What can category theory contribute to cognitive science? We argue that the category theory principle of construction via a universal mapping property affords a significant contribution. Such universal constructions explain why, not just how cognition is systematic/compositional, i.e. the “best” one can do in the given context. The significance of this principle is indicated by examples.

Visual perception, receiving a two-dimensional (2D) visual input, often constructs the three-dimensional (3D) perceptual image. Although there are generally multiple structures in the external world that give an equivalent two-dimensional retinal image, the perceptual process naturally and easily infers only one 3D structure as the solution. However, the following problems are not obvious at all: what kind of structure can be obtained as a 3D perceptual image from certain 2D information, and why do we get a three-dimensional perceptual image instead of a two-dimensional one. In the present study, we investigate this problem by untangling the Necker Cube phenomenon, and propose a novel theory of three-dimensional visual perception from the viewpoint of the efficiency of information coding. Among the possible structures that can yield the 2D retinal image of the Necker Cube, the structure of the typical three-dimensional perceptual image of the Necker Cube maximizes the symmetry (in group theory). This maximization of symmetry is characterized by the pairs of adjoint functors (in category theory). Therefore, according to this proposed theory, “the Necker Cube” in the three-dimensional space is perceived as the most efficient encoding of the two-dimensional retinal image.

Machine learning techniques have realized some principal cognitive functionalities such as nonlinear generalization and causal model construction, as far as huge amount of data are available. A next frontier for cognitive modelling would be the ability of humans to transfer past knowledge to novel, ongoing experience, making analogies from the known to the unknown. Novel metaphor comprehension may be considered as an example of such transfer learning and analogical reasoning that can be empirically tested in a relatively straightforward way. Based on some concepts inherent in category theory, we implement a model of metaphor comprehension called the theory of indeterminate natural transformation (TINT), and test its descriptive validity of humans' metaphor comprehension. We simulate metaphor comprehension with two models: one being structure-ignoring, and the other being structure-respecting. The former is a sub-TINT model, while the latter is the minimal-TINT model. As the required input to the TINT models, we gathered the association data from human participants to construct the “latent category” for TINT, which is a complete weighted directed graph. To test the validity of metaphor comprehension by the TINT models, we conducted an experiment that examines how humans comprehend a metaphor. While the sub-TINT does not show any significant correlation, the minimal-TINT shows significant correlations with the human data. It suggests that we can capture metaphor comprehension processes in a quite bottom-up manner realized by TINT.

The aim of this article is to provide references to cognitive scientists, who are interested in learning category theory and using it in their research. This article consists of the three sections, question-and-answers on category theory, utility of category theory on cognitive science, and tutorial materials. In the question-and-answers on category theory, we answered to questions, with which beginners of category theory may come up. In the utility of category theory on cognitive science, we raised the three items of utility of category theory in building cognitive models. The learning materials share the books, slides, and videos on the web, recommended to start with.

When conversational participants jointly attend to an object in the environment, how do they refer to it through their utterances and/or body movements to refer to it? Hindmarsh and Heath (2000) has proposed that participants employ the principle of recipient design of visual reference, wherein they design a visual reference as to provide recipients with clearer views of the relevant objects and scenes. In this study, focusing on a visual reference to an object occluded by a referrer/recipient's body or another object in the environment, we conducted detailed analyses of three segments collected from natural conversation data. The results show that some visual references were characterized as follows: (1) not providing a “clearer view”to the recipient, (2) not referring to the position or direction of the object directly, (3) specifying the spatial layout of the referrer, the recipient, an occluding object, and a referent. These results suggest that the recipient design of visual reference is sufficient, if an object is referred to visually, such that the recipient can recognize the spatial layout of the referrer, the recipient, and the object. Hence, even if clearer views are not available, a referent can be established visually if the recipient is provided with a guide to it.

The purpose of this study was to verify the possibility of setting multi-boundaries and to reconsider the meaning of intergenerational learning. The focus was on a school-community partnership organization called S Junior High School Support (S-Sapo). The data for the study are from fieldwork conducted in the organization from April 2016 to June 2020. The analysis is based on field notes and semi-structured interviews with volunteers who participated in the activity, either inside or outside of the school district. Two volunteers from different generations were selected as the subjects of the analysis. The selected senior volunteer showed substantial generativity, while the younger volunteer sought to absorb knowledge from the older generation. Through the collaborative practice encouraged by S-sapo, the younger subject adopted practical behaviors generated from the senior subject's job experiences by crossing syntactic boundaries. This phenomenon is called knowledge transfer and is generated from differences in the level of knowledge of the participants. The younger subject also succeeded in adopting perspectives that support these practical behaviors, reinterpreting the senior's perspective, and then applying it to practical contexts by crossing semantic boundaries. This phenomenon of adoption is generated by interactions among contexts. Crossing these boundaries enabled the younger subject to imagine her future from a present perspective. Such illumination of a future path can be regarded as intergenerational learning. By borrowing other people's perspectives in the process of an intergenerational exchange, one is able to envision steps for the future, even if they are somewhat vague. Thus, the intergenerational exchange works as social scaffolding for members of the younger generation.

In this study, the effects of a theatrical activity on social abilities were examined. The participants (N = 40) were randomly assigned to an experimental or control group. While the experimental group read a playscript, planned a performance, and performed, the control group read a playscript and summarized the content of the story. All the participants completed three social ability measures, namely, Reading the Mind in the Eyes Test, Yoni Test, and Situational Test of Emotional Understanding as well as a questionnaire, namely, Interpersonal Reactivity Index on three occasions: pre, post, and followup measurement. Moreover, they completed the Narrative Transportation Scale immediately after the intervention, which assessed the psychological state of immersion into the narrative world. It was predicted that the extent of narrative transportation the participants experience would moderate the effect of theatrical activity. The results revealed that highly transported participants in the experimental group scored significantly higher than those in the control group on various scales including Reading the Mind in the Eyes Test, Yoni Test, and the Interpersonal Reactivity Index Empathic Concern scale. Furthermore, after the intervention, the extent of transportation predicted these social abilities in the experimental group.

The use of diagrams, instead of linguistic and symbolic representations, in logical reasoning has been the subject of continuous research interest over the last few centuries. Especially in 1990s, such logic diagrams have attracted substantial research attention in the fields of logic, artificial intelligence and cognitive science. More recently, according to some applications to computer sciences, various diagrammatic systems have been developed to handle extended types of logical information and reasoning. The research direction of logic diagram design raises the question of to what extent diagrammatic systems can be expressive enough to handle the generality of logics as well as be natural and effective so that they are easy to understand. This paper surveys historical and recent developments of logic diagrams, focusing on the two types of visualization: Euler/Venn diagrams and graph representations. While a majority of previous researches on cognitive science of reasoning focus on linguistic forms, this survey provides a systematic overview of diagrammatic reasoning, contributing to understanding the theoretical basis of reasoning in visual forms.

In a story, the meaning of an event depends on its context or its relationship with other events. To elucidate the cognitive mechanism associated with the construction of such contextual meaning is a fundamental issue from the perspective of cognitive science and artificial intelligence. Hence, this study proposes a computational model that determines the similarity between two events occurring in different stories based on the analogical mapping between the events' contextual structures. The proposed model automatically constructs the contextual structure of a chosen event by associating relevant verbs gathered from precedent and subsequent events. Similarity between the selected events occurring in two stories is thereby estimated on the basis of the mapping between their contextual structures. To assess the proposed model, it was compared with human ratings of event similarity by incorporating twelve pairs of experimental stories. The obtained result indicates that a moderate or a strong positive correlation can be observed between the proposed model and the human ratings.

Negative attitudes toward gambling disorder have a lot of undesirable effects on the patients' mental health and prevent them from fundamental human rights. Labeling is the process of specifying the name of the illness and understanding the patient in terms of their connection with the illness. The relation between labeling and negative attitudes toward gambling disorder has been investigated in the literature, but studies have not shown consistent results. In our study, we mainly focused on the roles of two mediator variables —responsibility and controllability— and also examined the effects of perceivers' past contact experiences with people with mental illness and of their gender. A total of 440 Japanese participants took part in the experiment. Half of the participants read the vignette with labeling to the target, who was diagnosed as having moderate gambling disorder, while the other half of the participants read a vignette without such labeling. We analyzed the data with Bayesian path analysis where we determined the model based on goodness of fit indices. The results suggested that labeling possibly decreases people's negative attitudes through a decrease in the extent of responsibility attributed to the target. Contact experience did not affect the attitudes and women tended to show more negative attitudes toward the target than men. Despite the limitations, the findings of the study have implications for human rights education and to reduce the negative attitudes toward gambling disorder.

観察可能な行動を手がかりにしてその背後にある情報処理の仕組みをモデル化し，人の心を理解しようと試みるのが認知心理学の基本的なスタンスである．認知心理学のこのような考え方は，確率モデルでデータの生成過程を表現し，その確率モデルをデータに当てはめることによって現象の理解と予測を促す統計モデリングの手法と非常に相性が良い．特に近年，Stan (Stan Development Team, 2020) やJAGS (Plummer, 2020) といったベイズ推定を実行するための確率的プログラミング言語が登場し，従来よりも容易かつ柔軟に統計モデリングを実施できる環境が整ってきたことは認知心理学にとっても追い風である．このような状況を踏まえ，本稿ではベイズ統計モデリングの強みを生かした最近の認知心理学研究を紹介する^a．
a  行為主体がベイズの定理に基づいて信念を更新するとみなす意思決定の認知モデル(e.g., 中村, 2009; 繁桝, 1995) については本稿では扱わない．