destroy-D-TR-PRS-PL 1PL.INCL

‘We are tearing down the old house’.

However, the gender of a given noun is not necessarily constant: it can differ between languages of the same family, and it can also change over time (e.g. Van Epps et al., forthc.). Such change is well known in the Indo-European languages, as in Sanskrit cakrás (masc.) vs. Proto-Germanic *hweγwlán (neut.), from the same root, both meaning ‘wheel’. We also find alternations for shared concepts such as ‘moon’, which is masculine in the Germanic languages (viz. der Mond in German) but feminine in the Romance languages (la luna in Spanish and Italian; see, e.g., Matasović, 2014). Yet the motivation for these changes is still poorly understood: are they connected to a specific gender class, to particular semantic domains or words, or are they related to frequency and therefore usage? (e.g. Carling, forthc.).

In this paper we investigate the distribution and stability of gender classes across 18 Nakh-Dagestanian languages, in both inherited and borrowed lexemes (n = 1991). The dataset constitutes wordlists of 95 cultural vocabulary items, classified for semantic domain according to colexification and meaning change within the 104-lexeme culture list of the Mouton Atlas of Languages and Cultures (Carling, 2019). Gender class is standardised across the languages, based on the historical development of the gender markers within the family. The majority of lexemes are also coded either for cognacy or as loanwords, with the provenance of the latter indicated where known.

A preliminary descriptive statistical analysis shows that B gender is the most frequently occurring gender for all lexemes across all languages (24%), followed by N (the equivalent of B in languages with only three genders; 17%), D (14%), then J (7%). Loanwords are also over-represented in the B gender class, followed by D, and N (however, see Wichers Schreur, forthc. for a different situation in Tsova-Tush). It is unsurprising, then, that the gender classes are not evenly distributed amongst semantic classes, or over individual word meanings. Applying a preliminary analysis to the data, using a log-likelihood method to measure the overrepresentation of a specific gender in a subsection of the data (compared to the complete data), we observe significant over-representation of B gender in the semantic category ‘domestic animals’ (LL = 6.33) and ‘cattle’ (LL = 3.43). We envisage a similar scenario for the classes of ‘draught animals’, ‘game’, ‘poultry’ and ‘small animals’. In contrast, D gender is highly significantly over-represented for ‘metals’ (LL = 15.26) and ‘seasons’ (LL = 10.68). Weapons are also mostly classed as D gender (see Carling, forthc., on the over-representation of feminine gender in the ‘weapons’ category in Indo-European).

These over-represented gender classes will enable us to propose hypotheses concerning family-internal processes of gender assignment and change, based on semantic domain, borrowability and cognacy. In turn, these scenarios will contribute to our understanding of the principles of gender assignment, and the historical development of gender, in Nakh-Daghestanian specifically, as well as from a broader typological perspective.

References :

Carling, Gerd. Forthcoming. A dangerous story: the linguistic behaviour of the category sharp cutting implements. In: Romain Garnier (ed.), Loanwords and Substrata: Proceedings of the Colloquium held in Limoges (June 5th—7th 2018), Innsbruck: Innsbrucker Beiträge zur Sprachwissenschaft.

Carling, Gerd. 2019. Mouton Atlas of Languages and Cultures. Vol. 1: Europe and West, Central and South Asia. Berlin/Boston: Mouton de Gruyter.

Corbett, Greville G. 2013. Number of Genders. In: Matthew S. Dryer & Martin Haspelmath, Martin (eds.), The World Atlas of Language Structures Online. Leipzig: Max Planck Institute for Evolutionary Anthropology. (Available online at http://wals.info/chapter/30, Accessed on 2020-04-30.)

Corbett, Greville G. 1991. Gender. Cambridge: Cambridge University Press

Forker, Diana. 2020. A grammar of Sanzhi Dargwa. Berlin: Language Science Press.

Matasović, Ranko. 2014. Gender in Indo-European, Heidelberg: Winter.

Hauk, Bryn & Alice C. Harris. Forthcoming. Batsbi. In: Yuri Koryakov, Yury Lander & Timur Maisak (eds.), The Caucasian languages: An international handbook. Berlin: De Gruyter Mouton.

Klimov, G. A. 1978. Strukturnye obščnosti kavkazskix jazykov. Moscow: Nauka.

Van Epps, Briana, Gerd Carling, and Yair Sapir. In press. 'Gender assignment in six North Scandinavian languages: Patterns of variation and change', Journal of Germanic Linguistics.

Wichers Schreur, Jesse. Forthcoming. Nominal borrowings in Batsbi (Nakh-Daghestanian, Georgia) and their gender assignment. In: Forker, Diana & Lenore A. Grenoble (eds.) Language contact in the territory of the former Soviet Union.

Biography : 

Gerd Carling is Associate Professor of General Linguistics at Lund University, Sweden. Her research interests target historical linguistics and language evolution, in particular on Indo-European and South American languages. She has initiated a research program on language diversity and evolution at Lund University and founded the database infrastructure and lab DiACL – Diachronic Atlas of Comparative Linguistics, which harbors data from a large amount of languages worldwide. Recent publications include the volume Mouton Atlas of Languages and Cultures (De Gruyter 2019), A Dictionary and Thesaurus of Tocharian A (Brill/Harrassowitz, coming 2021) and several articles in journals such as Linguistic Typology, PLOS ONE, Language, Philosophical Transactions of the Royal Society B, Diachronica, and Journal of Indo-European Studies.

Kate Bellamy is currently a Marie Curie Postdoctoral Research Fellow at Lacito, CNRS (France). Her project focuses on gender assignment in mixed nominal constructions among bilingual communities in Mexico and Georgia. As well as code-switching and grammatical gender, Kate also conducts research into the Purepecha language from language-internal and language contact perspectives.

Jesse Wichers Schreur is a PhD student at the Goethe University Frankfurt and École Pratique des Hautes Études in Paris. His research focusses on languages contact in the Caucasus, and historical linguistics of both the Kartvelian and the East Caucasian families. Furthermore, he is working at Linked Open Dictionaries project, which aims to provide a computational toolset for linguists working with Caucasian dictionaries and corpora.