Sociophonetic Variation and Human Interaction with Digital Voice Assistants

Dr. Nicole Holliday, Pomona College

Abstract: As technology that relies on speech in increasingly integrated into modern American society, voice assistants are becoming a more significant part of our everyday lives. This talk will present the results of three studies that focus on social perception of voice assistants, voice quality variation another the assistants themselves, and how one assistant's "tone of voice" evaluation reinforces systematic linguistic bias. Results of the first study demonstrate how listeners  engage in racialized judgements of digital voice assistants and that these judgements interact with perceptions of the personality of such assistants, providing evidence that listeners personify these voices. Results of the second study shed light on the voice quality features that may trigger judgments of speaker race and personal characteristics, even when the speaker is

non-human. Finally, results of the third study show the ways in which speech recognition technology can reinforce and perpetuate bias against already marginalized groups of speakers. A more comprehensive understanding of how sociolinguistic variation interacts with the design of such assistants may help us to understand how listeners process variation and make judgments of voices, both digital and human. Additionally, a thorough analysis of how computational systems police speaker behavior can help us address systematic inequality as the linguistic line between

humans and computers becomes increasingly porous.

Speaker Biography: Nicole Holliday is an assistant professor of linguistics at Pomona College in Claremont, California. She received her Ph.D. in linguistics from New York University in 2016, where she wrote a dissertation entitled “Intonational Variation, Linguistic Style and the Black/Biracial Experience”. Her research focuses on what it means to sound black, both phonetically and socially, and from the perspectives of speakers and listeners, both human and computational. Her work has appeared in scholarly venues such as Journal of Sociolinguistics, Laboratory Phonology, and American Speech. She has made media appearances in outlets such as the New York Times, the Philadelphia Inquirer, and the Washington Post.

Keynote #1: Learner corpora in second language acquisition research

Dr. Nicole Tracy Ventura, West Virginia University

Abstract: This presentation will review how the disciplines of linguistics and language education have traditionally privileged language as an entity without regard for the bilingual speaker.  In so doing, we show how the named language perspective that the disciplines have traditionally adopted have been complicit in producing the racialization of minoritized speakers, as well as their failure in education. By centering minoritized bilingual speakers and the ways in which they do language, we disrupt understandings of language contact and additive bilingualism. These two notions have long created the image of minoritized bilinguals, and in particular Latinx students, as deficient. The concept of translanguaging, as well as translanguaging pedagogical practices, are then offered as alternative conceptualizations that can be helpful to decolonize linguistics and language education.

Speaker Biography: Nicole Tracy-Ventura is assistant professor of applied linguistics at West Virginia University. Her research focuses on second language acquisition, study abroad, task-based language teaching, and corpus linguistics. She is a founding member of the Languages and Social Networks Abroad Project (LANGSNAP) and co-author of the book Anglophone Students Abroad: Identity, Social Relationships, and Language Learning (with Rosamond Mitchell and Kevin McManus – Routledge 2017), which details the LANGSNAP results from 2011-2013. She is currently leading a follow-up study with the LANGSNAP participants investigating the variables that contribute to long-term retention of second languages. Nicole is also co-editor of the Routledge Handbook of Second Language Acquisition and Corpora (with Magali Paquot – 2021).

Keynote #2: Centering minoritized bilingual speakers: Alternative linguistic and educational understandings

Dr. Ofelia García and Dr. Ricardo Otheguy, CUNY Graduate Center

Abstract: This presentation will review how the disciplines of linguistics and language education have traditionally privileged language as an entity without regard for the bilingual speaker.  In so doing, we show how the named language perspective that the disciplines have traditionally adopted have been complicit in producing the racialization of minoritized speakers, as well as their failure in education. By centering minoritized bilingual speakers and the ways in which they do language, we disrupt understandings of language contact and additive bilingualism. These two notions have long created the image of minoritized bilinguals, and in particular Latinx students, as deficient. The concept of translanguaging, as well as translanguaging pedagogical practices, are then offered as alternative conceptualizations that can be helpful to decolonize linguistics and language education.

Dr. García Biography: Ofelia García is internationally known for her work on bilingual education, language policy, multilingualism, and sociology of language. Her concepts of dynamic bilingualism and translanguaging have had a significant impact on understandings of the complex language practices of bi/multilingual students in the twenty-first century. This was the topic of her book Bilingual Education in the 21st Century: A Global Perspective (2009). Most recently, she has focused on bilingualism in formal and informal U.S. educational contexts, and particularly in New York City, coediting, with Zeena Zakharia and Bahar Otcu, Bilingual Community Education and Multilingualism: Beyond Heritage Languages in a Global City (2012) and cowriting Additive Schooling in Subtractive Times: Bilingual Education and Dominican Immigrant Youth in the Heights (2011) with Lesley Bartlett. Among her other books are Educating Emergent Bilinguals (with J. Kleifgen), Handbook of Language and Ethnic Identity, Vol. I & II (with J. A. Fishman), Negotiating Language Policies in Schools (with K. Menken), Imagining Multilingual Schools (with T. Skutnabb-Kangas and M. Torres-Guzmán), and Translanguaging: Language, Bilingualism and Education (with L. Wei). García came to the Graduate Center from Columbia University’s Teachers College and was previously dean of the School of Education at the Brooklyn campus of Long Island University. She holds a Ph.D. in Hispanic and Luso-Brazilian Languages and Literatures from the Graduate Center. García was named to the Hunter College Hall of Fame in 2012. She was the 2012 Wits Claude Leon Distinguished Scholar (University of the Witwatersrand, South Africa) and received the 2008 NYSABE Gladys Correa Award. She has been a Fulbright Scholar, a Spencer Fellow of the U.S. National Academy of Education, and a Fellow of the Stellenbosch Institute for Advanced Study (South Africa).

Dr. Otheguy Biography: Ricardo Otheguy is well known for his work in theoretical and applied linguistics, which has appeared in major international journals such as Language, Language in Society, the Modern Language Journal, and the Harvard Educational Review. His publications in theoretical linguistics are in the areas of language contact, functional grammar, and the Spanish of the United States; in applied linguistics, his publications have been in the area of bilingual education and the teaching of Spanish to native speakers of Spanish. He is coauthor of Spanish in New York: Language Contact, Dialectal Leveling, and Structural Continuity (2012) and was founding editor of the journal Spanish in Context. Otheguy has developed textbook materials for the teaching of Spanish to Latino students in the United States and is coauthor of Tu Mundo: Curso para hispanohablantes. He has also written Spanish materials for English-speaking students and is coauthor of one the most widely used high school Spanish textbooks in the United States, Avancemos. Otheguy has participated in national and international conferences throughout the United States, Europe, and Latin America, and has lectured and conducted research in universities and research centers in numerous countries, including Cuba, Germany, Mexico, Puerto Rico, Spain, and Uruguay. Otheguy, who was appointed to the Graduate Center in 1998, is founding director of the CUNY Research Institute for the Study of Language in Urban Society, which conducts basic and applied research in urban linguistics, bringing to bear the research resources of the City University of New York on urban language issues. He holds a Ph.D. in linguistics from the Graduate Center (1976), as well as degrees and diplomas in Spanish from Louisiana State University, the City College of New York, and the University of Madrid, Spain.

• Otto Santa Ana - Chicana/o Studies, UCLA

• Otto Santa Ana - Chicana/o Studies, UCLA

• Justin Davidson - Spanish and Portuguese, UC Berkeley

Evaluating Minority-Language Agentivity in Catalan Contact Spanish:

Intervocalic Fricatives in Barcelona and Valencia

Dr. Justin Davidson, University of California, Berkeley

                        Though a probabilistic directionality has been posited to account for contact-induced innovation and change in multilingual communities, which predicts a greater strength of linguistic influence for majority languages than for minority languages (Thomason & Kaufman 1988; Thomason 2001; 2008; 2010), cases of contact-induced change observed in the majority language as affected by the minority language (i.e., minority-language source agentivity [Van Coetsem 2000]) remain less studied (Galindo Solé 2003) and often restricted to lexical phenomena (Poplack & Dion 2012; Poplack & Levey 2010). With respect to Catalan-Spanish contact, this asymmetry of contact influence has traditionally been evoked in reference to the fear that Catalan is over time becoming a mere dialect of Spanish (Prats, Rafanell, & Rossich 1990; see also Hualde 2015 for parallels with Basque and Spanish), culminating in the most recent proposal (Arnal 2011) that Catalan-Spanish contact effects are exclusively unidirectional, with modern speakers no longer exhibiting any Catalan phonetic features in their Spanish.

            In this talk, I explore these questions of directionality and minority-language agentivity through two empirical case studies performed in Barcelona and Valencia. The linguistic phenomenon of interest is the variable production of Spanish intervocalic alveolar /s/ (e.g. ‘the friends,’ Spanish las amigos [la.sa.mí.ɣas]; Catalan les amigues [lə.zə.mí.ɣəs]), characterized as a phonetic hallmark of Catalan Contact Spanish (Casanovas Català 1995). Sociolinguistic interviews paired with a word-list reading task in both languages were used to elicit approximately 10,000 intervocalic alveolar fricatives in Spanish and Catalan from 96 Catalan-Spanish bilinguals stratified by gender, age, and language dominance across each city. Fricative tokens, measured in Praat for percentage of each segment’s duration with acoustic voicing correlates (File-Muriel & Brown 2011), were subsequently subjected to mixed-effects regression modeling in R.

            Results reveal that whereas Catalan-like voicing patterns are prevalent in the Spanish of Barcelona, in Valencia the opposite directionality is observed, with Spanish-like voicing patterns being prevalent in Catalan. I account for the observed divergence in fricative production across the two speech communities through an analysis of the unique sociopolitical status of Catalan in each city, notably informed by speaker attitudes and language ideologies elicited in sociolinguistic interviews.

These findings highlight the confluence of linguistic and extralinguistic factors that underpin linguistic variation in multilingual speech communities, and ultimately serve to reinforce the bidirectional nature of language contact in majority- and minority-language settings.

References 

• Arnal, Antoni. 2011. Linguistic changes in the Catalan spoken in Catalonia under new contact conditions. Journal of Language Contact, 4: 5-25.

• Casanovas Català, Montse. 1995. La interferencia fonética en el español de Lleida: Algunos apuntes para su estudio. Sintagma, 7: 53-59.

• File-Muriel, Richard and Earl Brown. 2011. The gradient nature of s-lenition in Caleño Spanish. Language Variation and Change, 23: 223-243.

• Galindo Solé, Mireia. 2003. ‘Language contact phenomena in Catalonia: The influence of Catalan in spoken Castilian’. In Lofti Sayahi (ed.), Selected Proceedings of the First Workshop on Spanish Sociolinguistics. Somerville, MA: Cascadilla Press, 18-29.

• Hualde, José Ignacio. 2015. ‘Basque as an extinct language’. In Beatriz Fernández & Pello Salaburu (eds.), Ibon Sarasola, Gorazarre. Homenatge, Homenaje. Bilbao: University of the Basque Country, 319-326.

• Poplack, Shana and Nathalie Dion. 2012. Myths and facts about loanword development. Language Variation and Change, 24: 279-315.

• Poplack, Shana and Stephen Levey. 2010. ‘Contact-induced grammatical change’. In Peter Auer and Jürgen Erich Schmidt (eds.), Language and Space - An International Handbook of Linguistic Variation: Volume 1 - Theories and Methods, Berlin: Mouton de Gruyter, 391-419.

• Prats, Modest, August Rafanell, and Albert Rossich. 1990. El futur de la llengua catalana. Empúries.

• Thomason, Sarah G. 2001. Language Contact. Edinburgh University Press.

• Thomason, Sarah G. 2008. Social and linguistic factors as predictors of contact-induced change. Journal of Language Contact, 2: 42-56.

• Thomason, Sarah G. 2010. ‘Contact explanations in linguistics’. In Raymond Hickey (ed.), The Handbook of Language Contact. Malden, MA: Wiley-Blackwell Publishing, 31-47.

• Thomason, Sarah G. and Terrence Kaufman. 1988. Language Contact, Creolization, and Genetic Linguistics. University of California Press.

• Van Coetsem, Frans. 2000. A General and Unified Theory of the Transmission Process in Language Contact. Winter Publishing. 

• Bernard Comrie - Linguistics, UC Santa Barbara

• Lal Zimman - Linguistics, UC Santa Barbara

Possessive Chains and Possessor Camouflage

Dr. Bernard Comrie, University of California, Santa Barbara

Recursive possessive constructions produce possessive chains like English the girl’s father’s house. In most such constructions, from knowing the morphosyntax of possessor and possessum in the corresponding bipartite construction, e.g. the girl’s house, one can predict the morphosyntax of the intermediate possessor (father): It combines the distinctive properties of possessor and possessum. Many languages with typologically distinct possessive constructions conform to this generalization, e.g. Russian, Finnish, Abkhaz, Welsh, Turkish, Tsez, Standard Arabic. However, two independent cases are known in which the properties of an intermediate possessor are not predictable from the properties of possessor and possessum in the bipartite construction, a phenomenon that may be called Possessor Camouflage. In Sakha (Yakut) (and the closely related Dolgan), the bipartite possessive construction is head-marking, as in učūtal ǰie-te teacher house-3SG ‘the teacher’s house’. However, an intermediate possessor, in addition to being head-marked, must also be dependent-marked by means of the genitive case, as in kini ehe-ti-n oron-o s/he grandfather- 3SG-GEN bed-3SG ‘her grandfather’s bed’. The second case involves Scottish Gaelic and Irish, where dependent-marking (genitive case) in the bipartite possessive construction is lost on an intermediate possessor. Data will be presented and analyzed, and typological implications discussed. 

Bernard Comrie is Distinguished Faculty Professor of Linguistics at the University of California, Santa Barbara. His main interests are language universals and typology, historical linguistics (including in particular the use of linguistic evidence to reconstruct aspects of prehistory), linguistic fieldwork, and languages of New Guinea and of the North Caucasus.  

Voicing (trans)gender identity: 

Rethinking the relationship between gendered identities, bodies, and the voice

Dr. Lal Zimman, University of California, Santa Barbara

 The relationship between gender and the voice is of central importance to phoneticians, transgender people, and listeners who engage in the process of gender attribution based (partially) on auditory cues; in other words, almost everyone who’s hearing. This talk presents a reconceptualization of the relationship between the voice, gendered identities, and sexed bodies through a focus on transgender speakers. There is a long tradition in the social sciences of looking to transgender people as a means of understanding the practices we all engage in as part of constructing our gender (e.g. Garfinkel’s 1967 discussion of Agnes, an intersex trans woman). Linguists, however, have been relatively slow to take up this line of inquiry. This talk is part of a broad research program that explores the vast potential for linguistic insight from the study of trans people’s verbal practices. Following a brief overview of different strains of that program, this talk focuses on the gendering of the voice, which in linguistics remains a site of biological essentialism and gender binarism. Through an examination of transgender speakers in two Western United States urban centers, the central analysis highlights the ways speakers with non-normative bodies and identities navigate the complex interplay of embodiment and social practice. Specifically, three phonetic features are discussed: fundamental frequency, /s/, and creaky phonation in both read and spontaneous speech. Two primary arguments emerge from this work. First, the gendered phonetic styles trans speakers produce demand a more satisfactory account of the gendered voice that recognizes the effect not only of the body and of early socialization, but of speakers’ multidimensional alignments and disalignments with normative femininity and masculinity. Second, even this complexified theory of gender must be supplemented by discourse analysis of the analyzed speech, which demonstrates the crucial importance of interactional stance and agentive stylistic moves in the production of gender differences in the voice. Together, these arguments suggest that the role of gender in linguistics requires more careful and thorough theorization, particularly as the field advances in our recognition of transgender voices, in every sense of the word. 

Lal Zimman is an Assistant Professor of Linguistics and Affiliated Faculty in Feminist Studies at the University of California, Santa Barbara. His research is broadly focused on the linguistic practices of transgender speakers, in which he employs a range of qualitative and quantitative methodologies. 

• Judith Kroll - Psychology, UC Riverside

• Maria Manoliu - Professor Emerita, Linguistics & French, UC Davis

Bilingualism, Mind, and Brain

Dr. Judith Kroll, University of California, Riverside 

The use of two or more languages is common in most places in the world. Yet, until recently, bilingualism was considered to be a complicating factor for language processing, cognition, and the brain. In the past 20 years, there has been an upsurge of research that examines the cognitive and neural bases of second language learning and bilingualism and the resulting consequences for cognition and for brain structure and function over the lifespan. Contrary to the view that bilingualism adds complication to the language system, the new research demonstrates that all languages that are known and used become part of the same language system. A critical insight is that bilingualism provides a tool for examining aspects of the cognitive architecture that are otherwise obscured by the skill associated with native language performance in monolingual speakers. In this context, variation in language experience becomes a critical tool for investigating the constraints and plasticity associated with language learning across the lifespan. In this talk, I illustrate this approach to language processing and consider the consequences that bilingualism holds for cognition and the neural networks that support it more generally. 

'Animacy' versus 'Efficacy': Culture and Grammatical Gender in Latin

Dr. Maria Manoliu, University of California, Davis

If studies of non Indo-European languages point to the fact that the scale of Animacy differs from one culture to another (see Dahl 2000), historical grammars of Romance languages have interpreted the Latin grammatical genders in Latin according to Western European culture, as a way of encoding two main inherent semantic oppositions: [Animate vs. Inanimate]. But this model is far from accounting for the difference between the semantic features encoded by Latin neuter and non-neuter grammatical genders. Compare, for example, (1) and (2): 

(1) neut. saxum ‘stone, rock’ vs. masc. lapis ‘stone, landmark’, rupes ‘cliff’ 

(2) neut. mare ‘sea, salted water’ vs. masc. pontus ‘sea’ and fem. aqua ‘water’. 

The Animacy model cannot account for the masculine gender of pontus and neuter of mare or feminine of aqua. 

According to our hypothesis, the gender subclassification of nouns in Latin was rooted in an earlier Mediterranean culture, in which the cognitive category of ‘Efficacy’ reflected a perception of the ‘(in)capacity of doing, affecting other beings’ as an inherent property of objects. Compare the following distribution of nouns: 

(i) ‘capable of being effective’ :

(3) feminine: terra ‘earth’, arbor ‘tree’, aqua, like femina ‘woman’, masculine: ignis ‘fire’, ventus ‘wind’, like vir ‘man’ 

(ii) ‘incapable of being effective’: neuter

(4) saxum ‘stone’, templum ‘temple’, tempus ‘time’; most nouns referring to fruit: pirum ‘pear’, prunum ‘prune’, generics for species: animal ‘animal’), etc 

Since some entities can be active in some contexts, and passive in others, it is necessary to distinguish between the virtual and the actual properties of the referents. The ‘capacity of being a doer (active, effective)’ is a virtual property of the referent that might be encoded in an inherent semantic feature (seme) of the noun. The property of ‘being a doer’ is the actualization of this capacity in certain conditions.

• John Baugh  - Washington University in St. Louis

• Nick Ellis  - Psychology, University of Michigan

“Usage-based approaches to language acquisition”

Dr. Nick Ellis, University of Michigan

Usage-based approaches hold that we learn language through using language. Our linguistic ability emerges as a result of our cognitive learning mechanisms analyzing this experience. Corpus Linguistics provides the relevant evidence of usage. Cognitive Linguistics and Construction Grammar analyze the representations of language and how people acquire, represent, and process this knowledge.  

This presentation analyses (1) usage patterns of English verb- argument constructions (VACs) in terms of their grammatical form, semantics, lexical constituency, and distribution patterns in large corpora, (2) patterns of VAC usage in learner-directed speech and in first and second language acquisition, (3) psycholinguistic investigations of VAC processing in first and second language speakers. It demonstrates that: 

1. Language usage is highly patterned in ways that support learning.

2. Language acquisition is guided by this patterning.

3. Language users have rich implicit statistical knowledge of these patterns. 

• Nobuko Koyama - East Asian Languages and Cultures, UC Davis

• Matt Traxler - Psychology, UC Davis

• Robert Bayley - Linguistics, UC Davis