Canada

How do humans produce and comprehend language in pragmatic ways? A variety of models of pragmatic inferences have been proposed, and these models are typically evaluated on their ability to account for human inferences in reference game experiments. However, these experiments are not tailored to target theoretical differences between models or clearly tease apart model predictions. We propose an optimal experiment design approach to systematically construct reference games that can optimally differentiate between models of human pragmatic reasoning. We demonstrate this approach and apply it to four models that have been debated in the literature: Grammar-based, Iterated Best Response (IBR), Rational Speech Act (RSA), and a recent variant of RSA grounded in rate-distortion theory (RD-RSA). Using these optimal reference game experiments, we find empirical evidence favoring iterated rationality models over the grammar-based model, as well as support for the relevance of rate-distortion theory to human pragmatic inferences. These results suggest that our optimal reference game design framework may help adjudicate between computational theories of pragmatic reasoning.

CogSci 2022

Teasing apart models of pragmatics using optimal reference game design

technical paper

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. 





Third party punishment (TPP), or altruistic punishment, is specifically human prosocial behavior. TPP denotes the administration of a sanction to a transgressor by an individual that is not affected by the transgression. In some evolutionary accounts, TPP is considered crucial for the stability of cooperation and solidarity in larger groups formed by genetically unrelated individuals. Belief in free will (BFW), on the other hand, is the idea that humans have control over their behavior. BFW is a human universal notion that, in some studies, has been found to be supportive of prosocial behavior. In our study, we examined the effect of BFW on TPP under high and low affect scenarios through optical brain imaging (fNIRS). We hypothesized that in low affect cases, there would be a positive correlation between the strength of the BFW and the severity of the punishment inflicted. Obtained results and related statistical analyses indicate that participants with higher degree of BFW have more neural activation in their right dorsolateral prefrontal cortex (DLPFC) (hbo and hbt measures) in high affect scenarios, whereas the participants with lower degree of BFW have higher levels of neural activation in the medial PFC (hbo and hbt measures) in low affect scenarios. These empirical findings are in line with the research findings in the relevant academic literature and support the hypothesis that the degree of BFW influences punishment decisions.

The Neural Correlates of the Effect of Belief in Free Will on Third-Party Punishment: An Optical Brain Imaging (fNIRS) Study

Dual character concepts have recently received significant attention from experimental philosophers. They are usually defined as possessing two independent criteria for categorization, each associated with its own sense. One of these criteria is descriptive, while the other is normative. Almost all examples discussed in the literature involve a positively evaluated normative element (e.g., the concept of scientist is associated with the pursuit of empirical truth, something that we value). But can dual character concepts also involve morally bad elements? We report the first evidence suggesting that they can. Our study contrasted pairs of positive and negative concepts (e.g., friend/enemy, joy/sadness) and neutral concepts (e.g., baker). We found that negatively valenced concepts such as traitor, enemy, superficial, or sadness have a dual character. We also observed that positive concepts were significantly more likely to have a dual character than negative ones, but the effect size of this difference was small.

Are There Bad Dual Character Concepts?

Morality is a critical aspect of life––it influences how we think, design systems, and even the stories we tell. Looking to the popularity of true crime stories and characters like Dexter Morgan, it seems that our preferences are toward exploring moral ambiguity and moral badness. Across two experiments, we examine what moral ambiguity is and what kinds of moral information spark curiosity and explanation-seeking. In Experiment 1, we manipulate moral ambiguity to mean someone with conflicting moral character, and we predict those individuals will trigger curiosity more than morally consistent people. Results suggest that both morally ambiguous and immoral minds pique curiosity for explanations. In Experiment 2, we find that when ambiguity is instead operationalized as what is typical or average, we are curious about morally deviant things. This research points to critical differences in the kinds of moral minds we are curious to learn more about.


What is moral ambiguity and when does it trigger curiosity?

Artificial Intelligence (AI) can be harnessed to create sophisticated social and moral scoring systems – enabling people and organisations to form judgements of others at scale. While this capability has many useful applications – e.g., matching romantic partners who are aligned in their moral principles, it also raises many ethical questions. For example, there is widespread concern about the use of social credit systems in the political domain. In this project, we approach this topic from a psychological perspective. With experimental evidence, we show that the acceptability of moral scoring by AI depends on its perceived accuracy, and that perceived accuracy is compromised by people's tendency to see themselves as morally peculiar, and thus less characterizable by AI. That is, we suggest that people overestimate the peculiarity of their moral profile, believe that AI will neglect this peculiarity, and resist for this reason the introduction of moral scoring by AI.


The Quantified Moral Self

What is the relationship between trust and perceived agency? The present study experimentally investigated the effect of people’s perception of a robot’s compliance (and resistance) to social norms on their evaluation of a robot’s perceived agency, performance trust, and moral trust. Participants reported a norm-conforming robot to have higher perceived agency and a greater sense of trust than a robot that violated social norms. 
We also found that perceived agency, regardless of how much a robot followed norms, was correlated with trust. We interpret this finding as evidence that as people see a robot as having agency, they trust it more

Perceived Agency Changes Performance and Moral Trust in Robots

Dennett’s (1981) call to use computational frameworks in philosophical analysis can assist in clarifying the distinction between System-1 and System-2 metacognition. Metacognition has been described as “cognition about cognitive phenomena” (Flavell, 1979) and is renowned for its particularly difficult issues. Understanding its “mysterious mechanisms” (Brown, 1987) has been obscured by the subject’s abstractness and disputed terminology. Computation is uniquely suited to overcome these barriers as it forces definitional clarity. New research is supportive of the Common Model of Cognition as the best candidate for computationally modeling the functional organization of the human brain (Stocco et al., 2021). The Common Model represents a powerful tool for investigating unresolved cognitive issues as it is informed by a convergence of research and is well-defined. We propose that applying the Common Model as a philosophical tool can assist in clarifying the problematic distinction between System-1 and System-2 metacognition.


Clarifying Metacognition System 1 & 2 with the Common Model

The primary phenomenon within the field of collaborative memory is collaborative inhibition: the tendency for collaborative groups to underperform in free recall tasks compared to nominal groups of comparable size. Previously, we adapted the Search of Associative Memory (SAM; Raaijmakers & Shiffrin, 1981) model to collaborative free recall and found support for the retrieval disruption hypothesis (Mannering, Rajaram, Shiffrin, & Jones, 2022). In this paper, we investigate another possible cause of collaborative inhibition: memory homogenization. Previous modeling attempts have shown that memory homogenization may contribute to collaborative inhibition (Luhmann & Rajaram, 2015). In this paper, we prevent SAM models from learning during collaborative retrieval which subsequently prevents memory homogenization. We found that even when SAM model memories remained diversified, collaborative inhibition persisted—though the strength of the effect was diminished. These results suggest that both retrieval disruption and memory homogenization may contribute to collaborative inhibition in SAM.

Modeling the Effect of Learning During Retrieval on Collaborative Inhibition

Metacognition ("thinking about thinking") depends on representations and metarepresentations, which are usually treated as internal knowledge structures, rather than as external-oriented processes. We explore how multimodal patterns of external semiotic activity (not monomodal and internal symbolic-based processes) can provide a more accurate description of metacognition. To develop our ideas, we examine the act of marking-for-self in theatrical dance. To mark is to perform a dance phrase in a simplified, schematic, or abstract way. When marking, dancers use their bodies in motion to represent some aspects (properties, dynamics, or structures) of the complete dance-phrase they are thinking on. Marking-for-self occurs when dancers mark in their own idiosyncratic manner, a process that potentializes real-time reflection on its own dancing sequences and structures. According to our view, these representations are metacognitive embodied and externalized diagrammatic structures. We detail this phenomenon in light of recent developments in Cognitive Semiotics and diagrammatic reasoning approach.

Metacognition and marking-for-self in theatrical dance as a multimodal pattern of semiotic activity

Post-error slowing describes a delayed response in action after error commission. PES is a key metric of error monitoring that individuals possess self-reflective information processing, which is crucial for adaptive behaviour in learning. Studies seeking to ascertain the relationship between learning, feedback and PES remain rare. The current study aims to investigate the role task and feedback played in the magnitude of PES. Participants (N = 487, Mage = 23.77 years, SD = 5.8) performed three cognitive tasks: Flanker, Numeric and Science knowledge task. A 2 (Type of trial) × 3 (Task) × 2 (Feedback) mixed ANOVA estimated by REML was conducted; participant was treated as a random effect. We found a significant PES effect where the post-error RT is slower than the post-correct RT. We also found a significant effect of task, feedback, and Task × Feedback interaction.

Feedback-related and Task-dependent Post-error Slowing

While population-level models often provide a good fit to the data, they may mask meaningful individual differences. Exploring individual differences can also be beneficial for gaining a better understanding of the processes that underlie pragmatic phenomena. In this study, we investigate whether the substantial differences in performance on a pragmatic reference game can be traced back to cognitive or socio-pragmatic traits. We observe a significant effect of the ability to inhibit an intuitive response and of abstract reasoning ability. In contrast, we do not find evidence that socio-pragmatic abilities or working memory capacity influence pragmatic responding.

Premium content

Teasing apart models of pragmatics using optimal reference game design

Downloads

Next from CogSci 2022

The Neural Correlates of the Effect of Belief in Free Will on Third-Party Punishment: An Optical Brain Imaging (fNIRS) Study

Stay up to date with the latest Underline news!

PRESENTATIONS

CONFERENCES

COMPANY

RESOURCES