Abstract Fabricius argues that the results of a recent study of ours speak for a deflationary competence limitation account of Theory of Mind development. According to this account, children do not operate with proper mental state concepts until age 6–7 or older. However, the original findings are, in fact, incompatible with such competence limitation accounts, but compatible with the standard picture of Theory of Mind development: According to this picture, children acquire proper mental state concepts by age 4, but this competence is sometimes masked by performance factors.
Abstract Accounts of negation processing, including two‐step simulation, pragmatic, and monotonicity‐based accounts, differ in their predictions about whether double negation requires additional cognitive effort relative to affirmation. Using event‐related potentials, we investigated a Mandarin Chinese double‐negation structure (“méiyǒu X bù Y”) that is logically equivalent to its affirmative counterpart (“suǒyǒu X dōu Y”). Forty native speakers read context sentences (double negation vs. affirmation) followed by critical sentences containing semantically congruent or incongruent target words. Crucially, the target word appeared in a separate sentence, downstream from the negation operators; the design, therefore, tests whether a pragmatically licensed double negation establishes a discourse representation functionally equivalent to affirmation at the point of propositional evaluation. At the critical word, incongruent words elicited larger N400/P600 amplitudes, and these congruency effects did not differ between contexts, indicating equivalent representations and no detectable downstream cost. Post‐critical analyses, however, revealed context‐specific differences: the affirmation context showed a prolonged negativity to incongruency—plausibly reflecting sustained engagement with its more salient violation—whereas the double negation context did not. A representational similarity analysis (RSA) further showed greater neural tracking of incongruency magnitude in the double negation context at a later window (726–798 ms). Together, these findings indicate that a pragmatically licensed double negation supports downstream semantic integration as efficiently as affirmation, while later processing remains context‐sensitive. Because this early equivalence is also consistent with a monotonicity‐based account, we frame the early‐equivalence/late‐divergence pattern as a tentative dual‐stage characterization to be tested in future work, rather than a decisive adjudication between theories.
Abstract To investigate potentially shared syntactic mechanisms between language and music, we conducted a novel adaptation of the dual task paradigm without syntactic violations to assess the degree of bidirectional interaction between the two domains. Participants listened to music progressions while reading natural language sentences, and were then asked to provide, in randomized order, music closure judgments and language comprehension answers. We manipulated both language and music stimuli in terms of their syntactic structural complexity and syntactic working memory demands based on nonlocal dependencies between words or chords. We discovered a bidirectional yet unbalanced pattern of cross‐domain interaction whereby language exerted relatively more influences over music than vice versa, apparently resulting mostly from syntactic structural complexity instead of syntactic working memory demands, although we discuss the possibility that alternative heuristic strategies may have been employed during concurrent, resource‐demanding processes. We also found that cross‐domain interactions were often most prominent when syntactically critical words and keys temporally coincided across sequences. These results provide new evidence in support of theoretical models of shared resources between language and music.
The ability to detect statistical regularities in the input is widely assumed to support language acquisition, but its unique contribution to individual differences in language outcomes remains uncertain, especially in neurodevelopmental conditions. This study tested whether statistical learning predicts expressive vocabulary and grammatical skills in children with typical development and in groups with developmental language disorder, autism spectrum disorder or attention deficit hyperactivity disorder. A total of 102 Hungarian-speaking children aged 8 to 15 years completed tasks assessing statistical learning, auditory and visual perceptual speed, short-term and working memory, sentence repetition, and expressive vocabulary. Statistical learning showed modest associations with language measures, but the effects of perceptual speed and working memory were stronger. Verbal working memory consistently emerged as the strongest predictor of vocabulary and sentence repetition across all groups. At the same time, the relatively low reliability of the statistical learning measures may have attenuated the observed associations, potentially leading to an underestimation of the strength of their relationship with language. The pattern of associations did not differ systematically between diagnostic groups, consistent with the idea that variation in language outcomes reflects general cognitive differences that cut across traditional diagnostic categories. This finding is consistent with dimensional accounts of neurodevelopment, which propose that differences between conditions arise through quantitative rather than qualitative variation in core cognitive processes. However, the various dimensions were hard to disentangle at this sample size. By demonstrating this pattern for language abilities, the present study extends the dimensional perspective to the domain of language development and highlights the importance of accounting for shared cognitive mechanisms when studying typical and atypical language variation.
Given various grammatical options, language users prefer the word order choice that reduces the overall length of syntactic dependencies, a principle known as dependency length minimization (DLM). The origins of this preference remain an open question, particularly whether it originates from constraints on efficient information processing. Computational simulations provide a powerful approach to identifying the factors influencing the emergence of linguistic phenomena. However, previous simulations of DLM have not examined realistic interaction contexts and have produced mixed results. The present study investigates the emergence of DLM in artificial languages using a recently proposed language learning and communication framework based on recurrent neural networks (RNNs). In this framework, agents are trained to speak and interpret artificial languages and then use these languages to communicate. Using this framework, we study the impact of several factors related to processing limitations in a communicative setting, such as noise during listening, limited speaker capacity, and incremental sentence processing. Our results reveal a complex interplay among these factors in shaping word order preferences in neural agents. Specifically, in the full meaning space, agents regularize toward a single dominant word order, while in the half meaning space they show a short-before-long preference that only aligns with DLM in verb-initial languages. A consistent DLM preference emerges only when agents are subject to incremental processing pressure. These findings suggest that limitations in human cognitive processing may indeed play a role in shaping DLM. Our findings provide insights into the conditions under which neural models replicate human-like preferences and highlight the challenges of designing emergent communication models that capture human cognitive biases in language processing.
