Invited Speakers AT AIIB
Submitted papers and poster abstracts are listed here.
The schedule for the symposium is here.
Post symposium report for UKCRC
Most human infants, by 12 months of age, have developed a suite of triadic engagement skills, specifically, joint attention, social referencing, and intentional communication (e.g., Russell et al., 1997). Developmental psychologists emphasised the cognitive underpinnings of how infants become able to coordinate their engagement with people and their engagement with objects to form the referential triangle of infant- mother-object (triadic) engagement (e.g., Bard, 1992). Developmental psychology provided the conceptual frame for many studies of comparative cognition, and some comparative psychologists have asserted that this suite of triadic engagement skills specifies the beginning of the uniquely human 'social cognitive revolution'. New evidence emerging from the FEELIX GROWING project (Canamero, 2008: Hiolle et al., 2009) highlights the importance of emotion (e.g., Murray et al., 2009), and suggests a simple learning system can account for the development of triadic engagements (e.g., Boucenna et al, 2010). Evidence emerging from my developmental comparative studies (e.g., Bard & Leavens, 2009) and those of my colleagues (e.g., Leavens et al., 1996, 2005) indicate that emotion plays a crucially important role in the development of social cognition in young apes, as well (Bard, in press; Bard et al., 2005; Leavens et al., in press).
Bard, K.A. (1992). Intentional behavior and intentional communication in young free-ranging orangutans. Child Development, 63, 1186-1197.
Bard, K.A. (in press). Emotional engagement: how chimpanzee minds develop. To appear in de Waal, F. & Ferrari, P. (eds), The Primate Mind. Cambridge, MA: Harvard University Press.
Bard, K.A. & Leavens, D.A. (2009). Socio-emotional factors in the development of joint attention in human and ape infants. In L. Roska- Hardy & E. Neumann-Held (Eds.), Learning from animals? Examining the nature of human uniqueness (pp. 89-104). London: Psychology Press.
Bard, K.A. Leavens, D.A., Custance, D., Van?atova?, M., Keller, H., Benga, O., & Sousa, C. (2005). Emotion cognition: Comparative perspectives on the social cognition of emotion. Cognitie, Creier, Comportament (Cognition, Brain, Behavior), Special Issue: "Typical and atypical development", VIII, 351-362.
Boucenna, S., Gaussier, P., Hafemeister, L., & Bard, K. (submitted, Feb 2010). Social referencing. To appear in From Animals to Animats 11: Eleventh international conference on simulation adaptive behavior, Lecture notes on Computer Science, Springer.
Canamero, L. (2008). Development within the "Complete Creature" Paradigm. IEEE AMD Newsletter - The Newsletter of the Autonomous Mental Development Technical Committee, Vol. 5, No. 2, November 2008.
Hiolle, A., Bard, K.A., & Canamero, L. (2009). Assessing human reactions to different robot attachment profiles. Proceedings of the 18th IEEE International Symposium on Robot and Human Interactive Communication Toyama, Japan, (RO-MAN 09), Sept. 27-Oct. 2, 2009, p.251- 256.
Leavens, D.A., Bard, K.A. & Hopkins, W.D. (in press). BIZARRE chimpanzees do not represent "The Chimpanzee". Commentary on target article by Heinrich, Heine & Norensayan, 'The weirdest people in the world?', Behavioral and Brain Sciences, 33(1).
Leavens, D.A., Hopkins, W.D., & Bard, K.A. (1996). Indexical and referential pointing in chimpanzees (Pan troglodytes). Journal of Comparative Psychology, 110, 346-353.
Leavens, D.A., Hopkins, W.D., & Bard, K.A. (2005). Understanding the point of chimpanzee pointing: Epigenesis and ecological validity. Current Directions in Psychological Science, 14, 185-189.
Murray, J. C., Canamero, L., Bard, K. A., Davila Ross, M., and Thorsteinsson, K. (2009). The Influence of Social Interaction on the Perception of Emotional Expression: A case study with a robot head. In J.H. Kim et al (eds); FIRA 2009, Lecture Notes in Computer Science 5744, pp 63-72. Heidelberg, Berlin:Springer-Verlag.
Russell, C.L., Bard, K.A., & Adamson, L.B. (1997). Social referencing by young chimpanzees (Pan troglodytes). Journal of Comparative Psychology, 111, 185-193.
It has been standard in linguistics to argue that language acquisition is so difficult that it is only possible if most linguistically interesting structure is embedded in a genetically-specified "language module," "language instinct," or "universal grammar." But how could a biological basis for language have evolved, in the absence of a prior linguistic environment which could shape biological evolution. A natural, and popular, assumption is that the biological evolution of a language faculty must have co-evolved with the cultural evolution of language itself. Here I report work with Morten Christiansen, Florencia Reali, Andrea Baronchelli and Romualdo Pastor-Satorras, which places strong limitations on such co-evolution. Indeed, these simulations suggest that the traditional conception of a "universal grammar" can be ruled out on evolutionary grounds; and may have broader indications for claims that humans have biologically based adaptations to other aspects of culture.
It is now commonly accepted that the motor system makes use of so-called forward and inverse models in order to control the musculoskeletal system during rapid, skilled, motor behaviour. Inverse models are held to allow the system to determine the motor commands necessary to achieve a desired state, while forward models are held to allow the system to predict the expected sensory feedback of a motor command, allowing rapid error detection when actual and predicted feedback do not match. It has recently been suggested that these ideas from control theory might also be applied to the control of cognitive processes, allowing (for example) the cognitive system to anticipate processing conflict and pre-emptively minimise it by adjusting processing strategies or the allocation of processing resources. This paper reviews theories of cognitive control that are broadly consistent with the use of complementary forward and inverse models. It is argued that there is indeed a role for such models in cognitive control -- particularly in relation to a putative monitoring function -- but that the models involved are likely to be somewhat impoverished.
There is great interest in building intrinsic motivation into artificial systems using the reinforcement learning framework. Yet, what intrinsic motivation may mean computationally, and how it may differ from extrinsic motivation, remains a murky and controversial subject. We adopt an evolutionary perspective and define a new optimal reward framework that captures the pressure for good primary reward functions that lead to evolutionary success across environments. The emergent optimal reward functions thereby convert distal pressures on fitness into proximal pressures on behavior. The results of several computational experiments in this framework yield two interesting outcomes. First, optimal reward signals are adapted to both the internal structure of the agent and the external structure of the environment: the same fitness function and the same environment may yield different optimal reward for different agents. Second, optimal primary reward signals yield both emergent intrinsic and extrinsic motivation, leading to the conclusion that there are no hard and fast features distinguishing intrinsic and extrinsic reward computationally. Rather, the directness of the relationship between rewarding behavior and evolutionary success varies along a continuum. (This is joint work with Satinder Singh and Jonathan Sorg at University of Michigan, and Andrew Barto at University of Massachusetts.)
See alsoBounded optimality and bounded rational analysis
Our work on bounded optimality and bounded rationality provides a new way to model and explain human behavior by leveraging its adaptive nature.
In recent years, psycholinguistic theories have become increasingly concerned with the relationships between domain-specific linguistic and general cognitive processes. In my talk I will (a) discuss some of the relationships between vision, attention, and linguistic processing, and (b) demonstrate how eye tracking can be used to study these relationships.
In the first part of the talk I will review studies of spoken language comprehension demonstrating that the listener's eyes tend to be drawn to objects mentioned or implied in spoken utterances. I will propose that listeners aim to generate conceptual representations that, as much as possible, integrate visual and auditory information, that this process is facilitated by directing visual attention at the named or implied objects, and that the direction of visual attention is reflected in the eye movements.
In the second part of the talk I will review studies of language production demonstrating that speakers describing displays or events tend to fixate upon each object they name, and that their gaze remains focussed on the object until they have retrieved the sound form of its name. I argue that the tight link between eye gaze and speech output and the long gazes to the objects arise because directing one's visual attention at an object facilitates not only the recognition of the object but also the retrieval of associated information, including the object name. The easiest way of direction one's attention to an object is to fixate upon it.
In final part of the talk I will summarise some of the strengths and weaknesses of eye tracking as a tool for psycholinguistic research and suggest directions for future research.
Neurons engage in causal interactions with one another and with the surrounding body and environment. Neural systems can therefore be analyzed in terms of causal networks, without assumptions about information processing, neural coding, and the like. Causal networks therefore provide a lingua franca for translating insights obtained from computer simulations or robots into predictions and interpretations relevant to biological systems. My talk will be in three parts. First, I will describe the theoretical basis for 'causal network analysis' which rests on a combination of time-series analysis ("Granger causality") and network theory. I will then describe its application to a series of "brain-based devices", drawing implications for possible causal pathways in the hippocampus and for the relation between synaptic plasticity and behavioural learning. Finally, I will explore how causal networks shed new light on the possible biological mechanisms underlying conscious experience, with reference to the nascent project of 'artificial' or 'machine' consciousness.
Seth, A.K. (2009). The strength of weak artificial consciousness. (PDF) International Journal of Machine Consciousness.
Seth, A.K. (2008). Causal networks in simulated neural systems. (PDF) Cognitive Neurodynamics, 2:49-64.
In recent years the cognitive prowess of corvids has been amply demonstrated. Despite their evolutionary distance from primates and the correspondingly different organisation of their brains, corvids are capable of tool manufacture and use, physical and causal cognition, social cognition and deception, and mental time travel. But how are these capacities realised in their brains? A satisfactory answer to this question would surely meet the criterion of implementability. If we really knew what makes a crow tick, we would be able to build a clockwork crow. Conversely, if we could build a clockwork crow, then we might be in a position to advance hypotheses about what makes a real crow tick. But today's robots fall a long way short of corvid-level intelligence. In this respect, the prospects for AI-inspired biology look as forlorn as those for biologically-inspired AI. Perhaps the way to fill the explanatory hole is to look to the emerging field of neurodynamics for the right theoretical vocabulary.
Subset of Presentation
(Some pre-publication images have been removed. Anyone wanting the full set of slides should email the author.)
list of AISB2010 Convention plenary speakers
She has provided a paper relevant to the symposium here.
Are Animals Stuck in Time or Are They Chronesthetic Creatures?
N. S. Clayton, J. Russell, A. Dickinson
See also: bird tango (the movie).
Last updated: 24 Feb 2010; 27 Feb 2010;3 Mar 2010; 6 Mar 2010; 18
Mar 2010; 20 Mar 2010; 9 Apr 2010;15 Apr 2010; 23 Apr 2010
Installed: 24 Feb 2010
Maintained by Aaron Sloman
School of Computer Science
The University of Birmingham