EnglishbJapanese
Site Top
Instructions for Submissions -closed-
Submission Formats -closed-
Registration Information
Conference Schedule -new-
Conference Program -new-

Additional lodging
Conference Schedule
Conference Program

The 28th CogSci 2006
Cognitive Science Society
The 23th JCCS2006(in Japanese)
Japanese Cognitive Science Society

In this site, the PDF file is included in part. Adobe Reader of the Adobe Co. is necessary to
inspect it.

Workshops / Symposia

| workshop 1 | workshop 2 | symposium | workshop 3 | workshop 4 |
Workshop 1

How the mind and culture are shaped interactively: Bridging social and cognitive views

Organizer:
Hiroshi Yama (Kobe College, Japan)
Many of the cognitive psychologists have assumed that human mind is universal, and have inferred that, even if they observe cultural differences, they are the consequences of different cultural stimuli. Recently, some of social psychologists have argued an interactive view that mind and culture are shaped interactively, and this has given an impact to cognitive psychologists. However, although some social psychologists use cognitive tasks, their terminology and concepts are often different from those of cognitive psychologists'. The symposium of "Culture and Cognition" is organized in order to construct a bridge between the two sides. Three speakers are invited and each of them is to talk on cross-cultural studies between Westerners and Easterners. J-B. Van der Henst is to talk from the view of cognitive psychology, whereas T. Masuda and T. Hamamura are to talk from the view of social psychology.

Speakers

Jean-Baptiste VAN DER HENST (CNRS University of Lyon 1, France)
Title:How is inconsistency resolved on belief? : French and Japanese cases.
French and Japanese participants were asked to judge or decide when they received inconsistent information to their supposed believes. Japanese are more likely than French to think dialectically. This is consistent with recent cross-cultural studies of Westerners-Easterners comparison.

Takahiko MASUDA (University of Alberta, Canada)
Title: Cultural variation in attentional regulation between the Japanese and Westerners: An eye-tracking study
We simply presented a blinking circle, and found that the Japanese numbers of fixation points and their average distance from the center were significantly larger than those of Westerners. These results are consistent with the cross-cultural researches that suggests that East Asians are more likely than Westerners to be sensitive to the contextual information.

Takeshi HAMAMURA (University of British Colombia, Canada)
Title:Culture and regulatory focus: Cognitive and motivational implications
We found that Japanese performed better on a task if they were rewarded for not losing points, whereas European-Canadians performed better when they were rewarded for winning points. These results are discussed based on self-esteem and concern for "face".



Workshop 2

Language-specific semantics and categorization

Abstract
Until the early 1990's, the predominant view was that human languages share a significant number of universal properties and that these properties are shaped by universal cognition. In the last few years, however, with evidence of extensive and significant differences across languages particularly in the semantic domain of space, the debate on language and cognition has been re-kindled and has become, once again, a central issue in psychology, linguistics and cognitive science. In this debate, studies on languages such as Korean and Mandarin-Chinese (that are very different from much researched languages like English) have provided crucial information about the extent to which language influences cognition in children and adults. In this workshop, we present three papers showing influence of language on cognition, specifically categorization, with data involving English, Korean, and Mandarin-Chinese. These three papers will be followed by a discussion paper.

Presentation 1: Soonja Choi (San Diego State University)
Presentation 2: Marianella Casasola (Cornell University)
Presentation 3: Twila Tardif, Nina Simms, Chao Liu (University of Michigan)
Presentation 4: Discussion by Laraine McDonough (Brooklyn College and Graduate Center of CUNY)

Presentation 1: Soonja Choi
title:Influence of Language-specific input on spatial categorization: Categories of containment
Recent research on infant cognition has shown that preverbal infants (9-, 11-, and 14-month-olds), regardless of their language environments, can make a distinction between tight-fit and loose-fit containment relations. This distinction is systematically made in Korean (kkita 'fit tightly' vs. nehta 'put something loosely in a container'), but not in English (i.e. in). The questions are, whether this nonlinguistic sensitivity to differences in degree of fit is influenced by the language-specific input as infants learn their target language, and if yes, when the influence begins to take place. Using a preferential-looking method, Choi examined children's nonlinguistic sensitivity to the distinction between Tight-in and Loose -in in English- and Korean-learners at different ages: 18, 24, 29, and 36 months. The children's productive vocabulary was also measured through language survey filled out by the mothers. The preferential data showed a significant decrease in sensitivity in English learners to the differential feature (i. e. degree of fit) between the two relations from 18-/24-month-olds to 29-/36-month-olds. In contrast, 29-/36-month-old Korean learners maintained the distinction between the two relations. The decrease in sensitivity to the tight/loose feature in English learners corresponded with increase in vocabulary level and the production of the word 'in': E children who produced in or had high vocabulary level showed much less sensitivity to the difference between tight-in and loose-in than those who did not produce in or had low vocabulary level. Thus, there is an intimate relation between vocabulary learning and nonlinguistic sensitivities of the relevant domain.

Presentation 2: Marianella Casasola
title:How linguistic input shapes spatial categorization in infants and toddlers
This paper presents the author's recent findings on the effect of teaching a novel word on toddlers' spatial categorization. In this study, English-learning toddlers of 21 to 22 months were taught a novel spatial word for actions resulting in a tight-fit spatial relation, a relation that is lexically marked in Korean (i.e. kkita) but not in English. A second group of toddlers viewed the actions but were not provided with a novel word. Following this training session, toddlers' comprehension of the novel word was tested in a preferential-looking paradigm. Only toddlers who were taught the novel spatial word looked significantly longer at the tight-fit events during the test trials that presented the novel word than during control trials that presented neutral linguistic stimuli. The results indicate that toddlers can map and generalize a novel word onto actions resulting in a tight-fit relation given limited experience with the novel word. The results provide insight into how young word learners begin to form language-specific semantic spatial categories.
A second experiment further documents the influence of linguistic input on infants' spatial categorization. Korean- and English-learning infants of 10 and 18 months were tested on their categorization of containment and tight-fit relations. At 10 months, infants from both language environments formed the same spatial categories (i.e., containment but not tight-fit). At 18 months, however, infants formed only the spatial category consistent with the semantic pattern of their language. Together, the two studies begin to delineate the role of linguistic experience in shaping how infants and toddlers organize spatial events into categories.

Presentation 3: Twila Tardif, Nina Simms and Chao Liu
title:Removing the "Basic" and "Typical" in Noun Categories: ERP and Behavioral Studies with English and Chinese speakers
In response to a number of studies showing that English-speakers use more noun types and that Mandarin-speakers use more verb types when speaking (Tardif, Shatz, & Naigles, 1997; Tardif, in press). This study explores how object labels in Mandarin Chinese and English differ and examines the effects of these differences on categorization tasks. Basic-level labels (e.g., qi4che1 'car', huo3che1 'train', zi4xing2che1 'bicycle') in Chinese often include the superordinate category term (e.g., che1 'vehicle'), whereas in English, basic-level labels do not. Study 1 examined how this difference might affect the use of superordinate and basic level terms by 40-English and 20 Mandarin-speaking adults, and found that Mandarin speakers are more likely than English speakers to replace basic level terms with superordinate level terms. Also, English speakers were more likely than Mandarin speakers to replace basic level terms with other basic level terms from the same category. Study 2 examined typicality effects and basic-level effects in a timed typicality rating task with 27 English-speaking and 24 Mandarin-speaking adults with different types of categories. Across all category types, English speakers showed similar typicality effects for all category types, but Mandarin speakers showed typicality effects only for certain category types. These results were further replicated in Study 3, a yes-no category inclusion ERP task, with both English- and Mandarin-speakers showing significant N400 responses to out-of-category items, but only English-speakers showing N400 responses to atypical category members. The data across all three studies converge to suggest that in addition to the differences in noun and verb use, there may also be strong and significant differences in noun and verb category structures that lead to important differences in the processing of object and action concepts for speakers of these two languages.

Presentation 4: Laraine McDonough.
Discussion
In each of the three presentations, we find that languages differ in the kinds of semantic distinctions are made in the domain of spatial relations (Choi and Casasola) and object labels (Tardiff et al.). These findings urge us to reconsider assumptions we have held over the years about what infants need to know before acquiring language (linguistic precursors), the required flexibility they need to maintain during the acquisition process, and how language then quickly influences how infants structure their experiences in the world as evidenced not only in verbal but nonverbal tasks. These studies impress upon us that the transition from preverbal to verbal thought is a rich arena for researchers to test views on the relation between language and thought. In her discussion, Laraine McDonough will cover some common assumptions made about the preverbal infant mind, first language acquisition and how language influences cognition. Suggestions will be made to elicit discussion among the symposium panel and audience as to how these views can be updated, revised and/or completely changed to accommodate the data presented.



Long symposium

Android science

Organizers:
Karl F. MacDorman, Ph.D. (Cambridge & Osaka University) & Hiroshi Ishiguro , (Osaka University)

Android science exploits the fact that robots that look and act like people can elicit many kinds of social responses from human subjects that only people were able to elicit before. Thus, androids provide a well-controlled experimental apparatus for testing theories about human interaction, cognition, and development; and a test bed for theories about how neural or cognitive processes influence behavior. To make the android act in a way that people see as natural, we have to study the norms of human interaction. But to evaluate theories about how those norms function, we also need to implement them in an android. In that way, androids may prove to be a useful tool in unifying sciences that focus on behavior and those that focus on the neural underpinnings of behavior. We think that embodying cognitive, neural, or social theories in androids sets a higher bar for their evaluation because subjects apply their model of a human Other to androids, which they may not be doing to more mechanical-looking robots. (For more details, please see the call for papers at www.androidscience.com.)

What I would like to propose is that we devote up to three 90-minute CogSci/ICCS-2006 workshop slots to themes related to android science. One slot could be a panel discussion focused on the Uncanny Valley: the perceptual phenomenon that subjects may be more sensitive to imperfections in a robot that closely resembles a human being. The panel would consist mainly of neuroscientists and cognitive psychologists who have performed research in this area. I think Hiroshi might like to propose a theme for a second panel discussion -- for example, on explorations of nonverbal behavior in communication, such as his group's groundbreaking work on the social and cognitive functions of gaze. This panel would include roboticists and social psychologists. Finally, we could consider having an open "call for papers" like we did for the workshop in Stresa, and have the third workshop filled by three 30-minute presentations selected by peer review. Workshop Title: "Influence of culture and language on concepts and cognitive functions"




Workshop 3

Influence of culture and language on concepts and cognitivefunctions

Organizers:
Mutsumi Imai (Keio University at Shonan-Fujisawa, Japan)
Barbara Malt (Lehigh University, USA)

Speakers:

Sandra Waxman (Northwestern University)
Norbert Ross (Vanderbilt University)
Mutsumi Imai(Keio University at Shonan-Fujisawa, Japan)

Discussant:

Barbara Malt

Overview of the panel

Whether and to what extent concepts and basic cognitive functions are universal is of fundamental importance for understanding of the nature of human thought. On the one hand, a shared cognitive architecture might be argued to provide constraints on the nature of knowledge and thought processes that transcend cultural differences. On the other hand, much of human knowledge is acquired through the medium of language and in the context of specific cultural needs and interpretations, and thus language and culture may filter or shape the knowledge acquired and perhaps even the cognitive processes that typically operate on them. In this panel, we present three papers that examine these possibilities from different angles to deepen our understanding of the interplay among universally shared conceptual/cognitive constraints, culture, and language.

Speaker:
Sandra Waxman (Northwestern University)
title:Core knowledge, naming and the acquisition of the fundamental (folk) biologic concept 'ALIVE'
I'll begin by describing (in broad strokes) a collaborative research venture in which we focus on the acquisition of folkbiologic knowledge (e.g., 'normative' cognitive development). Adopting a cross-linguistic, cross-cultural, developmental perspective, we have asked a) what capacities infants and young children bring to the task of acquisition and b) how the environment (including the objects and events that populate the child's world, the language used to describe them, and the cultural practices invoked to highlight them) shapes the process of acquisition. After describing briefly the populations we have thus far included and our research strategy for identifying the contributions of language and culture in the acquisition of knowledge, I'll present evidence from categorization, naming and reasoning tasks in children and adults from Indonesia, Mexico, and the US. This work reveals important commonalities in early development and also illustrates an intimate connection between culture, language and conceptual organization in the evolution of knowledge. In closing, I will highlight the advantages of combining psychological, linguistic and anthropologic methods in developing theories of acquisition.

Speaker:
Norbert Ross (Vanderbilt University)
title:Cultural Aspects in the Development of Children's Understanding of Folkbiology
Much of the work in developmental psychology has focused on detecting universals of human thought. Recent developments in the field, however, make it necessary to extend our research with respect to two aspects: First, in order to understand cognitive universals we need to enlarge our data base beyond the standard populations (white urban middle class children in industrialized countries). Second, in order to understand children's development we have to better understand the context of knowledge acquisition. To do so it is not sufficient to add yet more independent variables (such as language or culture) to our analyses, instead we need to better understand the life context of the children we study. In this paper I will use data from Tzotzil Maya children in Mexico to outline a study of child development that is sensitive to the socio-cultural context in which the children grow up. In this approach culture is not an independent variable, but has more of a distributional character. The focus of the analysis is not on mean comparisons across different independent variables, but on tracing patterns of agreement and disagreement to wider social processes and the resulting background of the individual children. Consequently this approach brings together anthropology and psychology both with respect to topics and theories as well as methodologies.

Speaker:Mutsumi Imai (Keio University at Shonan-Fujisawa)
title:Influence of language, culture, and universal cognitive constraints on the everyday object concepts
Whether and to what extent our conceptual structure is universal is of great importance for our understanding of the nature of human concepts. Two major factors that might affect our concepts are language and culture. In this research, we investigated whether these two factors affect our concepts of everyday objects in any significant ways. Concerning the linguistic factor, we tested whether a grammatical categorization system by classifiers, which classify objects in the world very differently from the way nouns do, influence the conceptual structure of speakers of classifier languages. Concerning the culture factor, we tested a specific proposal by Nisbett and colleagues (e.g., Nisbett, Peng, Choi, & Norenzayan, 2001), which predicts that Westerners organize object concepts around taxonomic relations while Easterners organize them around thematic relations. For this purpose, we tested children and adults speaking two classifier languages, Chinese and Japanese, and those speaking German, a non-classifier language, on a range of tasks including categorization, similarity judgment, label extension, and inductive reasoning. The overall results indicate that the global structure of our concepts of everyday objects is very similar across the three different culture/language groups, while there was much variability in the participants' performance across different tasks. Some effect of the classifier system was found in Chinese but not in Japanese speakers. We discuss how language-specific linguistic properties might affect our concepts, categories, and cognitive processes, and how they interact with task-specific constraints as well as our universal cognitive dispositions.

Discussant:Barbara Malt (Lehigh University)
title:Language, culture, and cognition: Commentary
This talk will discuss the preceding three presentations, drawing from them lessons on when and how language and culture may influence mental representations and processes.



Workshop 4

Internal Observation as a Method for Constructive Science

Organizer: Hideyuki Nakashima (Future University - Hakodate)

Theme of the workshop:
The methodology for analytic Science is established in the 20th Century. It was most successful in the study of physics. However the methodology was not allmighty. For example, experimental psychology that followed the same direction hit its limitation and was gradually replaced by cognitive science. We found that this movement is also universally observed in other areas of science. Some of the examples are seen in artificial intelligence, architecture, biology and geology. We intend to extract similarity in the movements in those areas and establish a new methodology for science, which we call "constructive science" in contrast to "analytic science".

The new methodology we propose is tightly connected to the internal view point of a sytem. Theorist's view, enforced in analytic science, tries to separate a system from its observer so that observation does not interfere with the sytem. We believe this is an effective attitude toward science, but not the only one. When a system is too complex for analytical method, the system must be observed and tested in vivo. Observation cannot be separated from the functionality of the system. If an observer is separated from the system, he or she is detatched from the situation in which the sytem operates. It is not that the observer cannot recognize the situation itself, but that he or she may miss the connection between the system and the situation. We believe that the internal observation, or agents' view (as called in AI), is essential to understand and properly describe situatedness fo the system.

One method of constructive science that is consistent with the notion of "internal observation of a system" is metacognitive verbalization. That is a method in which an individual person, or people forming a community, externalizes by verbalization what she (or they) perceives, thinks and does and how she (or they) moves her (their) own body. The effectiveness of this method has gained growing interest in learning, especially acquisition of embodied expertise. Just as designers externalize still-on-going and thus ambiguous ideas by drawing sketches in order to develop ideas, metacognitive verbalization is not just for record or communication to others but more importantly for development of their own thoughts and body performance.

In this method, a cognitive system and its observer is the same and thus the "objective observation" principle of analytic science is violated. A person's verbalization, as soon as it is being conducted, will influence how her thoughts and perception to the world develop. Verbalization by people forming a community will change their way of thinking, perceiving and interacting with each other. Processes of development of these sorts contain important aspects of human intelligence that the conventional analytical science was unable to capture. We claim that human intelligence lies not just in knowledge obtained by analytical science, but in processes of development revealed by constructive science. The "internal observation" methodology such as metacognitive verbalization provides richer insights into human intelligence.

Utilization of repetitive externalization and internalization as an cycle is important for constructive activities such as design. As the nature of architectural design, for example, it is essential for an architect to use internal as well as external representation of the things and thoughts. The representation and the designer who produces it have reflective relations. The representation is fluid in the sense that the interpretation of it is situated and doesn't remain the same. They are far from objective. It is interesting that successful architects get along with such representation, which is hardly dealt with by analytic science. Consequently, some important parts in architectural design are unleashed as science. Only traditional methods of training may deal with them if constructive science describes the peripheral things surrounding what cannot be represented by 'public'language.

We found that Japanese culture is based on internal observers' view points, and the structure of Japanese language reflects this. In the workshop, we also try to understand the relationship between the view and the language (including the role of "verbalization").

In analytic science, the actor = the theorist, plays an explicit role. In constructive science, the actor is buried in the sytem and thus implicit. Analytic science tries to describe central phenomena as a rule, while constructive science rather focuses on describing peripheral phenomena to make borders clearer.


Related materials are found in writings of following researchres:
Atsunobu Ichikawa (ecologist)
Baron-Cohen (psychologist)
Bin Kimura (psycho-pathologist)
Brian Smith (computer scientist)
Taki Kanaya (linguist)
| Back | This Page Top | Site Top |