This paper discusses the concept of joint attention and the di#erent skills underlying its development. We argue that joint attention is much more than gaze following or simultaneous looking because it implies a shared intentional relation to the world. The current state-of-the-art in robotic and computational models of the di#erent prerequisites of joint attention is discussed in relation with a developmental timeline drawn from results in child studies.

in parsing

This paper explores the hypothesis that pointing gesture recognition can be learned using a reward based system. An experiment with two four-legged robots is presented.

Human children 18 – 24 months of age and 3 young chimpanzees interacted in 4 cooperative activities with a human adult partner. The human children successfully participated in cooperative problem-solving activities and social games, whereas the chimpanzees were uninterested in the social games. As an experimental manipulation, in each task the adult partner stopped participating at a specific point during the activity. All children produced at least one communicative attempt to reengage him, perhaps suggesting that they were trying to reinstate a shared goal. No chimpanzee ever made any communicative attempt to reengage the partner. These results are interpreted as evidence for a uniquely human form of cooperative activity involving shared intentionality that emerges in the second year of life. From soon after birth, human infants interact with other persons dyadically in coordinated, turn-taking sequences (Trevarthen, 1979). From about 6 to 9 months of age, infants ’ social interactions become more complex, as they often incorporate outside objects and so become triadic (Tomasello, 1995). Some of these triadic interactions are relatively extended and maintain a turn-taking structure, for example, rolling a ball back and forth or taking turns beating a drum (Gustafson, Green, & West, 1979; Ratner & Bruner, 1978). Most of these early triadic interactionsFsometimes called cooperative gamesFseem to rely on adult scaffolding in fairly ritualized situations, because the introduction of novel toys or a peer partner disrupts them almost totally until 18 months of age (Hay, 1979; Ross, 1982). In a series of longitudinal studies, Eckerman and colleagues have investigated the emergence of young children’s skills in cooperative games of a less ritualized

Abstract: It is widely believed that what distinguishes the social cognition of humans from that of other animals is the belief-desire psychology of four-year-old children and adults (so-called theory of mind). We argue here that this is actually the second ontogenetic step in uniquely human social cognition. The first step is one year old children’s understanding of persons as intentional agents, which enables skills of cultural learning and shared intentionality. This initial step is ‘the real thing ’ in the sense that it enables young children to participate in cultural activities using shared, perspectival symbols with a conventional/normative/reflective dimension—for example, linguistic communication and pretend play—thus inaugurating children’s understanding of things mental. Understanding beliefs and participating in collective intentionality at four years of age—enabling the comprehension of such things as money and marriage—results from several years of engagement with other persons in perspective-shifting and reflective discourse containing propositional attitude constructions. By all appearances, the cognitive skills of human beings are very different from those of other animal species, including our nearest primate relatives. Human

Abstract. In this paper we present a computational model of lexicon formation that extends the nature of form-meaning associations from previous language game models to be more “flexible”. The model more easily copes with complexity and it better captures the flexibility of form-meaning associations found in human languages. Experiments are conducted using embodied agents playing situated language games. We present and discuss some interesting properties of the emergent dynamics exhibited by our model: first, the model seems to be able to handle some difficulties earlier models struggled with. Second, meanings of words can shift between being very specific (names) and general (e.g. thing). And third, not the model itself but the distribution of object properties in the world biases the specificity of words. 1

Abstract. This paper is part of an ongoing research program to understand the cognitive and functional bases for the origins and evolution of spatial language. Following a cognitive-functional approach, we first investigate the cross-linguistic variety in spatial language, with special attention for spatial perspective. Based on this language-typological data, we hypothesize which cognitive mechanisms are needed to explain this variety and argue for an interdisciplinary approach to test these hypotheses. We then explain how experiments in artificial language evolution can contribute to that and give a concrete example. 1

Abstract — In the communicative eye gaze such as joint visual attention and gaze alternation, the importance of intentional agency is pointed out. Namely, infants understand others’ intentions and also intentionally gaze at others ’ gazing points. In this paper, for investigating the development of the intentional agency that the infant separates desired goals from the actions, we constructed a computational model that is able to acquire the ability of the gaze alternation to an object outside the visual field based on the intentional agency. In the construction and operation of the model, we confirmed that two functions play an important role in the visual joint attention. The one function performs discrimination between the caregiver and the object, the other does association evoked by visual stimuli by storing the relations between sensory states during looking at the caregiver’s face and that during looking at the objects as a frequency distribution. We discussed the process of the joint visual attention from the viewpoint of the intentional agency. Based on the discussion, we claims that the association can be thought of as a goal that is possessed as the agent’s internal state and the movement of the gaze point can be thought of as an action to achieve the goal.

It has been pointed out that gaze alternation by infants, which is a basis for social communication, is related to the process of the development of intentional agency. Intentional agency is defined as an act with a desired goal and a means. It has been pointed out that infants understand others ’ intentions based on intentional agency. In our recent work, we constructed a computational model which acquired gaze alternation behavior with elementary intentional agency. In this paper, through the analysis of the behavior and the internal states of the constructed model, we confirm that two mechanisms, discrimination and accumulation, play an important role in developing elementary intentional agency. The former discriminates between a caregiver and objects that are producing stimuli. The latter associates the caregiver with the objects by accumulating relationships between sensory states. We also argue that a nested structure, in which gaze alternation composed of a goal and a means is utilized as a means to achieve another goal, is an important feature in the development of intentional agency.

Abstract. Over the past several decades, psycholinguists have gained countless insights into the process of child language acquisition. Can these findings be used for the development of language competence in autonomous artificial systems? This paper reports on our attempt to apply insights from developmental psychology in order to enable artificial systems to acquire language. We consider a comprehensive chain of computational processes, starting from conceptualization and extending through language generation and interpretation, and show how they can be intertwined to allow for acquisition of complex aspects of grammar. 1