Neuroclassicism is the view that cognition is explained by “classical” computing mechanisms in the nervous system that exhibit a clear demarcation between processing machinery and read–write memory. The psychologist C. R. Gallistel has mounted a sophisticated defense of neuroclassicism by drawing from ethology and computability theory to argue that animal brains necessarily contain read–write memory mechanisms. This argument threatens to undermine the “connectionist” orthodoxy in contemporary neuroscience, which does not seem to recognize any such mechanisms. In this paper I argue (...) that the neuroclassicist critique rests on a misunderstanding of how computability theory constrains theorizing about natural computing mechanisms. (shrink)

The Miyashita experiments are very interesting and the results should be examined from a viewpoint of attractor dynamics. Amit's target article shows a path toward realistic modeling by artificial neural networks (ANN), but it is not necessarily the only one. I introduce another model that can explain a substantial part of the empirical observations and makes an interesting prediction. This model consists of such units that have nonmonotonic input-output characteristics with local inhibition neurons.

Inflectional morphology has been taken as a paradigmatic example of rule-governed grammatical knowledge (Pinker, 1999). The plausibility of this claim may be related to the fact that it is mainly based on studies of English, which has a very simple inflectional system. We examined the representation of inflectional morphology in Serbian, which encodes number, gender, and case for nouns. Linguists standardly characterize this system as a complex set of rules, with disagreements about their exact form. We present analyses of a (...) large corpus of nouns which showed that, as in English, Serbian inflectional morphology is quasiregular: It exhibits numerous partial regularities creating neighborhoods that vary in size and consistency. We then asked whether a simple connectionist network could encode this statistical information in a manner that also supported generalization. A network trained on 3,244 Serbian nouns learned to produce correctly inflected phonological forms from a specification of a word's lemma, gender, number, and case, and generalized to untrained cases. The model's performance was sensitive to variables that also influence human performance, including surface and lemma frequency. It was also influenced by inflectional neighborhood size, a novel measure of the consistency of meaning to form mapping. A word-naming experiment with native Serbian speakers showed that this measure also affects human performance. The results suggest that, as in English, generating correctly inflected forms involves satisfying a small number of simultaneous probabilistic constraints relating form and meaning. Thus, common computational mechanisms may govern the representation and use of inflectional information across typologically diverse languages. (shrink)

Every immediate memory is unique; it is therefore unlikely to consist of an attractor or even a combination of attractors. In the present state of knowledge about the chemistry of synaptic transmission, there is no reason to look beyond neurons that directly receive sensory afferents for the afterdischarges that correspond to active memories.

Artificial models of cognition serve different purposes, and their use determines the way they should be evaluated. There are also models that do not represent any particular biological agents, and there is controversy as to how they should be assessed. At the same time, modelers do evaluate such models as better or worse. There is also a widespread tendency to call for publicly available standards of replicability and benchmarking for such models. In this paper, I argue that proper evaluation ofmodels (...) does not depend on whether they target real biological agents or not; instead, the standards of evaluation depend on the use of models rather than on the reality of their targets. I discuss how models are validated depending on their use and argue that all-encompassing benchmarks for models may be well beyond reach. (shrink)

This article addresses the ability of Parallel Distributed Processing (PDP) networks to generate stagewise cognitive development in accordance with Piaget's theory of cognitive epigenesis. We carried out a replication study of the simulation experiments by McClelland (1989) and McClelland and Jenkins (1991) in which a PDP network learns to solve balance scale problems. In objective tests motivated from catastrophe theory, a mathematical theory of transitions in epigenetical systems, no evidence for stage transitions in network performance was found. It is concluded (...) that PDP networks lack the ability to recover the positive outcomes of analogous catastrophe analyses of real cognitive developmental data. In an attempt to further characterize the learning behaviour of PDP networks, we carried out a second simulation study using the discrimination‐shift paradigm. The results thus obtained indicate that PDP learning is compatible with the learning of stimulus‐response relationships, not with the acquisition of mediating rules such as conceived in (neo‐)Piagetian theory. In closing, we speculate about the feasibility of simulating stagewise development with alternative network architectures. (shrink)

The acquisition of English noun and verb morphology is modeled using a single-system connectionist network. The network is trained to produce the plurals and past tense forms of a large corpus of monosyllabic English nouns and verbs. The developmental trajectory of network performance is analyzed in detail and is shown to mimic a number of important features of the acquisition of English noun and verb morphology in young children. These include an initial error-free period of performance on both nouns and (...) verbs followed by a period of intermittent over-regularization of irregular nouns and verbs. Errors in the model show evidence of phonological conditioning and frequency effects. Furthermore, the network demonstrates a strong tendency to regularize denominal verbs and deverbal nouns and masters the principles of voicing assimilation. Despite their incorporation into a single-system network, nouns and verbs exhibit some important differences in their profiles of acquisition. Most importantly, noun inflections are acquired earlier than verb inflections. The simulations generate several empirical predictions that can be used to evaluate further the suitability of this type of cognitive architecture in the domain of inflectional morphology. (shrink)

The current renewal of connectionist techniques using networks of neuron-like units has started to have an influence on cognitive modelling. However, compared with classical artificial intelligence methods, the position of connectionism is still not clear. In this article artificial intelligence and connectionism are systematically compared as cognitive models so as to bring out the advantages and shortcomings of each. The problem of structured representations appears to be particularly important, suggesting likely research directions.

I consider the role of cognitive modeling in cognitive science. Modeling, and the computers that enable it, are central to the field, but the role of modeling is often misunderstood. Models are not intended to capture fully the processes they attempt to elucidate. Rather, they are explorations of ideas about the nature of cognitive processes. In these explorations, simplification is essential—through simplification, the implications of the central ideas become more transparent. This is not to say that simplification has no downsides; (...) it does, and these are discussed. I then consider several contemporary frameworks for cognitive modeling, stressing the idea that each framework is useful in its own particular ways. Increases in computer power (by a factor of about 4 million) since 1958 have enabled new modeling paradigms to emerge, but these also depend on new ways of thinking. Will new paradigms emerge again with the next 1,000‐fold increase? (shrink)

The study of human intelligence was once dominated by symbolic approaches, but over the last 30 years an alternative approach has arisen. Symbols and processes that operate on them are often seen today as approximate characterizations of the emergent consequences of sub- or nonsymbolic processes, and a wide range of constructs in cognitive science can be understood as emergents. These include representational constructs (units, structures, rules), architectural constructs (central executive, declarative memory), and developmental processes and outcomes (stages, sensitive periods, neurocognitive (...) modules, developmental disorders). The greatest achievements of human cognition may be largely emergent phenomena. It remains a challenge for the future to learn more about how these greatest achievements arise and to emulate them in artificial systems. (shrink)

The apparent very close similarity between the learning of the past tense by Adam and the Plunkett and Marchman model is exaggerated by several misleading comparisons--including arbitrary, unexplained changes in how graphs were plotted. The model's development differs from Adam's in three important ways: Children show a U-shaped sequence of development which does not depend on abrupt changes in input; U-shaped development in the simulation occurs only after an abrupt change in training regimen. Children overregularize vowel-change verbs more than no-change (...) verbs; the simulation overregularizes vowel-change verbs less often than no-change verbs. Children, including Adam, overregularize more than they irregularize; the simulation overregularized less than it irregularized. Interestingly, the RM model--widely criticized as being inadequate--does somewhat better, correctly overregularizing vowel-change verbs more often than no-change verbs, and overregularizing more often than it irregularizes. Although Plunkett and Marchman's (1993) state of the art model incorporated hidden layers and back-propagation, used a more realistic phonological coding scheme, and explored a broader range of parameters than Rumelhart and McClelland's model, their results are farther from psychological reality. It is unknown whether any connectionist model can mimic a child's performance without resorting to unrealistic exogenous changes in the training or input, but it is clear that adding a hidden-layer and back-propagation does not ensure a solution. (shrink)

Cognitive scientists must understand not just what the mind does, but how it does what it does. In this paper, I consider four aspects of cognitive architecture: how the mind develops, the extent to which it is or is not modular, the extent to which it is or is not optimal, and the extent to which it should or should not be considered a symbol‐manipulating device (as opposed to, say, an eliminative connectionist network). In each case, I argue that insights (...) from developmental and evolutionary biology can lead to substantive and important compromises in historically vexed debates. (shrink)

Rogers & McClelland's (R&M's) précis represents an important effort to address key issues in concepts and categorization, but few of the simulations deliver what is promised. We argue that the models are seriously underconstrained, importantly incomplete, and psychologically implausible; more broadly, R&M dwell too heavily on the apparent successes without comparable concern for limitations already noted in the literature.

By pointing to deep philosophical confusions endemic to cognitive science, Wittgenstein might seem an enemy of computational approaches. We agree that while Wittgenstein would reject the classicist’s symbols and rules approach, his observations align well with connectionist or neural network approaches. While many connectionisms that dominated the later twentieth century could fall prey to criticisms of biological, pedagogical, and linguistic implausibility, current connectionist approaches can resolve those problems in a Wittgenstein-friendly manner. We present the basics of a Vector Symbolic Architecture (...) formalism, inspired by Smolensky, and indicate how high-dimensional vectors can operate in a context-sensitive and object-independent manner in biologically plausible time scales, reflecting Wittgenstein’s notions of language-games and family resemblance; we show how “soft” symbols for such a formalism can be formed with plausible learning cycles using Sparse Distributed Memory, resolving disputes surrounding Wittgenstein’s private language argument; and show how connectionist networks can extrapolate meaningful patterns to solve problems, providing “ways to go on” without explicit rules, which indicates linguistic plausibility. Connectionism thus provides a systematicity and productivity that is more than a mere implementation of a classical approach, and provides Wittgenstein-friendly and Wittgenstein-illuminating models of mind and language for cognitive science. (shrink)

Many cognitive and evolutionary theories of religion argue that supernatural explanations are byproducts of our cognitive adaptations. An influential argument states that our supernatural explanations result from a tendency to generate anthropomorphic explanations, and that this tendency is a byproduct of an error management strategy because agents tend to be associated with especially high fitness costs. We propose instead that anthropomorphic and other supernatural explanations result as features of a broader toolkit of well-designed cognitive adaptations, which are designed for explaining (...) the abstract and causal structure of complex, unobservable, and uncertain phenomena that have substantial impacts on fitness. Specifically, we argue that (1) mental representations about the abstract vs. the supernatural are largely overlapping, if not identical, and (2) when the data-generating processes for scarce and ambiguous observations are complex and opaque, a naive observer can improve a bias-variance trade-off by starting with a simple, underspecified explanation that Western observers readily interpret as “supernatural.” We then argue that (3) in many cases, knowledge specialists across cultures offer pragmatic services that involve apparently supernatural explanations, and their clients are frequently willing to pay them in a market for useful and effective services. We propose that at least some ethnographic descriptions of religion might actually reflect ordinary and adaptive responses to novel problems such as illnesses and natural disasters, where knowledge specialists possess and apply the best available explanations about phenomena that would otherwise be completely mysterious and unpredictable. (shrink)

Segregated neural circuits that effect particular domain-specific behaviors can be differentiated from neuroanatomical structures implicated in many different aspects of behavior. The basal ganglionic components of circuits regulating nonlinguistic motor behavior, speech, and syntax all function in a similar manner. Hence, it is unlikely that special properties and evolutionary mechanisms are associated with the neural bases of human language.

Hebbian cell-assembly theory and attractor networks are good starting points for modeling cortical processing. Detailed cell models can be useful in understanding the dynamics of attractor networks. Cell assemblies are likely to be distributed, with the cortical column as the local processing unit. Synaptic memory may be dominant in all but the first couple of seconds.

Recent progress in artificial intelligence has renewed interest in building systems that learn and think like people. Many advances have come from using deep neural networks trained end-to-end in tasks such as object recognition, video games, and board games, achieving performance that equals or even beats that of humans in some respects. Despite their biological inspiration and performance achievements, these systems differ from human intelligence in crucial ways. We review progress in cognitive science suggesting that truly human-like learning and thinking (...) machines will have to reach beyond current engineering trends in both what they learn and how they learn it. Specifically, we argue that these machines should build causal models of the world that support explanation and understanding, rather than merely solving pattern recognition problems; ground learning in intuitive theories of physics and psychology to support and enrich the knowledge that is learned; and harness compositionality and learning-to-learn to rapidly acquire and generalize knowledge to new tasks and situations. We suggest concrete challenges and promising routes toward these goals that can combine the strengths of recent neural network advances with more structured cognitive models. (shrink)

Evolution is often characterized as a tinkerer that creates efficient but messy solutions to problems. We analyze the nature of the problems that arise when we try to explain and understand cognitive phenomena created by this haphazard design process. We present a theory of explanation and understanding and apply it to a case problem – solutions generated by genetic algorithms. By analyzing the nature of solutions that genetic algorithms present to computational problems, we show that the reason for why evolutionary (...) designs are often hard to understand is that they exhibit non-modular functionality, and that breaches of modularity wreak havoc on our strategies of causal and constitutive explanation. (shrink)