Research on the capacity to understand others' minds has tended to focus on representations ofbeliefs,which are widely taken to be among the most central and basic theory of mind representations. Representations ofknowledge, by contrast, have received comparatively little attention and have often been understood as depending on prior representations of belief. After all, how could one represent someone as knowing something if one does not even represent them as believing it? Drawing on a wide range of methods across cognitive science, (...) we ask whether belief or knowledge is the more basic kind of representation. The evidence indicates that nonhuman primates attribute knowledge but not belief, that knowledge representations arise earlier in human development than belief representations, that the capacity to represent knowledge may remain intact in patient populations even when belief representation is disrupted, that knowledge (but not belief) attributions are likely automatic, and that explicit knowledge attributions are made more quickly than equivalent belief attributions. Critically, the theory of mind representations uncovered by these various methods exhibits a set of signature features clearly indicative of knowledge: they are not modality-specific, they are factive, they are not just true belief, and they allow for representations of egocentric ignorance. We argue that these signature features elucidate the primary function of knowledge representation: facilitating learning from others about the external world. This suggests a new way of understanding theory of mind – one that is focused on understanding others' minds in relation to the actual world, rather than independent from it. (shrink)

The prominence of Bayesian modeling of cognition has increased recently largely because of mathematical advances in specifying and deriving predictions from complex probabilistic models. Much of this research aims to demonstrate that cognitive behavior can be explained from rational principles alone, without recourse to psychological or neurological processes and representations. We note commonalities between this rational approach and other movements in psychology – namely, Behaviorism and evolutionary psychology – that set aside mechanistic explanations or make use of optimality assumptions. Through (...) these comparisons, we identify a number of challenges that limit the rational program's potential contribution to psychological theory. Specifically, rational Bayesian models are significantly unconstrained, both because they are uninformed by a wide range of process-level data and because their assumptions about the environment are generally not grounded in empirical measurement. The psychological implications of most Bayesian models are also unclear. Bayesian inference itself is conceptually trivial, but strong assumptions are often embedded in the hypothesis sets and the approximation algorithms used to derive model predictions, without a clear delineation between psychological commitments and implementational details. Comparing multiple Bayesian models of the same task is rare, as is the realization that many Bayesian models recapitulate existing (mechanistic level) theories. Despite the expressive power of current Bayesian models, we argue they must be developed in conjunction with mechanistic considerations to offer substantive explanations of cognition. We lay out several means for such an integration, which take into account the representations on which Bayesian inference operates, as well as the algorithms and heuristics that carry it out. We argue this unification will better facilitate lasting contributions to psychological theory, avoiding the pitfalls that have plagued previous theoretical movements. (shrink)

Recent research in cognitive and computational neuroscience portrays the neocortex as a hierarchically structured prediction machine. Several theorists have drawn on this research to challenge the traditional distinction between perception and cognition – specifically, to challenge the very idea that perception and cognition constitute useful kinds from the perspective of cognitive neuroscience. In place of this traditional taxonomy, such theorists advocate a unified inferential hierarchy subject to substantial bi-directional message passing. I outline the nature of this challenge and then raise (...) two objections to the cognitive architecture it proposes as a replacement: first, standard ways of characterising this inferential hierarchy are in tension with the representational reach of conceptual thought; second, there is compelling evidence that commonsense reasoning is structured around highly domain-specific intuitive theories that are difficult to situate within a single hierarchy. (shrink)

Is language understanding a special case of social cognition? To help evaluate this view, we can formalize it as the rational speech-act theory: Listeners assume that speakers choose their utterances approximately optimally, and listeners interpret an utterance by using Bayesian inference to “invert” this model of the speaker. We apply this framework to model scalar implicature (“some” implies “not all,” and “N” implies “not more than N”). This model predicts an interaction between the speaker's knowledge state and the listener's interpretation. (...) We test these predictions in two experiments and find good fit between model predictions and human judgments. (shrink)

Cognitive scientists have been debating how the folk concept of intentional action works. We suggest a simple account: people consider that an agent did X intentionally to the extent that X was causally dependent on how much the agent wanted X to happen (or not to happen). Combined with recent models of human causal cognition, this definition provides a good account of the way people use the concept of intentional action, and offers natural explanations for puzzling phenomena such as the (...) side-effect effect. We provide empirical support for our theory, in studies where we show that people’s causation and intentionality judgments track each other closely, in everyday situations as well as in scenarios with unusual causal structures. Study 5 additionally shows that the effect of norm violations on intentionality judgments depends on the causal structure of the situation, in a way uniquely predicted by our theory. Taken together, these results suggest that the folk concept of intentional action has been difficult to define because it is made of cognitive building blocks, such as our intuitive concept of causation, whose logic cognitive scientists are just starting to understand. (shrink)

Inspired by, and in close relation with, the contributions of this special issue, Kuperberg elegantly links event comprehension, production, and learning. She proposes an overarching hierarchical generative framework of processing events enabling us to make sense of the world around us and to interact with it in a competent manner.

Bayesian models are often criticized for postulating computations that are computationally intractable (e.g., NP-hard) and therefore implausibly performed by our resource-bounded minds/brains. Our letter is motivated by the observation that Bayesian modelers have been claiming that they can counter this charge of “intractability” by proposing that Bayesian computations can be tractably approximated. We would like to make the cognitive science community aware of the problematic nature of such claims. We cite mathematical proofs from the computer science literature that show intractable (...) Bayesian computations, such as postulated in existing Bayesian models, cannot be tractably approximated. This does not mean that human brains do not (or cannot) implement the type of algorithms that Bayesian modelers are advancing, but it does mean that proposing that they do by itself does nothing to parry the charge of intractability, because the postulated algorithms are as intractable (i.e., require exponential time) as the computations they try to approximate. Besides our negative message for the community, our letter also makes a positive contribution by referring to a methodology that Bayesian modelers can use to try and parry the charge of intractability in a mathematically sound way. (shrink)

The past 15 years occasioned an extraordinary blossoming of research into the cognitive and affective mechanisms that support moral judgment and behavior. This growth in our understanding of moral mechanisms overshadowed a crucial and complementary question, however: How are they learned? As this special issue of the journal Cognition attests, a new crop of research into moral learning has now firmly taken root. This new literature draws on recent advances in formal methods developed in other domains, such as Bayesian inference, (...) reinforcement learning and other machine learning techniques. Meanwhile, it also demonstrates how learning and deciding in a social domain—and especially in the moral domain—sometimes involves specialized cognitive systems. We review the contributions to this special issue and situate them within the broader contemporary literature. Our review focuses on how we learn moral values and moral rules, how we learn about personal moral character and relationships, and the philosophical implications of these emerging models. (shrink)

An important, but relatively neglected, aspect of human theory of mind is emotion inference: understanding how and why a person feels a certain why is central to reasoning about their beliefs, desires and plans. The authors review recent work that has begun to unveil the structure and determinants of emotion inference, organizing them within a unified probabilistic framework.

In this commentary, I advocate for strict relations between Marr's levels of analysis. Under a strict relationship, each level is exactly implemented by the subordinate level. This yields two benefits. First, it brings consistency for multilevel explanations. Second, similar to how a sculptor chisels away superfluous marble, a modeler can chisel a computational-level model by applying constraints. By sculpting the model, one restricts the set of possible algorithmic- and implementational-level theories.

Why is interaction so simple? This article presents a theory of interaction based on the use of shared representations as “coordination tools” (e.g., roundabouts that facilitate coordination of drivers). By aligning their representations (intentionally or unintentionally), interacting agents help one another to solve interaction problems in that they remain predictable, and offer cues for action selection and goal monitoring. We illustrate how this strategy works in a joint task (building together a tower of bricks) and discuss its requirements from a (...) computational viewpoint. (shrink)

Chimpanzees and humans are close evolutionary relatives who behave in many of the same ways based on a similar type of agentive organization. To what degree do they experience the world in similar ways as well? Using contemporary research in evolutionarily biology and animal cognition, I explicitly compare the kinds of experience the two species of capable of having. I conclude that chimpanzees’ experience of the world, their experiential niche as I call it, is: intentional in basically the same way (...) as humans’; rational in the sense that it is self-critical and operates with logically structured causal and intentional inferences; but not normative at all in that it does not operate with “objective” evaluative standards. Scientific data do not answer philosophical questions, but they provide rich raw material for scientists and philosophers alike to reflect on and clarify fundamental psychological concepts. (shrink)

Judea Pearl has argued that counterfactuals and causality are central to intelligence, whether natural or artificial, and has helped create a rich mathematical and computational framework for formally analyzing causality. Here, we draw out connections between these notions and various current issues in cognitive science, including the nature of mental “programs” and mental representation. We argue that programs (consisting of algorithms and data structures) have a causal (counterfactual-supporting) structure; these counterfactuals can reveal the nature of mental representations. Programs can also (...) provide a causal model of the external world. Such models are, we suggest, ubiquitous in perception, cognition, and language processing. (shrink)

Behavior oftentimes allows for many possible interpretations in terms of mental states, such as goals, beliefs, desires, and intentions. Reasoning about the relation between behavior and mental states is therefore considered to be an effortful process. We argue that people use simple strategies to deal with high cognitive demands of mental state inference. To test this hypothesis, we developed a computational cognitive model, which was able to simulate previous empirical findings: In two-player games, people apply simple strategies at first. They (...) only start revising their strategies when these do not pay off. The model could simulate these findings by recursively attributing its own problem solving skills to the other player, thus increasing the complexity of its own inferences. The model was validated by means of a comparison with findings from a developmental study in which the children demonstrated similar strategic developments. (shrink)

When people make decisions, they act in a way that is either automatic (“rote”), or more thoughtful (“reflective”). But do people notice when others are behaving in a rote way, and do they care? We examine the detection of rote behavior and its consequences in U.S. adults, focusing specifically on pedagogy and learning. We establish repetitiveness as a cue for rote behavior (Experiment 1), and find that rote people are seen as worse teachers (Experiment 2). We also find that the (...) more a person's feedback seems similar across groups (indicating greater rote-ness), the more negatively their teaching is evaluated (Experiment 3). A word-embedding analysis of an open-response task shows people naturally cluster rote and reflective teachers into different semantic categories (Experiment 4). We also show that repetitiveness can be decoupled from perceptions of rote-ness given contextual explanation (Experiment 5). Finally, we establish two additional cues to rote behavior that can be tied to quality of teaching (Experiment 6). These results empirically show that people detect and care about scripted behaviors in pedagogy, and suggest an important extension to formal frameworks of social reasoning. (shrink)

What place should formal or computational methods occupy in social psychology? We consider this question in historical perspective, survey the current state of the field, introduce the several new contributions to this special issue, and reflect on the future.

To introduce the issue of the tracking and identification of human agents, I examine the ability of an agent to track a human person and distinguish this target from other individuals: The ability to perform person identification. First, I discuss influential mechanistic models of the perceptual recognition of human faces and people. Such models propose detailed hypotheses about the parts and activities of the mental mechanisms that control the perceptual recognition of persons. However, models based on perceptual recognition are incomplete (...) theories of person identification because they do not explain several identification behaviors that are fundamental to human social interactions. Furthermore, recognition-based models tend to appeal to the controversial concept of the “identity” of a person without explaining what determines personal identity and persistence. To overcome these limitations, I propose to integrate the face-recognition program into a broader causal-historical theory of identification. The causal-historical theory of identification complements models focused on perceptual recognition because it can account for the types of non-perceptual identification overlooked by the face-recognition program. Moreover, it can decompose the identification behaviors into tracking processes that succeed or fail to be sensitive to causal characteristics of a target. I illustrate these advantages with a discussion of the difference between the tracking of a person understood as either a causally continuous biological organism or a psychologically continuous mind. Finally, I argue that the causal-historical theory provides a theoretical framework for investigating the tracking of relations between a target and its contextual and historical attributes, such as a target's possessions. (shrink)

The commentaries suggest many important improvements to the target article. They clearly distinguish two varieties of rationalization – the traditional “motivated reasoning” model, and the proposed representational exchange model – and show that they have distinct functions and consequences. They describe how representational exchange occurs not only by post hoc rationalization but also by ex ante rationalization and other more dynamic processes. They argue that the social benefits of representational exchange are at least as important as its direct personal benefits. (...) Finally, they construe our search for meaning, purpose, and narrative – both individually and collectively – as a variety of representational exchange. The result is a theory of rationalization as representational exchange both wider in scope and better defined in mechanism. (shrink)

Cognitive Science, Volume 46, Issue 7, July 2022.

A key benefit of Bayesian reasoning is that it stipulates how to optimally integrate unreliable sources of information. The authors present evidence that humans use Bayesian inference to determine how much to trust advice from another person, based on information about that person's knowledge and strategy.

Human vision supports social perception by efficiently detecting agents and extracting rich information about their actions, goals, and intentions. Here, we explore the cognitive architecture of perceived animacy by constructing Bayesian models that integrate domain‐specific hypotheses of social agency with domain‐general cognitive constraints on sensory, memory, and attentional processing. Our model posits that perceived animacy combines a bottom–up, feature‐based, parallel search for goal‐directed movements with a top–down selection process for intent inference. The interaction of these architecturally distinct processes makes perceived (...) animacy fast, flexible, and yet cognitively efficient. In the context of chasing, in which a predator (the “wolf”) pursues a prey (the “sheep”), our model addresses the computational challenge of identifying target agents among varying numbers of distractor objects, despite a quadratic increase in the number of possible interactions as more objects appear in a scene. By comparing modeling results with human psychophysics in several studies, we show that the effectiveness and efficiency of human perceived animacy can be explained by a Bayesian ideal observer model with realistic cognitive constraints. These results provide an understanding of perceived animacy at the algorithmic level—how it is achieved by cognitive mechanisms such as attention and working memory, and how it can be integrated with higher‐level reasoning about social agency. (shrink)

Human and automated tutors attempt to choose pedagogical activities that will maximize student learning, informed by their estimates of the student's current knowledge. There has been substantial research on tracking and modeling student learning, but significantly less attention on how to plan teaching actions and how the assumed student model impacts the resulting plans. We frame the problem of optimally selecting teaching actions using a decision-theoretic approach and show how to formulate teaching as a partially observable Markov decision process planning (...) problem. This framework makes it possible to explore how different assumptions about student learning and behavior should affect the selection of teaching actions. We consider how to apply this framework to concept learning problems, and we present approximate methods for finding optimal teaching actions, given the large state and action spaces that arise in teaching. Through simulations and behavioral experiments, we explore the consequences of choosing teacher actions under different assumed student models. In two concept-learning tasks, we show that this technique can accelerate learning relative to baseline performance. (shrink)

The dominant practice of AI alignment assumes (1) that preferences are an adequate representation of human values, (2) that human rationality can be understood in terms of maximizing the satisfaction of preferences, and (3) that AI systems should be aligned with the preferences of one or more humans to ensure that they behave safely and in accordance with our values. Whether implicitly followed or explicitly endorsed, these commitments constitute what we term apreferentistapproach to AI alignment. In this paper, we characterize (...) and challenge the preferentist approach, describing conceptual and technical alternatives that are ripe for further research. We first survey the limits of rational choice theory as a descriptive model, explaining how preferences fail to capture the thick semantic content of human values, and how utility representations neglect the possible incommensurability of those values. We then critique the normativity of expected utility theory (EUT) for humans and AI, drawing upon arguments showing how rational agents need not comply with EUT, while highlighting how EUT is silent on which preferences are normatively acceptable. Finally, we argue that these limitations motivate a reframing of the targets of AI alignment: Instead of alignment with the preferences of a human user, developer, or humanity-writ-large, AI systems should be aligned with normative standards appropriate to their social roles, such as the role of a general-purpose assistant. Furthermore, these standards should be negotiated and agreed upon by all relevant stakeholders. On this alternative conception of alignment, a multiplicity of AI systems will be able to serve diverse ends, aligned with normative standards that promote mutual benefit and limit harm despite our plural and divergent values. (shrink)

Actions are usually generalized among social group members. Importantly, the efficiency of an action with respect to achieving an external target determines action understanding, and it may have different degrees of social relevance to social groups. Thus, this study explored the role of action efficiency in action generalization. We used computer animations to simulate actions in social groups initiated by visual action cues or category labels, and we measured differences in response times between identifying actions that were and were not (...) consistent with group members, without explicit requirements regarding generalization. It was found that in both visually introduced and explicitly labeled social groups, when the group members acted inefficiently toward the external object, perceivers identified group‐consistent actions faster than group‐inconsistent actions, indicating that the exemplars' common inefficient actions are generalized to the unknown ingroup member, accordingly facilitating the identification of expected consistent inefficient action (Experiment 1). As this effect was not present when removing social group cues, it was determined to be specific to social groups (Experiment 2). Importantly, such generalization was not observed when the identical action was deemed efficient toward the external object (Experiment 3) and was specific to the demonstration of the action being completed by multiple group members rather than being repeated twice by one group member, supporting the group‐based inference and ruling out the possibility of the increased memorability of inefficient actions leading to more generalization relative to efficient actions (Experiment 4). Therefore, the efficiency of an action bounds the generalization of the action across social group members through a process that is spontaneous and implicit. This constrained action generalization may be due to inefficient actions being represented as culture‐specific conventional forms. (shrink)