Cognitive Science[JOURNAL]
Sun
200 papers
Can Large Language Models Simulate Spoken Human Conversations?
Mayor E, Bietti LM, Bangerter A
Cogn Sci · 2025 Sep · PMID 40889249 · Full text
Large language models (LLMs) can emulate many aspects of human cognition and have been heralded as a potential paradigm shift. They are proficient in chat-based conversation, but little is known about their ability to si...
Large language models (LLMs) can emulate many aspects of human cognition and have been heralded as a potential paradigm shift. They are proficient in chat-based conversation, but little is known about their ability to simulate spoken conversation. We investigated whether LLMs can simulate spoken human conversation. In Study 1, we compared transcripts of human telephone conversations from the Switchboard (SB) corpus to six corpora of transcripts generated by two powerful LLMs, GPT-4 and Claude Sonnet 3.5, and two open-source LLMs, Vicuna and Wayfarer, using different prompts designed to mimic SB participants' instructions. We compared LLM and SB conversations in terms of alignment (conceptual, syntactic, and lexical), coordination markers, and coordination of openings and closings. We also documented qualitative features by which LLM conversations differ from SB conversations. In Study 2, we assessed whether humans can distinguish transcripts produced by LLMs from those of SB conversations. LLM conversations exhibited exaggerated alignment (and an increase in alignment as conversation unfolded) relative to human conversations, different and often inappropriate use of coordination markers, and were dissimilar to human conversations in openings and closings. LLM conversations did not consistently pass for SB conversations. Spoken conversations generated by LLMs are both qualitatively and quantitatively different from those of humans. This issue may evolve with better LLMs and more training on spoken conversation, but may also result from key differences between spoken conversation and chat.
A Diachronic Investigation of the Change in Form and Formational-Semantic Systematicity of the Chinese Sign Language Lexicon.
Zou Y, Lin H
Cogn Sci · 2025 Sep · PMID 40889233 · Publisher ↗
It has been argued in previous research that several competing pressures guide the directions of language evolution (economy vs. redundancy; arbitrariness vs. systematicity). For sign languages, however, the effects of c...
It has been argued in previous research that several competing pressures guide the directions of language evolution (economy vs. redundancy; arbitrariness vs. systematicity). For sign languages, however, the effects of competing pressures on their change of lexical systems remain largely unclear. In the present study, we focus on the diachronic change in form and formational-semantic systematicity of the Chinese Sign Language (CSL) lexicon. Drawing on two CSL lexicons (one from the 1960 dictionary and the other from the 2019 dictionary), we found that in the dimension of form, the CSL lexical system shows a trend toward monosyllabicity and symmetry. In terms of formational-semantic systematicity, we found that there is a significant correlation between form and meaning in both lexicons, but the effect of the arbitrariness constraint gets stronger over time. Our findings regarding the change in form indicate that the competing pressures between economy and redundancy have different effects on different parameters when shaping the lexical system of CSL. As for the correlation between form and meaning, our study provides insight as to how a balance between arbitrariness and systematicity is reached.
Causal Perception(s).
Kominsky JF, Wenig K
Cogn Sci · 2025 Sep · PMID 40884396 · Full text
In addition to detecting "low-level" features like shape, color, and movement, the human visual system perceives certain "higher-level" properties of the environment, like cause-and-effect interactions. The strongest evi...
In addition to detecting "low-level" features like shape, color, and movement, the human visual system perceives certain "higher-level" properties of the environment, like cause-and-effect interactions. The strongest evidence that we have true causal perception and not just inference comes from the phenomenon of retinotopically specific visual adaptation to launching, which shows that launching events have specialized processing at a point in the visual system that still uses the surface of the retina as its frame of reference. Using this paradigm, we show that the visual system adapts to two distinct causal features found in different types of interaction: a broad "launching-like" causality that is found in many billiard-ball-like collision events including "tool-effect" displays, "bursting," and event "state change" events; and an "entraining" causality in events where one object contacts and then moves together with another. Notably, adaptation to entraining is not based on continuous motion alone, as the movement of a single object does not generate the adaptation effect. These results not only demonstrate the existence of multiple causal perceptions, but also begin to characterize the precise features that define these different causal event categories in perceptual processing.
Syntactic Complexity Phenomena Are Better Explained Without Empty Elements Mediating Long-Distance Dependencies.
da Cunha Y, Gibson E
Cogn Sci · 2025 Aug · PMID 40819385 · Full text
We report the results of two acceptability judgment experiments on English materials, which were designed in order to help disentangle predictions of syntactic theories with transformations from nontransformational theor...
We report the results of two acceptability judgment experiments on English materials, which were designed in order to help disentangle predictions of syntactic theories with transformations from nontransformational theories. The materials in these experiments were motivated from examples from Pickering & Barry (1991), who provided intuitive evidence that there is little processing cost for connecting a fronted prepositional phrase to its verb, even if it is the second postverbal argument of a verb in the declarative form. For example, the PP on which connects to the verb put in the sentence This is the saucer on which Mary put the cup into which she poured the milk. If there is a transformation of phrases from declarative structures to interrogative structures (as proposed in Chomsky (1957) and all versions of related theories since), then there is a long-distance connection between the fronted PP and its base position following the NP object, for example, the cup into which she poured the milk, which is not complete until the end of the sentence. In contrast, in a theory without transformations, the PP can be directly associated with its role-assigning verb put when this verb is encountered. If there is cost for processing making dependency connections that is proportional to their distances, then transformational theories predict a large processing cost for this kind of structure, relative to controls. In contrast, nontransformational theories predict no large cost. The results of the two rating experiments consistently supported the predictions of the non-transformational theories relative to those of the transformational theories. We argue that, in line with other current evidence, the nontransformational theories appear to better support the available empirical data.
Neural Mechanisms of Decision Making Under Risk of Punishment: Insights From Rodent Models.
Pyon WS, Bizon JL, Setlow B
Wiley Interdiscip Rev Cogn Sci · 2025 · PMID 40817760 · Publisher ↗
There are few cognitive functions more essential than decision making, as better decisions improve our chances of survival. Cost-benefit decisions as they apply to most scenarios in the developed world can range from rel...
There are few cognitive functions more essential than decision making, as better decisions improve our chances of survival. Cost-benefit decisions as they apply to most scenarios in the developed world can range from relatively mundane to reasonably important; however, particularly risky choices such as speeding on our way to work or consuming suspicious foods can pose a genuine risk of significant harm or illness. How is it that our brains learn and evaluate these risks and rewards to arrive at decisions? Additionally, what drives some of us to continue despite, or avoid because of, potential adverse consequences? This review explores neural mechanisms underlying cost-benefit decision making, focusing on paradigms used in human and particularly rodent studies to model decision making under the risk of explicit punishments, such as pain, discomfort, or loss. The review focuses on several key brain regions (the prefrontal cortex, basolateral amygdala, and striatum), and their roles in the assessment of rewards, punishments (or risk thereof), and motivated behaviors. It also discusses pertinent literature on the role of dopamine arising from the ventral tegmental area, as a neuromodulator critical for learning and reinforcement in the context of risky decision making. This article is categorized under: Neuroscience > Behavior Economics > Individual Decision-Making Psychology > Reasoning and Decision Making.
Contrastive Verbal Guidance: A Beneficial Context for Attention To Events and Their Memory?
Singh A, Rohlfing KJ
Cogn Sci · 2025 Aug · PMID 40810767 · Full text
Research suggests that presenting an action via multimodal stimulation (verbal and visual) enhances its perception. To highlight this, in most studies, assertive instructions are generally presented before the occurrence...
Research suggests that presenting an action via multimodal stimulation (verbal and visual) enhances its perception. To highlight this, in most studies, assertive instructions are generally presented before the occurrence of the visual subevent(s). However, verbal instructions need not always be assertive; they can also include negation to contrast the present event with a prior one, thereby facilitating processing-a phenomenon known as contextual facilitation. In our study, we investigated whether using negation to guide an action sequence facilitates action perception, particularly when two consecutive subactions contrast with each other. Stimuli from previous studies on action demonstration were used to create (non)contrastive actions, that is, a ball following noncontrastive and identical (Over-Over or Under-Under) versus contrastive and opposite paths (Over-Under or Under-Over) before terminating at a goal location. In Experiment 1, either an assertive or a negative instruction was provided as verbal guidance before onset of each path. Analyzing data from 35 participants, we found that, whereas assertive instructions facilitate overall action recall, negating the later path for contrastive actions is equally facilitative. Given that action goal is the most salient aspect in event memory due to goal-path bias in attention, a second experiment was conducted to test the effect of multimodal synchrony on goal attention and action memory. Experiment 2 revealed that when instructions overlap with actions, they become more tailored-assertive instructions effectively guide noncontrastive actions, while assertive-negative instruction particularly guides contrastive actions. Both studies suggest that increased attention to the goal leads to coarser perception of midevents, with action-instruction synchrony modulating goal bias in real-time event apprehension to serve distinct purposes for action conceptualization. Whereas presenting instructions before subactions attenuates goal attention, overlapping instructions increase goal attention and reveal the selective roles of assertive and negative instructions in guiding contrastive and noncontrastive actions.
Learning Partial Word Meanings From Referentially Ambiguous Naming Events.
Schoener N, Johnson SC, Suanda SH
Cogn Sci · 2025 Aug · PMID 40810271 · Publisher ↗
Both classic thought experiments and recent empirical evidence suggest that children frequently encounter new words whose meanings are underdetermined by the extralinguistic contexts in which they occur. The role that th...
Both classic thought experiments and recent empirical evidence suggest that children frequently encounter new words whose meanings are underdetermined by the extralinguistic contexts in which they occur. The role that these referentially ambiguous events play in children's word learning is central to ongoing debates in the field. Do children learn words from referentially ambiguous events via an incremental learning process? Or, do children learn words primarily from the rare referentially transparent events they experience? Across two experiments with adults as model word learners, the current work asks whether the answer to these questions depends in part on how word learning is assessed. Participants were asked to learn the meanings of novel words solely from their referentially ambiguous contexts. When learning was assessed by asking participants to identify the exact meanings of those novel words, participants struggled mightily. However, when learning was assessed by asking the same participants to identify which of two new contexts the novel word most likely occurred in, even those who failed the exact meaning assessment succeeded. These data suggest that although referentially ambiguous events may fall short in allowing learners to identify a word's exact meaning, they nevertheless lead learners into the right regions of semantic space. These findings are a reminder of the pervasiveness of partial word learning effects in vocabulary acquisition and highlight that the resolution to the debate over the role of referentially ambiguous events in learning may depend on how learning is defined.
Learned Insignificance of Credibility Signs.
Kainz V, Sulik J, Utz S … +1 more , Enßlin T
Cogn Sci · 2025 Aug · PMID 40810257 · Full text
A large part of how people learn about their shared world is via social information. However, in complex modern information ecosystems, it can be challenging to identify deception or filter out misinformation. This chall...
A large part of how people learn about their shared world is via social information. However, in complex modern information ecosystems, it can be challenging to identify deception or filter out misinformation. This challenge is exacerbated by the existence of a dual-learning problem whereby: (1) people draw inferences about the world, given new social information; and simultaneously (2), they draw inferences about how credible various sources of information are, given social cues and previous knowledge. In this context, we investigate how social influence and individual cognitive processing interact to explain how one might lose the ability to reliably assess information. Crucially, we show how this happens even when individuals engage in rational belief updating and have access to objective cues of deception. Using an agent-based model, the Reputation Game Simulation, we show that mere misinformation is not the problem: The dual-learning problem can be solved successfully with limited Bayesian reasoning, even in the presence of deceit. However, when certain agents consistently engage in fully deceptive behavior, intentionally distorting information to serve nonepistemic goals, this can lead nearby agents to unlearn or discount objective cues of credibility. This is an emergent delusion-like state, wherein false beliefs resist correction by true incoming information. Further, we show how such delusion-like states can be rehabilitated when agents who had previously lost the ability to discern cues of credibility are put into new, healthy-though not necessarily honest-environments. Altogether, this suggests that correcting misinformation is not the optimal solution to epistemically toxic environments. Though difficult, socially induced cognitive biases can be repaired in healthy environments, ones where cues of credibility can be relearned in the absence of nonepistemic communication motives.
Teaching Recombinable Motifs Through Simple Examples.
Ham H, Zhao B, Griffiths TL … +1 more , Vélez N
Cogn Sci · 2025 Aug · PMID 40810254 · Full text
A hallmark of effective teaching is that it grants learners not just a collection of facts about the world, but also a toolkit of abstractions that can be applied to solve new problems. How do humans teach abstractions f...
A hallmark of effective teaching is that it grants learners not just a collection of facts about the world, but also a toolkit of abstractions that can be applied to solve new problems. How do humans teach abstractions from examples? Here, we applied Bayesian models of pedagogy to a necklace-building task where teachers create necklaces to teach a learner "motifs" that can be flexibly recombined to create new necklaces. In Experiment 1 (N = 151), we find that human teachers produce necklaces that are simpler (i.e., have lower algorithmic complexity) than would be expected by chance, as indexed by a model that samples uniformly from all necklaces that contain the target motifs. This tendency to select simpler examples is captured by a pedagogical sampling model that tries to maximize the learner's belief in the true motifs by prioritizing examples that have fewer alternative interpretations. In Experiment 2 (N = 295), we find that simplicity is beneficial. Human learners recover the underlying motifs better when teachers produce simpler sequences, as predicted by the pedagogical sampling model. However, humans learn best from human teachers rather than from model-generated examples, which suggests that human teachers have additional expectations about how learners will interpret examples that are not captured by standard models of teaching. Our work provides a principled framework to understand when and why teachers use simple examples to convey abstract knowledge.
Deliberation in Guesstimation.
Salikutluk V, Jäkel F
Cogn Sci · 2025 Aug · PMID 40802877 · Full text
In many real-world settings, people often have to make judgments with incomplete information. Estimating unknown quantities without using precise quantitative modeling and data is called guesstimation, which is often nee...
In many real-world settings, people often have to make judgments with incomplete information. Estimating unknown quantities without using precise quantitative modeling and data is called guesstimation, which is often needed in forecasting settings. Furthermore, research in education found that solving guesstimation problems builds general problem-solving skills. In this paper, we present an empirical investigation on how people solve guesstimation problems. We study their problem-solving behavior with think-aloud methods, and we identify solution strategies that are frequently used. In a two-response paradigm, we first ask for gut-feeling answers to guesstimation questions and then allow deliberation before a second answer is given. Comparing the quality of these two answers reveals that deliberation improves the answer quality significantly. In a second experiment, we additionally elicit participants' confidence about their deliberated answers by asking for an entire distribution instead of just a point estimate. We find that participants are generally overconfident in their answers. We discuss guesstimation tasks as suitable test-beds for studying human deliberative judgments in general and in the more specific context of improving forecasting through appropriate artificial intelligence tools.
Seeking the Category: The Pragmatic Function of Formal Explanations and the Role of Cognitive Reflection.
Aslanov I, Kotov A, Guerra E … +2 more , Fedoriaieva A, Kotova T
Cogn Sci · 2025 Aug · PMID 40802863 · Publisher ↗
Formal explanations are statements that explain properties of an object by referring to its category. This study investigates the role of pragmatics in the evaluation of formal explanations. Across six experiments, we ex...
Formal explanations are statements that explain properties of an object by referring to its category. This study investigates the role of pragmatics in the evaluation of formal explanations. Across six experiments, we examined how a questioner's knowledge of category identity and an explanation's capability to specify a category affect satisfaction with such explanations. Experiments 1a and 1b demonstrate that participants find formal explanations less satisfactory when the questioner is already aware of the category identity. Experiments 2a and 2b show that participants assumed a questioner was unaware of an object's category if they were satisfied with the formal explanation. In Experiment 3, open-ended responses revealed that satisfied questioners were perceived as seeking to learn a category identity, while dissatisfied ones were assumed to have other motives. Finally, Experiment 4 compares tautological formal explanations (where a label points to all categories possessing a particular feature at once) and nontautological ones (where a label points to one of several competing categories), and examines the role of cognitive reflection in their evaluation. It demonstrates that people with high cognitive reflection are more sensitive to pragmatic context and value a formal explanation more if it can identify a specific category. This study shows that formal explanations are satisfactory when they fulfill a specific pragmatic function, namely, helping to define a category when the questioner knows only its feature. It also shows that people prone to automatic intuitive responses are less likely to consider this function and tend to evaluate formal explanations independently of this part of the pragmatic context.
The Inevitable and Unpredictable Role of Large Language Models in Education: A Commentary on Huettig and Christiansen (2024).
Guerra E, Peña M, Araya R
Cogn Sci · 2025 Aug · PMID 40798925 · Publisher ↗
This commentary critically complements a recent proposal that cognitive science can leverage Large Language Models (LLMs) to counter declining literacy. While recognizing the educational potential of LLMs, we highlight a...
This commentary critically complements a recent proposal that cognitive science can leverage Large Language Models (LLMs) to counter declining literacy. While recognizing the educational potential of LLMs, we highlight a significant trade-off: their inherent design reduces users' direct engagement with written text, undermining deeper literacy skills, especially in young learners. Acknowledging this tension is essential for developing pedagogically sound interventions. Cognitive scientists and educators must jointly anticipate which cognitive capacities may weaken, identify critical skills needed in emerging multimodal contexts, and collaboratively devise instructional strategies to preserve the cognitive benefits humans derive specifically from sustained interaction with written language.
Iconic Words Are Associated With Iconic Gestures.
Wilding E, Winter B, Littlemore J … +1 more , Perlman M
Cogn Sci · 2025 Aug · PMID 40762325 · Full text
Iconicity ratings studies have established that there are many English words which native speakers judge as "iconic," that is, as sounding like what they mean. Here, we explore whether these iconic English words are more...
Iconicity ratings studies have established that there are many English words which native speakers judge as "iconic," that is, as sounding like what they mean. Here, we explore whether these iconic English words are more likely to be accompanied by iconic gestures. We report a large-scale quantitative study comparing the gesture rate of words rated as high in iconicity (e.g., swoosh, puffy, crispy) to those rated as low in iconicity (e.g., ordain, rejoin, grateful), balancing for perceptual strength, part-of-speech, and syllable length. Five thousand seven hundred and twenty-five tokens from the TV News Archive were coded for whether speakers produced a gesture with the word, and whether the gesture was iconic. The results show that high iconicity words have a higher overall gesture rate (69%) than low iconicity words (56%): specifically, high iconicity words have a higher iconic gesture rate (24% vs. 11%). This effect is more pronounced among verbs than adjectives, which we hypothesize may be due to the dynamic nature of verbs. We also find that this result persists when controlling for perceptual and action strength ratings, suggesting that word-level iconicity is a more important predictor than sensorimotor strength of whether a speaker will use an iconic gesture. We find that some high iconicity words are more likely to occur with iconic gestures when they come with markers of syntactic isolation, suggesting that morphosyntactic behavior is also relevant to iconic gesture production. Our findings demonstrate that iconicity in spoken communication is inherently multimodal, manifesting in both speech and gesture simultaneously, and that iconicity is often psychologically active when speakers use conventionalized iconic words.
The Lifetime of Sequential Memory Traces in the Absence of Language.
Ordonez Magro L, Pinto Arata L, Fagot J … +2 more , Grainger J, Rey A
Cogn Sci · 2025 Aug · PMID 40762318 · Full text
Statistical learning allows us to implicitly create memory traces of recurring sequential patterns appearing in our environment. Here, we study the dynamics of how these sequential memory traces develop in a species of n...
Statistical learning allows us to implicitly create memory traces of recurring sequential patterns appearing in our environment. Here, we study the dynamics of how these sequential memory traces develop in a species of nonhuman primates (i.e., Guinea baboons, Papio papio) that, unlike humans, cannot use language and verbal recoding strategies to strengthen these memory traces. We test a group of Guinea baboons in a Hebb visuo-motor pointing task in which a target sequence is repeated with random sequences inserted between repetitions. In this study, we systematically manipulate the interval between two repetitions of the target sequence by varying the number of interposed random sequences. We found that baboons can learn repeated visuo-motor sequences, even when the repetitions are separated by six random sequences. Our results also suggest that the learning curve of the target sequence best fits a logarithmic function. The present study, therefore, provides a quantitative assessment of the development of a sequential memory trace as a function of repetition spacing and without the use of verbal recoding strategies.
Seeking Meaning: Incorporating Linguistic Information in Cross-Situational Verb Learning.
Chen CH, Zhang Y, Yu C
Cogn Sci · 2025 Aug · PMID 40762294 · Full text
Learning the meaning of a verb is challenging because learners need to resolve two types of ambiguity: (1) word-referent mapping-finding the correct referent event of a verb, and (2) word-meaning mapping-inferring the co...
Learning the meaning of a verb is challenging because learners need to resolve two types of ambiguity: (1) word-referent mapping-finding the correct referent event of a verb, and (2) word-meaning mapping-inferring the correct meaning of the verb from the referent event (e.g., whether the meaning of an action word is TURNING or TWISTING). The present work examines how adult learners solve this challenge by utilizing both in-the-moment linguistic information within individual learning situations and cross-situational statistical information across multiple learning situations. We investigate how different cues provided in the moment affect information selection and how cross-situational learning as a general computational mechanism allows for information integration over time. Two experiments were designed based on a Human Simulation Paradigm, in which adult learners were presented with a sequence of short videos from parent-toddler toy play and asked to guess a mystery verb the parent produced in each video. In Experiment 1, we compared individual learning situations containing linguistic information to the exact same learning scenes without linguistic information and found that linguistic information helped learners narrow down the meaning of a verb embedded in individual situations, which was consistent with prior research. In Experiment 2, the videos sharing the same target verb were presented in a blocked design to incorporate cross-situational statistics for the same verb. We measured the variability, convergence, and accuracy of participants' guesses. Within-trial linguistic information allowed learners to quickly narrow down their search space and focus on a few relevant aspects in a scene, while cross-situational learning allowed them to fine-tune their learning further across trials. Our findings support a unified account wherein within-trial linguistic information and cross-situational statistical information are integrated for more efficient verb learning.
Speech Disfluencies and Hand Gestures as Metacognitive Cues.
Yılmaz B, Furman R, Göksun T … +1 more , Eskenazi T
Cogn Sci · 2025 Aug · PMID 40762291 · Publisher ↗
How language interacts with metacognitive processes is an understudied area. Earlier research shows that people produce disfluencies (i.e., "uh" s or "um" s) in their speech when they are not sure of their answers, indic...
How language interacts with metacognitive processes is an understudied area. Earlier research shows that people produce disfluencies (i.e., "uh" s or "um" s) in their speech when they are not sure of their answers, indicating metacognitive monitoring. Gestures have monitoring and predictive roles in language, also implicating metacognitive processes. Further, the rate of speech disfluencies and gestures change as a function of the communicational setting. People produce fewer disfluencies and more gestures when they can see the listener than when the listener is not visible. In the current study, 50 participants (32 women, Mage = 21.16, SD = 1.46) were asked 40 general knowledge questions, either with a visible (n = 25) or nonvisible (n = 25) listener. They provided feelings-of-knowing (FOK) judgment immediately after seeing the question and were asked to think aloud while pondering their answers. Then, they provided retrospective confidence judgments (RCJs). Results showed that gestures and speech disfluencies were not related either to the accuracy or the FOK judgments. However, both gestures and speech disfluencies predicted RCJs uniquely and interactively. Speech disfluencies negatively predicted RCJs. In contrast, hand gestures were positively related to RCJs. Importantly, the use of gestures was more strongly related to RCJs when disfluencies were also higher. No effect of communicational setting on the rate of gestures or speech disfluencies was found. These results highlight the importance of multimodal language cues in the elaboration of metacognitive judgments.
Is Mind-Reading Involved in Ownership Judgments?
Blazsek R, Heintz C
Cogn Sci · 2025 Aug · PMID 40762282 · Full text
How do people determine who owns what? While existing research has identified a number of psychological and behavioral sources of ownership judgments, the role of mental state attribution has received less attention. We...
How do people determine who owns what? While existing research has identified a number of psychological and behavioral sources of ownership judgments, the role of mental state attribution has received less attention. We conducted three online experiments (N = 1246) examining if ownership judgments rely on mind-reading: the capacity to infer others' intentions, beliefs, and knowledge states. Using vignettes, we tested if ownership judgments are sensitive to variations in contextual cues (Study 1), beliefs about the permissibility of taking items (Study 2), and knowledge about social norms (Study 3). We also tested if the moral aspects of a scenario affect judgments of rightful ownership transfer. Our findings indicate that ownership judgments indeed vary in response to these factors, and that they do not vary on par with moral judgments. These findings are best explained in terms of mind-reading and support the argument that ownership is fundamentally a social phenomenon: not a relationship between people and resources but rather between people about resources.
Water Is and Is Not HO, Depending on Who You Ask: Conceptualizations of Water Vary Across Chemists and Laypeople.
Mazzuca C, Arcovito M, Falcinelli I … +2 more , Fini C, Borghi AM
Cogn Sci · 2025 Aug · PMID 40762281 · Full text
Conceptual representations can be shaped by multiple factors, including expertise. In this study, we tested whether the concept of water is represented differently across laypeople and chemists, focusing on psychological...
Conceptual representations can be shaped by multiple factors, including expertise. In this study, we tested whether the concept of water is represented differently across laypeople and chemists, focusing on psychological essentialism. Essentialized categories are thought to be determined by internal factors (e.g., chemical composition). Previous research suggests laypeople do not essentialize "water." Here, we sought to verify whether extensive experience with chemicals might lead to more essentialist conceptions. In the first two experiments, participants provided HO estimates, typicality, centrality, and frequency ratings for water examples, which showed that chemists partially incorporate HO in their conceptual representation of "water." Experiment 3 underlined qualitative differences in the semantic organization of "water" across the two groups using similarity ratings. Experiment 4 consolidated these results with a sentence acceptability task, underlying the importance of chemical composition in determining what counts as "water" for chemists. Finally, Experiment 5 showed that laypeople consider both "HO" and "water" as more abstract compared to chemists. Our results provide evidence on the variability of both psychological essentialism and conceptual representation overall, which can vary as a function of expertise.
Computational Sentence-Level Metrics of Reading Speed and Its Ramifications for Sentence Comprehension.
Sun K, Wang R
Cogn Sci · 2025 Jul · PMID 40692511 · Full text
The majority of research in computational psycholinguistics on sentence processing has focused on word-by-word incremental processing within sentences, rather than holistic sentence-level representations. This study intr...
The majority of research in computational psycholinguistics on sentence processing has focused on word-by-word incremental processing within sentences, rather than holistic sentence-level representations. This study introduces two novel computational approaches for quantifying sentence-level processing: sentence surprisal and sentence relevance. Using multilingual large language models (LLMs), we compute sentence surprisal through three methods, chain rule, next sentence prediction, and negative log-likelihood, and apply a "memory-aware" approach to calculate sentence-level semantic relevance based on convolution operations. The sentence-level metrics developed are tested and compared to validate whether they can predict the reading speed of sentences, and, further, we explore how sentence-level metrics take effects on human processing and comprehending sentences as a whole across languages. The results show that sentence-level metrics are highly capable of predicting sentence reading speed. Our results also indicate that these computational sentence-level metrics are exceptionally effective at predicting and explaining the processing difficulties encountered by readers in processing sentences as a whole across a variety of languages. The proposed sentence-level metrics offer significant interpretability and achieve high accuracy in predicting human sentence reading speed, as they capture unique aspects of comprehension difficulty beyond word-level measures. These metrics serve as valuable computational tools for investigating human sentence processing and advancing our understanding of naturalistic reading. Their strong performance and generalization capabilities highlight their potential to drive progress at the intersection of LLMs and cognitive science.
Cognitive Control Skills Are Related to Ambiguity Awareness in French-Learning 5-to-6-Year-Olds: Implications for Reading Development.
Bigot V, Trueswell J, Carvalho A
Cogn Sci · 2025 Jul · PMID 40692485 · Full text
Five-to-six-year-olds' abilities to detect and solve ambiguities in spoken language have been found to be a predictor of their later reading abilities in first-to-third grade. However, the origins of this relationship re...
Five-to-six-year-olds' abilities to detect and solve ambiguities in spoken language have been found to be a predictor of their later reading abilities in first-to-third grade. However, the origins of this relationship remain unclear. Success in ambiguity detection may be reflective of overall language attainment, which varies with socioeconomic status (SES) and is known to predict reading. Yet, it is also possible that children's ability to detect ambiguity is explained by domain-general cognitive control skills, which can also vary with SES and predict literacy attainment. In this cross-sectional study, we examined within the same children the contributions of overall language knowledge, SES, and cognitive control skills to their ability to detect ambiguities in speech. Five-to-six-year-old French-learning preschoolers (n = 38) performed three different tasks: ambiguity detection, a cognitive control (Flanker/No-Go) task, and standard assessments of vocabulary and oral language comprehension in French (BSEDS). Years of maternal education after the end of high school were used as a proxy of family SES. Individual differences in the ability to detect ambiguity were strongly related to children's cognitive control abilities, as indexed by congruency effects in the Flanker task. No relations with SES or language assessment were observed. These results lend support to the idea that children's reading development may hinge upon their ability to deal effectively with temporary lexical, syntactic, and semantic ambiguities that pervade real-time sentence interpretation and that their ability to deal with representational conflict in speech is reflective of their domain-general cognitive control skills.
About
- Frequency
- Sun
- Papers found
- 200
- RSS feed
- Subscribe