Canada

Intuitive explanations for natural phenomena are typically our default explanations, even after we have learned more accurate, scientific explanations (Shtulman &amp; Valcarcel, 2012). The current study examined whether priming students with scientific images improves their ability to verify counterintuitive scientific statements, like “bacteria need nutrients” and “bubbles have weight.” Participants (100 college undergraduates) verified scientific statements interspersed with images relevant to the predicates of those statements; the images depicted either schematic diagrams (scientific primes) or everyday scenes (intuitive primes). Scientific primes increased the accuracy of participants’ responses, relative to intuitive primes, but not the speed of those responses, indicating that scientific primes facilitate a preference for scientific ideas over intuitive ones but do not eliminate the initial conflict between them.

CogSci 2022

Priming Counterintuitive Scientific Ideas

Intuitive explanations for natural phenomena are typically our default explanations, even after we have learned more accurate, scientific explanations (Shtulman & Valcarcel, 2012). The current study examined whether priming students with scientific images improves their ability to verify counterintuitive scientific statements, like “bacteria need nutrients” and “bubbles have weight.” Participants (100 college undergraduates) verified scientific statements interspersed with images relevant to the predicates of those statements; the images depicted either schematic diagrams (scientific primes) or everyday scenes (intuitive primes). Scientific primes increased the accuracy of participants’ responses, relative to intuitive primes, but not the speed of those responses, indicating that scientific primes facilitate a preference for scientific ideas over intuitive ones but do not eliminate the initial conflict between them.

poster

Join us for CogSci2022! Our 44th Annual Conference will be centered around the theme of diversity, with an accompanying range of keynotes and invited speakers who themselves highlight the diversity of our field. Our three symposia are Cognition across Diverse Populations, Cognition within Diverse Societies, and Global Challenges for Cognitive Science. The conference is being offered in a fully hybrid environment for onsite and online attendees.
 
 [**The Program at a Glance can be found here**](https://cognitivesciencesociety.org/wp-content/uploads/2022/07/Cogsci-2022-PAAG-Detailed-July-12final-1.pdf)
 [**The Detailed Program can be found here**](https://cognitivesciencesociety.org/wp-content/uploads/2022/08/DeArmond-CogSci2022-program-v113-final.pdf)

This conference requires a registration to attend. Please use the link below to secure your ticket. The website will be available to all registered attendees on July 25th. 

[Register to Attend](https://cognitivesciencesociety.org/registration/)

Registration Is Required

The 44th Annual Meeting of the Cognitive Science Society highlighted research on the theme Cognitive Diversity in addition to the full breadth and diversity of research topics offered by the society’s membership. 





This study investigates how neural networks reveal developmental trajectories of child language, focusing on the Agent-First strategy in comprehension of an active transitive construction in Korean. We develop three models (LSTM; BERT; GPT-2) and measure their classification performance on the test stimuli used in Shin (2021) involving scrambling and omission of constructional components at varying degrees. Results show that, despite some compatibility of these models’ performance with the children’s response patterns, their performance does not fully approximate the children’s utilisation of this strategy, demonstrating by-model and by-condition asymmetries. This study’s findings suggest that neural networks can utilise information about formal co-occurrences to access the intended message to a certain degree, but the outcome of this process may be substantially different from how a child (as a developing processor) engages in comprehension. This implies some limits of neural networks on revealing the developmental trajectories of child language.

Limits on Neural Networks: Agent-First Strategy in Child Comprehension

The contingency symmetry inference, the inference to generalize a learned contingency to a reverse direction, is known to be extremely difficult for non-human animal species (Lionello-DeNolf, 2009). In contrast, humans are known to have the “affirming the consequent fallacy”, which reverses the antecedent and the consequence (if P then Q: Q therefore P). The contingency symmetry bias has been long discussed in relation to the ontogenesis of language learning, as word learning requires understanding of bidirectional relationship between symbols and objects. But how this bias emerges has not been known. This research tested whether 8-month-old human infants have this bias on a matching-to-sample task. The results demonstrated the possession of this bias in human infants before they start active word learning. This bias is likely a uniquely human cognitive bias, which may explain why only humans have language.

The contingency symmetry bias as a foundation of word learning: Evidence from 8-mont-olds in a matching-to-sample task

This study examines (1) whether infant-directed speech (IDS) facilitates children’s word learning compared to adult-directed speech (ADS); and (2) the link between the prosody of IDS in word-learning contexts and children’s word learning from ADS and IDS. Twenty-four Dutch mother-child dyads participated when children were 18 and 24 months old. We collect mothers’ ADS and IDS at both ages and test children’s word learning from ADS and IDS at 24 months using an Intermodal Preferential Looking Paradigm (IPLP). We find that Dutch 24-month-old children could reliably learn novel words from both ADS and IDS, and IDS had a facilitative effect. Also, children’s word learning from IDS (but not ADS) is predicted by IDS pitch range when mothers introduce unfamiliar words to children at 18 months. Our findings contribute to an understanding of the role of IDS prosody in language development, highlighting both individual differences and contextual differences in IDS prosody.

Prosodic input and children’s word learning in infant- and adult-directed speech

People understand abstract ideas (e.g., positive/negative valence) through concrete concepts (e.g., up/down; Lakoff & Johnson, 2013). Empirical research has shown that upward/downward motor actions stimulate positive/negative feelings and memories (Casasanto & Dijkstra, 2010), and congruent motor actions facilitate word learning (Casasanto & de Bruin, 2019). Although prior studies reveal a close link between language and perceptual experiences, no study has tested whether the visual perception of upward/downward movements enhances the learning of words whose meaning involves either higher/lower spatial position (e.g., cloud, road), positive/negative emotional valence (e.g., joy, grief), or higher/lower social status (e.g., doctor, unemployed). The effects of directional congruency in word learning are discussed based on the results of an experiment in which Japanese speakers learned 54 English-based pseudowords presented with automatic visual movements that were congruent (e.g., upward-positive) or incongruent (e.g., downward-positive) with the pseudowords’ assigned meaning, or controls (e.g., rightward/leftward-positive).

Visual perception of vertical movements in word learning

Picture books are a popular medium through which to promote language acquisition in young children. However, not much is known about how the pictorial context in which words are introduced in such books impacts word learning in toddlers, or how joint book reading further mediates this relationship. The present study introduced words to 19-23-month-old toddlers through books in either contextually rich, semantically relevant illustrations, or on a white background in isolation. Children and their parents participated in three lab visits during which a range of language and environmental measures were taken. Parents read our intervention materials at home between the first and second visits. We found that the pictorial context in which vocabulary words are presented was significantly related to language measures throughout our study. Further, this context also influences parents’ reading techniques, with longer interactions and more target words produced when reading contextually illustrated books. Our minimal book intervention shows promise in promoting vocabulary development in typically talking toddlers.

Learning from Word Books: Does the Type of Illustration Matter?

To explain the importance of interaction for a truthful performance in acting, the present study captures the characteristics of interaction and attempts to probe the underlying intrapersonal changes through interaction during an acting course which emphasizes paying attention to a partner. Novice participants tend to change their way of communication as the course progresses, the pattern of which further differs from that of professional actors. While actors devote themselves more to the connection with their partner and demonstrate more balanced communication, novices rely on general inference to speculate about others’ affective states. This study offers a new perspective to elucidate the construction of interaction in acting, and emphasizes the significance of involvement in interaction when applying acting approaches to general training with the aim of improving social understanding.

Interaction in Acting Training and its Manifestations in Novices and Actors

Word learning is characterized by a bias for mapping meanings at the “basic”-level such as apple, as opposed to a subordinate-level like red apple (Markman, 1990). The fact that learners nevertheless acquire subordinate nouns has been attributed to properties of the referential world that co-occur with the word (e.g., Xu and Tanenbaum, 2007b; Spencer et al., 2011). However, learners may also make inferences about the informativity of labels as intentional linguistic acts. We investigated whether learners exploit information about semantic contrast to generalize word meanings beyond the basic level. Experiment 1 found that the introduction of a labelled alternative at the subordinate level (green apple) eliminated the basic-level bias. Experiment 2 found that the presence of the alternative exemplar without a label merely suppressed the bias. We propose that the acquisition of subordinate-level meanings is facilitated by expectations of informativity which allow learners to enter the relevant alternatives into consideration.

The acquisition of subordinate nouns as pragmatic inference: Semantic alternatives modulate subordinate meanings

We used agent-based modelling to study the emergence of linguistic conventions for basic word order (the order of subject, object and verb) in different populations. As a starting point, we take word order variation based on semantic properties, as observed in improvised gesture experiments. In our first simulation we explore the relative contributions of two pressures, one for semantically conditioned variation, and the other structural priming (which takes place when two individuals engage in communication), and show that a relatively increasing influence of structural priming best explains an increase in word order regularity. Next we implement a larger simulation, investigating how properties of the population affect regularization of word order. Our models compare population sizes with different population densities, and show that the speed of regularization in languages is heavily influenced by population density, and population size has little effect.

Modelling the Emergence of Linguistic Conventions for Word Order: The Roles of Semantics, Structural Priming, and Population Structure

A quantum walk model is developed for emotion transmission in serial reproduction of narratives. The readers' emotions are represented by density operators, and the influences of the narratives on the readers' emotions are modeled by applying the controlled unitary operators to the density operators. The performance of the quantum model is evaluated on a large corpus of narratives, compared to that of the Bayesian Markov chain model. The quantum model not only outperforms the Bayesian model for all five emotion transmissions presented in the corpus but can also account for order effects in serial reproductions. These results suggest a promising first step towards extending quantum-like models to explain group-level cognition.

Premium content

Priming Counterintuitive Scientific Ideas

Next from CogSci 2022

Limits on Neural Networks: Agent-First Strategy in Child Comprehension

Stay up to date with the latest Underline news!

PRESENTATIONS

CONFERENCES

COMPANY

RESOURCES