United States

Recent years have seen a surge of interest in probabilistic measures of argument strength. Such measures have been used to elucidate longstanding questions of both theoretical and practical interest around fallacies of argumentation, scientific argument, legal argument, or argument and evidence aggregation. The measures typically used have drawn, in one form or other, on conditional probability as a central component for measuring argument quality, strength or confirmation. Work in this area has also consistently highlighted the value of Bayesian Belief Networks for computing argument strength in multi-argument (variable) contexts. However, to date, there has been no consideration of whether the key structural property underlying such networks – the notion of conditional independence- has a bearing on argument quality. In this paper, we consider the potential impact of causal structure on the utility of arguments.

CogSci 2023

Causal Structure and Argumentative Value

### Welcome to The 63rd Annual Meeting of the Association for Computational Linguistics (ACL 2025)

Message from the General Chair: 
*It is my great pleasure and honor to welcome you to the 63rd Annual Meeting of the Association for Computational Linguistics (ACL 2025), held in beautiful Vienna, Austria, from July 27 to August 1, 2025. ACL2025continues our field’s tradition of excellence in scholarship, innovation, and inclusivity, and I am deeply grateful to the many volunteers who have worked tirelessly to bring this event to life.* 
[Read more](https://drive.google.com/file/d/1GI_hvOpjswAuYdUTromfeDiPpCcqidwg/view?usp=sharing)

To access this event page, you need to log in with the **email address you registered with**. Access credentials will be sent to your email from Underline - subject line "Welcome to ACL 2025". Please be sure to check your spam email folder if you do not see an email confirmation right away.

Please log in

To access this event page, you need to log in with the **email address you registered with**. 

ACL 2025

Welcome to The 63rd Annual Meeting of the Association for Computational Linguistics (ACL 2025)

poster

The 45th Annual Meeting of the Cognitive Science Society will take place in Sydney, Australia as a modified hybrid conference. Presentations from the main-session room will be live-streamed every day for all events. This will include all invited talks and award sessions. Recordings of all CogSci 2023 sessions will be made available for on-demand viewing after the conference. 
The in-person component will be held at the International Convention Centre Sydney at Darling Harbour. The Conference will take place Wednesday July 26th – Saturday July 29th, 2023 (Wednesday, July 26th is the Workshops Day).

Join us!

 [**The Program at a Glance can be found here**](https://cognitivesciencesociety.org/wp-content/uploads/2023/07/COgSci-PAAG-V2.pdf)
 [**The Detailed Program can be found here**](https://online.fliphtml5.com/ebtyf/mswz/)
 

#### **CogSci 2023 Program Book**
<iframe style="width:700px;height:400px" src="https://online.fliphtml5.com/ebtyf/tzgk/" seamless="seamless" scrolling="no" frameborder="0" allowtransparency="true" allowfullscreen="true" ></iframe>

The 45th Annual Meeting of the Cognitive Science Society presents the latest research across cognitive science and highlights the theme of Cognition in Context.

The investigation of reasoning involves the study of what is rational (e.g., adequate, optimal) as well as the study of human reasoning (e.g., common intuitions on plausibility). While the first aspect is typically covered by logic or formal epistemology, the second is a matter of empirical psychological research. However, many approaches relate both aspects. This paper discusses these hybrid approaches, their motivation, general critiques that have been raised against them as well as the kind of philosophical commitment behind choosing such kind of framework.


Hybrid Frameworks of Reasoning: Normative-Descriptive Interplay

Since Gottlob Frege’s disambiguation of laws of thought, logicians and psychologists alike have adhered to a strict division of labour, which has created a gap between reasoning as a psychological phenomenon and logic as a formal discipline. However, logic is a benchmark of reasoning according to the standard picture of rationality, and it is hard to accept that logic has no positive special role to play in reasoning. Intuitively, logic and reasoning are connected; however, attempts at bridging the gap are often vulnerable to accusations of psychologism. 
Jean Piaget developed a compass to navigate the hazards of psychologism, and he made logic the mirror of thought rather than vice versa. Metaphors are helpful but ultimately poor substitutes for explanations; in this paper, I therefore intend to elaborate on Piaget’s intimations on the relationship between logic and reasoning to negotiate the pitfalls of psychologism and bridge the gap in mind.

Logic and Psychology—Minding the Gap with Jean Piaget

It is common for cases where people “fail to deny” unflattering claims to be taken to imply those claims are true. Yet, the ‘argument from ignorance’ – arguing in favour of something due to a lack of evidence against it – is often deemed a fallacy. Why is abduction from missing evidence permissible in some cases but not others? We present a framework of factors which disambiguate these cases, using a Bayesian Network model. We suggest that a source’s silence often reflects a latent conflict between their intrinsic motives about what they would like their audience to believe and their extrinsic motives about complying with the institutional constraints on their speech, such as accuracy. Thus, silence often implies that the source does not believe what they would ideally like to say is true, which licenses a probabilistic inference about what is true. We present data from two studies suggesting people reason about silence in this way.

The Epistemic Weight of Silence

This study identified a novel and robust reasoning error. Lay reasoners significantly deviate from the prediction of Bayesian inference by consistently underestimating the added probative value of corroborating testimonial reports. Most surprisingly, however, is that in certain contexts the sum of corroborating evidence is considered to be significantly less valuable than a single report. There is a selective devaluing of corroborating testimony when a highly reliable report is corroborated by less reliable, but credible, reports. This intuitive error is not explained by an inaccurate understanding of individual cues of reliability and number of reports, but specifically when it is required to integrate these both cues. Findings indicate the operation of alternative reasoning strategies, resulting in errors at individual and group level.

Evaluating testimony from multiple witnesses: consistent undervaluing and selective devaluing of corroborating reports

Theories can be designed to predict novel evidence or to accommodate known evidence. Despite the lively debate in philosophy of science whether prediction may hold a superior value over accommodation, people’s intuitions about this issue have not been empirically examined. Within a medical scenario, we assess individuals’ sensitivity to this dilemma. Overall, we find tentative evidence that people favour the predictive account and regard the predicting theorist (i.e., doctor) as more reliable in contrast to the accommodating counterpart. Strikingly, discrepant preference patterns emerged out of their verbal reasoning data echoing the distinct philosophical stances surprisingly well. Possible reasons why people’s reasoning systematically diverges despite the general preference for prediction are discussed.


Do people prefer prediction over accommodation? An empirical study

Intuitive conceptual systems impact how university students learn formal biological science. One such conceptual system--anthropocentric thinking--involves anchoring our understanding of the biological world in human terms. Prior research indicates that anthropocentric thinking, like other intuitive conceptual systems, can engender misconceptions, but also varies with richness of informal biological experience. To examine the impact of informal experience on anthropocentric thinking and formal biology learning, we assessed students from a rural institution (n=212) and an urban institution (n=188) on their use of anthropocentric reasoning and their agreement with biological misconceptions using both qualitative coding of free responses and quantitative analysis of close-ended measures. Our findings indicate linkages between anthropocentric thinking and misconceptions, as well as differences between urban and rural students. This work demonstrates context-based variability in anthropocentric thinking, highlights potential mechanisms driving this variability, and underlines the implications of intuitive thinking for STEM learning.

Relations Between Anthropocentric Thinking and Biological Misconceptions in Students attending Urban and Rural Universities

We investigate people’s use of strategies for sampling data in an active learning task. In the spirit of resource-rational analysis, we argue that people may often use effective heuristics to guide sampling in lieu of more computationally expensive optimization strategies, but that they do not simply continue relying heuristics that are shown to be ineffective, but instead flexibly shift to new strategies. By clustering participants’ sampling behaviour, we show that people engage in varied sampling strategies, shifting more often when encountering some function families. Although not all strategies improved participants’ performance on a subsequent prediction task, when the function family changed, participants quickly updated their beliefs about the likely function family on subsequent trials. Although sparse opportunities to learn may limit participants’ ability to fully learn the underlying functional relationships, people’s ability to dynamically shift their active learning behaviour may help them understand the abstract features of complex relationships.

Characterizing Shifts in Strategy in Active Function Learning

Theory of Mind enables us to attribute mental states to others. But we not only make inferences about mental states (like what someone believes or wants), but about mental processes (like if someone is distracted or whether they remember something). Here, we present a computational formalization of these kinds of inferences. We propose that inferences about mental processes are structured around a principle of rational mental effort: the expectation that other people allocate mental resources rationally so as to minimize thinking costs incurred while pursuing their goals. We develop this theory into a computational model in the context of the Rush Hour puzzle game. In two behavioral experiments testing different inferences about mental processing, we find that our model predicts participant judgments. This work advances our understanding of the richness of the human mind's ability to think about other minds, and even about thinking itself.

Thinking about Thinking as Rational Computation

Problem solving under time and material constraints assumes that humans face both their own internal computational limitations and constraints of the external task environment. To find optimal solutions under these constraints, the agent must adapt their behavior by using various strategies such as offloading (i.e., using external materials to aid their performance). We designed a novel tower building task to investigate adaptive use of strategies under constraints. The task worked as designed: Participants found the optimal solution most often on the least difficult scenario and least often on the most difficult scenario. Surprisingly, offloading led to no significant differences in performance. However, on the most difficult scenario, some participants found the optimal solution using a prospective, concurrent, or retrospective strategy based on experience with the constraints of the task environment. We conclude that optimality can be understood as a trend over time and investigated in studies that allow multiple attempts.

Adaptivity and optimization under constraints

Comprehensive datasets used for modeling endeavours in syllogistic reasoning
research usually contain only a single conclusion per task for each subject.
However, this means that no information about the other conclusions is provided,
preventing the distinction between conclusions that were rejected, conclusions
deemed to be valid but not the preferred conclusion and conclusions that were
not considered at all. In this work, we present a multiple choice dataset
containing all conclusions that participants considers valid. The data is
compared to datasets with other response designs, and an extensive evaluation
is performed to assess the impact of the response design on the predictive 
performance of cognitive models. Finally, our results are discussed and put
into perspective.


Premium content

Downloads

Next from CogSci 2023

Hybrid Frameworks of Reasoning: Normative-Descriptive Interplay

Similar lecture

Towards Comprehensive Argument Analysis in Education: Dataset, Tasks, and Method