Last update: Mar 5, 1997
Larry Cauller NeuroInteractivism: Explaining emergence without representationRecently established principles of neural connectionism promote a neurointeractivist paradigm of brain and behavior which emphasizes interactivity between neurons within cortical areas, between areas of the cerebral cortex, and between the cortex and the environment. This paradigm recognizes the closed architecture of the behaving organism with respect to motor/sensory integration within a dynamic environment where the majority of sensory activity is the direct consequence of self-initiated motor actions. We have found that top-down cortical inputs to primary sensory areas which predict discrimination behavior in monkeys selectively activate the cortico-bulbar neurons that mediate directed movements. Unlike the widely distributed axons and long-lasting excitatory synaptic effects of the top-down projections which generate the associative context for motor/sensory interactivity, the bottom-up sensory projections are spatially precise and activate a brief excitation followed by a long-lasting inhibition. Therefore, the sensory consequences of a motor action are the major sources of negative feedback which complete an interactive cycle of associative hypothesis testing: a winner-take-all motor/sensory pattern initiates an action within a top-down associative context; the bottom-up sensory consequences of that action interfere with top-down sensory predictions for refinement of the associative hypothesis; then the testing cycle repeats as the sensory inhibition releases the next motor/sensory winner-take-all action. Given this interactivity, perception is a proactive behavior rather than information processing, neurons simply respond to their inputs rather than encode sensory properties, so there is no need to impose representationalism because neural activity patterns are dynamical attractors associating sensory predictions with motor actions. The associative hypothesis is the neurointeractive equivalent to awareness and hypothesis testing is the basis for attention. The formation of action/prediction associations by interactivity explains early development from self-organized cortical attractor spaces in utero, to the emergence of self-identity in the newborn which learns to predict the immediate effects of self-action (i.e., listening to its own speech sounds), to the emergence of recognition by prediction of ecological contingencies, to the emergence of speech by prediction of motherís responses to infant speech. Ultimately, our scientific paradigm likewise emerges by neurointeractivity as we learn to see the world in a way that explains more of the effects of our actions.
Georg Dorffner The connectionist route to embodiment and dynamicismThis talk discusses the possible contributions of connectionism to new trends in cognitive science, in particular to cognitive models that focus on issues of embodiment, and on viewing cognition from a dynamical point of view. It will be argued that connectionism has had its fair share in driving cognitive science towards new horizons, and that its potentials are far from being fully exploited -- despite the fact that connectionism has seem to become an established subfield of cognitive science more in the realm of classical encoding theories.
The most important contribution of connectionism is its provision of concrete modeling frameworks that help in replacing traditional conceptions of representation by an approach that views representation in the constructivist sense of behavior-guiding structures, which are inextricably tied to an agent's bodily actions in an environment. Although hardly exploited in recent models, taking connectionism seriously almost naturally leads to such a novel approach. This talk will discuss the major reasons why this is so.
A second contribution becomes obvious when one realizes that connectionism again provides a natural basis for shifting one's view of cognition from a computational to a dynamical point of view (such as exemplified in the work of Tim van Gelder). Concrete issues such as complexity in dynamic state spaces -- and why one would want to realize them in a connectionist network -- will be discussed.
As a concrete example, the issue of categorization will be presented. Categorization can be viewed as an essential component for concept formation in connectionist models. It can also be seen as an important mechanism enhancing the capabilities of autonomous agents. Finally it can be viewed as the process of inducing certain kinds of attractors in a dynamic state space. Thus, categorization is one example of how results from more traditionally applied connectionism, results from work on embodied autonomous agents, and results from work on dynamical approaches to cognitive modeling can be brought together.
Ernst von Glasersfeld Piaget's legacy: Cognition as adaptive activityIn the visual arts, "representation" usually means a copy or reproduction of some original. In that context it is clear that the original is always something the representer has seen, something that is the product of ordinary visual perception. With the introduction of the term in philosophical writings, the spurious question has arisen whether or not representations could reproduce, replicate, or correspond to things-in-themselves. The question was long ago given a negative answer on logical grounds by neurophysiology. Most arguments on the topic could have been avoided if one had followed Mark Baldwin, the pioneer of cognitive psychology, and had used the term "representation" which has the added advantage of being a viable translation of the German "Vorstellung".
Stevan Harnad Keeping a grip on the real/virtual distinction in this representationalist ageRepresentation is a three part relation. There is (1) an object (the "representation"), plus (2) whatever the object is a representation of, plus (3) whoever the object is a representation to. Only cognitive scientists have to worry about all three of these parts. Other disciplines, such as archeology and theology can concern themselves only with one or two of them. Cognitive scientists deal with internal representations (hence the object is inside a brain or a machine) and they must avoid the symbol grounding problem, which is to dub the internal object a representation merely because it represents something to the cognitive scientist. Current undergraduates, motherboard-bred, are losing their grip on the distinction between the virtual and the real, ready to believe such incoherencies as that they themselves might merely be objects in someone else's virtual world. The best cure is to abandon representationalist talk and mentalism altogether, and focus only on the internal structures and processes that make it possible to do what robotic models can do. Leave the question of whether or not they mean something till the work of scaling up to our full robotic capacity is complete.
Wolf Singer The observer in the brainConsciousness has many different connotations some of which are amenable to treatment within neurobiological description systems while others are not. It is possible to define in neurobiological terms the brain states associated with consciousness. It is also conceivable that neurobiology will ultimately provide a reductionistic explanation of mechanisms which enables the brain 1.) to construct from the sparse and diverse signals of its sensors coherent models of its environment, including the organism itself, and to generate abstract descriptions, 2.) to iterate the same strategy to monitor its own states, thereby generating meta descriptions, 3.) to weigh the combined results of these analyses in order to reach decisions and to generate adapted behavioural responses, and 4.) to communicate through various channels at different levels of abstraction the results of these cognitive processes to other brains. Since it became clear that the concept of the Cartesian theatre is untenable, that processes in the brain are highly distributed and that there is no single convergence center where the results of the numerous parallel operations are brought together for joint interpretation and decision making, analysis of processes that are in principle amenable to neurobiological explanation is in itself a major challenge.
Problems of different nature are encountered if one attempts a reductionistic explanation of the subjective connotations of consciousness, the qualia and self-awareness, attributes that are assessed by introspection and by extrapolation from one's own awareness of mental states to that of others. I shall defend the position that these aspects of consciousness cannot be understood as emergent properties of individual brains alone but come into existence only through communication among brains whose cognitive abilities must be sufficiently developed to generate a theory of mind, i.e. to generate models of presumed states of the respective other brain. Thus, self-awareness and the ability to experience qualia as subjective reality would have to be considered as cultural achievements or, and this is equivalent, as the result of experiencing the possibility of dialogues of the kind: "I know that you know that I know". Hence, these aspects of consciousness come into existence only through a social learning process in which brains experience a class of mental phenomena that emerge only from mutual reflection. These phenomena are ontologically different from those qualified above as amenable to direct neurobiological investigation because unlike the latter they are the result of a dialogue among brains that got increasingly refined during cultural evolution. This is probably the reason why these phenomena appear as not deducible from analysis of individual brains in the same way as one can analyse the neuronal substrate of pattern recognition, memory or motor control. My proposal is that the phenomena that give rise to the so called "hard problems" in the philosophy of consciousness, the self-awareness and the qualia, can be understood as emergent properties of brains without having to take a dualistic position; however, because these phenomena have a social or cultural origin and hence both a historical and interpersonal dimension, they cannot be understood as an emergent property of an isolated brain alone and hence transcend the reach of conventional neurobiological approaches.
Sverre Sjölander How animals handle reality: the adaptive aspect of representationA comparative look at the animal kingdom reveal that there are as many animal realities as there are basic animal types. A borderline between simple reactions and higher cognitive functions cannot be drawn, instead we find a plethora of ways of relating to reality, where the common denominator is that the methods of representation allow the animal to act in an adaptive way. Animal representation does not need reality - but it is a way of doing something to reality which in the long run ensures survival and reproduction. If reality is represented this way or that way, or at all, is of no importance, as long as the animal can act adaptively. It is even questionable if the concept of representation is helpful in understanding the cognition of lower animals, where interaction with reality merely consists of preprogrammed reactions devoid of understanding. Nevertheless, a focus on representation as a biological adaptation is helpful when discussing our own way of relating to reality, since the mechanisms at work must have evolved in a stone age ecology, geared at living in that particular form of reality. Some of our more notorious failures at handling our present-day world may be more easy to understand if this heritage is taken into account.