Abstract not found

In classic generative phonology, linguistic competence in the area of sound structure is modeled by a phonological grammar. The theory takes a grammatical form because it posits an inventory of categories

Standard feedforward and recurrent networks cannot support strong systematicity when constituents are presented as local input/output vectors (Phillips, 1998). To explain systematicity connectionists must either: (1) develop alternative models; or (2) justify the assumption of similar (non-local) constituent representations prior to the learning task. I show that the second commonly presumed option cannot account for systematicity, in general. This option, termed first-order connectionism, relies upon established spatial relationships between common-class constituents to account for systematic generalization: Inferences (functions) learned over, e.g., cats extend systematically to dogs by virtue of both being nouns with similar internal representations so that the function learned to make inferences employing one simultaneously has the capacity to make inferences employing the other. But, humans generalize beyond common-class constituents. Cross-category generalization (e.g., inferences that require treating mango as a colour, rather than a fruit) makes having had the necessary common context to learn similar constituent representations highly unlikely. At best, the constituent similarity proposal encodes for one binary relationship between any two constituents, at any one time. It cannot account for inferences, such as transverse patterning that require identifying and applying one of many possible binary constituent relationships that is contingent on a third constituent (i.e., ternary relationship). Connectionists are, therefore, left with the first option which amounts to developing models with the symbol-like capacity to explicitly represent constituent relations independent of constituent contents, such as in tensor-related models. However, rather just simply impl...

Fodor and Pylyshyn argued that connectionist models could not be used to exhibit and explain a phenomenon that they termed systematicity, i.e., compositional syntax and semantics for mental representations and structure sensitivity of mental processes. This inability, they argued, was particularly serious since it meant that connectionist models could not be used as alternative models to classical symbolic models to explain cognition. In this paper, a connectionist model is used to identify some properties which show that connectionist networks supply means for accomplishing a stronger version of systematicity than Fodor and Pylyshyn opted for. Specifically, it is argued that context-dependent systematicity is achievable within a connectionist framework. The arguments put forward rest on a particular formulation of content and context of connectionist representation, firmly and technically based on connectionist primitives in a learning environment. The perspective is motivated by the fundamental differences between the connectionist and classical architectures, in terms of prerequisites, lower-level functionality and inherent constraints. 2 1

Brain/mind modularity is a contentious issue in cognitive science. Cognitivists tend to conceive of the mind as a set of distinct specialized modules and they believe that this rich modularity is basically innate. Cognitivist modules are theoretical entities which are postulated in "boxes-and-arrows" models used to explain behavioral data. On the other hand, connectionists tend to think that the mind is a more homogeneous system that basically genetically inherits only a general capacity to learn from experience and that if there are modules they are the result of development and learning rather than being innate. In this chapter we argue for a form of connectionism which is not anti-modularist or anti-innatist. Connectionist modules are anatomically separated and/or functionally specialized parts of a neural network and they may be the result of a process of evolution in a population of neural networks. The new approach, Evolutionary Connectionism, does not only allow us to simulate how genetically inherited information can spontaneously emerge in populations of neural networks, instead of being arbitrarily hardwired in the neural networks by the researcher, but it makes it possible to explore all sorts of interactions between evolution at the population level and learning at the level of the individual that determine the actual phenotype. Evolutionary Connectionism shares the main goal of Evolutionary Psychology, that is, to develop a psychology informed by the importance of evolutionary process in shaping the inherited architecture of human mind, but differs from Evolutionary Psychology for three main reasons: (1) it uses neural networks rather than cognitive models for interpreting human behavior; (2) it adopts computer simulations for testing evolutionary scenarios; (3) it has a less pan-adaptivistic view of

Abstract not found

from ungrammatical sentences given appropriate task structure: reply to Marcus

2> j -- output j ) * input i b. Since no input is ever provided for input position 4, d 4j will always be zero. B. Distributed role vectors i. Critique: There was no reason the network should have generalized to position 4. Based strictly on the training data, there is no reason to assume that the systematic output is more correct than the associationist output. Assumptions that such generalization should occur are interpreter--imposed. Suppose the first three units above represented the letters that made up some word, while the last unit represented the grammatical class of the same word. One would not assume that knowledge about mapping letters should transfer to the mapping of grammatical class information. We, the interpreters, are using more information in evaluating the network than we are providing it with. ii. What if the network was "told" that the untrained position was similar to the trained positions? The network

The Syntagmatic Paradigmatic (SP) model is a distributed, instance-based account of sentence processing. Built on the Minerva II model of episodic memory (Hintzman 1988), it characterizes sentence processing as the retrieval of sets of associative constraints from long-term memory and the resolution of these constraints in working memory. In common with connectionist approaches, the SP model provides a data-driven account of language learning and does not make strong a priori assumptions concerning the nature of syntactic knowledge.

supported principles of change (ESPs) and not credential trademarked therapies or other treatment packages. Behavior Modification, 27, 300–312] recommended that empirically supported principles be listed instead of empirically supported treatments because the latter approach enables the creation of putatively new therapies by adding functionally inert components to already listed effective treatments. This article attempts to facilitate inquiry into empirically supported principles by reviewing possible mechanisms responsible for the effectiveness of systematic desensitization and exposure therapy. These interventions were selected because they were among the first empirically supported treatments for which some attempt was made at explanation. Reciprocal inhibition, counterconditioning, habituation, extinction, two-factor model, cognitive changes including expectation, self-efficacy, cognitive restructuring, and informal network-based emotional processing explanations are considered. Logical problems and/or available empirical evidence attenuate or undercut these explanations. A connectionist learning-memory mechanism supported by findings from behavioral and neuroscience research is provided. It demonstrates the utility of preferring empirically supported principles over treatments. Problems and limitations of connectionist explanations are presented. This explanation warrants further consideration and should stimulate discussion concerning empirically supported principles.