Brains, it has recently been argued, are essentially prediction machines. They are bundles of cells that support perception and action by constantly attempting to match incoming sensory inputs with top-down expectations or predictions. This is achieved using a hierarchical generative model that aims to minimize prediction error within a bidirectional cascade of cortical processing. Such accounts offer a unifying model of perception and action, illuminate the functional role of attention, and may neatly capture the special contribution of cortical processing to (...) adaptive success. This target article critically examines this approach, concluding that it offers the best clue yet to the shape of a unified science of mind and action. Sections 1 and 2 lay out the key elements and implications of the approach. Section 3 explores a variety of pitfalls and challenges, spanning the evidential, the methodological, and the more properly conceptual. The paper ends (sections 4 and 5) by asking how such approaches might impact our more general vision of mind, experience, and agency. (shrink)

Parallel distributed processing is transforming the field of cognitive science. Microcognition provides a clear, readable guide to this emerging paradigm from a cognitive philosopher's point of view. It explains and explores the biological basis of PDP, its psychological importance, and its philosophical relevance.

As Ruben notes, the macrostrategy can allow that the distinction may also be drawn at some micro level, but it insists that descent to the micro level is ...

A Neurocomputationial Perspective illustrates the fertility of the concepts and data drawn from the study of the brain and of artificial networks that model the...

Chomsky proposes a reformulation of the theory of transformational generative grammar that takes recent developments in the descriptive analysis of particular ...

This paper offers both a theoretical and an experimental perspective on the relationship between connectionist and Classical (symbol-processing) models. Firstly, a serious flaw in Fodor and Pylyshyn’s argument against connectionism is pointed out: if, in fact, a part of their argument is valid, then it establishes a conclusion quite different from that which they intend, a conclusion which is demonstrably false. The source of this flaw is traced to an underestimation of the differences between localist and distributed representation. It has (...) been claimed that distributed representations cannot support systematic operations, or that if they can, then they will be mere implementations of traditional ideas. This paper presents experimental evidence against this conclusion: distributed representations can be used to support direct structure-sensitive operations, in a man- ner quite unlike the Classical approach. Finally, it is argued that even if Fodor and Pylyshyn’s argument that connectionist models of compositionality must be mere implementations were correct, then this would still not be a serious argument against connectionism as a theory of mind. (shrink)

In artificial intelligence, recent research has demonstrated the remarkable potential of Deep Convolutional Neural Networks (DCNNs), which seem to exceed state-of-the-art performance in new domains weekly, especially on the sorts of very difficult perceptual discrimination tasks that skeptics thought would remain beyond the reach of artificial intelligence. However, it has proven difficult to explain why DCNNs perform so well. In philosophy of mind, empiricists have long suggested that complex cognition is based on information derived from sensory experience, often appealing to (...) a faculty of abstraction. Rationalists have frequently complained, however, that empiricists never adequately explained how this faculty of abstraction actually works. In this paper, I tie these two questions together, to the mutual benefit of both disciplines. I argue that the architectural features that distinguish DCNNs from earlier neural networks allow them to implement a form of hierarchical processing that I call “transformational abstraction”. Transformational abstraction iteratively converts sensory-based representations of category exemplars into new formats that are increasingly tolerant to “nuisance variation” in input. Reflecting upon the way that DCNNs leverage a combination of linear and non-linear processing to efficiently accomplish this feat allows us to understand how the brain is capable of bi-directional travel between exemplars and abstractions, addressing longstanding problems in empiricist philosophy of mind. I end by considering the prospects for future research on DCNNs, arguing that rather than simply implementing 80s connectionism with more brute-force computation, transformational abstraction counts as a qualitatively distinct form of processing ripe with philosophical and psychological significance, because it is significantly better suited to depict the generic mechanism responsible for this important kind of psychological processing in the brain. (shrink)

The introduction of connectionist or parallel distributed processing (PDP) systems to model cognitive functions has raised the question of the possible relations between these models and traditional information processing models which employ rules to manipulate representations. After presenting a brief account of PDP models and two ways in which they are commonly interpreted by those seeking to use them to explain cognitive functions, I present two ways one might relate these models to traditional information processing models and so not totally (...) repudiate the tradition of modelling cognition through systems of rules and representations. The proposal that seems most promising is that PDP-type structures might provide the underlying framework in which a rule and representation model might be implemented. To show how one might pursue such a strategy, I discuss recent research by Barsalou on the instability of concepts and show how that might be accounted for in a system whose microstructure had a PDP architecture. I also outline how adopting a multi-leveled view of the mind, where on one level the mind employed a PDP-type system and at another level constituted a rule processing system, would allow researchers to relocate some problems which seemed difficult to explain at one level, such as the capacity for concept learning, to another level where it could be handled in a straightforward manner. (shrink)

Despite the considerable attention that the systematicity argument has enjoyed, it is worthwhile examining the argument within the context of similar explanatory arguments from the history of science. This kind of analysis helps show that Connectionism, qua Connectionism, really does not have an explanation of systematicity. Second, and more surprisingly, one finds that the systematicity argument sets such a high explanatory standard that not even Classicism can explain the systematicity of thought.

The prominence of Bayesian modeling of cognition has increased recently largely because of mathematical advances in specifying and deriving predictions from complex probabilistic models. Much of this research aims to demonstrate that cognitive behavior can be explained from rational principles alone, without recourse to psychological or neurological processes and representations. We note commonalities between this rational approach and other movements in psychology – namely, Behaviorism and evolutionary psychology – that set aside mechanistic explanations or make use of optimality assumptions. Through (...) these comparisons, we identify a number of challenges that limit the rational program's potential contribution to psychological theory. Specifically, rational Bayesian models are significantly unconstrained, both because they are uninformed by a wide range of process-level data and because their assumptions about the environment are generally not grounded in empirical measurement. The psychological implications of most Bayesian models are also unclear. Bayesian inference itself is conceptually trivial, but strong assumptions are often embedded in the hypothesis sets and the approximation algorithms used to derive model predictions, without a clear delineation between psychological commitments and implementational details. Comparing multiple Bayesian models of the same task is rare, as is the realization that many Bayesian models recapitulate existing (mechanistic level) theories. Despite the expressive power of current Bayesian models, we argue they must be developed in conjunction with mechanistic considerations to offer substantive explanations of cognition. We lay out several means for such an integration, which take into account the representations on which Bayesian inference operates, as well as the algorithms and heuristics that carry it out. We argue this unification will better facilitate lasting contributions to psychological theory, avoiding the pitfalls that have plagued previous theoretical movements. (shrink)

Ramsey (1997) argues that connectionist representations ‘do not earn their explanatory keep’. The aim of this paper is to examine the argument Ramsey gives to support that conclusion. In doing so, I identify two kinds of explanatory need—need relative to a possible explanation and need relative to a true explanation and argue that internal representations are not needed for either connectionist or non‐connectionist possible explanations but that it is quite likely that they are needed for true explanations. However, to show (...) that the latter is the case requires more than a consideration of the form of explanation involved. (shrink)

Marcus et al.’s experiment (1999) concerning infant ability to distinguish between differing syntactic structures has prompted connectionists to strive to show that certain types of neural networks can mimic the infants’ results. In this paper we take a closer look at two such attempts: Shultz and Bale [Shultz, T.R. and Bale, A.C. (2001), Infancy 2, pp. 501–536] Altmann and Dienes [Altmann, G.T.M. and Dienes, Z. (1999) Science 248, p. 875a]. We were not only interested in how well these two models (...) matched the infants’ results, but also whether they were genuinely learning the grammars involved in this process. After performing an extensive set of experiments, we found that, at first blush, Shultz and Bale’s model (2001) replicated the infant’s known data, but the model largely failed to learn the grammars. We also found serious problems with Altmann and Dienes’ model (1999), which fell short of matching any of the infant’s results and of learning the syntactic structure of the input patterns. (shrink)

A set of hypotheses is formulated for a connectionist approach to cognitive modeling. These hypotheses are shown to be incompatible with the hypotheses underlying traditional cognitive models. The connectionist models considered are massively parallel numerical computational systems that are a kind of continuous dynamical system. The numerical variables in the system correspond semantically to fine-grained features below the level of the concepts consciously used to describe the task domain. The level of analysis is intermediate between those of symbolic cognitive models (...) and neural models. The explanations of behavior provided are like those traditional in the physical sciences, unlike the explanations provided by symbolic models. (shrink)

Computer simulations show that an unstructured neural-network model [Shultz, T. R., & Bale, A. C. (2001). Infancy, 2, 501–536] covers the essential features␣of infant learning of simple grammars in an artificial language [Marcus, G. F., Vijayan, S., Bandi Rao, S., & Vishton, P. M. (1999). Science, 283, 77–80], and generalizes to examples both outside and inside of the range of training sentences. Knowledge-representation analyses confirm that these networks discover that duplicate words in the sentences are nearly identical and that they (...) use this near-identity relation to distinguish sentences that are consistent or inconsistent with a familiar grammar. Recent simulations that were claimed to show that this model did not really learn these grammars [Vilcu, M., & Hadley, R. F. (2005). Minds and Machines, 15, 359–382] confounded syntactic types with speech sounds and did not perform standard statistical tests of results. (shrink)

This paper advocates explicitness about the type of entity to be considered as content- bearing in connectionist systems; it makes a positive proposal about how vehicles of content should be individuated; and it deploys that proposal to argue in favour of representation in connectionist systems. The proposal is that the vehicles of content in some connectionist systems are clusters in the state space of a hidden layer. Attributing content to such vehicles is required to vindicate the standard explanation for some (...) classificatory networks’ ability to generalise to novel samples their correct classification of the samples on which they were trained. (shrink)

Human agents draw a variety of inferences effortlessly, spontaneously, and with remarkable efficiency – as though these inferences were a reflexive response of their cognitive apparatus. Furthermore, these inferences are drawn with reference to a large body of background knowledge. This remarkable human ability seems paradoxical given the complexity of reasoning reported by researchers in artificial intelligence. It also poses a challenge for cognitive science and computational neuroscience: How can a system of simple and slow neuronlike elements represent a large (...) body of systemic knowledge and perform a range of inferences with such speed? We describe a computational model that takes a step toward addressing the cognitive science challenge and resolving the artificial intelligence paradox. We show how a connectionist network can encode millions of facts and rules involving n-ary predicates and variables and perform a class of inferences in a few hundred milliseconds. Efficient reasoning requires the rapid representation and propagation of dynamic bindings. Our model (which we refer to as SHRUTI) achieves this by representing (1) dynamic bindings as the synchronous firing of appropriate nodes, (2) rules as interconnection patterns that direct the propagation of rhythmic activity, and (3) long-term facts as temporal pattern-matching subnetworks. The model is consistent with recent neurophysiological evidence that synchronous activity occurs in the brain and may play a representational role in neural information processing. The model also makes specific psychologically significant predictions about the nature of reflexive reasoning. It identifies constraints on the form of rules that may participate in such reasoning and relates the capacity of the working memory underlying reflexive reasoning to biological parameters such as the lowest frequency at which nodes can sustain synchronous oscillations and the coarseness of synchronization. (shrink)

Recently, connectionist models have been developed that seem to exhibit structuresensitive cognitive capacities without executing a program. This paper examines one such model and argues that it does execute a program. The argument proceeds by showing that what is essential to running a program is preserving the functional structure of the program. It has generally been assumed that this can only be done by systems possessing a certain temporalcausal organization. However, counterfactualpreserving functional architecture can be instantiated in other ways, for (...) example geometrically, which are realizable by connectionist networks. (shrink)

In this paper I assess the explanatory role of internal representations in connectionist models of cognition. Focusing on both the internal‘hidden’units and the connection weights between units, I argue that the standard reasons for viewing these components as representations are inadequate to bestow an explanatorily useful notion of representation. Hence, nothing would be lost from connectionist accounts of cognitive processes if we were to stop viewing the weights and hidden units as internal representations.

One of the hallmarks of human cognition is the capacity to generalize over arbitrary constituents. Recently, Marcus (1998, 1998a, b; Cognition 66, p. 153; Cognitive Psychology 37, p. 243) argued that this capacity, called universal generalization (universality), is not supported by Connectionist models. Instead, universality is best explained by Classical symbol systems, with Connectionism as its implementation. Here it is argued that universality is also a problem for Classicism in that the syntax-sensitive rules that are supposed to provide causal explanations (...) of mental processes are either too strict, precluding possible generalizations; or too lax, providing no information as to the appropriate alternative. Consequently, universality is not explained by a Classical theory. (shrink)

In 1988 Fodor and Pylyshyn issued a challenge to the newly-popular connectionism: explain the systematicity of cognition without merely implementing a so-called classical architecture. Since that time quite a number of connectionist models have been put forward, either by their designers or by others, as in some measure demonstrating that the challenge can be met (e.g., Pollack, 1988, 1990; Smolensky, 1990; Chalmers, 1990; Niklasson and Sharkey, 1992; Brousse, 1993). Unfortu- nately, it has generally been unclear whether these models actually do (...) have this implication (see, for instance, the extensive philosophical debate in Smolensky, 1988; Fodor and McLaughlin, 1990; van Gelder, 1990, 1991; McLaughlin, 1993a, 1993b; Clark, 1993). Indeed, we know of no major supporter of classical orthodoxy who has felt compelled, by connectionist models and argu- ments, to concede in print that connectionists have in fact delivered a non-classical explanation of systematicity. (shrink)

ABSTRACTWhile over the last few decades AI research has largely focused on building tools and applications, recent technological developments have prompted a resurgence of interest in building a genuinely intelligent artificial agent – one that has a mind in the same sense that humans and animals do. In this paper, I offer a theoretical and methodological framework for this project of investigating “artificial minded intelligence” that can help to unify existing approaches and provide new avenues for research. I first outline (...) three desiderata that a framework for AMI research should satisfy. In Section 1, I further motivate the desiderata as well as the need for a new framework. Section 2 develops a general methodological approach, the “generative methodology,” and Section 3 develops a version of this methodology, the “Competence Framework for AI research”. (shrink)

Classicists and connectionists alike claim to be able to explain systematicity. The proposed classicist explanation, I argue, is little more than a promissory note, one that classicists have no idea how to redeem. Smolensky's (1995) proposed connectionist explanation fares little better: it is not vulnerable to recent classicist objections, but it nonetheless fails, particularly if one requires, as some classicists do, that explanations of systematicity take the form of a‘functional analysis’. Nonetheless, there are, I argue, reasons for cautious optimism about (...) the prospects of a connectionist explanation. (shrink)

If connectionism is to be an adequate theory of mind, we must have a theory of representation for neural networks that allows for individual differences in weighting and architecture while preserving sameness, or at least similarity, of content. In this paper we propose a procedure for measuring sameness of content of neural representations. We argue that the correct way to compare neural representations is through analysis of the distances between neural activations, and we present a method for doing so. We (...) then use the technique to demonstrate empirically that different artificial neural networks trained by backpropagation on the same categorization task, even with different representational encodings of the input patterns and different numbers of hidden units, reach states in which representations at the hidden units are similar. We discuss how this work provides a rebuttal to Fodor and Lepore's critique of Paul Churchland's state space semantics. (shrink)

In this paper I defend the propriety of explaining the behavior of distributed connectionist networks by appeal to selected data stored therein. In particular, I argue that if there is a problem with such explanations, it is a consequence of the fact that information storage in networks is superpositional, and not because it is distributed. I then develop a ``proto-account'''' of causation for networks, based on an account of Andy Clark''s, that shows even superpositionality does not undermine information-based explanation. Finally, (...) I argue that the resulting explanations are genuinely informative and not vacuous. (shrink)

This paper presents considerations in favour of the view that traditional (classical) architectures can be seen as emergent features of connectionist networks with distributed representation. A recent paper by William Bechtel (1988) which argues for a similar conclusion is unsatisfactory in that it fails to consider whether the compositional syntax and semantics attributed to mental representations by classical models can emerge within a connectionist network. The compatibility of the two paradigms hinges largely, I suggest, on how this question is answered. (...) Focusing on the issue of syntax, I argue that while such structure is lacking in connectionist models with local representation, it can be accommodated within networks where representation is distributed. I discuss an important paper by Smolenski (1988) which attempts to show how connectionists can incorporate the relevant syntactic structure, suggesting that some criticisms levelled against that paper by Fodor & Pylyshyn (1988) are wanting. I then go on to indicate a strategy by which a compositional syntax and semantics can be defined for the sort of network that Smolenski describes. I conclude that since the connectionist can respect the central tenets of classicism, the two approaches are compatible with one another. (shrink)

Fodor's and Pylyshyn's stand on systematicity in thought and language has been debated and criticized. Van Gelder and Niklasson, among others, have argued that Fodor and Pylyshyn offer no precise definition of systematicity. However, our concern here is with a learning based formulation of that concept. In particular, Hadley has proposed that a network exhibits strong semantic systematicity when, as a result of training, it can assign appropriate meaning representations to novel sentences (both simple and embedded) which contain words in (...) syntactic positions they did not occupy during training. The experience of researchers indicates that strong systematicity in any form is difficult to achieve in connectionist systems.Herein we describe a network which displays strong semantic systematicity in response to Hebbian, connectionist training. During training, two-thirds of all nouns are presented only in a single syntactic position (either as grammatical subject or object). Yet, during testing, the network correctly interprets thousands of sentences containing those nouns in novel positions. In addition, the network generalizes to novel levels of embedding. Successful training requires a, corpus of about 1000 sentences, and network training is quite rapid. The architecture and learning algorithms are purely connectionist, but ‘classical’ insights are discernible in one respect, viz, that complex semantic representations spatially contain their semantic constituents. However, in other important respects, the architecture is distinctly non-classical. (shrink)

The past decade has witnessed the emergence of a novel stance on semantic representation, and its relationship to context sensitivity. Connectionist-minded philosophers, including Clark and van Gelder, have espoused the merits of viewing hidden-layer, context-sensitive representations as possessing semantic content, where this content is partially revealed via the representations'' position in vector space. In recent work, Bodén and Niklasson have incorporated a variant of this view of semantics within their conception of semantic systematicity. Moreover, Bodén and Niklasson contend that they (...) have produced experimental results which not only satisfy a kind of context-based, semantic systematicity, but which, to the degree that reality permits, effectively deals with challenges posed by Fodor and Pylyshyn (1988), and Hadley (1994a). The latter challenge involved well-defined criteria for strong semantic systematicity. This paper examines the relevant claims and experiments of Bodén and Niklasson. It is argued that their case fatally involves two fallacies of equivocation; one concerning ''semantic content'' and the other concerning ''novel test sentences''. In addition, it is argued that their ultimate construal of context sensitive semantics contains serious confusions. These confusions are also found in certain publications dealing with "latent semantic analysis". Thus, criticisms presented here have relevance beyond the work of Bodén and Niklasson. (shrink)

In his discussion of results which I (with Michael Hayward) recently reported in this journal, Kenneth Aizawa takes issue with two of our conclusions, which are: (a) that our connectionist model provides a basis for explaining systematicity “within the realm of sentence comprehension, and subject to a limited range of syntax” (b) that the model does not employ structure-sensitive processing, and that this is clearly true in the early stages of the network's training. Ultimately, Aizawa rejects both (a) and (b) (...) for reasons which I think are ill-founded. In what follows, I offer a defense of our position. In particular, I argue (1) that Aizawa adopts a standard of explanation that many accepted scientific explanations could not meet, and (2) that Aizawa misconstrues the relevant meaning of ‘structure-sensitive process’. (shrink)

In their provocative 1988 paper, Fodor and Pylyshyn issued a formidable challenge to connectionists, i.e. to provide a non‐classical explanation of the empirical phenomenon of systematicity in cognitive agents. Since the appearance of F&P's challenge, a number of connectionist systems have emerged which prima facie meet this challenge. However, Fodor and McLaughlin (1990) advance an argument, based upon a general principle of nomological necessity, to show that one of these systems (Smolensky's) could not satisfy the Fodor‐Pylyshyn challenge. Yet, if Fodor (...) and McLaughlin's analysis is correct, it is doubtful whether any existing connectionist system would fare better than Smolensky's. In the view of Fodor and McLaughlin, humans and classical architectures display systematicity as a matter of nomological necessity (necessity by virtue of natural law), but connectionist architectures do not. However, I argue that the Fodor‐Pylyshyn‐McLaughlin appeal to nomological necessity is untenable. There is a sense in which neither classical nor connectionist architectures possess nomological (or‘nomic’) necessity. However, the sense in which classical architectures do possess nomic necessity applies equally well to at least some connectionist architectures. Representational constituents can have causal efficacy within both classical and connectionist architectures. (shrink)

In "Representations without Rules, Connectionism and the Syntactic Argument'', Kenneth Aizawa argues against the view that connectionist nets can be understood as processing representations without the use of representation-level rules, and he provides a positive characterization of how to interpret connectionist nets as following representation-level rules. He takes Terry Horgan and John Tienson to be the targets of his critique. The present paper marshals functional and methodological considerations, gleaned from the practice of cognitive modelling, to argue against Aizawa's characterization of (...) how connectionist nets may be understood as making use of representation-level rules. (shrink)