Steels, L. (2001b). The methodology of the artificial. Behavioral and brain sciences, 24(6).  Home/Search   Document Not in Database   Summary   Related Articles   Check

....utterances. Why are we alone in this position Human language is a result of three adaptive systems: learning, genetic evolution, and cultural evolution. Over the past half century cognitive scientists has addressed the problem of learning. The past ten years has seen a resurgence of interest in the evolutionary basis of language (Pinker Bloom, 1990). Only recently has the cultural evolution of language been seriously analysed. Hare Elman (1994) outlined perhaps the first iterated learning model. The iterated learning model seeks to model the evolution of language through generations of language users, solely on the basis of each agent ....

The evolution of syntactic communication, Nature, 404, 495-498. Pinker, S., & Bloom, P. (1990). Natural Language and Natural Selection. Behavioral and Brain Sciences, 13.

....neuroscience. This objective implies that subsymbolic connectionism has to address the question of how the symbolic description of psychological processes, in terms of rules and representations, provided by traditional cognitivism can be related to a non symbolic one in terms of brain mechanisms. [Haugeland 1978] has called this the problem of complete reduction. A substantial contribution to this connectionist ambition has been made by [Smolensky 1987, Smolensky 1988] who has tried to explicate the subsymbolic connectionist paradigm. I will evaluate the solution of the problem of complete reduction ....

....that in order to understand cognition knowledge of the algorithms that the brain is supposed to execute is all that matters. This implies that we have to drop the ideal that there is a unity of science in which all sciences are in principle reducible to physics (e.g. Putnam 1960, Fodor 1974, Haugeland 1978]) It The mind brain dilemma 137 is assumed that there is one special science of cognition which develops a formal logical description of the rules and representations constituting it. The level of implementation is considered of minor importance. Winograd and Flores 1986] p.15, describe the ....

Haugeland J.: The nature and plausibility of cognitivism, Behavioral and Brain Sciences 2, 215--260, 1978.

....(e.g. 24, 20, 11, 25, 62] is a young discipline within the biological sciences: it is where ethology (the study of behaviour) meets neuroscience. Its youth puts it in a situation not dissimilar to cognitive science; its precise aims and methods are still a subject of debate (see e.g. [31, 25]) Put most simply, neuroethology is the study of the neural mechanisms of animal behaviour [31, p.384] More specifically, the goal of neuroethology is to relate behaviour to activity within interconnected groups of nerve cells, where behaviour is defined as patterns of movement coordinated in ....

G. Hoyle. The scope of neuroethology. The Behavioral and Brain Sciences, 7:367--412, 1984.

....games the average success varies mostly between 0.96 and 1. Figure 7 shows the total number of feature sets used as a function of the number of games played for both methods. For the Simple Prototype method, the number of feature sets is rather capricious, and does not seem to converge during the 10,000 games that have been observed. With the Adaptive Subspace method the number of feature sets used does converge, and is substantially smaller. 7. Conclusions Two methods for discrimination between objects based on sensor data have been compared. Both methods are based on the principle that an ....

....and is substantially smaller. 7. Conclusions Two methods for discrimination between objects based on sensor data have been compared. Both methods are based on the principle that an agent adapts its categorization of the world to increase its ability to distinguish different objects from each other. The development 0 0.2 0.4 0.6 0.8 1 0 2000 4000 6000 8000 10000 Number of discrimination games Figure 5 Success using the Simple Prototype method. 0 0.2 0.4 0.6 0.8 1 0 2000 4000 6000 8000 10000 Number of discrimination games Figure 6 Success using the Adaptive Subspace method. of discrimination success for both ....

The development of features in object concepts. Behavioral and Brain Sciences, In Press. Steels, L. (1996a). Emergent adaptive lexicons. In Maes, P., editor, From Animals to Animats 4: Proceedings of the Fourth International Conference On Simulating Adaptive Behavior, Cambridge MA. The MIT Press.

....1. We do not pretend at all that robotic experiments model in any realistic way children nor the environments in which they typically operate. Our goal here is to examine specific assumptions about the emergence of communication by building artificial systems, so realism is not at issue [Steels, 2001b] 2. It is an extraordinary challenge to build and maintain physical robots of the required complexity. For practical reasons (limitations of camera resolution, memory and processing power available on board) we cannot always use the best known algorithm available today. This puts limits on what ....

Steels, L. (2001b). The methodology of the artificial. Behavioral and brain sciences, 24(6).

....from the modeled system to the modeling system. terminology trap of worrying whether cognition (or computation for that matter) is really continuous or really discrete; what matters is: which sort of model is better, continuous or discrete, at the appropriate level of analysis 7 (See also MacLennan 1990c. Elsewhere (MacLennan in press c, in press d) I ve proposed a methodological principle aimed at avoiding this trap. It is called the Complementarity Principle, and states that continuous and discrete models should be complementary, that is, an approximately discrete continuous model should make ....

....There is as yet no corresponding theory of universal analog machines, though various notions of universality have been proposed and are being explored along with the related computability questions (e.g. Blum 1989; Blum al. 1988; Franklin Garzon 1990; Garzon Franklin 1989, 1990; MacLennan 1987, 1990d, in press a, in press b; PourEl Richards 1979, 1981, 1982; Stannett 1990; Wolpert MacLennan submitted) 3.2.8 Formality. Each variable in the equations may represent a physical quantity or a pure number. In conventional terms, each variable has an associated dimension. 15 If all the ....

[Article contains additional citation context not shown here]

MacLennan, B. J. (1990c) The discomforts of dualism. Behavioral and Brain Sciences 13: 673--674.

No context found.

D. Premack. The codes of man and beast. Behavioral and Brain Sciences, 6:125--167, 1983.

No context found.

Kupfermann I., Weiss K.R.: The command neuron concept, Behavioral and Brain Sciences 1, 3--39, 1978.

No context found.

b Over the rainbow: The classification of unique hues. Behavioral and Brain Sciences 20 (2): 204-205. Miller, George A. and P. Johnson-Laird

No context found.

On the proper treatment of connectionism. Behavioral and Brain Sciences 11, 1-23. Squire, L. R.

No context found.

D. Premack. The codes of man and beast. Behavioral and Brain Sciences, 6:125--167, 1983.

Online articles have much greater impact   More about CiteSeer.IST   Add search form to your site   Submit documents   Feedback  

CiteSeer.IST - Copyright Penn State and NEC