Periodic Reporting for period 1 - ConfliGram (Conflicting Grammars: Gender among bilinguals in Mesoamerica and the Caucasus)
Reporting period: 2019-07-01 to 2021-06-30
The majority of the world's population speaks at least two languages, therefore it is vital that we use multilingual data to develop sound theories of language comprehension, processing and use. How individuals combine their languages in the same sentence or conversation - a practice known as code-switching - also helps us to understand the limits and possibilities inherent to human language.
The main aim of the “ConfliGram” project was to investigate how an individual who speaks two languages, one possessing grammatical gender and one without, combines the two languages in a mixed nominal construction. To do this, I collected naturalistic and experimental data from two language pairs: Purepecha-Spanish, whose speakers live in Michoacán in western Mexico, and Tsova-Tush--Georgian, whose speakers can be found in one village in eastern Georgia. In both cases, one of the languages can be considered a minority language, Purepecha (isolate) and Tsova-Tush (Nakh-Daghestanian), while the other is the more dominant, national language, namely Spanish (Indo-European) and Georgian (Kartvelian) respectively. In the case of Purepecha-Spanish, the national language possesses a binary gender system (masculine/feminine), whereas in the Tsova-Tush--Georgian case, it is the minority language that has a gender system (five genders: masculine animate, feminine animate and three 'neuter' genders).
Previous research shows that speakers of different language pairs deal with this grammatical conflict in different ways. This project therefore sought to answer the following research questions:
i. What strategy or strategies do Purepecha-Spanish and Tsova-Tush--Georgian use for assigning gender to otherwise genderless nouns?
ii. What role do age, gender, age of acquisition, language dominance and the sociolinguistic status of the languages play in the strategy or strategies adopted by these bilinguals? How do social and linguistic factors interact to affect and predict gender assignment strategies?
iii. What implications do the findings of this project have for current theories of code-switching, in terms of production and comprehension? How can they help to refine current predictions of code-switching and bilingual grammar?
By recognising code-switching as a natural, yet complex, phenomenon, we can also help to destigmatise it, as well as its occurrence in minority language contexts.
The main aim of the “ConfliGram” project was to investigate how an individual who speaks two languages, one possessing grammatical gender and one without, combines the two languages in a mixed nominal construction. To do this, I collected naturalistic and experimental data from two language pairs: Purepecha-Spanish, whose speakers live in Michoacán in western Mexico, and Tsova-Tush--Georgian, whose speakers can be found in one village in eastern Georgia. In both cases, one of the languages can be considered a minority language, Purepecha (isolate) and Tsova-Tush (Nakh-Daghestanian), while the other is the more dominant, national language, namely Spanish (Indo-European) and Georgian (Kartvelian) respectively. In the case of Purepecha-Spanish, the national language possesses a binary gender system (masculine/feminine), whereas in the Tsova-Tush--Georgian case, it is the minority language that has a gender system (five genders: masculine animate, feminine animate and three 'neuter' genders).
Previous research shows that speakers of different language pairs deal with this grammatical conflict in different ways. This project therefore sought to answer the following research questions:
i. What strategy or strategies do Purepecha-Spanish and Tsova-Tush--Georgian use for assigning gender to otherwise genderless nouns?
ii. What role do age, gender, age of acquisition, language dominance and the sociolinguistic status of the languages play in the strategy or strategies adopted by these bilinguals? How do social and linguistic factors interact to affect and predict gender assignment strategies?
iii. What implications do the findings of this project have for current theories of code-switching, in terms of production and comprehension? How can they help to refine current predictions of code-switching and bilingual grammar?
By recognising code-switching as a natural, yet complex, phenomenon, we can also help to destigmatise it, as well as its occurrence in minority language contexts.
I have presented findings from my code-switching studies, as well as recent work on the stability of the gender systems in the Nakh-Daghestanian languages, at nine international conferences. The most recent talk actually takes the form of a short documentary film and includes native Purepecha speakers, speaking in Purepecha about their experiences with the Intercultural Orchestra of Leon.
From late October to early December 2019 I conducted linguistic fieldwork on Purepecha in the region of Michoacán known as the Cañada, or Valley of the Eleven Villages. I collected around ten hours of naturalistic speech from individuals in the villages of Carapan, Ichán, Santo Tomás, Tacuro and Zopoco. This trip enabled me to re-connect with previous participants, as well as to forge new connections with native speakers, including both students and a teacher. I also presented on the importance of documenting a (minority) language to native speakers interested in language revitalisation work.
In July 2019, I conducted fieldwork on Tsova-Tush and Georgian in the village of Zemo Alvani, Georgia. Together with my co-author Jesse Wichers Schreur, we recorded director-matcher and acceptability judgement tasks in the homes of our participants. Wichers Schreur and I have expanded this work on gender assignment to include research into the stability of the gender systems in Nakh-Daghestanian languages, and developing a predictive model of gender assignment in the Nakh languages, in collaboration with Dr. Marc Allassonnière-Tang (CNRS/University of Lyon 2) and Neige Rochant (Lacito). Data analysis is on-going, with work on writing up the results as a journal article due to begin in July.
In order to enhance my skills and knowledge base, I followed, or continue to follow a number of online courses, including language competence and statistical analysis. I have also attended the Code-switching Reading Group at the University of Cambridge (UK) and the Heritage Languages Reading Group, Anglia Ruskin University (UK) in order to improve my subject knowledge and to broaden my network. In both cases, my online participation included presenting articles and chairing sessions.
The other main project output is a corpus of naturalistic Purepecha speech, which will shortly become available in the Pangloss archive. It comprises around ten hours of speech from eleven individuals, including seven women, aged 27 to around 80. This will be a valuable resource for Purepecha speakers and learners, as well as for the academic community.
From late October to early December 2019 I conducted linguistic fieldwork on Purepecha in the region of Michoacán known as the Cañada, or Valley of the Eleven Villages. I collected around ten hours of naturalistic speech from individuals in the villages of Carapan, Ichán, Santo Tomás, Tacuro and Zopoco. This trip enabled me to re-connect with previous participants, as well as to forge new connections with native speakers, including both students and a teacher. I also presented on the importance of documenting a (minority) language to native speakers interested in language revitalisation work.
In July 2019, I conducted fieldwork on Tsova-Tush and Georgian in the village of Zemo Alvani, Georgia. Together with my co-author Jesse Wichers Schreur, we recorded director-matcher and acceptability judgement tasks in the homes of our participants. Wichers Schreur and I have expanded this work on gender assignment to include research into the stability of the gender systems in Nakh-Daghestanian languages, and developing a predictive model of gender assignment in the Nakh languages, in collaboration with Dr. Marc Allassonnière-Tang (CNRS/University of Lyon 2) and Neige Rochant (Lacito). Data analysis is on-going, with work on writing up the results as a journal article due to begin in July.
In order to enhance my skills and knowledge base, I followed, or continue to follow a number of online courses, including language competence and statistical analysis. I have also attended the Code-switching Reading Group at the University of Cambridge (UK) and the Heritage Languages Reading Group, Anglia Ruskin University (UK) in order to improve my subject knowledge and to broaden my network. In both cases, my online participation included presenting articles and chairing sessions.
The other main project output is a corpus of naturalistic Purepecha speech, which will shortly become available in the Pangloss archive. It comprises around ten hours of speech from eleven individuals, including seven women, aged 27 to around 80. This will be a valuable resource for Purepecha speakers and learners, as well as for the academic community.
In order to progress beyond the current state of knowledge regarding code-switching (CS) and grammatical gender, as well as to further our understanding of the languages under investigation, I have published or submitted a total of nine journal articles and book chapters.
The Purepecha course I have been following with the Instituto de Enseñanza de Lenguas Originarias (Mexico) has been particularly fruitful, since I take a more pedagogical role, supporting the instructor in teaching different grammar topics, including writing and distributing handouts. The instructor and I have submitted an abstract for a chapter on variation and change in the Purepecha diaspora to a special volume edited by colleagues at the Centre for Indigenous American Studies at Leiden University, and he has also invited me to co-author a Purepecha textbook for second language learners. The development of more, and more varied, teaching materials is important for minority languages, since it offers greater legitimacy as well as supporting a growing number of learners with different backgrounds and needs.
Reagarding scientific results, the Tsova--Tush-Georgian study supports the prediction that L1 speakers of a gendered language prefer a translation equivalent strategy. Moreover, phonological criteria display a stronger role in gender assignment than previously found. Noun frequency in also partially explains inter-speaker and inter-stimulus variation and consistency, providing a plausible pathway from code-switches to borrowings. Moreover, no role was found for the variables of age or gender. The results did, however, contradict the previously posited relationship between the default strategy and language dominance and code-switching frequency. Order of acquisition of the languages involved also seems to play a role, as previously suggested: the translation equivalent is preferred by speakers who acquired the gendered language first. The key take-home message is that the field still lacks a huge amount of data in order to be able to make predictions with wide-ranging coverage. The field needs to expand into more language dyads, as well as extend research on individual language dyads, using multiple methodologies and in communities differing in their code-switching frequency. The results of this project also indicate that both linguistic and extra-linguistic factors need to be included in any possible predictive model of gender assignment.
The Purepecha course I have been following with the Instituto de Enseñanza de Lenguas Originarias (Mexico) has been particularly fruitful, since I take a more pedagogical role, supporting the instructor in teaching different grammar topics, including writing and distributing handouts. The instructor and I have submitted an abstract for a chapter on variation and change in the Purepecha diaspora to a special volume edited by colleagues at the Centre for Indigenous American Studies at Leiden University, and he has also invited me to co-author a Purepecha textbook for second language learners. The development of more, and more varied, teaching materials is important for minority languages, since it offers greater legitimacy as well as supporting a growing number of learners with different backgrounds and needs.
Reagarding scientific results, the Tsova--Tush-Georgian study supports the prediction that L1 speakers of a gendered language prefer a translation equivalent strategy. Moreover, phonological criteria display a stronger role in gender assignment than previously found. Noun frequency in also partially explains inter-speaker and inter-stimulus variation and consistency, providing a plausible pathway from code-switches to borrowings. Moreover, no role was found for the variables of age or gender. The results did, however, contradict the previously posited relationship between the default strategy and language dominance and code-switching frequency. Order of acquisition of the languages involved also seems to play a role, as previously suggested: the translation equivalent is preferred by speakers who acquired the gendered language first. The key take-home message is that the field still lacks a huge amount of data in order to be able to make predictions with wide-ranging coverage. The field needs to expand into more language dyads, as well as extend research on individual language dyads, using multiple methodologies and in communities differing in their code-switching frequency. The results of this project also indicate that both linguistic and extra-linguistic factors need to be included in any possible predictive model of gender assignment.