The Acquisition of Anti-Agreement by Future Heritage Speakers of Tashelhiyt Berber

• Imane Bou-Saboun (University of Maryland, College Park)

Heritage speakers show high variability in the target-like attainment of agreement morphology in their heritage language (Polinsky, 2018). In this study, I consider Tashelhiyt Berber (TB), a VSO language with rich agreement inflection. A variant of its agreement paradigm is Anti-Agreement (AA). AA arises when subject φ-features on the verb are suppressed upon extraction of the subject in relative clauses, wh-questions and clefts (Ouhalla 1993, 2005; Ouali 2011). Heritage speakers of TB whose dominant language is Moroccan Arabic (MA) overmark verbal agreement even in subject extraction, regardless of proficiency (Bou-Saboun, HLRI 2021). Two main hypotheses were advanced to account for the HS struggle with AA.

Hypothesis 1: AA is acquired late by all children, including those raised in a mostly monolingual environment. H1 predicts that TB-dominant preschoolers would have difficulty with AA-licensing constructions.

Hypothesis 2: AA is acquired early but later undergoes attrition. H2 predicts that preschoolers comprehend AA constructions. The two examples below illustrate a subject relative clause (1) that triggers AA and an object relative clause (2) where subject-verb agreement is maintained.

(1). Tafruxt lli i-ull-n __ ikzin. Girl.FS REL AA-lifted-AA __ dog.FS

‘The girl that lifted the dog’.

(2). Ikzin lli t-ull tfruxt ____. Dog.FS REL 3SGF-lifted girl.CS ____.

‘The dog that the girl lifted'.

AA, instantiated into the participial circumfix i-X-n (Ouhalla 1993, 2005), indicates that a subject was extracted. TB-learning children should have access to a clear linguistic cue for distinguishing between extraction types in the form of AA. I ran an experiment (N=86, 26-62-month-olds) involving future TB heritage speakers in Morocco, i.e. children being raised in a predominantly TB-speaking environment until school age, who will be Arabic-dominant. I aimed to test their ability to distinguish between subject and object extraction. This ability was taken as a proxy for AA knowledge. Older children were more likely to interpret subject A’-constructions as agent constructions. No effect was observable in the object condition, where children performed at chance regardless of age. I used a pointing task paradigm (Perkins & Lidz 2020). Each participant watched a 5 minute video divided into 6 trials. The crucial testing point was a split screen with 2 looped videos of 1 transitive action simultaneously performed by puppets. The 4 conditions were object/subject relative clauses and object/subject wh-questions. I measured pointing towards the correct target of the instructions. For the results, a GLM model (Gelman & Hill 2006) fitted to experimental items showed no main effect of extraction type, question or age. However, results suggest an interaction between subject and age (β̂=-0.06; CI=[-0.02; 0.13]; P(β>0)=0.93) entailing more agent responses in subject conditions as the participant age increases. In conclusion, TB-learning children struggle with the subject-object distinction until about 33 months of age. Therefore TB heritage speakers might struggle with AA in adulthood since systematic exposure to MA, a societally dominant language with no AA, started around the time they were still consolidating TB syntax.

 The Impact of Interactional Context on Heritage Bilingual Speakers’ Language Regulation and Cognitive Control during the Pandemic

• Ariel Chan (University of California, Los Angeles)

Recent studies have shown that changes in language environments have both positive and negative effects on bilingual speakers’ English and heritage language development respectively during the pandemic (Sun et al, 2022; Wang, 2020). Other research has shown that language-related uncertainties predict bilinguals’ language fluency and engagement in proactive cognitive control strategies (Gullifer & Titone, 2021). To build on previous findings, the current study examines language processing and cognitive control of Cantonese heritage speakers in the U.S., focusing on how their interactional context during the pandemic affected their cognitive control and regulation of their heritage and dominant languages. Data from this study came from a larger study that examined language processing in Cantonese-English bilinguals with varied language experiences and cultural identities. In the present study, we focused on 32 Cantonese-English heritage bilinguals. All heritage bilinguals completed a Language Experience and Proficiency questionnaire (LEAP-Q, Marian et al. 2007) on language use, a General Ethnic Identity questionnaire (Tsai et al. 2000) on their American, Chinese, and Hongkonger identities, and a pandemic questionnaire on their interactional context during the pandemic. They also completed a comprehension task on code-switching, a verbal fluency task to access language proficiency and regulation, and an AX-continuous performance task (AX-CPT) to assess proactive and reactive cognitive control. Online experimental and questionnaire platforms, Finding Five and Qualtrics, were used to collect data in 2021. Bilinguals’ language use and interactional context were established from LEAP-Q and the pandemic questionnaire while their language proficiency and regulation from the verbal fluency task. Code-switched word processing and cognitive control were measured based on the reaction times and accuracy rates in the language comprehension task and AX-CPT, the measure of cognitive control. For language regulation, heritage bilinguals were more dominant in English than Cantonese. Interestingly, their verbal fluency scores in Cantonese and English were not significantly different from those of immersed bilinguals, who were raised in a Cantonese-dominant context and were more dominant in Cantonese than English. Heritage bilinguals’ verbal fluency performance was also not a predictor of their cognitive control and code-switched word processing time, while enhanced proactive cognitive control was linked to their bicultural identity and not to language regulation. Data analysis is ongoing; we expect that the interactional context, as computed by the pandemic questionnaire, will better predict heritage bilinguals’ verbal fluency and cognitive control performances as it captures heritage bilinguals’ social network and language use at work and at home during the pandemic.

 On the Role of Continuous Exposure in Heritage Speakers’ Adolescence: Two Groups of Chinese-dominant Tibetan Heritage Speakers in China

•  Yunchuan Chen (Duke University)

Q-Neg sentences are like All teachers did not use Sandy’s car. There are two possible readings in English and Tibetan: (i) for every teacher, he/she did not use Sandy’s car (all>not, surface scope (SS)); and (ii) it is not the case that all teachers used Sandy’s car (not>all, inverse scope (IS)). However, for the equivalent Chinese sentence, the IS reading is prohibited and only the SS reading is allowed. Thus, Chinese is more restrictive than English/Tibetan concerning the interpretation of Q-Neg sentences. XX (2022) examined how Chinese-dominant Tibetan heritage speakers interpret Q-Neg sentences in both Tibetan and Chinese and they found three different groups: (i) Group I, which accepts IS in both languages; (ii) Group II, which accepts IS in Tibetan and rejects IS in Chinese; and (iii) Group III, which rejects IS in both languages. According to XX, all heritage speakers in the study shared the following characteristics: a. they were born into Tibetan-speaking parents and spoke Tibetan at home; b. they started learning Mandarin Chinese from age 3 to 7 when entering kindergarten/elementary school and were able to speak Chinese fluently by Grade 6; c. after finishing elementary school around age 12-13 in Tibet, they were selected to attend boarding schools in Mandarin-speaking cities of inland China for secondary education. This is when Mandarin Chinese fully took over Tibetan to become their dominant language. Based on these results, we undertook a study of the bilinguals who do not go to inland China for the secondary education. We recruited 21 such Tibetan-Chinese bilinguals to complete the same experiment conducted in XX (2022).

They shared all characteristics with XX’s heritage group except for the following: rather than go to inland China for their secondary education, they all attended Tibetan-track programs in their local schools until Grade 12, where almost all subjects were taught in Tibetan. In addition, they actively used Tibetan in their communities. Thus, they had constant exposure to Tibetan during adolescence, unlike XX’s heritage group. Their experimental data showed that 20 out of 21 (95.24%) participants consistently accepted IS of the Tibetan Q-Neg sentence. Out of the 20 participants, 10 consistently accepted IS in Chinese while 10 consistently rejected it. Only one participant consistently rejected IS in both languages, like Group III in XX (2022). To measure their respective language proficiency, we also asked the participants to complete proficiency tests in both languages. Forty L1 Chinese speakers completed the Chinese test as a baseline, and we adopted 40.5 (out of 50 points) as the cutoff point to select participants whose Chinese proficiency falls into the native range. All our bilinguals scored above 40.5, meaning that their Chinese proficiency is native. Their cores were then converted to z-scores and a point-biserial correlation was run between them and the participants’ categorical judgment on the Chinese IS condition. The result showed a small correlation: rpb (18)=-.255 but it is not significant: p = .27.

Heritage Language Development In the School-Age Period: When Do Basque-Spanish Bilinguals Acquire Subject Pronoun Expression In Their Two Languages?

• Eider Etxebarria (Northwestern University)
• Silvina Montrul (University of Illinois at Urbana-Champaign)

This study examines the bilingual development of children acquiring Basque, a national minority language, and Spanish. Investigating the linguistic development of school-age bilingual children is critical to understand typical milestones of bilingual language acquisition and to evaluate potential developmental delays in the heritage/minority language. We examined subject pronoun expression (SPE) in Basque and Spanish, two null subject languages (NSL). Bilingualism research on SPE has traditionally focused on NSLs in contact with English, a non-NSL, and gave rise to the Interface Hypothesis (e.g., Tsimpli et al., 2004; Sorace & Filiaci, 2006; Serratrice, 2007; Argyri & Sorace 2007). However, recent research on contact situations between two NSLs (Giannakou, 2018; Rodríguez-Ordóñez & SainzMaza-Lecanda, 2018) suggests that the Interface Hypothesis (Sorace and colleagues, 2006, 2009, 2011) cannot accurately account for adult bilingual performance in contact situations between two NSLs. This research is furthered by investigating SPE in Basque-speaking children and adults and by testing these bilinguals in the two languages. Since the pragmatic distribution of null and overt subjects lies at the syntax-discourse interface, and this takes time to develop, we asked: 1) What is the overall rate and distribution of null and overt subjects in monolingual and bilingual children’s production in Spanish and Basque? 2) At what age do monolingual and bilingual children’s subject pronoun production patterns match adult-like patterns in both these languages? The production of third-person singular null and overt subject pronouns was tested in Spanish and Basque by 136 Spanish monolingual and Basque-Spanish bilingual children (ages 6, 8, and 12), as well as 60 Spanish monolingual adults (both from the monolingual region of Madrid and bilingual region of the Basque Country) and Basque-Spanish bilingual adults. All participants completed an oral pronoun elicitation production task, where bilinguals completed it in both languages (Spanish and Basque). Subject pronoun expression was tested in same- and switch-reference contexts (i.e., Éli ya ha comido. Øi no tiene hambre (Spanish); Beraki dagoeneko jan egin du. Øi ez da gose (Basque) ‘He already ate. (He) is not hungry’ vs. Yoi trabajo como periodista. Ellaii es maestra (Spanish); Niki kazetari bezala lan egiten dut. Beraii irakaslea da (Basque) ‘Ii work as a journalist. Sheii is a teacher’). The results indicated that bilingual children show adult-like pragmatic distribution and use of SPE in NSLs by age 12/14. Monolingual and bilingual children reach the same developmental milestones synchronously in Spanish, and bilinguals also reach the same developmental milestones synchronously in both Basque and Spanish. Our findings revealed that null/overt pronoun rates and distributions in contact situations between two NSLs are driven by pragmatic and crosslinguistic effects related to the referential properties of pronouns in the two languages (see also Giannakou 2018). Our results contribute to theoretical debates on the acquisition of interface phenomena in bilingualism and confirm the importance of academic and societal support of the heritage/minority language for its successful maintenance and transmission.

Resultatives in the Dominant and Heritage Language

•  Martine Gallardo (University of Illinois at Urbana-Champaign)

Heritage speakers often exhibit structural differences from baseline grammars across linguistic domains (Montrul, 2010). The present study examines Spanish heritage speakers’ (SHS) knowledge of argument structure; specifically, SHSs’ knowledge of resultatives in English and Spanish, in comparison with monolingually raised native speakers of Spanish (MRS) and English (MRE). Crucially, resultatives occur in English but not in Spanish (Pylkkänen, 2008). Resultatives predicate a state of an NP, which is a result of the action described by the verb (Hovav & Levin, 2001). Examples (1) and (2) show resultatives.

(1) John hammered the metal flat.

(2) The lake froze solid.

Resultatives are subject to a Direct Object Restriction (Levin & Hovav, 1995), meaning they can only be predicated of objects. As such, they are grammatical with transitive verbs, as in (1) and unaccusative verbs, as in (2). They are ungrammatical with unergatives, as in (3), unless a fake reflexive object is added as in (4).

(3) *She sang hoarse.

(4) She sang herself hoarse.

Given SHSs’ dominance shift to English, they are predicted to have nativelike knowledge of English resultatives. This prediction is supported by previous research which found that SHSs have nativelike knowledge of the dominant language (Montrul, 2006; Montrul & Ionin, 2012). Assuming dominant language influence, SHSs are predicted to accept resultatives incorrectly in Spanish. With respect to dominance, greater English dominance should result in greater acceptance of resultatives in Spanish, while greater Spanish dominance should result in greater rejection of resultatives in Spanish. 19 Spanish heritage speakers, 25 monolingually raised Spanish native speakers, and 29 monolingually raised English native speakers completed the study (Table 1). Spanish speaking groups completed the Bilingual Language Profile (Birdsong et al., 2012), Spanish and English cloze tests, and 2, 56-item acceptability judgment tasks, one each in English and Spanish. For the Spanish task, conditions were *Resultative and Resultative which contained items like (1) and their grammatical Spanish counterparts. For the English task, conditions were Transitive Resultative, items like (1), and *Unergative Resultative, items like (3). The MRE group completed only English tasks. According to the statistical results, overall predictions were partially supported. For the Spanish task, there were significant differences between the SHS and the MRS groups. The SHS group accepted significantly more ungrammatical resultatives in Spanish, indicating dominant language influence. For the English task, results diverged from predictions. There were no significant differences found between the SHS and MRS groups. However, there were significant differences found between the SHS group and the MRE group, in the transitive resultative condition. Interestingly, this difference was due to the SHS group demonstrating more categorical judgments in English. This finding confirms the tendency attested in heritage phonetics/phonology in which HSs, “tend to amplify properties that separate their two languages.” (Polinsky, 2018, p. 162). Finally, across both tasks dominance and proficiency were not significant predictors, which may be attributable to the nature of our sample: the high level of English proficiency of the MRS group and the relatively balanced bilingualism of the SHS group.

Japanese Heritage Language Development and Language Policies in Multilingual Families

•  Asako Hayashi-Takakura (University of California, Los Angeles)

This paper is a summary of parents’ concerns about Japanese heritage language learners’ dual language development and language policies of multilingual families. Presenters will report the results of interviews with parents who do not share the same first language between them. Then, we discuss language resources and educational opportunities, that is, community-based programs and two-way immersion for Japanese as heritage language (JHL) learners. First, we will report the questions and concerns from Japanese parents at the JHL development workshop held in November, 2022. From questions from about seventy participants, there were three main themes: 1) they wanted to know how they should set the family language policy for JHL development in English speaking environments; 2) they were concerned about JHL children’s Japanese literacy development and roles of parents and schools; 3) they sought educational opportunities for JHL children other than Japanese weekend school or private tutoring. After the workshop, we conducted follow-up interviews with volunteers. The interviews focused on exploring their family language policy and their language use at home. We will share findings obtained from parents of intermarriage couples with the audience. Then, we will discuss community-based education and two-way immersion for JHL learners in K-12 settings outside Japan. Parents of JHL learners often believe that Hoshuukoo (Japanese government sponsored weekend schools) is the only option for their children to learn the Japanese language in the United States. The Hoshuukoo were originally established for the school-age children of sojourners from Japan. However, the majority of students in Hoshuukoo are now JHL learners who are 2nd generation Japanese-Americans. Therefore, the Hoshuukoo curriculum that adopts textbooks and curriculum for Japanese school children in Japan does not match the JHL development and maintenance. As a result, a number of Hoshuukoos in the United States have opened special classes, or parent volunteers have founded their own classes for JHL learners to meet their learning needs. Some interviewed participants showed such negative attitudes toward Hoshuukoo as their concerns and stress of condensing so much material to learn after school and on the weekend. At the conclusion of our presentation, we will discuss new opportunities for JHL learners’ dual language development: Japanese immersion programs in public school systems. There are forty-five schools that offer Japanese immersion programs in the United States as of the 2022-23 school year. These schools provide a strong form of bilingual education to benefit both majority-language and minority-language children. Two-way immersion education in public schools is unique in that it contributes to making up for the gap between mainstream education and community-based heritage language instruction. Two school districts in Los Angeles county have offered Japanese partial-immersion dual language programs. The Los Angeles Unified School District (LAUSD), the second largest school district in the US, started two new Japanese immersion programs: one in Gardena, where a number of Japanese-Americans reside in 2021, and the other in West LA in 2022. We will discuss how Japanese immersion education will help JHL learners and families in supporting the development of their dual languages.

Heritage Armenian as Dominant Code of Performing Masculinity

•  Shushan Karapetian (USC Institute of Armenian Studies)
•  Sarkis Tricha (USC Institute of Armenian Studies)

This study explores the sociocultural and sociolinguistic landscapes of the coming-of-age trajectory of Armenian men who immigrated to Los Angeles from the Republic of Armenia during the tumultuous decades before, during, and after the collapse of the Soviet Union (1980s-2000). Through a series of in-depth oral history interviews, we investigate their self-identified masculinity to explore how they construct and express their social roles and identities as men, with particular focus on the sociolinguistic behaviors and attitudes expected of them within their communities. We trace their performance of masculinity through the influences of various codes and their evolution over time and space. These include the glorification, romanticization, and adoption of the Soviet-era code of the “thieves in law,” from its origins in gulags and prisons, to its adoption into the (Soviet and post-Soviet) Armenian context, and finally, its exportation and adaptation into the urban Los Angeles youth subculture, replete with its visual and spoken language.

This poster will bring to light the resilience of Armenian as a heritage language in specific male-inhabited domains that attempt to “preserve” continuity in the social code. We demonstrate that performance of this code and performance of the heritage language become synonymous with presenting a proper Armenian masculinity, affecting social hierarchy, specifically during codified rituals of conflict negotiation. We describe how the men of our cohort are both socialized into this ritualized masculine code in LA and how they socialize others through explicit linguistic training in order to maintain status in conflict negotiation settings. The exclusive mobilization, functionality, status, and indeed dominance of the Armenian language as the linguistic form through which masculinity is conceptualized, projected, and performed, particularly in a diasporic context, adds a rich and novel dynamic to the landscape of heritage languages.

 Is Heritage Spanish a Distinct Variety of Spanish?: Evidence from Heritage Speakers’ Perception of Heritage Accent

• Joo Kyeong Kim (University of California, Los Angeles)
• Jenny Choi (University of California, Los Angeles)
• Jack Carter (University of California, Los Angeles)
• Ji Young Kim, PhD (University of California, Los Angeles)

Heritage speakers (HSs) are generally considered to sound more native-like than second language learners (L2s), despite having an accent that diverges from that of non-heritage native speakers (NHNSs) (Au et al., 2008; Kupisch et al., 2014; Lloyd-Smith et al., 2020). However, perception of the so-called “heritage accent” (Polinsky and Kagan, 2007) is largely based on NHNSs’ impression, while there is little understanding of how HSs perceive the speech of other HSs. In order to fill this gap, the present study investigates perceived global accent in heritage Spanish with both HSs and NHNSs as raters. Specifically, we seek to elucidate whether HSs recognize a heritage accent and whether they consider it to be native-sounding.

A perception experiment was conducted online using speech samples of three speaker groups: six HSs of Mexican Spanish, six L2s with English or Korean as their first language, and six NHNSs of Mexican Spanish. The speakers were asked to describe and narrate a wordless picture book. Two 10-second samples were extracted from each speaker, resulting in a total of 36 speech samples. All samples were free of morphosyntactic errors and speech rate was controlled across groups. The country of residence was the US for the HSs, the US or Korea for the L2s, and Mexico for the NHNSs. Two rater groups, HSs and NHNSs, participated in the experiment. After listening to each speech sample, the raters evaluated the speaker’s nativeness on a scale of 1 to 6 (= completely native) and predicted the speaker’s country of residence. Participant recruitment and data collection were administered using online platforms. Raters’ accent ratings and accuracy of identifying speakers’ country of residence (COR) were analyzed using (generalized) linear mixed effects models in R (R Core Team, 2021).Preliminary results, based on data obtained from 14 HS and 27 NHNS raters, showed that both rater groups rated the HS samples higher than the L2 samples and lower than the NHNS samples, which is consistent with the findings of previous heritage accent research. With regard to the accuracy of speakers’ COR, our data demonstrated that the raters had more difficulty identifying HSs’ COR than that of the NHNSs. While no rater group effect was found, it had significant interactions with the speaker groups, which were led primarily by HS raters’ more accurate identification of the COR of the HS samples than that of the NHNS raters. Our findings suggest that HSs recognize the speech of other HSs better than NHNSs and consider it native-sounding, but less so than the speech of NHNSs. In other words, certain phonetic and/or phonological features are shared among HSs, indicating that heritage Spanish is a distinct variety of Spanish. Nonetheless, heritage Spanish is judged by its speakers as less native-like than (non-heritage) Mexican Spanish, which is corroborated by a strong negative correlation between HS raters’ accent ratings and their response rates identifying the US as the COR, a strong positive correlation between HS raters’ accent ratings, and their response rates identifying a Spanish-speaking country as the COR.

 Project Designs for Chinese Heritage Language Learners

• Min Min Liang (Massachusetts Institute of Technology)

Since heritage languages learners often possess diverse linguistic skills before they enter formal classrooms, their needs are quite different from second language learners. Therefore, teaching heritage language requires a holistic macro-based concept to expand students’ competence while taking advantage of their prior knowledge (funds of knowledge.) Appropriate projects that highlight heritage language students’ linguistic competency and individual learners’ strengths and characteristics are crucial in teaching. Specific project designs will address the latent dimension of Chinese heritage learners. However, questions arise when language instructors attempt to design projects. Will this project evaluate skills or content knowledge? Is the project focused on content that has been taught or are students using their knowledge from other aspects of their life? Will this project promote learners’ motivation, target language cultural awareness and metacognition skills? How does it influence future learning and teaching? This presentation will address these issues for Chinese heritage language learners and demonstrate different project designs for the beginning level class that enable students to recognize their untapped potential, to utilize their funds of knowledge and motivate students to be active learners.A well-designed project not only helps students to learn in a virtual learning environment but also maximize their learning experience in the Face-to-Face learning environment. In this presentation, the speaker will showcase two different projects that focus on speaking, listening, reading and writing skills in a Chinese heritage language classroom. Constructive guidance will be provided to demonstrate: 1) how to design a test/task that will demonstrate students’ funds of knowledge; 2) how to facilitate students’ interaction and collaboration abilities. These formats will help students to recognize their untapped potential, to utilize their funds of knowledge and motivate them to be active learners. The relationship between motivation and achievement is a strong one. Students who are motivated to appreciate where they have advanced will learn more in a positive manner.

 Language Practices and Attitudes toward Russian and Ukrainian Heritage Languages and their Functional and Symbolic Roles among HL Learners

• Svitlana Malykhina (Boston University)

The war in Ukraine has triggered a humanitarian crisis and a new wave of immigration, which changes the dynamics of native language use and HL maintenance. New immigrants are unlikely to use Ukrainian or Russian in the same way that they use English because these languages serve different purposes for them and have specific functions, which are suited for different kinds of interaction. Ukrainian and Russian Diasporas in the US offer unique cultural and linguistic resources for promoting learning of heritage language locally. While public (media) discussions about the patterns of interaction in Ukrainian, Russian, and English are sometimes binary, in reality, personal narratives can hold contested attitudes and multiple opinions about various practices of linguistic interaction in educational and family setting. Further analysis of the driving mechanism for language maintenance enables us to analyze the dynamics of enrollment in HL classes in Ukrainian and Russian studies that we are currently witnessing. To conduct an empirical study based on qualitative analysis of language practices and attitudes, a three-part questionnaire was designed. The questions in Part A are aimed at eliciting information to draw a linguistic profile of HL students of the new wave of forced immigration. Part B contains close-ended questions that aim at eliciting data on participants’ practical considerations and different degrees of expectation, as well as reasons for sustainable learning of Ukrainian or Russian for use in the language classroom and beyond it. -This part of the questionnaire focuses on motivations for enrolling, continuing in, or not taking HL classes. With respect to close-ended questions, participants were asked to rank them on a 4-point Linkert scale, ranging from “strongly agree” to “strongly disagree.” Part C includes open-ended questions about several recurring metaphors used to discuss native language (Ukrainian or Russian respectively) that have emerged in diaspora and are identified in social media discussions. This study 1) aims to analyze changes in the use of and attitudes towards two major Slavic languages – Ukrainian and Russian in the U.S., and 2) examine and analyze the attitudes of new US immigrants to the options of studying heritage language, culture, folklore, literature, linguistics, and history.

 “I Can’t Find a Home in Both Worlds”: Positionings of Vietnamese Heritage Speakers

• Ha Nguyen (University of Hawai'i at Manoa)

Research in heritage language education has highlighted the importance of a sociolinguistically informed curriculum to cater for the needs of heritage speakers due to various issues related to language variation and varieties, dialects, and vernaculars (He, 2010; Leeman & Showstack, 2022). At US universities, heritage speakers make up the majority of learners in Vietnamese classes, yet not much is understood about their identity and sociolinguistic awareness. While most Vietnamese heritage speakers (VHSs) come from families that speak the Southern variety of Vietnamese (or the Southern dialect), the dialect of instruction is usually Northern, also known as the “standard” variety. The classroom is therefore the place for VHSs’ second dialect acquisition (Fairclough, 2016). In addition, recent heritage language scholarship espouses the importance of teaching heritage speakers about social, political and ideological issues regarding language use (Leeman & Showstack, 2022). Taking up this call, the current study introduces elementary Vietnamese students at the University of Hawaiʻi to issues of language variation, ideology and language subordination in Vietnamese. We seek to answer two questions:What are VHS’s attitudes towards learning about Vietnamese variation and language subordination?How do VHSs position themselves and are positioned vis-à-vis the Vietnamese language, culture and community?We collected data from multiple sources for triangulation purposes: (1) recordings of class discussions, (2) student work, and (3) semi-structured interviews with interested participants. Six students took part in the interviews. We performed open coding (Saldaña, 2013) before condensing codes into themes to answer the first research question. For the second question, we use Positioning (Bamberg, 1997) to elucidate the multiple ways VHSs (re)identify with the language, culture and community as well as ways in which these positionings were delegitimized (Bucholtz and Hall, 2005).The findings suggested, in line with Beaudie et al (2021), that learners show critical language awareness toward Vietnamese variations. Learners’ experiences did suggest a connection between dialect, social class, and ideology. We also found that identity issues among VHS are complex. While each student differed in their Vietnamese knowledge and use, their family tie plays a significant role in the way they position themselves as Vietnamese American/Asian American. VHSs also value the ability to read and write in Vietnamese. Such skills affected the extent to which VHSs identified as (native) speakers of Vietnamese. Similar to previous studies (Tran, 2021), the Vietnamese language was considered a tool for maintaining family ties in the US for some while for others, it’s a crucial part of their cultural identity. Furthermore, students’ narratives illuminated how their legitimacy as Vietnamese (speakers) were challenged, even when they were complimented. We showed that delegitimation (Bucholtz and Hall, 2005) used by both the students themselves and their family frequently and seemed to exacerbate VHSs’ identity struggle.

How Do Classifiers Facilitate Vietnamese Heritage Speakers’ Predictive Processing?

• Hoan Nguyen (University of Hawaii at Manoa)

The investigation of heritage speakers’ (HS) online language processing has gained increased attention as a means of tapping into their implicit language knowledge (Montrul 2023). Yet there is still a dearth of studies examining HSs’ predictive sentence processing. Recent visual-world eye-tracking studies with Spanish and Polish HSs showed that these HSs were able to use grammatical gender cues to predict upcoming nouns (Fuchs 2022a, b). Little is known, however, about how HSs of classifier languages such as Vietnamese use prenominal classifiers to facilitate their processing (Ito et al. 2020). Classifiers in Vietnamese immediately precede the noun (1). The two most commonly used classifiers, con and cái, are generally associated with animate and inanimate referents, respectively (Dao 2012; Pham & Kohnert 2008), with the exception of a few frequent inanimate nouns (e.g., dao 'knife') that can co-occur with the animate classifier con.

(1) Đâu là con chó/dao?

where is CL-animate dog/knife

Here we investigate to what extent heritage and home-country raised native speakers of Vietnamese use con and cái to anticipate both typically (e.g., dog) and atypically (e.g., knife) associated animate and inanimate nouns. In this study, heritage (N=11, data collection on-going) and home-country-raised (N=16) adult speakers of Vietnamese in the U.S. completed a visual-world eye-tracking experiment in which they listened to sentences containing typical animate and inanimate nouns preceded by con/cái and sentences with atypical inanimate nouns preceded by the animate classifier con. For typical nouns, the visual scene contained either only one (DIFFERENT-condition) or three (SAME-condition) referents matching the classifier. For atypical nouns, the visual scene contained the target, an animate competitor, and an inanimate distractor. Participants also completed offline tasks assessing their knowledge of classifier-noun associations and general proficiency. Analyses of fixations on the target referent during two 360-ms windows corresponding to the duration of the classifier and the noun, respectively, showed that for typical nouns, in the noun window, an interaction emerged between condition and group (b = -0.19, p=.02), with follow-up models indicating more looks to the target on DIFFERENT vs. SAME trials in both groups, but a smaller effect in the HS (b=-0.18, p=.03) than in the home-country group (b=-0.37, p<.001). No significant effects emerged in the classifier window. For atypical nouns, we compared looks to the (inanimate) target versus the animate competitor during the same time windows. In the classifier window, an interaction between AOI (target, competitor) and group (b=-0.26, p=0.008) emerged; follow-up models showed that home-country speakers (b=-0.19, p<0.001), but not HS (b=0.07, p=0.37), were more likely to look at the animate competitor than the target. In the noun window, a significant interaction (b=0.28, p=0.001) reflected the opposite pattern, i.e., HS looking more at the animate competitor than at the target (b=-0.24, p = 0.001) and home-country speakers now showing a non-significant trend towards looking at the target (b=0.05, p=0.33). These findings suggest that both home-country and heritage speakers use classifiers to create expectations about the animacy of upcoming nouns in real-time processing, yet HS appear to do so at a slight delay, and may take longer to recover from an incorrect prediction.

 What Eye Movements Tell About Reading in Russ. as a HL: Implications

• Olga Parshina (University of Delaware)

The study presents the first step in drawing pedagogical suggestions on teaching reading to Russian Heritage Speakers (RHS) based on the findings in the four empirical studies that focused specifically on investigating RHSs’ reading abilities. These studies to the best of our knowledge, present the first systematic eye-tracking investigation of reading abilities in Russian as a heritage language and Russian as a second language in comparison to monolingual adults and young literate children. Study 1 starts with establishing benchmark eye-movement characteristics in reading isolated Russian sentences by four groups of participants: 30 RHS, 30 second language learners (RL2), 30 monolingual adult speakers, and 30 monolingual 8-year-old children. Study 2 takes a step further by using a novel (in bilingualism) scanpath approach to explore the global reading patterns that heritage speakers adopt in reading sentences and how these patterns compare to the reading strategies adopted by the other three groups. Finally, Studies 3 and 4 examine the role of language prediction in facilitating sentence processing in reading in the Russian heritage language, the same way it does in monolingual sentence comprehension. Based on the overall picture of reading abilities in RHS and RL2 that emerged from the above-mentioned experimental research, we then discuss broader pedagogical implications for classroom settings suggesting both general approaches to developing reading in Russian heritage classes and practical ideas for pedagogical activities.

 Direct Object Clitics in Child Heritage Speakers of Romanian

• Mihaela Pirvulescu (University of Toronto Mississauga)

This study targets bilingual (Romanian-English) heritage speakers of Romanian (HS) growing up in Toronto Canada (16 children 8-11 years old). Although the children learned Romanian as their L1 and use it at home they are growing up in an English-dominant environment. We know that HS due to less language exposure and use are more likely to exhibit language loss or attrition than older bilingual children who had a longer period of monolingualism in their heritage language (Montrul 2008; Flores 2010; Montrul and Bateman 2020). We also know that the domain of pronominal object clitics is vulnerable to reduced language input and use (Pirvulescu et al. 2014); at the same time Romanian is a language where these clitics are acquired very early and with minimal errors (e.g. Tomescu and Avram 2019).In this context we ask the following questions: 1) do HS of Romanian use object clitics in their production and if yes are they using the correct form? 2) do HS of Romanian correctly interpret clitic features? 3) How is the use and interpretation of clitic (features) related to the variables Working Memory and Use of Romanian at home?We measured two modalities (comprehension and production) as it is well known that heritage speakers‚Äô language proficiency and accuracy depend on task modality (Pérez-Cortés et al. 2019). We compare results from a Clitic Elicitations Task and a Comprehension Task (picture choice) aimed at testing sensitivity to phi-features (gender and number) in clitics. In the Comprehension Task half of the items are direct object clitic pronouns and half are strong pronouns. We administered a Working Memory Test and a language questionnaire provided information about the amount of input and use. We compare the HS group with a group of multilingual children with L1 Romanian as dominant language from Bucharest Romania. Tables 1 and 2 show that while Romanian HS perform extremely well they do perform differently from the Romanian Dominant children; the latter group perform at or near ceiling while Romanian HS children show comparatively lower clitic production with clitic omission and errors of gender being the main divergence. The HS group displays errors in production as well as in comprehension (no statistically significant difference between the two). The variable Use of Romanian at home significantly predicts (correct) clitic production (F(213) = 3.79 R2 = .369 p = .017). Both variables Working Memory and Use of Romanian at home significantly predict correct clitic comprehension (F(213) = 10.13 R2 = .609 p < .05). For the HS group there is a significant difference in the comprehension task between the correct clitic vs strong pronouns responses children being much more accurate with strong pronouns than with clitics (t(15) = -4.563 p < .001). These results confirm previous results on age effects in heritage language acquisition and at the same time even though results are similar across tasks they seem to indicate that errors do arise from computational difficulties and difficulties in accessing linguistic representations due to cognitive load (e.g. Pérez-Cortés et al. 2019).

 Challenges Associated with Designing an Inclusive Environment for Heritage Speakers in a Military Command Language Program

• Irina Poliakova (Defense Language Institute Foreign Language Center)
• Olga Mukhortova (Defense Language Institute Foreign Language Center)

Heritage learners are increasingly interested in Russian language study. Multiple scholars (O. Kagan, M. Polinsky, I. Dubinina, O. Kisselev, O. Laleko) in this emerging field supply instructors with solutions how to teach heritage speakers and make inclusive environment for them in the classroom. Often instructors, who have no opportunity to offer a separate course specifically for heritage speakers, have to come up with techniques of differentiation approach in order to effectively deliver the material and let students gain their skills. This mixed environment provides students with rich and unique context, beneficial for both, heritage and L1 students, changing their attitudes towards the scope of the language, seeing more practicality in learning the language and seeing the language as more real because it gives real examples of communication in inclusive environment. In addition to mastering the language (which could be a challenge for both groups of students), mastering inculturation (which is easier for heritage speakers due to their exposure to it) is an issue in the mixed classroom. A salient challenge of serving both FL and HL learners in Russian -as-a foreign-language contexts entails the inter- and intra-group dynamics of combined FL/HL classrooms where affective factors can influence each group’s learning in distinct ways. To create inclusive environment for heritage speakers at the military command language program is even more challenging. The scope of my paper is to discuss approaches and techniques, designing the mixed environment in a military command language program.

 The Role of Cognitive and Social Factors in Processing Non-Binary Spanish Pronouns

• Alexandra Roman Irizarry (University of California, Irvine)
• Judith Kroll(University of California, Irvine)
• Julio Torres (University of California, Irvine)

Speakers of grammatically gendered languages are sensitive to grammatical gender violations. What is less known is how bilinguals with a binary grammatical gender system process gender-neutral forms. In Spanish, speakers use –x (e.g., Latinx) or –e (Latine) to name non-binary identities. From the literature, we know that Spanish heritage bilinguals differ in their processing of grammatical gender from Spanish-English bilinguals from Spanish speaking countries. In this study, we ask how heritage bilinguals and non-heritage bilinguals differ when it comes to processing non-binary morphemes –x and –e. To further test how bilinguals use cognitive resources to process their two languages, we tested how Spanish-English bilinguals process pronouns with non-binary morphemes –x and –e in contrast to processing pronouns with Spanish traditional morphemes –o and –a. We also tested how different social factors (Spanish language dominance and attitudes towards gender diversity) and cognitive factors (working memory and cognitive control), mediated the extent to which non-binary morphemes can be processed. Finally, we tested whether processing non-binary morphemes differs from processing grammatical gender violations. Participants were 37 Spanish-English bilinguals, 14 of whom were Spanish heritage bilinguals, while 23 were Spanish-English bilinguals born and raised in Spanish speaking countries. They completed two self-paced tasks, the Bilingual Language Profile, the Gender/Sex Diversity Beliefs Scale, the Operation Span Task, and the AX-Continuous Performance Task. In the first self-paced task, they read sentences that contained pronouns with –o, –a, –x, and –e. In the second self-paced task, they read sentences with grammatical gender violations between the head noun of the sentence and the pronoun of the subordinate clause. In general, no significant differences were found among the two groups of bilinguals. However, pronouns with –x had significantly longer reading times than pronouns with –o, while pronouns with –e were not statistically different from those with –o. This suggests that Spanish-English bilinguals can process pronouns with –e as the generic plural in a similar way that they can process pronouns with –o. With respect to language dominance, as Spanish dominance increased, average reading times were shorter across all conditions. Regarding gender attitudes, people with more traditional views on gender had longer reading times on pronouns with –x, –e, and –a. For the cognitive measures, higher working memory was associated with shorter reading times on –a. Although cognitive control measures were not statistically significant, proactive control was trending towards shorter reading times across all conditions. Finally, regarding grammatical gender violations, there was a trend for longer reading times on incongruent trials. Taken together, the results suggest that while Spanish-English bilinguals appear to process morpheme –x as a grammatical gender violation, morpheme –e is not processed as such. The processing difficulty of non-binary morphemes seems to increase for those who have more traditional beliefs on gender. An emerging trend suggests that cognitive factors are associated with more efficient processing of gender morphemes. More data are needed to confirm this trend. In short, the findings show support for the use of non-binary morphemes –e over non-binary morpheme –x in Latin* populations.

 Language Use of Japanese-English Bicultural Families and Children’s Oral Proficiency in Heritage Japanese

• Sachiko Roos (University of Hawaii at Manoa)

The development and maintenance of a heritage language is affected by various aspects of the heritage speaker's surroundings, including their family and the community they live in. In this exploratory study, we examine the relationships between proximal (language use between family members) and distal (prominence of the language in the community) input factors on the one hand, and children’s oral lexical proficiency in the heritage and the community language on the other, in Japanese-English speaking intercultural families raising their children in the U.S. Twenty-one children (aged 6-12) from families in Hawaii (strong Japanese community presence, N=11) and Washington (N=10) completed an oral picture naming task (HALA, adapted from O’Grady et al., 2009) in Japanese and English, and participated in a semi-structured interview about their family language use. Mothers and fathers separately completed a survey containing questions about their family language use, policy, and affiliation with the heritage community. The collected data allow us to explore the following questions:RQ1: To what extent do mother’s, father’s, and child’s reports about language use align with one another? RQ2: To what extent does family language use and other external factors correlate with lexical proficiency in Japanese (HALA-Japanese score), English (HALA-English score), and language dominance (HALA-English minus HALA-Japanese score)? RQ3: Are there any differences in these patterns between families in Hawaii and Washington? Results regarding RQ1 indicated that mothers’ and fathers’ reports about child’s language input from parents were strongly correlated (r>.7), whereas children’s and parents’ reports about children’s language use with parents were only moderately correlated (.4< r <.6). With regard to RQ2, we observed significant correlations between (native Japanese-speaking) mothers’ use of Japanese with the child and the child’s HALA-Japanese (rs=.772, p<.001) and HALA-English (rs=-.595, p=.004) scores, while (native English-speaking) fathers’ language use correlated significantly only with HALA-English (rs=-.579, p=.006) but not HALA-Japanese (rs=.229, p=.32) scores. Thus, only the quantity of Japanese used by the mother, not by the father, correlated with the child’s performance in Japanese. However, when looking at children’s dominance, i.e., their relative vocabulary skills in the two languages, both the quantity of Japanese used by the mother (rs=.854, p<.001) and by the father (rs=.464, p=.03) contributed to the strength of children’s dominance. In other words, families in which both the mother and the father regularly used Japanese with the child were more likely to have children whose relative vocabulary knowledge in Japanese vs English was stronger, yet this relation was more prominent with mothers than fathers. These findings present evidence that language choices by both parents matter for bicultural children’s bilingual development. Meanwhile, no notable differences emerged between families in Hawaii and Washington (RQ3), indicating that proximal input factors weigh more heavily than distal ones, at least in this small sample. Further quantitative and qualitative analyses are under way to further explore potential contributors to the variability in language outcomes among heritage Japanese-speaking children in the U.S.

 Lexical Entrainment in Bilinguals’ L1 and L2

• Yongjia Song (University of California, Irvine)

Lexical entrainment is a phenomenon in which people tend to re-use the words used by conversational partners (Brennan & Clark 1996). It is explained as either an automatic reaction caused by priming (Pickering and Garrod, 2004), or a strategic behavior that two interlocutors achieve conceptual agreements for communicative purposes (Brennan & Clark, 1996). Past studies suggest that speakers tend to entrain more when interacting with listeners with less language competence, such as computers (Branigan et al., 2011), children (Cai et al., 2021), and non-native partners (Cai et al., 2021; Suffill et al., 2021). However, few studies have explored how the features of speakers themselves determine the pattern of entrainment. Although speakers’ language proficiency is claimed to be less impactful on entrainment than listeners’ proficiency (Suffill et al., 2021, Zhang and Nicol, 2022), there are robust individual differences among speakers in the rates that they entrain (Tobar-Henríquez et al., 2020), suggesting unexplained variation on the part of speakers. Therefore, we target bilingual groups and explore individual differences in lexical entrainment by looking at their entrainment behavior. We conducted an online experiment with 100 bilinguals who are proficient in English. The experiment included LEAP-Q to assess language background, an English picture matching and naming task (Branigan et al., 2011) to measure entrainment frequency, and a verbal fluency (VF) task to measure language dominance/proficiency. In the picture matching and naming task, participants first saw a low frequency word from a partner and two pictures, and they matched the word with the target picture. Then, after two filler trials, participants saw the target picture again and were asked to name the picture. Entrainment frequency was calculated as the rate at which participants used the words they previously saw from their partner. Sixty seven participants were included in the analysis. The results of self-reported proficiency don’t show a significant correlation with entrainment frequency, but the VF results showed that English bilinguals who are more dominant/proficient in English tended to entrain less (β=-0.014, p < 0.05). We then ran a parallel study in Mandarin with 19 bilingual participants, but didn’t observe the similar proficiency relationship. However, by comparing the Mandarin results with results from 13 Mandarin-English bilinguals in the English study, we found that Mandarin-English speakers entrained more in L2 (English) than in L1 (Mandarin) (p<0.01). Unlike the conclusion from previous studies, our results indicate that speakers’ language dominance/proficiency has an effect on lexical entrainment, especially for L2 speakers. It may be that lower proficiency and operating in an L2 yield more entrainment because, with less experience, lower proficiency speakers are more susceptible to the influence of new input, or it may be that they are less confident in their language ability and therefore more likely to mirror a competent partner. Planned analyses will compare heritage speakers with the other bilinguals in our study in an attempt to tease apart automatic mechanisms (priming) vs. higher-level strategic behavior (confidence and language identity) as the driver of our results.

 Motivation, Grit, and Executive Function in Heritage Speakers and Language Learners

• Nicholas Sulier (University of California, Irvine)
• Judith Kroll (University of California, Irvine)
• Julio Torres(University of California, Irvine)

Cognitive control is a process integral to bilingual language use and second language development, and recent discoveries have illustrated that individual differences and socio-cultural experiences can shape how bilinguals and language learners engage these control processes (Gullifer et al., 2018; Pulido, 2021; Beatty-Martinez et al., 2019). While these recent investigations have shed light on the importance of examining individual differences in executive function while taking context into consideration, little work has examined if individual affective factors, like motivation, grit, and L2 grit, mediate cognitive engagement within and across contexts of language use and language learning. Thus, the current study first explores how language learning motivation (Dörnyei, 2009), grit (Duckworth, 2009) and L2 grit (Teimouri et el., 2022) mediate the engagement of cognitive control in two understudied populations of language users – L2/L3 learners of Spanish and heritage speakers of Spanish. Second, it examines how these psychological and cognitive factors influence both English and Spanish proficiency. Sixteen Spanish-English heritage bilinguals and thirty-one Spanish language learners enrolled university courses completed a battery of online tasks measuring language learning motivation, grit, and L2 grit (Selves Questionnaire, 8-item Grit Scale, L2 Grit Scale), cognitive control (AX-CPT), working memory (O-SPAN) and language proficiency (Elicited Imitation Tasks in English and Spanish). Preliminary analyses on both groups show positive correlations between motivation and Spanish proficiency. Interestingly, working memory was found to predict English proficiency across groups, but not Spanish proficiency. Finally, no relationship was found between the psychological variables and cognitive control, and null results were found for domain general grit. These preliminary results shed light on the important role that motivation plays during language learning and use in both language learners and heritage speakers. Despite overall similar effects of this psychological variable on language proficiency, we speculate that motivation might modulate language processing differently for heritage speakers than it does for learners. Specifically, diversity of cognitive experience, when combined with personal and cultural motives to learn or maintain a home language, may have unique consequences on grammatical processing in heritage speakers.

 Toward the Validation of a Linguistic Insecurity Instrument: A Mixed Methods Study

• Nicole Vargas Fuentes (University of California, Irvine)
• Julio Torres (University of California, Irvine)

Linguistic insecurity refers to a speaker’s confidence in their ability to communicate in different scenarios in relation to other members of their linguistic community. Previous literature (e.g. Goble 2016) suggests that heritage speakers may feel more linguistic insecurity when interacting with members that are deemed as “keepers” of the language such as older family/community members or language professors. Given the crucial role linguistic insecurity plays in heritage language classrooms, Ortega (2022) called for the need for validated instruments to measure linguistic insecurity. To that aim we report on the initial development and validation of a questionnaire to quantify linguistic insecurity among heritage language learners. Using a 7-point Likert scale that ranges from Always confident to Not confident at all, the questionnaire elicits heritage language learners’ ratings on their confidence communicating across 30 language scenarios with different speakers in their social networks (e.g., family members) classroom settings (e.g., giving an oral presentation) and general language contexts (e.g., writing a post on social media). We report on quantitative and qualitative data from two initial phases of validation. Phase 1 included examining responses to the 30 scenarios for reliability and item-rest correlations for each item to make necessary adjustments to the questionnaire. Phase 2 required a new group of participants to complete the revised questionnaire while thinking aloud. Based on a cluster analysis on their linguistic insecurity scores participants were grouped to engage in focus group discussions to probe deeper into their questionnaire responses. Qualitative data from the think-alouds and focus group discussions were then analyzed as a way to triangulate findings from participants’ ratings on the Likert scale items in the questionnaire. Reliability item and exploratory factor analyses were also conducted to examine the psychometric properties of the questionnaire. Finally, next steps in validating the questionnaire will be discussed.

Reference

Ortega L. (2021). Epílogo. El contexto sociopolítico y el español como lengua de herencia. In Aproximaciones al estudio del español como lengua de herencia (pp. 275-290). Routledge.