Quartz, S. & Sejnowski, T.J. (1997). The neural basis of cognitive development: A constructivist manifesto.

“Origins of communicative disorders ” to Elizabeth Bates, and by a grant from the John D. and Catherine T. MacArthur Foundation. We are grateful to Larry Juarez and Meiti Opie The effects of focal brain injury are investigated in the first stages of language development, during the passage from first words to grammar. Parent report and/or free speech data are reported for 53 infants and preschool children between 10- 44 months of age. All children had suffered a single, unilateral brain injury to the left or right hemisphere, incurred before six months of age (usually in the pre- or perinatal period). This is the period in which we should expect to see maximal plasticity, but it is also the period in which the initial specializations of particular cortical regions ought to be most evident. In direct contradiction of hypotheses based on the adult aphasia literature, results from 10- 17 months suggest that children with righthemisphere injuries are at greater risk for delays in word comprehension, and in the gestures that normally precede and accompany language onset. Although there were no differences between left- vs. right-hemisphere injury per se on expressive language, children whose lesions include the left temporal lobe did show significantly greater delays in expressive vocabulary and

Introduction Language is puzzling. On the one hand, there are compelling reasons to believe that the possession of language by humans has deep biological roots. We are the only species that has a communication system with the complexity and richness of language. There are cases of non-human primates who can be taught (sometimes only with heroic effort) some aspects of human language, but their performance comes nowhere close to those of a six-year old child. Second, although languages differ, but there are also striking similarities across widely divergent cultures. Finally, there are significant similarities in the patterns of language acquisition across very different linguistic communities. These (and other considerations as well) all suggest that species-specific biological factors play a critical role in human's ability to acquire and process language. So what is puzzling? First, it is not at all clear what the biological foundations are. What precisely do we mean when w

A constructivist neural network is presented that models the acquisition of the past tense of English verbs. The network constructs its architecture in response to the learning task, corresponding to neurobiological and psychological evidence. The model outperforms other connectionist and symbolic models in learning and in displaying psychologically realistic learning and generalization behavior. It is argued that the success of the network is due to its constructivist nature, and that the distinction between fixed architecture and constructivist models is fundamental. Given this distinction, constructivist systems constitute better models of cognitive development. 1. Introduction The acquisition of the English past tense has in the past years become a touchstone for different theories of language acquisition and of cognition in general. Different theories and models have not only been used in the debate between proponents of symbolic and connectionist accounts of language learning, ...

Abstract: Modularity in the human brain remains a controversial issue, with disagreement over the nature of the modules that exist, and why, when and how they emerge. It is a natural assumption that modularity offers some form of computational advantage, and hence evolution by natural selection has translated those advantages into the kind of modular neural structures familiar to cognitive scientists. However, simulations of the evolution of simplified neural systems have shown that, in many cases, it is actually non-modular architectures that are most efficient. In this paper, the relevant issues are discussed and a series of simulations are presented that reveal crucial dependencies on the details of the learning algorithms and tasks that are being modelled, and the importance of taking into account known physical brain constraints, such as the degree of neural connectivity. A pattern is established which provides one explanation of why modularity should emerge reliably across a range of neural processing tasks.

A constructivist neural network model is presented that learns the past tense of English verbs. The model builds its architecture in response to the learning task in a way consistent with neurobiological and psychological evidence. The model outperforms existing connectionist and symbolic past tense models in terms of learning and generalization behavior, and it displays a U-shaped learning curve for many irregular verbs. The trained model develops a modular architecture with dissociations between regular and irregular verbs, and lesioning the different pathways leads to results comparable with neurological disorders. It is argued that the successof the model is due to its constructivist nature, and that the distinction between fixedarchitecture and constructivist models is fundamental. Given this distinction, constructivist systems provide more realistic models of cognitive development. Introduction Models of learning the English past tense have in the past ten years become represent...

It seems to be a common feeling that animals learn to see, and this feeling, together with the reemergence of computer learning paradigms that mimic many forms of human learning, has raised hopes that learning is the key to the computer vision problem. Indeed, it seems clear that Nature does not "program" all our visual capabilities into the genome, and we certainly know that programming a computer with a closed-form solution to the vision problem is a daunting task. This aim of this informal and elementary report (basically a term paper) is to cast doubt on the idea that biological systems learn to see. The complex process of development, beginning at fertilization and ending with a mature individual, could be considered to have genetic ("nature") and learning ("nurture") processes as logical endpoints or opposite poles. This report mostly considers what goes on between those endpoints, and is meant to raise the possibility that some of the least understood processes in biology are re...

Abstract. A viable evolutionary cognitive psychology requires that specific cognitive capacities be (a) heritable and (b) ‘quasi-independent ’ from other heritable traits. They must be heritable because there can be no selection for traits that are not. They must be quasi-independent from other heritable traits, since adaptive variations in a specific cognitive capacity could have no distinctive consequences for fitness if effecting those variations required widespread changes in other unrelated traits and capacities as well. These requirements would be satisfied by innate cognitive modules, as the dominant paradigm in evolutionary cognitive psychology assumes. However, those requirements would also be satisfied by heritable learning biases, perhaps in the form of architectural or chronotopic constraints, that operated to increase the canalization of specific cognitive capacities in the ancestral environment (Cummins and Cummins 1999). As an organism develops, cognitive capacities that are highly canalized as the result of heritable learning biases might result in an organism that is behaviourally quite similar to an organism whose innate modules come on line as the result of various environmental triggers. Taking this possibility seriously is increasingly important as the case against innate cognitive modules becomes increasingly strong.

This paper shows how the relationship between two arrays of artificial neurons, representing different cortical regions, can be learned. The algorithm enables each neural network to self-organise into a topological map of the domain it represents at the same time as the relationship between these maps is found. Unlike previous methods learning is achieved without a separate training phase

Fodor (1983) claims that the modularity of mind (the relatively encapsulated, insulated, special-purpose nature of the psychological mechanisms of perception) helps undermine relativism in various forms. I shall show first, that the modular vision of mind provides insufficient support for the rejection of (most forms of) relativism, and second, that an alternative (‘neural constructivist’) model may, in fact, provide a better empirical response to the relativist challenge.