This talk examines how corpus linguistics and artificial intelligence treasure the potential to reshape contemporary language learning ecologies. It argues that the rapid normalisation of generative AI has intensified the need for pedagogical models that combine low-friction access to language support with transparent methods grounded in attested usage. Drawing on ecological perspectives and recent empirical research, the talk shows how AI-driven environments expand opportunities for language learning while creating risks related to opacity and over-reliance. Corpus linguistics, data-driven learning and corpus literacy offer a complementary foundation by providing traceable evidence, reproducible analyses, and practices that foster learners’ critical judgement. Two convergence scenarios are proposed: AI as an extension of DDL, and corpus literacy as the operational core of critical AI literacy. Together, these scenarios illustrate how open-box pedagogies can reconcile responsiveness and accountability, ensuring that AI-mediated learning remains anchored in transparent processes and empirically grounded language knowledge.

This study conducted an approximate replication of Teravainen-Goff (2023) to validate the Intensity and Perceived Quality of Engagement Scale for university students in the Japanese EFL context. Teravainen-Goff (2023) developed this scale based on an action-oriented definition of engagement and proposed a novel approach to measuring engagement among secondary school language learners in the UK. The study identified an 18-item, five-factor structure from a pool of 36 items through exploratory factor analysis (EFA). In this replication, we examined the validity and reliability of Teravainen-Goff’s scale in a different context, focusing on the replicability of the EFA results. We undertook this replication because engagement is context-dependent and EFA results can vary across samples. We compared the factorial structure with that of the initial study while modifying the target language and participant demographic. Results revealed a 22-item, six-factor structure with good fit. Although the same underlying factors emerged, several notable differences were observed. This approximate replication provided stronger evidence for the psychometric properties of the scale in a new context. Transparent documentation of modifications to the initial study and systematic comparison offered a promising approach to building robust evidence for engagement research and improving the rigour of questionnaire-based research overall.

Previous research has highlighted that supplementing standard group-level event-related potential analyses with assessments of individual variation can enhance our understanding of language-related brain activity. The present study pursues this approach by examining bilingual speakers’ brain responses to morphologically complex word forms in both their native (German, L1) and their second language (English, L2). We tested 108 bilingual speakers using an ERP violation paradigm examining overapplications of regular verb inflections (‘regularizations’) and of irregular ones (‘irregularizations’). We found a striking L1/L2 contrast within the same bilingual speakers, a left-anterior negativity for regularizations in the L1 and a positivity (P600) for both violation types in the L2. Consistent with previous research, individuals’ brain responses were found to vary along negativity-/positivity-dominant effects. However, the crucial L1/L2 contrast in participants’ brain responses to regularizations was stable across individual differences. We conclude that linguistic constraints, that is, violation type and language status (L1 vs. L2), limit individual variability.

Research on understanding the effects of language experiences upon executive control processes has turned away from static measures of language use to using more continuous measures such as proficiency, language switching and exposure. The present work utilizes language entropy, a measure that indexes the social and linguistic diversity of daily-life contexts (e.g., a classroom, cafeteria, home) of language use, to delineate the mechanisms through which contextual and social effects influence executive control. Results from existing studies utilizing entropy primarily examine bilingual contexts; however, this study focuses on multilingual university students in Ahmedabad, India. Participants (N = 56) provided entropy data from the Language History and Background Questionnaire and executive control measures from the AX-CP Task for proactive control and the n-back Task for working memory. Entropy measures proved very predictive for participants’ current language use patterns, but did not significantly predict any aspect of AX-CPT or n-back Task performance. Implications for context-specific stimulus categorization and the adaptive control hypothesis are discussed.

In three empirical studies, we compare one syntactic and one semantic approach to agreement preferences in so-called pancake constructions (pcs) in Swedish, as in Senap är starkt ‘Mustard is strong’. pcs are either substance-denoting, naming an inherent property of the subject, or situation-denoting, naming a property of the subject that is linked to some event. These two types were found to differ in predicative agreement patterns when their subjects were modified (e.g. Skånsk senap är … ‘Scanian mustard is’). The studies also indicate that the presence of a modal verb can affect agreement patterns differently in the two types: substance-denoting pcs were affected by modification and modality to a much larger extent than situation-denoting ones. We conclude that the two approaches can explain some patterns, but leave others unexplained, and the results lend partial support to analyses that make a syntactic difference between the two types of pcs.

Recent research has shown that adult learners can rapidly acquire novel words of a foreign language by tracking cross-situational statistics, but learning is substantially reduced when the target words are phonologically similar and contain non-native contrasts. We expand on this research by investigating whether perceptual discrimination training on non-native target contrasts facilitates cross-situational learning of new words (CSWL). Our design combines perceptual training and CSWL to test the transfer of perceptual gains to lexical learning—an approach that integrates methods from L2 speech and statistical learning. In two studies, we tested English-native and Portuguese-native speakers’ learning of 24 Portuguese pseudowords via a CSWL task. In Study 1, we examined baseline learning in both language groups without prior training. In Study 2, English-native speakers were assigned to one of three conditions: phonetic training with an AX discrimination task, phonetic training with an oddity discrimination task, or no phonetic training prior to the CSWL task. Results confirmed that adults can learn non-native words from cross-situational statistics, and that phonological overlap between words decreases learning. Perceptual training improved the discrimination of target contrasts, but this did not transfer to statistical learning of words that contain these contrasts. These findings suggest that phonetic training alone may not be sufficient for vocabulary acquisition, suggesting the need for instructional approaches that integrate phonetic training with more explicit teaching methods or meaning-based practice.

Scalar words such as warm may give rise to inferences such as warm but not hot. Under standard accounts, such scalar implicatures are derived by negating stronger alternatives. In processing, weaker scale-mates (warm) prime stronger ones (hot), suggesting that the latter are used in implicature processing (De Carvalho et al., 2016. Frontiers in Psychology, 7, 1500; Ronai & Xiang, 2023. Experiments in Linguistic Meaning, 2, 229–240). We test whether the activation of alternatives holds when no implicature is expected to arise and examine what kinds of alternatives form the basis from which scalar implicature derivation proceeds. We employ two manipulations: negation and antonymy. In line with an account derived from the theoretical treatments of implicature (e.g., Horn, 1972. On the semantic properties of logical operators in English), negating scale-mates cancelled the activation of strong terms (hot). Contrary to these accounts, however, antonyms activated the same targets. In a joint analysis, we found that negation interacted with both scale-mate primes and antonym primes. We explain these findings within the Alternative Activation Account (Gotzner, 2017. Alternative sets in language processing: How focus alternatives are represented in the mind), which assumes an initial activation of a broad cohort of associated expressions and their subsequent grammatical and contextual narrowing.

This study examines the utterance-initial prosodic marking of sarcasm in English and its perception in listeners who did and listeners who did not self-identify as being on the autism spectrum. We ask (i) whether speakers use prosody to mark sarcasm in the early, ‘pre-target’ portion of an utterance (that is, in the portion before a ‘target’ word most closely associated with the sarcastic intent occurs), (ii) whether individuals vary in how they mark sarcasm, (iii) whether listeners reliably recognize sarcasm from pre-target prosody alone, and (iv) whether recognition accuracy varies by speaker or self-identified autistic traits. Eight American English speakers were recorded producing utterances presented in contexts conducive to either sarcasm or sincerity. Pre-target parts were presented in a two-alternative forced-choice experiment to individuals who either did (n=51) or did not (n=44) self-identify as being on the autism spectrum, and were examined for syllable duration and f0-related properties (maximum, minimum, range, and wiggliness). Results show that speakers distinguish sarcasm and sincerity in the pre-target region with duration being the most salient marker. Most listeners recognize sarcasm from pre-target fragments, but there is variation in how well each speaker is perceived. Whether the listener self-identified as being on the autism spectrum or not does not predict sarcasm and sincerity recognition accuracy. The results provide evidence that utterance-initial prosody contributes to sarcasm recognition, with the proviso that speaker and listener variation be taken into account.

Task-based language teaching is believed to facilitate language learning opportunities that arise when performing tasks. Although the synergies between task and learner variables in this process rose to prominence recently, little has been undertaken to explore the individual difference-task interaction in textual meaning-making activities. This study thus explored how second language (L2) writing performance under different task complexity conditions was impacted by L2 writing willingness to communicate (WTC) and L2 writing proficiency. Participants with upper-intermediate English proficiency were recruited following a within-between-participant factorial research design. The results confirmed that WTC significantly influenced syntactic complexity, accuracy, and fluency in L2 writing, suggesting that conative individual differences (IDs) might play a more prominent role in L2 writing than cognitive IDs. Among the sub-components of WTC, motivational predispositions performed better than emotional, cognitive, and writing-specific features in affecting L2 writing performance. Additionally, WTC played a more pronounced role in the complex task, supporting the claim of Robinson’s cognition hypothesis that ID effects are more evident in complex tasks than simple tasks. However, no interaction between L2 writing WTC and proficiency was found. Theoretical and pedagogical implications were offered on considering both L2 WTC and task complexity in task-based writing instruction.

The current study presents an HPSG analysis for deliminative verbal reduplication in Mandarin Chinese. We provide a detailed description of the phenomenon. After discussing reduplication’s interaction with verb classes and aspect markers, we argue that it is better analyzed as a morphological rather than a syntactic process. We put forward a lexical rule for verbal reduplication in Mandarin Chinese, and the different forms of reduplication are captured in an inheritance hierarchy. The interaction between verbal reduplication and aspect marking is handled by multiple inheritance. This analysis covers all forms of deliminative verbal reduplication in Mandarin Chinese and has none of the shortcomings of previous analyses.