°µÍø½ûÇø

This study explores how multilingual speakers in India navigate language choice and code switching across social domains. Drawing on Green’s Inhibitory Control Model (1998) and Grosjean’s Language Mode Framework (2008), it examines the cognitive and pragmatic mechanisms underlying language control among university students who use multiple South Asian languages. Using mixed methods through surveys and semi structured interviews, the research investigates when and why speakers switch between English and regional languages in home, academic, and public settings. Findings reveal that language switching is fluid and context sensitive. Informal conversations allow spontaneous shifts driven by emotional resonance and ease of expression, while academic contexts elicit conscious inhibition of non target languages. Participants’ language choices reflect both cognitive regulation and social sensitivity, aligning with bilingual mode theory. The study contributes to understanding multilingual cognition and offers insights for educators fostering inclusive and linguistically responsive learning environments.

This study investigates how Hindi-acquiring children interpret and produce long-distance wh-questions to test whether comprehension and production rely on a shared grammatical system. Building on Lütken et al. (2020), we conducted a within-subjects comparison using both matrix-manner and embedded-object questions, contrasting English-like movement with Hindi scope-marking constructions (e.g., bacce-ne kyaa sochaa ki kis-ko dekhaa?). Results from 5–8-year-olds reveal that participants were highly conservative in production but showed significant non–adult-like comprehension, suggesting a developmental asymmetry between the two modalities. Children who produced target-like scope-marking forms did not necessarily interpret them accurately, indicating partial grammatical knowledge and modality-specific constraints. The findings contribute to ongoing debates that comprehension and production may draw on distinct but interacting components of the developing grammar.

Research on reading and cognitive control has predominantly focused on readers of alphabetic scripts, often overlooking individuals who read other scripts or multiple scripts. This study investigates Stroop performance across Devanagari (abugida) and Perso-Arabic (abjad) scripts in Hindi-Urdu biliterates to examine how orthographic transparency and script directionality influence inhibitory control. Thirty biscriptal adults will complete an online colour-word Stroop task with four counterbalanced blocks: two pure-script (Devanagari, Perso-Arabic) and two mixed-script conditions. Reaction times (RTs) and accuracies across congruent and incongruent trials will be analyzed using linear mixed models with block type and congruency as fixed effects. Faster RTs and higher accuracy are expected for Devanagari due to its greater orthographic transparency, while script-switching in mixed blocks is predicted to increase cognitive load. By isolating script-level effects within a single spoken language, this study contributes to understanding how biscriptality shapes cognitive control mechanisms beyond linguistic boundaries.

We report a self-paced reading study on Malayalam (Dravidian, India) on pronouns in the subject and object position where we investigate potential difference in coreference of pronouns across a clausal boundary. We observe effects of positional parallelism, a cross-linguistically attested phenomenon, in Malayalam as well. To our best knowledge, this is the first work on pronominal reference resolution in Malayalam.

A word-learning/minimal pair experiment with Tamil-speaking children (n=15) asked whether the introduction of word-onset voicing (/b-/), which is non-contrastive in Tamil, affects the production of oro-laryngeal timing of short voice-onset time onsets (/p-/) in known words. Results suggest that, rather than fortifying known words by increasing voice-onset time, speakers contrast the minimal pair by extending prevoicing in the new words. Results are discussed in terms of contrastive hyperarticulation and phonological pressures that may (or may not) have contributed to the debuccalization of /p/ in the history of pre-modern Kannada.

This study examines children’s understanding of Marathi relative clause structures (RCs). Marathi, a split-ergative Indo-Aryan language, has two distinct RC constructions, a gapped RC that is more frequent, but restricted in use, and a correlative clause which is less frequent but available for all combinations of relativizing position (SUBJ vs. OBJ), aspect (imperfective vs. perfective), and verb type (eventive vs. stative).

Noun bias as a fundamental strategy in language acquisition: A longitudinal study focusing on minimally verbal Assamese-speaking children diagnosed with autism. This study investigates the tendency of young children to preferably map nouns than actions (verbs) which is also a well established feature of typical language acquisition. It has been proposed as a fundamental learning approach that supports vocabulary growth in young children. This study has methodically examined whether the noun bias functions in a comparable manner among children with Autism Spectrum Disorder (ASD), uncovering significant theoretical and empirical insights regarding the language acquisition process in autistic children. This research examines two arguments concerning the noun bias observed in children with autism through a longitudinal study conducted over a span of 6 months, aiming to determine whether the results remain consistent throughout the study period or if they exhibit an increase over time. Children diagnosed with Autism Spectrum Disorder (ASD) often demonstrate delays or irregularities in language development, and their difficulties in social communication indicate possible variations in their attention to and processing of linguistic information. In light of these distinctions, it prompts the inquiry of whether noun bias operates similarly in the language acquisition process among children with autism. A total of 82 minimally verbal children diagnosed with autism spectrum disorder (MV-ASD) were examined to assess their receptive and expressive vocabulary profiles. These children were sourced from four private centers located in and around Guwahati: Ashadeep, Emmanuel Early Intervention (Geetanagar branch), Emmanuel Early Intervention (Rukminigaon branch), and Pratidhwani in Rangia, Assam. Each child had received an ASD diagnosis through the Autism Diagnostic Observation Schedule (ADOS). However, participants in the ASD group were eligible to participate if they met the ASD diagnostic criteria established by licensed psychologists or psychiatrists trained in the clinical application of the ADOS-2 instruments. The exclusion criteria encompassed a history of psychiatric or medical conditions, the presence of other neurological impairments, significant hearing impairments, and children whose parents were not willing to provide consent, all of whom were excluded from the study. The participants are first screened and then data were collected by face to face interviews, participant observant and tools such as Behavior Checklist (ABC), Childhood Autism rating Scale (CARS) and Receptive-Expressive Emergent Language Test (REEL) are used. Participants were instructed to respond to specific verbs, with their reactions recorded as 0 for negative responses and 1 for positive ones. In the case of nouns, participants were asked to identify common objects, and their responses were documented. The items utilized in the noun category are :- (Head) /mur/, (Tummy) /pet/, (Pencil) / pencil/, (Water bottle) /pani botol/. The actions commanded in the verb category are :- (Cry) /kanda/ , (Arms up) /hat daŋga/, (Jump) /zopiua/ and (Sleep) /xua/. The research was carried out over duration of six months, during which responses were documented to determine whether the outcomes vary over time and to assess if children with autism have a greater understanding of nouns compared to verbs, or vice versa. The results showed that the children with autism tend to preferentially associate words with objects instead of actions. This suggests that, despite variations in social interaction and overall cognitive characteristics, they employ this lexical bias at an early stage of vocabulary acquisition. Fig 1 illustrates this distinction clearly. This discovery is significant as it indicates that noun bias is not merely a behavior acquired through social interactions or direct interventions, but may instead represent fundamental strategies in language acquisition that go beyond the social challenges typically associated with autism. Furthermore, the findings revealed that this bias was evident shortly after diagnosis and initial engagement with Speech Therapy and Applied Behavior Analysis (ABA) therapy, suggesting that extensive language therapy had not yet impacted this pattern. Also, the results documented over a six-month period also demonstrate that there are no differences in the outcomes, with the children consistently exhibiting noun bias throughout the study which further stands as a testimony that noun bias is a fundamental strategy in language acquisition. Fig 2 Previous research indicated that this inclination appeared solely following significant language intervention, portraying an idea that therapies emphasizing labeling and object naming could artificially enhance noun usage. Nevertheless, this study demonstrates that the noun bias occurs independently of treatment influences, implying the presence of a fundamental language-learning strategy rather than a byproduct of therapeutic intervention. The manifestation of noun bias in children diagnosed with ASD suggests that specific essential lexical learning processes are preserved, devoid of the neurological and social difficulties associated with autism. This bias functions as a strong cognitive and linguistic tool that allows these children to systematically arrange their vocabularies by associating new labels with objects, which are arguably the most consistent and perceptually prominent referents in their surroundings. This aligns with the wider theoretical perspective that certain fundamental language acquisition processes are common across different etiologies, including among groups experiencing delayed or atypical development.

This study explores children’s ability to reconstruct NULL arguments based solely on verbal agreement. In Marathi, an Indo-Aryan language with three grammatical genders (Masc, Fem, Neu), verbal agreement is triggered by the highest null case-marked argument (typically NOM). In eventive verbs which exhibit NOM-ACC marking, this results in subject agreement. However, in stative verbs, which take DAT-NOM marking, this results in object agreement. Marathi frequently has NULL arguments, and Marathi listeners are able to retrieve the full argument structure of the utterance based on gender agreement of the verb and context.

Research has shown effects of preemption and entrenchment on children’s production of grammatical structures and overgeneralization errors. Ambridge et al (2020) describe ‘entrenchment’ as the phenomenon where the overall frequency of a verb root regardless of the structure in which it occurs prevents it from being used in a novel structure that has not been attested before. Preemption (Goldberg, 2011) requires the verb to occur in a competing structure with the same or similar meaning to decrease its possibility of occurring in a particular structure. The present study tests this prediction for children acquiring Marathi with a focus on intransitive verbs and their causativization. Data was collected from Marathi children and their mothers to arrive an understanding of naturalistic language the children are exposed to. Each child's peformance was compared with their own mother, rather than an average, against which overgeneralization errors were recorded. The study did not find significant effects of preemption or entrenchment, but there was an effect of canonical causativizing pattern on children's overgeneralization behavior. Overall, the study attempted to implement different strategies to account for the input.

This study examines how typological features shape child language acquisition by comparing Tamil, an agglutinative language, with English, a fusional one. I propose the Transparency Facilitation Hypothesis (TFH): transparent one-to-one morpheme–meaning mappings ease acquisition relative to opaque morphology. Using CHILDES longitudinal data (Tamil: Vanitha, 1;2–2;6; English: Adam, 2;3–2;6), results show a striking gap: by 25 months, the Tamil child produced adult-like, multi-clausal utterances across tense, aspect, negation, and case, whereas the English child reached comparable milestones over a year later. These findings highlight how morphological transparency accelerates language development and challenge universalist models of acquisition.

Children with hearing loss are at risk for delays in language development, even with early intervention through hearing aids or cochlear implants. This study examined whether supplemental listening and spoken language therapy leads to improvements in behavioral language outcomes and cortical maturation in children with treated hearing loss. Thirty-nine children were assessed at baseline and after approximately six months of therapy using the Preschool Language Scales (PLS-5) and cortical auditory evoked potentials (CAEP). Results revealed language gains exceeding expected maturation, with children showing nine to ten months of progress in total language and expressive communication over six months. CAEP latencies decreased at accelerated rates, and improvements in CAEPs were correlated with behavioral gains. Findings suggest that supplemental therapy promotes neuroplasticity and language growth in children with hearing loss. These effects are likely to be independent of the specific spoken language, though future research should confirm this across diverse linguistic populations.

The study unequivocally advocates for a fundamental pedagogical paradigm shift. It calls for educational systems to formally acknowledge and strategically integrate the immense academic and cognitive value inherent in students' natural multilingualism. By actively supporting and leveraging students' existing linguistic repertoires, educators can not only improve immediate comprehension but also profoundly enhance students' long-term critical thinking, sophisticated problem-solving abilities, and overall academic performance. This approach is vital for cultivating adaptable, critically engaged learners in linguistically rich contexts, moving beyond conventional monolingual biases towards a more inclusive and effective educational framework.

This case study examines the prevalence, characteristics, and classroom manifestations of expressive language disorders among Tamil-speaking children aged 6–8 years in a government primary school setting, with a particular focus on teacher identification practices and the classroom-based expression of language difficulties. A mixed-methods design was employed over five weeks in one government primary school, involving 35 children aged 6–8 years. Data collection integrated three complementary approaches: (1) Individual language assessments culturally adapted for Tamil linguistic structure, including picture naming tasks using Grade 1 textbook vocabulary, sentence construction activities with culturally familiar action scenarios, and narrative elicitation using locally relevant story sequences, all administered in colloquial Tamil; (2) Systematic classroom observations using ethnographic methods to document spontaneous verbal participation, peer interaction patterns, communication breakdowns, and compensatory strategies across instructional contexts, including Tamil language lessons, general class discussions, and social interactions; and (3) Semi-structured interviews with teachers (n=3) to examine identification practices and support needs. Assessment materials incorporated familiar cultural references and local dialectal variations to ensure ecological validity.

This study investigates how caregiver prompt types shape pragmatic responsiveness in autistic children, with implications for multilingual South Asian contexts. Drawing on 180 caregiver–child turns from the ASDBank (CHILDES) and 120 turns from typically developing peers, caregiver prompts were coded as directives, wh-questions, or yes/no questions, and child responses as appropriate, echoic, off-topic, or no response. Results show a significant association between prompt type and response type (χ² = 9.52, p < .01). Autistic children were most responsive to directives (87.1%), followed by yes/no questions (81.0%) and wh-questions (63.9%). While they produced more appropriate responses overall (65.0%) than TD peers (41.7%), they also uniquely showed echoic and non-responses, especially under cognitively demanding wh-questions. These findings highlight how pragmatic difficulty in autism is shaped not only by internal factors but also by the inferential demands of prompts, with direct implications for multilingual, code-switching caregiving contexts.

This study examined how age of language acquisition (simultaneous vs. sequential) and word order similarity (same vs. different) influence attention and cognitive flexibility in multilinguals. Ninety-nine participants from the Central University of Punjab were categorized into four groups based on acquisition type and word order. Attention was assessed using an OpenSesame-based word recognition task measuring accuracy and reaction time, while cognitive flexibility was evaluated with the Stroop Color–Word Test. A two-way MANOVA revealed significant main effects of acquisition type and word order, as well as a significant interaction between them. Simultaneous multilinguals, particularly those with exposure to different word orders, demonstrated higher accuracy, faster reaction times, and better Stroop performance. These results suggest that early multilingual exposure and structural diversity in languages enhance executive functioning, likely due to increased practice in managing competing linguistic systems and greater cognitive demands of switching between distinct syntactic structures.

My work on Parafoveal Preview Benefit in Urdu Word Recognition has been accepted for a talk. The work is about how reading Urdu words are benefitted through parafovea. We found parafoveal processing to be inhibitory than facilitatory in Urdu.

How speakers process morphology in their language could be determined by typological factors, such as the degree of morphological complexity that the language allows. Language-specificities in parsing could therefore emerge out of cross-linguistic differences in morphology. To isolate the role of morphological typology in shaping processing dynamics during visual word recognition, we investigate the potential differences in morphological priming across and within two morphologically complex languages that differ in their degree of complexity – Malayalam, being agglutinating, and Bangla, being fusional. Through a masked-priming lexical decision task, where we manipulate prime types and prime durations, we find a morphological priming effect in Bangla compared to none in Malayalam at shorter prime durations (48ms), and robust priming effects in both languages at later prime durations (600ms) while other variables are controlled for. We speculate that this points to early effects being more nuanced and language-specific, with robust morphological facilitation emerging at later stages of parsing.

This case-study examined the acquisition of Tamil (L2) by adult male and female Nepali migrant workers in the multilingual city of Chennai in Tamil Nadu, India, where they have been employed in the service industry for more than 10 years. Focusing on their use of verb and case morphology, the study asks: Can acquisition of a new language during adulthood in uninstructed settings with exposure only to spoken (colloquial Tamil) input succeed? The results indicated that the male migrant workers were overwhelmingly successful in acquiring the complex verb and case morphology of Tamil based solely on naturalistic exposure to spoken Tamil, in a context characterized by positive attitudes to bi/multilingualism. The female migrant workers were also successful, though less so, especially in relation to verb agreement inflection, due to differences in the language input and interactions in the workplace. The psycholinguistic and sociocultural implications of the findings will be discussed.

Cross-linguistic aphasia research, largely based on European languages, shows that typological variation shapes agrammatic symptoms. Core deficits of agrammatism include fragmented utterances and sentences, deficits in functional morphology, and scarcity of verbs. However, detailed profiles from South Asian languages are lacking. This paper presents findings from the Aphasia in South Asian Languages (ASAL) project https://doi.org/10.1080/02687038.2025.2548249, which investigates agrammatism in under-studied clinical populations speaking Indo-Aryan (Hindi-Urdu, Bengali) and Dravidian (Tamil, Kannada, Malayalam) languages. Using  narrative data from Hindi and Bengali speakers with post-stroke agrammatic aphasia, we observe consistent difficulties in fluency, sentence completeness, structural complexity, pronoun use, and noun and verb inflections. Language-specific patterns reflect morphosyntactic differences: Bengali speakers showed no verb inflection errors, while 40–50% of Hindi speakers exhibit errors in number and gender morphology. These differences highlight the varying complexity of agreement systems in the two languages. Our study offers the first systematic cross-linguistic documentation of agrammatism in these populations using a uniform methodology, with implications for improved assessment and intervention strategies.

The Implicit Association Test (IAT) is a central tool in social cognition research, designed to uncover implicit biases that individuals may not consciously report. Despite its influence, debates persist over its validity, particularly in cross-linguistic contexts. Research shows that the language of administration can alter IAT results, revealing that linguistic factors shape cognitive associations rather than merely reflecting them. This study examines one overlooked linguistic confound — the Gender Congruency Effect (GCE) — in grammatically gendered languages. Using Dari (genderless) and Pashto (gendered), we developed parallel Gender–Career IATs administered to 96 native speakers. Results from the Pashto sample revealed significantly faster responses when stimulus words matched the grammatical gender of category labels (t(1298) = 3.41, p = .001). These findings demonstrate that grammatical gender can systematically bias IAT performance. We advocate for linguistically informed adaptation protocols to ensure the validity of cognitive and psycholinguistic research across languages.

Through a truth-value judgment task, we evaluated children’s ability to reconstruct NULL arguments based on verbal agreement. Fifty-three children in Pune were tested in both TRUE and FALSE conditions across two verb types (eventive vs. stative), finding that while younger children struggle with object agreement, by age 5 children are at ceiling for both verb types.

Spontaneous speech is rarely fluent. Disfluencies, including pauses, filled pauses (such as "uh" and "um"), repetitions, and false starts, are frequent. They reflect planning delays, working memory load, structural, lexical, or phonological complexity, and/or change in discourse. For a speaker, lying carries cognitive effort. There is an inhibition of the truth and additional construction/execution of the lie, effectively affecting speech. Congruently, literature suggests that pitch and disfluencies serve as cues to deception. Between two experiments in Hindi, we find that individuals use more filled pauses when requiring more processing/inhibition. In a naturalistic setting, the average filled pause rate was over 3 times higher in truth-tellers, while the silent pause rate was narrowly higher in liars. Conversely, in the false opinion task, filled pause rate was significantly different across conditions, indicating greater fluency when supporting a statement. There was no difference in pitch observed with respect to deception.

The study focuses on how the cognition of speech-rhythms is made evident through the productions of children and adults speaking Bangla through the phonological construct of vowel lengthening and quantification of pauses. It investigates Bangla speech-rhythm by recording and analysing the spontaneous speech of Bangla speaking adults and one-to-eleven-year-old children. In this study based on the data obtained from eighty-two children with approximately 2411 utterances and thirty adults with approximately 900 utterances, we show and discuss that the overall rhythmic organization of Bengali is acquired very early by Bengali-speaking children, likely within 12-15 months of birth. Vowel lengthening patterns may also be acquired at an equivalent time or slightly later, but full acquisition of lengthening as the correlate of metrical structure appears to take till six years for complete acquisition. The systematic repetition of stress at approximately equal intervals and vowel lengthening yields the perceived prosody and meter in Bangla.

In India’s diverse classrooms, students frequently employ translanguaging, code-switching, and code-mixing to comprehend complex educational content. This bilingual and biliterate practices foster active and adaptable brains, yet remain largely unacknowledged or misunderstood by institutions. This study investigates the impact of these implicit strategies on comprehension and cognitive development among 15 undergraduate students from The English and Foreign Languages University (EFLU), Hyderabad, chosen for their diverse linguistic backgrounds. Through detailed classroom observations and semi-structured interviews, the research revealed that despite occasional vocabulary limitations, these strategies were essential for understanding difficult concepts within a multilingual learning environment. The findings suggest that implicit multilingual practices, while often stigmatized by institutions favouring standard language use, significantly contribute to enhanced cognitive flexibility and critical thinking skills.

Early language learning is foundational for typical cognitive and social development, and is an important predictor of outcomes like occupational SES and educational attainment, and a critical tool for diagnosing developmental disorders. Hence, it is important to understand how language development progresses. The McArthur-Bates Communicative Development Inventory (CDI) is a parent-report vocabulary checklist and gold standard tool for measuring early language learning. It is used in English-speaking contexts, but it has not been adapted to one of the major languages in the world, Hindi. The project aims to develop and document early language learning in Hindi using the CDI with 200 Hindi-speaking families in India, and aims to assess the reliability and validity of the newly developed Hindi CDI.

Word recognition in bilinguals is facilitated by the distinctiveness of L2 lexical items and not by its semantic relatedness. The retrospective semantic matching model(RSMM) and the inhibitory control model(ICM) hypothesizes that semantic relatedness and L2 disadvantage facilitates processing. While, episodic distinctiveness theory(EDT), negative priming(NP) and the proactive control strategy(PCS) predicts that processing is inhibited by semantic relatedness and L2 advantage. The experiment(n=50), with dual lexical decision tasks, had a 2 x 4 factorial design- (i.)SemanticRelatedness (ii.)Trial Type. The fastest RT was observed in Semantically unrelated, L2 repetition trials. The principal outcome of this study is that semantic relatedness caused inhibition in word recognition whereas the L2 advantages caused a facilitation.The findings were reinforced by episodic distinctiveness theory, negative priming and proactive control strategy. How semantic effects interacted with L2 effects in the bilingual brain was highlighted.

Differential object marking is gradient, not binary: ERP evidence from Malayalam

To this end, we conducted an ERP study on the processing of DOM in Malayalam, an agglutinative South Dravidian language. Malayalam is considered a true, one-directional DOM language, where animacy plays a central role in determining object case marking (Asher, 2013). However, there are contexts in which animacy interacts with verb type to condition the case marking of direct objects. Taking all these instances into account, we employed a 2x2x2 design, manipulating verb type (experiencer vs. action verb), case marking of the object noun at NP2 (accusative marked vs. unmarked), and animacy of NP2 (animate vs. inanimate). EEG from 36 first-language speakers of Malayalam (mean age: 27.08; 17 female; 19 male) was recorded when they read the critical sentences (Table1) and performed acceptability judgement and probe detection tasks. Results showed that behavioural acceptability ratings for violation conditions involving animate object arguments - regardless of verb type - were significantly lower (mean < 6.79%) than those for their correct counterparts (mean > 93.35%). Similarly, For sentences with inanimate arguments, violation conditions received lower acceptability ratings than their corresponding control conditions (ACT: 56.4 vs. 80.3; EXP: 50.2 vs. 77.9). ERP analysis at the verb revealed that all the violation conditions involving unmarked object arguments evoked a negativity effect in the 400-600 ms time window compared to their control counterparts, regardless of the animacy of the object and the verb type. Additionally a positivity effect ensued in the 700-900 ms time window for experiencer verbs when the preceding object argument was unmarked and inanimate. The N400 effect found in our study can be interpreted as the reflection of the violation of interpretively relevant case marking (Frisch & Schlesewsky, 2001; Choudhary et al., 2009). The P600 effect found for the experiencer verb construction with nominative inanimate object argument may reflect a blocked expectation for a more transitive action verb. This blocking occurs when argument distinctness is maximally fulfilled but a less transitive experiencer verb is used instead (Dröge et al., 2014; Gattei et al., 2015). Interestingly, for action verbs, the parser did not distinguish between unmarked and accusative inanimate object arguments. In contrast, for experiencer verbs, the neurophysiological data showed a clear distinction between unmarked and accusative objects regardless of animacy. This indicates that experiencer verbs impose stricter case-marking requirements, regardless of the animacy of the object argument. Taken together, our results indicate that the parser processes differential object marking differently depending on whether the verb is an experiencer or an action verb.

Spatial attention is likely involved in many aspects of reading. More specifically, how it influences single word processing is yet not clear despite many studies investigating this issue. Several different manipulations of attention have been used to investigate whether word processing is automatic or requires attention. Most of these studies however, are done with alphabetic languages written in Roman orthography. Recently some researches have emphasized that findings from other orthographic languages should also be incorporated to have a truly clear idea of how attention interacts with reading. Therefore, in the current study we investigated the relationship between spatial attention and word-reading using Hindi language which is written in Devanagari orthography. We used a spatial cuing paradigm and hypothesized faster reaction time and higher accuracy for condition when attention is engaged (valid condition) compared to when it is not (invalid condition). Twenty-two participants participated in the experiment. As expected, the results showed faster reaction time and higher accuracy for targets presented during valid condition compared to invalid condition. However, there was a lexicality effect was also found. For non-words the trend of findings were opposite to what has been observed previously in Roman orthographic languages. Overall, we conclude that orthography influences the way attention interacts with word processing. We are conducting further experiments to bring more clarity to these issues.

This experimental study explores the effectiveness of rhythm metrics in capturing prosodic variation in second language (L2) speech, produced by Bangla speakers from diverse dialectal backgrounds. Speech rhythm has been studied as a diagnostic feature of language typology, and recent advances in prosodic modeling suggest that when normalized for speech rate, rhythm metrics can effectively capture L1 prosodic influence in L2 production. The study synthesizes acoustic analyses and machine learning approaches for dialect classification and rhythm modeling, focusing on features like ΔV, rPVI-V, and nPVI-V in identifying L1 dialectal influence in spontaneous L2 English speech. Vowel-based metrics show consistent predictive power for dialectal classification, particularly among speakers from Madaripur and Dhaka. The study contributes to both theoretical and applied phonetics by demonstrating the utility of rhythm metrics in revealing dialectal transfer in L2 speech.

This study examines phonological awareness and reading development in Assamese middle school learners (Grades VI–VIII), an understudied South Asian language. Ninety-six children (56 females, 40 males; mean age = 12.59) participated. Reading skills were assessed using an Assamese-adapted version of the Test of Word Reading Efficiency (Torgesen et al., 1999), with 108 real words and 66 nonwords selected for familiarity and syllable structure. Phonological awareness tasks were based on PAT-2 (Robertson & Salter, 2007), including syllable- and phoneme-level deletion, segmentation, blending, and rhyme discrimination. ANOVA and regression analyses revealed grade-level progression in real word reading but not nonword reading. Children consistently outperformed on syllable-level tasks compared to phoneme-level tasks. Phoneme segmentation showed improvement between Grades VI and VII, but no difference between Grades VII and VIII, suggesting a developmental plateau. Nonword decoding was influenced by both syllable and phoneme awareness, indicating dual phonological representations, while real word reading relied more on whole-word processing.