ABSTRACT. Introduction. Anecdotal reports of syntax teaching paint it as uniquely unwelcoming among the subfields of linguistics. People describe experiences of gatekeeping that mirror those reported for science, mathematics, and other ‘hard’ subjects (Partee 2011; Ceci et al. 2014; Cheryan et al. 2017; Charity Hudley et al. 2020; Dockum & Green 2024). These effects are particularly felt by women and minoritised students, and are thus of concern from the perspective of a more inclusive academic environment. There is some work describing ‘best practice’ or particular linguistics teaching techniques (e.g., Bunger 2017[1983]; Kuiper 2011; Costa-Silva & Lee-Schoenfeld 2024). But no research has directly investigated how syntax is taught or experienced across multiple institutions: are negative experiences in the syntax classroom simply the result of a few bad apples, or is there something systemic about how syntax is taught that results in bad experiences? We report on an in-progress study of student experiences in undergraduate syntax classrooms. Our research questions are: What do students expect before taking syntax, and are these expectations met? Do their opinions change over the course of a term? What kind of students like/dislike or succeed/fail in syntax as a subject? This work forms part of a larger project investigating the experiences of teaching and learning syntax across Canada, the UK, and the US, from the perspective of both instructors and students. Our findings so far bear out the anecdotal reports, based on the experiences of instructors (and their own past experiences as students). Here, we report on a survey of current undergraduate students enrolled in syntax classes at both the start and end of term.

Methods. We conducted a two-part online Qualtrics survey of undergraduate syntax students, recruited by contacting syntax instructors who taught an undergraduate syntax class in the 2025 spring term (further recruitment is planned for the 2025–26 academic year). Participating instructors took a brief survey collecting information about themselves and the course. The first part of the student survey, delivered at the start of term, asked students about their academic background and interests, demographic information, and expectations at the start of their syntax course. The second part of the survey, delivered at the end of term, asked about their experiences in the syntax class, including whether the course lived up to their initial expectations. The full list of questions will be available at our online repository.

Participants. We recruited 7 instructors who distributed an invitation to the survey to their classes, yielding 44 students who participated in the first part of the survey (an average of 6.2 students per instructor, though with a very wide range from 2 to 15). Of the student respondents, 29 (65.9%) selected White as their race/ethnicity, 7 East Asian, 4 South Asian, and 4 Hispanic/Latino/a/x (multiple selection was possible); 32 were women, 8 were men, and 3 selected more than one gender or ‘neither men nor women’ (1 declined to answer).

Survey 1. Asked about whether they expected to do well in syntax class, 26 students responded yes, 3 responded no, and 15 did not answer. When asked in a free-response question what they expected, by far the most common response was that they expected the class to be ‘hard’ (n=12, 27%). Examples are shown in (1)-(3) below: I have heard that Syntax is the ‘Boogeyman’ of Ling courses. However, with my current professor, I know I will gain the clarity I need to understand this field better and be able to gain momentum in crossing bridges to interdisciplinary studies. I have heard it is difficult and makes little sense I heard a lot about syntax trees and how they're hard The next most common theme in free-response answers to this question was about the instructor (mostly positive) as in (1); examples are shown in (4)-(5): I've heard the professor is good Not much, but I am familiar with the lecturer and like her teaching style and work In addition to questions about syntax class, Survey 1 also asked students about their academic interests outside of linguistics.

Prospective results of Survey 2. Survey 2 will be distributed in late April/early May and the responses to the two surveys compared. We capture any changes over the course of the term by repeating the questions about whether students like syntax and whether they think they are good at it. We also ask questions about their instructors: are they fair, strict, helpful, etc., to correlate with their experiences. Questions about identity-related experiences probe the inclusiveness of the classroom environment, and questions about resources (textbooks and beyond) used will give us an overview of what syntax looks like across the different contexts.

Discussion. Even before the results of Survey 2 are integrated into our analysis, some interesting findings have emerged. We see comments about the perceived difficulty of the subject, and that it is a hard subject or ‘mathematical’, echoing the results of our previous survey of instructor attitudes. We are encouraged to see positive comments about the instructors; of course, students who like their professor are more likely to take a survey on their request. The two surveys together will allow us to identify patterns in students’ experiences: differences across teaching styles, changes in attitudes throughout the course, and the usefulness of various resource types. We will be in a position to make recommendations for other syntax instructors, and we will have a better sense of who ends up in syntax classes, especially in the US where linguistics is a ‘discovery major’. The findings will also inform the upcoming focus group phase of the larger research project.

Bunger, Ann (ed.). 2017. Special Volume: Reissue of Innovations in Linguistic Education (IULC Working Papers). IULC Working Papers 17(2). Ceci, Stephen J., Donna K. Ginther, Shulamit Kahn & Wendy M. Williams. 2014. Women in Academic Science: A Changing Landscape. Psychological Science in the Public Interest 15(3), 75–141. doi:10.1177/1529100614541236. Charity Hudley, Anne H., Christine Mallinson & Mary Bucholtz. 2020. Toward racial justice in linguistics: Interdisciplinary insights into theorizing race in the discipline and diversifying the profession. Language 96(4), e200–e235. doi:10.1353/lan.2020.0074. Cheryan, Sapna, Sianna A. Ziegler, Amanda K. Montoya & Lily Jiang. 2017. Why are some STEM fields more gender balanced than others? Psychological Bulletin 143(1), 1–35. doi:10.1037/bul0000052. Costa-Silva, Jean & Vera Lee-Schoenfeld. (2024.) Syntactically branching out beyond the traditional classroom: A report on the discovery method. Language 100(3), e99–e123. Dockum, Rikker & Caitlin M. Green. 2024. Toward a big tent linguistics: Inclusion and the myth of the lone genius. Inclusion in Linguistics. Oxford University Press. Kuiper, Koenraad. 2011. Teaching Linguistics: Reflections on Practice. Equinox Publishing Ltd. Partee, Barbara H. 2011. On teaching formal semantics. Teaching Linguistics: Reflections on Practice, 40–50. Equinox Publishing Ltd.

ABSTRACT. Introductory syntax forms a core part of undergraduate linguistic training. While it is necessary in order to engage with literature across a range of sub-disciplines, syntax is also seen by many students as unnecessarily difficult or obtuse. This perception is often compounded by traditional “one and done” assessment structures, which provide limited opportunities for feedback or development. Combined, these factors tend to lead to high levels of anxiety and a focus on “just getting the mark”. Allowing re-sits or extra credit can result in unwanted increases in marking time, administration and potentially inequality. Instead, this presentation reports initial results of a project meant to address these issues through implementation of an alternative assessment structure, Skills-Based Grading (SBG), in a large UK undergraduate linguistics module. Our findings include lower student anxiety and greater focus on content over marks, without drastic changes to lecturer or tutor workloads. This system can be implemented even within the self-imposed constraints of UK marking and moderation regimes.

ABSTRACT. Research into Residence Abroads (RAs) has typically focused on students with European L1s (see Borràs and Llanes, 2021 for a state-of-the-art). Furthermore, many employ only offline measures of linguistic development. The current study addresses this gap with a longitudinal investigation into the development of L1 Mandarin speakers’ online processing of both syntax and phonology during a RA at a UK university. This study is – to the best of our knowledge – the first to use eye-tracking data to investigate the development of multiple linguistic levels in a study abroad context. At the beginning of their RA (timepoint 1), 30 L1 Mandarin speakers (and 30 L1 English controls) completed two Visual World eye-tracking experiments testing syntactic and phonological processing in English. The syntax experiment tested predictive processing of number with the determiners this/these/that/those, each followed by 8 singular and 8 plural nouns, as well as a the+noun condition to test sensitivity to nominal number marking independent of determiner number marking (English and Mandarin differ regarding nominal number morphology). The phonology experiment tested processing of 12 minimal pairs contrasting /i/ vs. /ɪ/ (e.g., sheep vs. ship), a tense-lax contrast which is known to be difficult for L1 Mandarin learners of English (Yang et al., 2016), as well as 12 word-final /s/ vs. /z/ (e.g., rice vs. rise) minimal pairs. Given that the /i/–/ɪ/ contrast helps to distinguish the English determiners these and this (along with /s/–/z/), we hypothesised that participants who struggle with the processing of this vowel contrast would also struggle to use these two determiners predictively in online processing. Results of the syntax experiment reveal that for the congruent conditions (e.g., that + singular noun, these + plural noun), L1 Mandarin speakers correctly predict the upcoming noun. The exception is when the determiner is this (Figure 1). Generalised Additive Mixed Models confirmed that the difference between the proportion of looks to the target vs. competitor is significantly lower for the this condition compared to the other three determiners. Further analyses suggest this may be linked to a phonological processing difficulty: in the phonology experiment, the difference between the proportion of looks to target vs. competitor is significantly higher for /i/ than for /ɪ/ (i.e., the L1 Mandarin speakers find the processing of the vowel in this more challenging than the vowel in these). This is likely due to the phonetics of the Mandarin /i/ vowel mapping on to English /i/ more straightforwardly than /ɪ/. Indeed, a significant correlation between the processing of /ɪ/ and the predictive processing of this is found even after partially out participants’ proficiency scores. We argue that the confusion of /ɪ/, in this, with the tense vowel, /i/, may lead to partial activation of these, inhibiting the prediction of a singular noun from the determiner. No difference was observed between the processing of /s/ and /z/. Furthermore, results of the syntax experiment suggest that even once L1 Mandarin speakers have heard the final consonant cluster of the noun in the this + singular noun condition i.e., the absence of plural marking, they never fully discount the plural noun. This is again confirmed by statistical modelling and we explore possible explanations for this. Group-level results from timepoint 2 (end of RA) indicate no significant improvement to the predictive processing of the this condition. We argue this appears to be due to there also being no group-level improvement in the phonological processing experiment at the end of the RA. Analyses of individual results, however, do suggest that certain participants whose phonological processing had improved also improved their predictive processing of syntax. We explore the potential effects of background variables such as the quantity and quality of English exposure (in particular exposure to native vs. non-native English whilst in the UK) on the development of processing from timepoint 1 to timepoint 2. References Borràs, J. and Llanes, À. (2021). Re-examining the impact of study abroad on L2 development: a critical overview. The Language Learning Journal, 49(5). doi: 10.1080/09571736.2019.1642941 Yang, X., Shi, F., Liu, X. & Zhao, Y. (2016). Learning styles and perceptual patterns for English /i/ and /ɪ/ among Chinese college students. Applied Psycholinguistics, 37. doi: 10.1017/S014271641500020X

ABSTRACT. Introduction. This study investigates the impact of pre-reading morphological instruction (PRMI) on the acquisition and retention of syntactically complex, morphologically rich vocabulary among EFL learners. Building upon the foundation of the original research by Yuan & Tang (2025), which demonstrated the effectiveness of PRMI over traditional word meaning explanations (PRWME) and reading-only (RO) conditions, this extension specifically focuses on complex syntactic word structures such as nominalizations and participial forms. Previous research underscores the limitations of incidental learning through reading alone, emphasizing the need for explicit intervention, especially for morphologically and syntactically complex vocabulary. Morphological awareness is theorized to enhance lexical quality by binding orthographic, phonological, grammatical, and semantic properties. Given that complex words often operate within complex syntactic structures (e.g., redistribute resources and accusation that he lied), morphological training may facilitate not just word recognition but also syntactic integration. Methods. 93 Chinese EFL college learners with comparable vocabulary sizes (M = 30.71/140) were randomly assigned to three conditions: PRMI (n=31), PRWME (n=31), and RO (n=31). 17 morphologically and syntactically complex target words containing Latin bound roots were selected (e.g., accusation, redistribute, advocate). PRMI group received 45 minutes of morphological analysis instruction weekly for six weeks. PRWME group received word meaning explanations. RO group engaged only in reading without prior word exposure.Immediate and delayed posttests measured two dimensions of orthography and form-meaning mapping. Linear mixed-effects models (LMMs) and ANOVA comparisons using R were applied. Cohen’s d measured effect sizes. Results. The PRMI group achieved the highest mean scores in both orthographic and form-meaning tests at both time points. Although all groups showed a decline from immediate to delayed posttests, the PRMI group exhibited the smallest decrease, suggesting stronger retention. Standard deviations indicated that the PRMI group also had more consistent performance compared to PRWME and RO groups, as shown in Table 1. Additionally, LMMs showed significant effects for group (F(2, 90) = 92.45, p < .001) and time (F(1, 91) = 38.62, p < .001), as well as a group × time interaction (F(2, 90) = 7.54, p < .001). PRMI significantly outperformed both PRWME and RO. Effect size analyses (Cohen’s d) confirmed the practical significance of these differences. Figure 1 and Figure 2 illustrated that PRMI learners consistently outperformed PRWME and RO learners, maintaining higher scores over time. Discussion and Conclusion. The results demonstrate that pre-reading morphological instruction benefits both the immediate acquisition and longer-term retention of morphologically and syntactically complex vocabulary. PRMI learners achieved higher scores and exhibited smaller declines over time compared to PRWME and RO groups. One reason for the superior performance of PRMI learners is likely the emphasis on morphological decomposition, which may have facilitated deeper lexical processing and enhanced memory retention. Additionally, since many target words appear in complex syntactic environments, morphological training may have improved learners’ sensitivity to grammatical structures. These findings support integrating morphological instruction into language teaching, especially for academic vocabulary development. Future research could extend these results by examining productive language skills and investigating long-term effects over an academic year.

Selected references. Yuan, X. & Tang, X. (2025). The effect of pre-reading morphological intervention on the acquisition and retention of morphologically complex words. System.

ABSTRACT. Within the minimalist framework, head movement has been considered to take place at PF (e.g., Chomsky 2001), as it does not instigate any LF effect. Chomsky (2001) admits, however, that incorporation in the sense of Baker (1988) is relevant to narrow syntax. The principal goal of this study is to show that the alternation between Source PP and accusative DP observed with the EXIT-class of verbs in Japanese, e.g., deru “exit,” saru “leave,” and hanareru “part,” involves head movement of the Source postposition to an EXIT-class verb. The movement in question can be recast in Chomsky’s (2021) theory as “Internal Merge.” To begin our discussion, (1) and (2) illustrate that with EXIT-class verbs, Source can be expressed either by PP or accusative DP: (1) Ken-ga heya-kara/-o de-ta. Ken-NOM room-from/ACC exit-PAST “Ken exited the room.” (2) Tanaka-sensee-ga daigaku-kara/-o sat-ta. Tanaka-teacher-NOM university-from/ACC leave-PAST “Prof. Tanaka left the university (e.g., due to dissatisfaction).” The alternation between PP and accusative DP in (1) and (2) can be compared to the alternation observed in Kinyarwanda (Kimenyi 1980, cited in Baker 1988: 238). (3) a. Abaana b-iica-ye ku meeza. b. Abaana b-iica-ye-ho ameeza children SP-sit-ASP on table children SP-sit-ASP-APPL table “The children are sitting on the table.” “The children are sitting on the table.” Baker (1988) claims that the affixal preposition -ho in (3b) moves to the verb, which derives an applicative construction. The preposition-less object is called an applied object and has been shown to exhibit behavior similar to that of a direct object (Marantz 1984, Baker 1988). As regards Japanese facts such as (1) and (2), (4) indicates that the presence of the accusative object, entotu “chimney,” requires an agentive subject, not the non-agentive subject, kemuri “smoke.” Miyake (1996) argues that this can be taken to show that the Exit-class verb involving P-to-V movement assigns an external θ-role to the subject and that the accusative object behaves as a direct object, just as the applied object in (3). Note that, unlike languages such as Kinyarwanda, the incorporated P in Japanese is morphologically null. (4) a. Kemuri-ga entotu-kara/*-o de-ta. smoke-NOM chimney-from/*-o go.out-PAST “Smoke went out of the chimney.” There is empirical support for the null Source P and its movement, based on Takita & Goto’s (2013) argument on PP-internal N'-deletion (NP-ellipsis hereafter (see Lobeck 1990, Saito & Murasugi 1990)). Note first that the target NP identical to the antecedent NP can be elided, as illustrated by the double strikethrough in (5): (5) [DP Kenk-no [NP tk taido]]-wa yoi-ga, [DP Junk-no [NP tk taido]]-wa 　yoku-nai. Ken-GEN attitude-TOP good-though Jun-GEN attitude-TOP good-not “Though Ken’s attitude is good, Jun’s is not.” This NP-ellipsis can also be applied to PP-internal NPs, as long as the identity constraint is satisfied, as in (6b) and (6c), in which the target NPs identical to the antecedent NP in (6a) are elided (Takita & Goto 2013). (6) a. Antecedent: Kinseezin-wa [kinoo-no [NP kaseezin-no koogeki]-kara] seekansita. Venusian-TOP yesterday-GEN Martian-GEN attack-from 　survived “(lit.) Venusians survived [from yesterday’s attack by Martians].” b. Target: Suiseezin-wa [kyoo-no [NP kaseezin-no koogeki]-kara] toosoosita. Mercurian-TOP today-GEN Martian-GEN attack-kara 　 ran.away “(lit.) Mercurians ran away [from today’s attack by Martians].” c. Target: Suiseezin-wa [kyoo-no [NP kaseezin-no koogeki]-e] taioosita. Mercurian-TOP today’s Martian-GEN attack -to responded “(lit.) Mercurians responded [to today’s attack by Martians].” Interestingly, when quantifiers are present, the identity constraint extends to Ps. That is, ellipsis is blocked in cases where Ps are not identical. Consider (7) (Takita & Goto 2013). Crucially, in (7c), the P -e “to” in the target differs from the P -kara “from” in the antecedent, and the ellipsis results in ungrammaticality. Recall that this difference in Ps does not affect the grammaticality of (6c) under ellipsis: (7) a. Antecedent: Kinseezin-wa [subete-no [NP kaseezin-no koogeki]-kara] seekansita. Venusian-TOP all-GEN Martian-GEN attack-from survived “(lit.) Venusians survived [from all attacks by Martians].” b. Target: Suiseezin-wa [hotondo [NP kaseezin-no koogeki]-kara] toosoosita. Mercurian-TOP most Martian-GEN attack -from ran.away “(lit.) Mercurians ran away [from most attacks by Martians].” c. Target: *Suiseezin-wa [hotondo [NP kaseezin-no koogeki]-e] taioosita. Mercurian-TOP most Martian-GEN attack-to responded “(lit.) Mercurians responded [to most attacks by Martians].” Accounting for the contrast between (6c) and (7c), Takita & Goto (2013) have proposed that in the presence of quantifiers in the antecedent and the target, as in (7), P moves to Q, and the complement of Q undergoes ellipsis, as illustrated in (9b) and (9c). Importantly, the identity of the trace (or copy) of P in the target with that in the antecedent is required in the same way as the trace (or copy) of V in the elided VP must be identical to that of the antecedent VP in V-stranding VP-ellipsis as observed in such languages as European Portuguese, Hebrew, Irish and Japanese (Goldberg 2005, Funakoshi 2016), in which VP is elided after its V head raises to a higher functional head. In (9c), the trace of P2, which is not identical to that of P1 in the antecedent, violates the identity constraint. In contrast, in the absence of quantifiers, Ps are not involved in ellipsis, as shown in (8), hence not subject to the identity constraint. (8) a. Antecedent : [FP [NP …] P1] (9) a. Antecedent : [QP [FP [NP …] tk ] Q-P1k] b. Target: [FP [NP …] P1] b. Target: [QP [FP [NP …] tk ] Q-P1k] c. Target: [FP [NP …] P2] c. Target: *[QP [FP [NP …] tk ] Q-P2k] Provided (i) that Takita & Goto’s (2013) analysis is correct and (ii) that (1) and (2) with the accusative DPs involve null Ps that undergo movement, it should follow that the identity constraint can be satisfied concerning the trace of the overt Source P -kara in the antecedent and that of the null Source P in the target. Consider (10): (10) a. Kinseezin-wa [QP subete-no [FP kaseezin-no [NP koogeki ] tk] Q-karak] nogareta. Venusian-TOP all-GEN Martian-GEN attack 　　 -from escaped “(lit.) Venusians escaped from all attacks by Martians.” b. Suiseezin-wa [QP hotondo [FP kaseezin-no [NP koogeki ] tk ] Q-t'k]]-o nogareta-∅k. 　 Mercurian-TOP most Martian-GEN attack -ACC escaped 　 “(lit.) Mercurians escaped most attacks by Martians.” The prediction is borne out: Ellipsis is licensed in (10b). Note that the reverse order, namely, placing the accusative DP in the antecedent and the -kara phrase in the target, also results in grammaticality. Note that our analysis is further supported by (11). In (11a), with the presence of the quantifier, the P -de “with” in the antecedent moves to Q, leaving a trace, while (11b) does not contain an EXIT-class verb, and neither a null Source P nor its trace is involved in the derivation. Thus, the identity constraint is not satisfied, and the ellipsis is not licensed. (11) a. Kinseezin-wa [QP subete-no [FP kaseezin-no [NP koogeki ] tk] Q-dek] hiheesita. Venusian-TOP all-GEN Martian-GEN attack -with got.exhausted “(lit.) Venusians got exhausted [with all attacks by Martians]. b. *Suiseezin-wa [QP hotondo [FP kaseezin-no [NP koogeki ] ]-Q]-o hihansita. Mercurian-TOP most Martian-GEN attack -ACC criticized “(lit.) Mercurians criticized [most attacks by Martians].” (cf. Takita & Goto 2013)

ABSTRACT. This paper examines the complex relationship between language policy and classroom practice in Nigerian primary education, focusing on the impact of language of instruction (LOI) on educational experiences and attainment. Nigeria's language-in-education policy states that instruction should be in the language of the environment for the first three years of primary education before transitioning to English. However, implementation varies widely, creating tensions between policy requirements and educational needs.

Drawing on data from a nine-week ethnographic study across four primary schools in Lagos State, Nigeria, this research provides empirical insights into the classroom realities of multilingual education. The dataset includes fourteen hours of classroom observations, interviews with 25 teachers, 92 pupils, 18 parents/guardians, and 2 head teachers, complemented by 16 focus groups and questionnaires. This diverse methodological approach enables triangulation of perspectives from key stakeholders navigating the linguistic complexities of Nigerian education.

Preliminary findings reveal significant discrepancies between official language policy and actual classroom practices. Teachers frequently employ strategic translanguaging practices—moving between English and indigenous languages—to facilitate comprehension, despite official English-only mandates. This pragmatic approach responds to the immediate needs of learners with varying degrees of English proficiency but creates tensions with assessment practices that remain predominantly English-focused.

The study also identifies divergent attitudes among stakeholders. While many teachers recognise the pedagogical value of incorporating local languages, they express concerns about deviating from policy mandates. Parents often display complex language ideologies that simultaneously value indigenous language preservation while prioritizing English for socioeconomic advancement. Pupils demonstrate remarkable linguistic agency, strategically employing their multilingual resources to navigate classroom discourse.

These findings contribute to broader educational linguistics debates about language policy implementation in postcolonial contexts, highlighting the need for approaches that legitimise students' full linguistic repertoires while supporting academic achievement. The research has significant implications for language-in-education policy in Nigeria and similar multilingual settings, suggesting that rigid monolingual policies may impede rather than enhance educational outcomes. A second phase of data collection planned for April-May 2025 will further explore these dynamics through additional classroom observations and interviews with Local Government Education Authority (LGEA) officials responsible for policy implementation. This will provide crucial insights from policymakers, creating a comprehensive picture of the policy-to-practice pipeline while focusing on assessment practices and literacy development in multilingual classroom environments.

ABSTRACT. India’s ambitious new National Education Policy 2020 (NEP2020)—its first since 1986—sets out the lofty goal of overhauling all levels of the education system to address inequalities and ensure inclusive education, in line with UN Sustainable Development Goal 4: “ensure inclusive and equitable quality education” for all by 2030 (United Nations, 2015:19). In reaching for this rather tight deadline, its language-in-education section prescribes the classroom inclusion of ‘local/regional’ languages (GOI, 2020, p.13) as part of a three-language formula for education. This formula embodies the “unity in diversity” motto of India’s first Prime Minister Jawaharlal Nehru, which Axel (2002) argues continues to inform the Indian Government in modern times.

In accordance with these policy goals, the present research aimed to analyse whether minoritised languages are likely to be included within the educational framework that NEP2020 provides, and if not, then how might this impact their speakers. What are their attitudes to such educational inclusion in the first place? These are important questions to answer given NEP2020’s own acknowledgement that language-inclusive education is “the best tool for achieving economic and social mobility, inclusion, and equality” (pg.4) and that “young children learn and grasp nontrivial concepts more quickly in their home language/mother tongue” (p. 13). These claims are indeed consistent with the current academic literature on mother-tongue-based multilingual education (Benson, 2019; Malone, 2019).

In order to interrogate these questions, a bidirectional methodology, combining ‘top-down’ critical policy analysis (O’Connor and Rudolph, 2023)—via the empirical analysis of NEP2020 and previous Indian education policy papers—with the ‘bottom-up’ analysis of speaker attitudes—via both quantitative (questionnaire) and qualitative (interview) means—is used. The present research uses a case study of Awadhi, a minoritised language—often pejoratively called “dehati” (‘villager’), including by its speakers—spoken in Uttar Pradesh (UP) (Saxena, 1971). UP is within the ‘Hindi Belt’, where Hindi—India’s most widely spoken language—is the dominant lingua franca, subsuming many regional Indo-Aryan languages, including Awadhi. These are largely thought of as dialects or corruptions of Hindi (“boli”), rather than languages in their own right (“bhasha”), and as such are subsumed under ‘Hindi’ by institutions including the Indian Census; almost none are officially recognised at the state or national levels.

The policy analysis found that, despite its ambitious talk, NEP2020’s language-in-education provisions are unlikely to be successfully implemented regarding minoritised languages like Awadhi. The policy lacks crucial roadmaps to implementation—without which “even the most well-intended policies are liable to failure” (Mahapatra and Anderson, 2023, p. 2)—including financial planning and mechanisms to check or encourage its success. Its vague language and conflations of key terms allow ample room for corner-cutting at every stage of implementation (Cairney, 2019), particularly in the Hindi Belt, where many minoritised languages are officially subsumed under Hindi. In fact, the policy not only glosses over key questions of when and how its aims will be executed, but also whether they should be: the precarious phrase “wherever possible” is used five times throughout NEP2020 (pg. 13, 29, 54); four of these are in its language-in-education section, leaving further room for corner cutting during implementation.

Meanwhile, Awadhi speakers’ attitudes (currently largely omitted from the discourse) were elicited, finding that—although both men and women strongly support its sociocultural value—while men highly supported Awadhi’s educational use, women were strongly against, much preferring the prestige, high-status Hindi and English used in all official contexts (Deo, 2018), including education. Their reasons were primarily those of future employment, concurring with Nepalese findings on Awadhi (Thakur and Yadav, 2013). Interviewees recalled being struck and humiliated by their teachers as punishment for speaking their language at school as children, and parents were told their children would have no prospects using Awadhi, experiences concurring with the literature on minoritised languages elsewhere (Agnihotri, 2020).

Given that first-language education has been found worldwide to be more effective than the sole use of national languages (Heugh, 2005), these policy failures mean a golden opportunity is being missed to not only improve the education system for minoritised speakers—especially important for Awadhi, considering UP’s low literacy and high school drop-out rates, particularly among girls (World Bank, 2019)—but also counteract their current marginalisation, both in education and the broader sociolinguistic inequalities produced as a result. Reversing this would take a sizeable shift in thinking from unsympathetic stakeholders, but this necessity is not acknowledged in the policy and possible strategies to accomplish this are therefore absent.

India has made clear its intention to nurture a new, multilingual generation. However, policy failures and unequal sociolinguistic contexts mean minoritised language speakers will likely be overlooked during NEP2020 implementation. If India is to achieve true educational inclusion for its next generations, let alone by 2030, policymakers should use these findings to draft implementation guidance specifically addressing its issues, sufficiently plan and fund class-room use, and better understand women’s concerns—often overlooked during policymaking (Weldon, 2002)—regarding pervasive negative attitudes to work towards educational equity. The exclusion of minoritised languages must be addressed before it becomes too late (Volker and Anderson, 2015); as Mahatma Gandhi stated, while advocating for education to be provided in nondominant Indian languages, “the question of vernaculars as media of instruction is of national importance… [the] neglect of the vernaculars means national suicide” (1922, p. 307).

Bibliography: Agnihotri, R.K., 2020. Examining the Linguistic Dimension of Draft National Education Policy, 2019. Econ. Polit. Wkly. 55, 43–49. Axel, B.K., 2002. From the margins: historical anthropology and its futures. Duke University. Benson, C., 2019. L1-based multilingual education in the Asia and Pacific region and beyond. The Routledge International Handbook of Language Education Policy in Asia, p.29–41. Cairney, P., 2019. Understanding Public Policy: Theories and Issues. Bloomsbury Publishing. Deo, A. (2017) Dialects in the Indo‐Aryan Landscape. The Handbook of Dialectology. Gandhi, M., 1922. Speeches and writings of MK Gandhi. GA Natesan. GOI, 2020. Government of India. National Education Policy 2020. Heugh, K., 2005. Mother-Tongue Education is Best. HSRC Rev. 3(3):6-7. Mahapatra, S.K., Anderson, J., 2023. Languages for learning: a framework for implementing India’s multilingual language-in-education policy. Curr. Issues Lang. Plan. 24, 102–122. Malone, S., 2019. Promoting Multilingual Education. MTB MLE Resource Kit. UNESCO. O’Connor, K., Rudolph, S., 2023. Critical Policy Analysis in Education. Saxena, B.R., 1971. Evolution Of Awadhi. Allahabad. Motilal Banarsidass Publication. Thakur, I. and Yadav, S. (2013) A Sociolinguistic Survey of Awadhi. TU, Kathmandu. United Nations (2015). The 2030 Agenda for Sustainable Development. A/RES/70/1. Volker, C.A., Anderson, F.E., 2015. Education in languages of lesser power: Asia-Pacific perspectives. John Benjamins Publishing Company. Weldon, S.L., 2002. Beyond Bodies: Institutional Sources of Representation for Women in Democratic Policymaking. J. Polit. 64, 1153–1174. World Bank (2019) Literacy rate, adult total (% of people ages 15 and above) in 2019.

ABSTRACT. In this study, we present the languages of Chokri, Sopvoma [Mao], and Tenyidie [Angami] to show that tone can be used to study the genealogical relationships between languages, and also track the historical development of the tone systems over time. This approach also provides for the proposal of the tone system of the proto-language that the three languages are assumed to be descended from. Chokri, Sopvoma, and Tenyidie are all endangered languages of the Tibeto-Burman family spoken in north-east India in the states of Nagaland and Manipur. They are spoken in geographically adjacent areas and all fall under the Angami-Pochuri sub-classification of the family. However, there are no finer classifications relating the languages under this sub-family. All these three languages have at least four level tones used for lexical differentiations, but have very different tone systems when it comes to morphological interactions or even simply in the lexical entries of the cognates. Having four level, lexical tones in the inventory for a language is generally considered rare in the world’s languages and perhaps, therefore, tone is not usually used as a cue for historical studies. In this study we find that there are enough tonal cues to establish a relationship between the three languages, and they support non-tonal cues. For example, in this study we collected around eighty words from a curated wordlist (adapted from the Leipzig-Jakarta wordlist, Haspelmath & Tadmor 2009) and find that in Sopvoma and Tenyidie, where majority of the words can be identified as cognates, tones can be variedly different. But, upon closer inspection, it is seen that around 75% of the cognate pairs can be seen to have tones on the same register (tone height space). That is, in most of the pairs, the case is that when one of them is Extra-High for tone, the other is High, if not the same (tones in the higher register); and the same is for Mid and Low (tones in the lower register). On the other hand, for Chokri and Tenyidie, all the words documented can be identified as cognates, and at the same time, barring a few exceptions, tones in the lower register are the same―that is, Mid and Low tones correspond exactly between the pairs. Extra-High and High tones do not correspond that well, however, but like Sopvoma and Tenyidie, the pairs remain in the same tonal space, the higher register. The tone data here suggest that Chokri and Tenyidie are closer to each other than Sopvoma and Tenyidie are (which is also evidenced by the non-tonal cue of have more cognate pairs in the pairs). This study also provides clues to historical development of tone where it may be surmised that the proto-language had a two-tone system from which it branched into Sopvoma on one side and Chokri and Tenyidie on the other side. The latter side then developed into a three-tone system, giving rise to the tone pattern in Chokri and Tenyidie. Ultimately, all the languages developed into to a four-tone system one way or the other, which is seen in all the three languages in the current tine. A similar study for the comparison of tone in Chokri and Sopvoma is underway.

REFERENCE: Haspelmath, Martin and Uri Tadmor (eds.), 2009. Loanwords in the World’s Languages: A Comparative Handbook. Berlin and New York: Mouton de Gruyter.

ABSTRACT. See Attached File

ABSTRACT. See pdf.

ABSTRACT. Variation in Present-Day Swahili: Emerging Dialect Areas

General Session

Swahili is a major East African lingua franca spoken by more than 100 million people from diverse communities across a wide range of socio-cultural contexts. While variation in Swahili, particularly in relation to phonology and morphology, has long been noted (e.g. Steere 1870, Stigand 1915, Bakari 1985, King’ei 2000, Rugemalira 2010), less attention has been given to structural variation and to the emergence of new varieties (cf. Shinagawa & Nassenstein 2019). In particular, the effects of post-independence language policies and nation-building, multilingualism and language contact - both structural and sociolinguistic - in the emergence of new dialects has only recently begun to be analysed.

This study addresses some of these challenges by drawing on structural and sociolinguistic data from a comparative study on Swahili variation. We present data collected across four locations in Kenya (Kilifi, Kisumu, Lamu, Nairobi) and six locations in Tanzania (Dar es Salaam, Iringa, Moshi, Mtwara, Mwanza and Tanga). We show examples of morphosyntactic variation, by focusing on three variables: Habitual marking, locatives, and noun classes and agreement.

For example, in (1-3) we see instances of variation in habitual marking in different Swahili varieties. Example (1) shows the use of standard habitual marker hu- in a grammaticalised auxiliary construction. In (2) the habitual meaning is encoded by the suffix -ag-, as previously described for ‘colloquial mainland Swahili’ (Rugemalira 2010) and in (3) the suffix -ang- is used employed in an example from Nairobi Swahili.

(1) Kanisa hu-w-a tu-na-onge-a Ki-swahili na Ki-ingereza. 5.church HAB-be-FV SM1PL-PRS-speak-FV 7-Swahili and 7-English ‘At church we normally speak Swahili and English’ (Kilifi, KFMB03_230410_SQ01)

(2) Si-lal-ag-i mchana SM1SG.NEG-sleep-HAB-PRS.NEG 9.afternoon ‘I don’t usually sleep in the afternoon’ (Iringa, IRMA01_220621_EL01))

(3) A-na-ku-j-ang-a like annually SM1-PRS-STM-come-HAB-FV like annually ‘He usually comes [to visit us] annually’ (Nairobi, KWGMB29_221206_IN27)

The morphosyntactic data are linked to results from sociolinguistic surveys, which show different language practices across communities, varying attitudes towards Swahili, and different perceptions of this variation.

Morphosyntactic and sociolinguistic data from our data suggest the emergence of five broad dialectal areas: Kenya Mainland, Tanzania Mainland, Coastal Dialects, Western Swahili and Nairobi as a distinct urban dialect region. Our data also show that while larger geographical dialect groupings are emerging, micro- and macro-variation is also found within and across these zones.

Results reported in the talk will help to better understand the emerging dialectal variation in Swahili while at the same time providing novel and complex evidence for dialectology and variationist theory and methodology which has had a ‘long-standing bias towards speech communities in the Western and especially Anglophone societies’ (Kasstan 2017, Adli & Guy 2022: 10). We show how the study of dialectology and variation linguistics stands to derive considerable benefits from the study of African languages.

References Adli, A., & Guy, G. R. 2022. Globalising the study of language variation and change: A manifesto on cross- cultural sociolinguistics. Language and Linguistics Compass, 16(5-6), e12452. Bakari, Mohamed. 1985. The Morphophonology of the Kenyan Swahili Dialects. Berlin: Dietrich Reimer. Kasstan, J. 2017. New speakers: Challenges and opportunities for variationist sociolinguistics. Language and Linguistics Compass, 11(8), e12249. King’ei, K. 2000. Problems of Acceptability of Standard Swahili Forms by Non-Kiunguja Native Kiswahili speakers. In: K. Kahigi, Y. Kihore and M. Mous (eds.), Lugha za Tanzania. Languages of Tanzania, pp. 81–88. CNWS Publications, Universiteit Leiden. Rugemalira, Josephat M. 2010. The -ag- TAM marker and the boundary between cliticization and inflection in Bantu. In Karsten Legère and Christina Thornell (eds.) Bantu Languages: Analyses, Description and Theory. Cologne: Rüdiger Köppe, 229-237. Steere, Edward. 1919. A handbook of the Swahili language as spoken at Zanzibar. London: Society for Promoting Christian knowledge. Stigand, Chauncy Hugh. 1915. Dialect in Swahili: A Grammar of Dialectic Changes in the Kiswahili Language. Cambridge: Cambridge University Press. Shinagawa, Daisuke & Nassenstein, Nico. 2019. On variation in Swahili: Current approaches, trends and directions. Swahili Forum 26: 1-45.

ABSTRACT. Gesture has been of interest to science for decades, with McNeill (1992)’s characterisation a turning point for a shared terminology. The last 10 years have seen attempts to integrate the psychology literature for gesture with linguistic analysis; the study of the formal semantics and pragmatics of gesture being key (Schlenker, 2019; 2020; Francis et al., 2023). Consideration of superlinguistic phenomena (Patel-Grosz et al., 2023) is central to understanding the human faculty for language, for if they too have formal properties, a theory of domain-general signs is possible. While a grammar of gesture has been proposed with respect to its relation with speech (Fricke, 2013), the literature currently lacks a precise characterisation of the properties of the gesture system itself. I propose several features deserve closer inspection: gesture roles, role hierarchies, and reference establishment.

Many gesture ‘types’ have been offered throughout their study (Ekman and Friesen, 1969; Kendon, 2004; McNeill, 1992; Wundt, 1973; inter alia.). However, most suffer a lack of mutual exclusivity, inhibiting generalisations of categorisability of gesture, in part due to the conceptualisation of the gesture system as contingent on language (Körner et al., 2022), and in part due to limitations in methodological design and amounts of data (Holler, 2015). My analysis from examples I will present from a proprietary corpus of over 8,000 gestures – 49 25-minute dyadic semi-naturalistic conversations with a confederate – suggests a set of mutually exclusive communicative roles a gesture can play: (a) reference establishment, (b) intersubjective acknowledgement, (c) literal/abstract imagistic representation, (d) emphatic timing, and (e) as a conventional lexeme. Critically, I observe that these roles (reminiscent of syntactic categories) represent a complex system with some apparent properties: arbitrariness of linguistic content and kinematic form, subconscious awareness of gestural intention, individual variability for identical function, and hierarchical organisation.

My formal analysis begins with hierarchical organisation. For most gestures, a one-to-one relationship with a role is easy to establish. For a few, hierarchies of roles exist. This presupposes that there is a foremost intention a speaker has that corresponds to one specific role, but that others may be relevant in fully understanding the gesture production’s meaning(s). Consider (1) and (2). In (1), an utterance is produced with a role (c) two-handed gesture that occurs during underlined elements, where the hands are outstretched with palms facing one another. In (2) this same kinematic form is maintained in up-down motions of a role (d) gesture during the underlined elements in the second clause:

(1) This is a very big idea (2) This is a very big idea and we must do it now

One explanation for (2) is that the resetting of gesture phrases over an utterance is not synchronised with added intentions. Instead, the gesture embodies the added intention within the kinematic restrictions of the required current role on top of the kinematic restrictions of the previous intention’s corresponding role. This leads to the preservation of handshapes from prior, and (possibly) concurrently relevant, intentions’ roles within a new intention’s realisation. In other words, (2) evidences a kind of kinematic productivity where different intentions’ roles cause typically discrete kinematic forms to merge (Chomsky, 1995) into novel, superficially blended phrases. Moreover, such phrases carry dependency relations, where the temporally prior ‘head’’s kinematics predominate through the utterance over a later gesture’s form. With such a formal model, it may be possible to generate syntactic rules based on utterance context and biomechanical limitations for the interaction between temporally prior roles/kinematics with the gestures of utterance-final added intentions.

A second aspect of the gesture system I suggest be subject to formal analysis concerns a property of spatial reference assignment: their temporal decay. They are typically achieved by abstract pointing, which establish points in time, points in space, and discourse indexes. Consider (3), whose common ground contains knowledge of who the object of the verb is. In (3), each underlined element is accompanied by a one-handed pointing gesture to any two different positions around the speaker’s vicinity.

(3) On Tuesday I saw him by Tesco

Example (3) evidences the speaker assigning discourse referents in abstract space which can then be deictically called upon later in the discourse (a phenomenon widely documented in research into sign and gesture; Lillo-Martin and Klima 1990; Schlenker, 2020). I propose that this may be amenable to formal modelling. Reference assignment entails the establishment of a locus. I suggest each locus may be mathematically modelled as temporally decaying in abstract space around the speaker as a function of discourse relevance and the speaker forgetting, qua the decay of radioactive atoms or the decrease in concentration of a medication in the human body. This pertains to the Ebbinghaus (1880) forgetting curve, explaining the logarithmic decay of memories, and future experimental work is required to probe the half-life of gestural loci. The talk will conclude with avenues for further research across linguistics, psychology, and mathematics.

ABSTRACT. In this paper, I want to address criticism of the so-called ‘Gesture School’ in sign language linguistics (Dotter, 2018; Wilcox et al., 2024). I discuss the rise of the ‘Gesture School’ framework within cognitive-functional linguistics and place it in historical context, and how it has also influenced formal linguistic analyses of sign languages. I want to publicly acknowledge on some of the problems this approach inadvertently created and the misunderstandings between scholars working within this framework and some of those working in both cognitive- functional and generative approaches. I partly do this here by revisiting the much-discussed relationship between sign languages, gesture, and spoken language and by reflecting on how proposed boundaries between ‘linguistic’ and ‘non-linguistic’ phenomena have been investigated in our respective fields. I agree that any notion of a binary ‘linguistic’ versus ‘non-linguistic’ distinction is problematic (cf., Goldin-Meadow & Brentari, 2017), reflects academic ideologies about what ‘language’ is (Kusters & Sahasrabudhe, 2018), and that the relationship between gesture, speech, and sign is better examined instead along several dimensions (Okrent, 2002; Coppola & Senghas, 2017; Schembri & Cooperrider, 2021; Cohn & Schilperoord, 2024). Sometimes scholars wish to portray language as categorical, conventionalised, and compositional in nature, whereas gesture is considered as a phenomenon that is gradient, less conventionalised, and/or less integrated into grammar (Brentari & Goldin-Meadow, 2017; Kita & Emmorey, 2023; Cohn & Schilperoord, 2024). Some work has shown, however, that these divisions do not map readily onto gesture and sign (e.g., Kendon, 2004; Occhino & Wilcox, 2017). Another set of useful dimensions are the degree to which aspects of human communication are multimodal (involving more than one perceptual modality, such as audition and vision), multichannel (involving more than one means of production, such as the mouth, the hands, the head etc.), and multi-semiotic (involving different ways of making meaning through symbols, icons, and indexes, either separately or in combination, e.g., Enfield, 2009; Ferrara and Hodge, 2018). I will explore previous work on pointing actions in sign languages and gesture, as well as on verb directionality, along these six dimensions to show how these help us tease apart both similarities and differences, and to ask new questions (Cormier et al., 2013; Schembri et al., 2018; Fenlon et al., 2019). I argue that these distinctions help us break down ‘sign language’ and ‘gesture’ (and ‘linguistic’ and ‘non-linguistic’) into smaller, and, most importantly, less baggage-laden concepts. Investigating sign, gesture, and language along these many dimensions will reveal more about the nature of human communication than the traditional ‘linguistic’ versus ‘non-linguistic’ binary. Like Hou ( 2022) who has called for new approaches to understand the interplay between the structure and function in sign language directional verbs and a movement away from the focus on theoretical debates about whether this system does or does not constitute a type of verb agreement, I would like to see us move away from essentialist discussions about ‘sign language’ versus ‘gesture’ and move towards a more data-driven exploration of these two different aspects of multimodal language systems.

ABSTRACT. (Abstract submitted for the themed session: Gesture and the Linguistic System)

See PDF attached

ABSTRACT. In a public consultation in 2023, the UK Department for Education, sought feedback on the proposal subject content for the GCSE BSL curriculum for England. In his feedback, Adam identified a number of issues around key concepts and terminology in sign language linguistics, and also raised concerns about the curriculum proposing a focus on ‘standard BSL’ (no recognised ‘standard’ form exists, but there is widespread use of a variety heavily influenced by southern English sign varieties on television and in social media). In this talk Adam will discuss the evolution of the BSL GCSE and its current trajectory.

ABSTRACT. Eva Duran Eppler, Reader in Linguistics at the University of Roehampton and Chair of the Committee for Linguistics in Education, will provide an overview of the DfE's Curriculum and Assessment Review (CAR) process. She will outline the aims of the CAR, present a digest of responses to the CAR by 19 language-related stakeholder organisations, and the DfE's Interim Report to stimulate discussion on the latter's marked silence on Knowledge about Language (KAL) and Language Awareness (LA) and what this may mean for linguistics in education.

ABSTRACT. In this talk, I will discuss the motivation behind the creation of the awareness day, the enthusiasm it generated among language specialists and aficionados of diverse backgrounds, and the resources which were offered to schools. I will also speak about plans for National Linguistics Day 2025 and highlight the tremendous scope of the initiative for future development.

ABSTRACT. In this talk, I discuss the role of gestures in the formal treatment of negation in Romance, a topic which has not been investigated before. Specifically, I present a case study on the manual gesture [CHIN-FLICK] ([CF]) in Neapolitan (e.g., southern Italo-Romance) and I argue that in this language, when [CF] accompanies speech, it can be considered as a gestural Negative Polarity Item (NPI), similar to the English `not...at all'. Crucially, [CF] has two striking properties. First, to the best of my knowledge, [CF] is the first reported example of a gestural NPI (cf. gestural questions-markers, focus-markers, topic-markers, etc.). Second, similar to some spoken NPIs found in Dutch, Greek, and English (Den Dikken 2002, 2006; Hoeksema 2007; Goncharov and Zeijlstra 2022) [CF] is licensed parasitically, in that it relies on the successful licensing of some other NPI in the derivation. This broadens the empirical database of parasitic NPI licensing, a poorly-understood phenomenon. Capitalising on new evidence from Neapolitan, I demonstrate that, although rare, parasitic licensing is attested in Romance and also in the visual-gestural modality. This is overall consistent with the "Grammatical Integration Hypothesis" (Colasanti 2023a,b), by which some gestures are morphemes realised at PF in the visual-gestural rather than in the auditory-spoken modality.

ABSTRACT. Schlenker’s formal analysis has revealed that post-speech and pro-speech gestures evidence a rich inferential typology available in the manual modality (Schlenker, 2019). Experimental evidence places the understudied interactive gesture (Bavelas et al., 1992) as one of the most important gestures by frequency of use in dyadic conversation (Williamson, in prep.). We analyse the properties of co-speech interactive gestures in terms of their possible use cases and use conditions to make two proposals. First, that interactive gestures encompass a variety of uses with a formally intriguing property: acknowledgment of an interlocutor’s mental contents. Second, that interactive gestures pertain only to the intersubjectivity of the interlocutors and are divorced from the semantics of the utterance. Nevertheless, co-speech interactives present with use conditions (Kaplan, 1999) that are determined by their temporal alignment with elements of that utterance.

Gestures have been traditionally demarcated qua McNeill (1992): iconic, metaphoric, deictic, and beat. While these are the most frequently analysed gesture ‘types’, scarce attention (Bavelas et al., 1992; de Beer et al., 2019) has been paid to interactive gestures, which seem to have no bearing on an utterance’s semantic content but instead intersubjectively pertain to an interlocutor’s mental contents. These gestures are characterised by the movement of a hand or hands to point towards the interlocutor, with the palm open and facing at some upward angle. Originally, Bavelas et al. (1992) identify four use cases: citing the other’s previous contribution; seeking agreement, understanding, or help; the delivery of new vs. shared information; and events around the speaking turn.

We report data collected from 49 healthy participants in dyadic, naturalistic and experimental, conversations with a confederate. Conversations lasted 25 minutes and consisted of a 5-minute warm-up followed by 20 minutes of conversation prompted by question cards. This dataset reveals that interactive gestures constitute one of the most frequent gesture types, with hundreds of examples. With this, we modify Bavelas et al.’s (1992) original classification of interactive gestures into playing two fundamental roles: (1) acknowledging (previous content of any participant, information that is in the common ground, a response that is anticipated, or a communicative intent), and (2) transferring (information that is new, a softening or conceding of opinion, a request for approval or agreement). Further, contrasting Bavelas et al. (1992), we suggest that interactive gestures’ temporal alignment with utterance elements undermines the proposal that they are entirely divorced from communicative content qua pragmatic intention – they evidence the capacity to acknowledge possible inferences an interlocutor would make, albeit bound by use conditions (Kaplan, 1999) based on a gesture’s temporal alignment with utterance meaning.

Observations of the semantic and pragmatic contributions of co-speech gestures have been a topic of significant interest in the formal literature (Ebert & Ebert, 2014; Esipova 2019; Schlenker, 2019). We propose a formal analysis of the use conditions of co-speech interactive gestures in this vein. In a prototypical usage, a co-speech interactive gesture can be aligned by the speaker over a whole utterance to acknowledge their interlocutor’s prior knowledge of the proposition entailed by their utterance, as in (1). Underlining indicates the temporal alignment of an interactive gesture (e.g., an outstretched hand to the interlocutor). (1)

A and B are discussing their housemate Chris. Earlier in the conversation, A had mentioned that they saw Chris’ car in the driveway of their house. Later, A says: A: “I saw that Chris’ car was in the driveway...” Entailment: A saw that Chris’ car was in the driveway. Information acknowledged: B knows that A saw that Chris’ car was in the driveway.

We propose that this gesture can target other utterance content depending on its temporal alignment with the utterance. For example, in (2) a presupposition is targeted:

(2) A and B are discussing their housemate Chris. They both know that Chris has recently bought a Ferrari, and both know that the other knows this. A has seen Chris’ Ferrari in the driveway of their house, but B has not, and A knows that B has not. A: “I saw that Chris’ Ferrari was in the driveway….” Presupposition: Chris has a Ferrari. Information acknowledged: B knows that Chris has a Ferrari.

Following Schlenker (2019), we propose an analysis of the possible linguistic inferences that interactive gestures can refer to and generate. We conclude that interactive gestures cannot interact directly with utterance semantics, but may interact relatively freely with utterance-level pragmatic information, preceding information from the wider discourse, and, perhaps most commonly, the intersubjectivity of the interlocutors. However, given the availability of targeted temporal alignment over specific elements of utterances as in (2), we suggest that a formal syntactic analysis of this gesture in terms of expressive pragmatic features should be available (following Gutzmann, 2019). We test our predictions for a formal analysis of co-speech interactive gestures with our corpus data.

Finally, we conclude with the observation that how speakers must temporally align interactive gestures with use-conditions alludes to the fact that the linguistic and gestural output originate from the same conceptual source (McNeill, 2015; Hagoort & Özyürek, 2024). More precisely, it appears that the preparation of semantics and intersubjective behaviour is psycholinguistically contemporaneous and coordinated for realisation cross-modally. Exciting avenues for experimental work exist to test these hypotheses.

ABSTRACT. What experimental evidence can tell us about homesign creation (To be considered for the themed session ‘Gesture and the Linguistic System’) Deaf and hard of hearing children (DHH) born to a hearing, non-signing family often innovate their own gestural systems known as homesigns. These are of great interest to language development researchers because they exhibit certain fundamental properties of natural languages such as recursion, combinatoriality, and argument structure (Flaherty et al., 2021) despite their creators not having access to a (fully) usable linguistic model. The standard assumption in homesign research is that caregivers unknowingly provide their children with input by producing gestures alongside their speech. These have been shown to lack linguistic structure; thus, it is further assumed that the children systematize the inconsistent input they were exposed to, innovating linguistic structure. Although this assumption has been discussed at length from a theoretical standpoint, it has never been, to my knowledge, discussed or examined from a neurocognitive perspective. The present research is an original review that synthesizes experimental evidence from a variety of methods including event-related fMRI, artificial language learning, and silent gesture aimed at investigating the internal processes that would be necessary for the standard assumption to be possible. The findings of the reviewed experiments can be classified into three categories: (i) the linguistic dimension of gestures (ii) the effects of cross-modal reorganization following congenital deafness, and (iii) children’s increased capacity to regularize inconsistencies in their input and innovate linguistic structure. The nature of gestures, particularly when produced simultaneously with speech, has been the subject of a large body of research (see Özyürek, 2014 for a comprehensive review) where they are often argued to supplement the semantic, pragmatic, and syntactic content of spoken utterances. As such, their processing interacts with that of speech and largely takes place in the language areas of the brain, specifically, in the inferior frontal and superior temporal gyri (IFG and STG), which correspond to Broca’s and Wernicke’s areas, respectively (Wolf et al., 2017). While it is known that hearing children also make use of gestures during acquisition (Özçalışkan & Dimitrova, 2013), homesigners do not have access to the spoken language information that gestures complement, they only have access to the gestures. Could they be able to derive more information from these gestures than their hearing peers and, if yes, what enables them to do so? Sensory loss induces an adaptive mechanism of the brain known as cross-modal reorganization, where areas designated to the processing of the lost modality are repurposed, leading to enhanced performance in the remaining modalities (Bonna et al., 2021). Crucially, this is the result of sensory loss itself, not of individual experience, e.g., sign language exposure in childhood (Hall et al., 2019), which might in fact improve spoken language outcomes in DHH children with cochlear implants (Davidson et al., 2014). In the congenitally deaf, the primary auditory cortex is repurposed, engaging in the processing of visual stimuli, resulting in both attentional and perceptual enhancements, specifically in the peripheral visual field (Scott et al., 2014). fMRI studies have shown that non-linguistically structured manual communication (gestures) activate brain areas that are traditionally involved in the processing of language (e.g., STG) to a significantly greater extent in early deaf individuals than in hearing individuals (Simon et al., 2020). This suggests that gestures can be processed as linguistic material, especially in the early deaf, where cortical activation patterns during the processing of meaningful gestures (emblems) have been shown to be near-identical to those during the processing of American Sign Language (ASL) signs (Husain et al., 2009). Regarding the ability to innovate structure, a common experimental method is silent gesture (e.g., Bohn et al., 2019; Özçalışkan et al., 2016). Bohn et al. (2019) found that children spontaneously created core linguistic properties, including conventionality and compositionality when asked to describe a picture scene without using spoken language. Compositional gestures improved comprehension compared to holistic ones and subtle differences were ignored with participants copying each other’s gestures and converging on one form, demonstrating an increased capacity for regularization (Bohn et al., 2019). Experiments employing different methodologies, such as miniature artificial languages (Hudson Kam & Newport, 2009) and pattern recognition and rearrangement tasks (Kempe et al., 2015) report comparable findings regarding children’s capacity for regularizing inconsistent input and converging on forms that are more easily transmissible and reproducible. In comparison, adults are more likely to reproduce inconsistencies until the number of alternate forms is too great or they occur in low frequency (Hudson Kam & Newport, 2009). This project highlights not only the necessity of interdisciplinarity in research, but also of the use of novel experimental methodologies in scientific inquiry. Technological advances have allowed us to obtain evidence for hypotheses that form the foundation of a major current within the field of linguistics (see Finkl et al., 2020 for experimental evidence of a ‘core language network’ separated from externalization channels). Similarly, the collection of evidence presented here has provided support for claims that have been central to homesign research for at least the past 20 years (see Goldin-Meadow, 2005) without having been explored to their full extent. References: Bonna, K., Finc, K., Zimmerman, M., Bola, M., Mostowski, P., Szul, M., Rutkowski, P., Duch, W., Marchewka, A., Jednoróg, K., & Szwed, M. (2021). Early deafness leads to re-shaping of functional connectivity beyond the auditory cortex. Brain Imaging and Behavior, 15, 1469-1482., Davidson, K., Lillo-Martin, D., & Chen Pichler, D. (2014). Spoken English language development among native signing children with cochlear implants. J Deaf Stud Deaf Educ. 18(2), 238-250., Finkl, T., Hahne, A., Friederici, A. D., Gerber, J., Mürbe, D., & Anwander, A. (2020). Language without speech: Segregating distinct circuits in the human brain. Cerebral Cortex, 30, 812-823., Flaherty, M., Hunsicker, D. & Goldin-Meadow, S. (2021). Structural biases that children bring to language learning: A cross-cultural look at gestural input to homesign. Cognition, 211, 1-11., Goldin-Meadow, S. (2005). The Resilience of Language: What gesture creation in deaf children can tell us about how all children learn language. Psychology Press., Hall, M. L., Hall, W. C., & Caselli, N. K. (2019). Deaf children need language, not (just) speech. First Language, 39(4), 367-395., Hudson Kam, C. L. & Newport, E. L. (2009). Getting it right by getting it wrong: When learners change languages. Cognitive Psychology, 59(1), 30-66., Husain, F. T., Patkin, D. J., Thai-Van, H., Braun, A. R., & Horwitz, B. (2009). Distinguishing the processing of gestures from signs in deaf individuals: An fMRI study. Brain Research, (1276), 140-150., Kempe, V., Gauvrit, N., & Forsyth, D. (2015). Structure emerges faster during cultural transmission in children than in adults. Cognition, 136, 247-254., Özçalışkan, Ş. & Dimitrova, N. (2013). How gesture input provides a helping hand to language development. Seminars in Speech and Language, 34(4), 227-236. Özçalışkan, Ş., Lucero, C., & Goldin-Meadow, S., (2016). Does language shape silent gesture? Cognition, 148, 10-18., Özyürek, A. (2014). Hearing and seeing meaning in speech and gesture: insight from brain and behaviour. Phil. Trans. R. Soc. B, (369):20130296, 1-10., Scott, G. D. Karns, C. M., Dow, M. W., Stevens, C., & Neville, H. J. (2014). Enhanced peripheral visual processing in congenitally deaf humans is supported by multiple brain regions, including primary auditory cortex. Frontiers in Human Neuroscience, 8:177, 1-9., Simon, M., Lazzouni, L., Campbell, E., Delcenserie, A., Muise-Hennessey, A., Newman, A. J., Champoux, F., & Lepore, F. (2020). Enhancement of visual biological motion recognition in early-deaf individuals: Functional and behavioral correlates. PLoS ONE, 15(8), 1-17., Wolf, D., Rekittke, L. M., Mittelberg, I., Klasen, M., & Mathiak, K. (2017). Perceived conventionality in co-speech gestures involves the fronto-temporal language network. Frontiers in Human Neuroscience, 11:573.

ABSTRACT. (Abstract submitted for the themed session: Gesture and the Linguistic System)

Please see PDF attached.