Representation in Connectionism

In this paper, we consider how certain longstanding philosophical questions about mental representation may be answered on the assumption that cognitive and perceptual systems implement hierarchical generative models, such as those discussed within the prediction error minimization framework. We build on existing treatments of representation via structural resemblance, such as those in Gładziejewski :559–582, 2016) and Gładziejewski and Miłkowski, to argue for a representationalist interpretation of the PEM framework. We further motivate the proposed approach to content by arguing that it (...) is consistent with approaches implicit in theories of unsupervised learning in neural networks. In the course of this discussion, we argue that the structural representation proposal, properly understood, has more in common with functional-role than with causal/informational or teleosemantic theories. In the remainder of the paper, we describe the PEM framework for approximate Bayesian inference in some detail, and discuss how structural representations might arise within the proposed Bayesian hierarchies. After explicating the notion of variational inference, we define a subjectively accessible measure of misrepresentation for hierarchical Bayesian networks by appeal to the Kullbach–Leibler divergence between posterior generative and approximate recognition densities, and discuss a related measure of objective misrepresentation in terms of correspondence with the facts. (shrink)

Two different but closely related issues in current cognitive science will be considered in this essay. One is the controversial and extensively discussed question of how connectionist and symbolic representations of knowledge are related to each other. The other concerns the notion of connectionist learning and its relevance for the understanding of the distinction between propositional and nonpropositional knowledge. More specifically, I shall give an overview of a result in Rantala and Vadén establishing a limiting case correspondence between symbolic and (...) connectionist representations and, on the other hand, study the problem, preliminarily investigated in Rantala, of how propositional knowledge may arise from nonpropositional knowledge. I shall also try to point out that on some more or less plausible assumptions, often made by cognitive scientists, these results may have some significance when we try to comprehend the nature of human knowledge representation. Some of these assumptions are rather hypothethical and debatable for the time being and they will become justified in the future only if there will be more progress in the empirical and theoretical research on the brain and on artificial networks. The assumptions concern, besides some questions of the behavior of neural networks, such things as the relevance of pattern recognition for modelling human cognition, in particular, knowledge acquisition, and the relation between emergence and reduction. (shrink)

Purely parallel neural networks can model object recognition in brief displays – the same conditions under which illusory conjunctions have been demonstrated empirically. Correcting errors of illusory conjunction is the “tag-assignment” problem for a purely parallel processor: the problem of assigning a spatial tag to nonspatial features, feature combinations, and objects. This problem must be solved to model human object recognition over a longer time scale. Our model simulates both the parallel processes that may underlie illusory conjunctions and the serial (...) processes that may solve the tag-assignment problem in normal perception. One component of the model extracts pooled features and another provides attentional tags that correct illusory conjunctions. Our approach addresses two questions: How can objects be identified from simultaneously attended features in a parallel, distributed representation? How can the spatial selectional requirements of such an attentional process be met by a separation of pathways for spatial and nonspatial processing? Our analysis of these questions yields a neurally plausible simulation of tag assignment based on synchronizing feature processing activity in a spatial focus of attention. (shrink)

A semantic network or net is a graphic notation for representing knowledge in patterns of interconnected nodes and arcs. Computer implementations of semantic networks were first developed for artificial intelligence and machine translation, but earlier versions have long been used in philosophy, psychology, and linguistics. What is common to all semantic networks is a declarative graphic representation that can be used either to represent knowledge or to support automated systems for reasoning about knowledge. Some versions are highly informal, but other (...) versions are formally defined systems of logic. Following are six of the most common kinds of semantic networks, each of which is discussed in detail in one section of this article. 1. Definitional networks emphasize the subtype or is-a relation between a concept type and a newly defined subtype. The resulting network, also called a generalization or subsumption hierarchy, supports the rule of inheritance for copying properties defined for a supertype to all of its subtypes. Since definitions are true by definition, the information in these networks is often assumed to be necessarily true. (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)

Localizing content in neural networks provides a bridge to understanding the way in which the brain stores and processes information. In this paper, I propose the existence of polytopes in the state space of the hidden layer of feedforward neural networks as vehicles of content. I analyze these geometrical structures from an information-theoretic point of view, invoking mutual information to help define the content stored within them. I establish how this proposal addresses the problem of misclassification and provide a novel (...) solution to the disjunction problem, which hinges on the precise nature of the causal-informational framework for content advocated herein. (shrink)

There is a gap between two different modes of computation: the symbolic mode and the subsymbolic (neuron-like) mode. The aim of this paper is to overcome this gap by viewing symbolism as a high-level description of the properties of (a class of) neural networks. Combining methods of algebraic semantics and non-monotonic logic, the possibility of integrating both modes of viewing cognition is demonstrated. The main results are (a) that certain activities of connectionist networks can be interpreted as non-monotonic inferences, and (...) (b) that there is a strict correspondence between the coding of knowledge in Hopfield networks and the knowledge representation in weight-annotated Poole systems. These results show the usefulness of non-monotonic logic as a descriptive and analytic tool for analyzing emerging properties of connectionist networks. Assuming an exponential development of the weight function, the present account relates to optimality theory – a general framework that aims to integrate insights from symbolism and connectionism. The paper concludes with some speculations about extending the present ideas. (shrink)

One of the main challenges that Jerry Fodor and Zenon Pylyshyn (Cognition 28:3–71, 1988) posed for any connectionist theory of cognitive architecture is to explain the systematicity of thought without implementing a Language of Thought (LOT) architecture. The systematicity challenge presents a dilemma: if connectionism cannot explain the systematicity of thought, then it fails to offer an adequate theory of cognitive architecture; and if it explains the systematicity of thought by implementing a LOT architecture, then it fails to offer an (...) alternative to the LOT hypothesis. Given that thought is systematic, connectionism can offer an adequate alternative to the LOT hypothesis only if it can meet the challenge. Although some critics tried to meet the challenge, others argued that it need not be met since thought is not in fact systematic; and some claimed not to even understand the claim that thought is systematic. I do not here examine attempts to answer the challenge. Instead, I defend the challenge itself by explicating the notion of systematicity in a way that I hope makes clear that thought is indeed systematic, and so that to offer an adequate alternative to the LOT hypothesis, connectionism must meet the challenge. (shrink)

Any creature that must move around in its environment to find nutrients and mates, in order to survive and reproduce, faces the problem of sensorimotor control. A solution to this problem requires an on-board control mechanism that can shape the creature’s behaviour so as to render it “appropriate” to the conditions that obtain. There are at least three ways in which such a control mechanism can work, and Nature has exploited them all. The first and most basic way is for (...) a creature to bump into the things in its environment, and then, depending on what has been encountered, seek to modify its behaviour accordingly. Such an approach is risky, however, since some things in the environment are distinctly unfriendly. A second and better way, therefore, is for a creature to exploit ambient forms of energy that carry information about the distal structure of the environment. This is an improvement on the first method since it enables the creature to respond to the surroundings without actually bumping into anything. Nonetheless, this second method also has its limitations, one of which is that the information conveyed by such ambient energy is often impoverished, ambiguous and intermittent. (shrink)

In Connectionism and the Philosophy of Psychology, Horgan and Tienson (1996) argue that cognitive processes, pace classicism, are not governed by exceptionless, “representation-level” rules; they are instead the work of defeasible cognitive tendencies subserved by the non-linear dynamics of the brain’s neural networks. Many theorists are sympathetic with the dynamical characterisation of connectionism and the general (re)conception of cognition that it affords. But in all the excitement surrounding the connectionist revolution in cognitive science, it has largely gone unnoticed that connectionism (...) adds to the traditional focus on computational processes, a new focus – one on the vehicles of mental representation, on the entities that carry content through the mind. Indeed, if Horgan and Tienson’s dynamical characterisation of connectionism is on the right track, then so intimate is the relationship between computational processes and representational vehicles, that connectionist cognitive science is committed to a resemblance theory of mental content. (shrink)

“There’s Plenty of Room at the Bottom”, said the title of Richard Feynman’s 1959 seminal conference at the California Institute of Technology. Fifty years on, nanotechnologies have led computer scientists to pay close attention to the links between physical reality and information processing. Not all the physical requirements of optimal computation are captured by traditional models—one still largely missing is reversibility. The dynamic laws of physics are reversible at microphysical level, distinct initial states of a system leading to distinct final (...) states. On the other hand, as von Neumann already conjectured, irreversible information processing is expensive: to erase a single bit of information costs ~3 × 10−21 joules at room temperature. Information entropy is a thermodynamic cost, to be paid in non-computational energy dissipation. This paper addresses the problem drawing on Edward Fredkin’s Finite Nature hypothesis: the ultimate nature of the universe is discrete and finite, satisfying the axioms of classical, atomistic mereology. The chosen model is a cellular automaton with reversible dynamics, capable of retaining memory of the information present at the beginning of the universe. Such a CA can implement the Boolean logical operations and the other building bricks of computation: it can develop and host all-purpose computers. The model is a candidate for the realization of computational systems, capable of exploiting the resources of the physical world in an efficient way, for they can host logical circuits with negligible internal energy dissipation. (shrink)

During the last decades, many cognitive architectures (CAs) have been realized adopting different assumptions about the organization and the representation of their knowledge level. Some of them (e.g. SOAR [35]) adopt a classical symbolic approach, some (e.g. LEABRA[ 48]) are based on a purely connectionist model, while others (e.g. CLARION [59]) adopt a hybrid approach combining connectionist and symbolic representational levels. Additionally, some attempts (e.g. biSOAR) trying to extend the representational capacities of CAs by integrating diagrammatical representations and reasoning are (...) also available [34]. In this paper we propose a reflection on the role that Conceptual Spaces, a framework developed by Peter G¨ardenfors [24] more than fifteen years ago, can play in the current development of the Knowledge Level in Cognitive Systems and Architectures. In particular, we claim that Conceptual Spaces offer a lingua franca that allows to unify and generalize many aspects of the symbolic, sub-symbolic and diagrammatic approaches (by overcoming some of their typical problems) and to integrate them on a common ground. In doing so we extend and detail some of the arguments explored by G¨ardenfors [23] for defending the need of a conceptual, intermediate, representation level between the symbolic and the sub-symbolic one. In particular we focus on the advantages offered by Conceptual Spaces (w.r.t. symbolic and sub-symbolic approaches) in dealing with the problem of compositionality of representations based on typicality traits. Additionally, we argue that Conceptual Spaces could offer a unifying framework for interpreting many kinds of diagrammatic and analogical representations. As a consequence, their adoption could also favor the integration of diagrammatical representation and reasoning in CAs. (shrink)

Embedded and embodied approaches to cognition urge that (1) complicated internal representations may be avoided by letting features of the environment drive behavior, and (2) environmental structures can play an enabling role in cognition, allowing prior cognitive processes to solve novel tasks. Such approaches are thus in a natural position to oppose the ‘thesis of linguistic structuring’: The claim that the ability to use language results in a wholesale recapitulation of linguistic structure in onboard mental representation. Prominent examples of researchers (...) adopting this critical stance include Andy Clark, Michael Wheeler, and Mark Rowlands. But is such opposition warranted? Since each of these authors advocate accounts of mental representation that are broadly connectionist, I survey research on formal language computation in artificial neural networks, and argue that results indicate a strong form of the linguistic structuring thesis is true: Internal representational systems recapitulate significant linguistic structure, even on a connectionist account of mental representation. I conclude by sketching how my conclusion can nonetheless be viewed as consistent with and complimentary to an embedded/embodied account of the role of linguistic structure in cognition. (shrink)

We describe and try to motivate our project to build systems using both a knowledge based and a neural network approach. These two approaches are used at different stages in the solution of a problem, instead of using knowledge bases exclusively on some problems, and neural nets exclusively on others. The knowledge base (KB) is defined first in a declarative, symbolic language that is easy to use. It is then compiled into an efficient neural network (NN) representation, run, and the (...) results from run time and (eventually) from learning are decompiled to a symbolic description of the knowledge contained in the network. After inspecting this recovered knowledge, a designer would be able to modify the KB and go through the whole cycle of compiling, running, and decompiling again. The central question with which this project is concerned is, therefore, How do we go from a KB to an NN, and back again? We are investigating this question by building tools consisting of a repertoire of language/translation/network types, and trying them on problems in a variety of domains. (shrink)

After people learn to sort objects into categories they see them differently. Members of the same category look more alike and members of different categories look more different. This phenomenon of within-category compression and between-category separation in similarity space is called categorical perception (CP). It is exhibited by human subjects, animals and neural net models. In backpropagation nets trained first to auto-associate 12 stimuli varying along a onedimensional continuum and then to sort them into 3 categories, CP arises as a (...) natural side-effect because of four factors: (1) Maximal interstimulus separation in hidden-unit space during autoassociation learning, (2) movement toward linear separability during categorization learning, (3) inverse-distance repulsive force exerted by the between-category boundary, and (4) the modulating effects of input iconicity, especially in interpolating CP to untrained regions of the continuum. Once similarity space has been "warped" in this way, the compressed and separated "chunks" have symbolic labels which could then be combined into symbol strings that constitute propositions about objects. The meanings of such symbolic representations would be "grounded" in the system's capacity to pick out from their sensory projections the object categories that the propositions were about. (shrink)

Some regularities enjoy only an attenuated existence in a body of training data. These are regularities whose statistical visibility depends on some systematic recoding of the data. The space of possible recodings is, however, infinitely large – it is the space of applicable Turing machines. As a result, mappings that pivot on such attenuated regularities cannot, in general, be found by brute-force search. The class of problems that present such mappings we call the class of “type-2 problems.” Type-1 problems, by (...) contrast, present tractable problems of search insofar as the relevant regularities can be found by sampling the input data as originally coded. Type-2 problems, we suggest, present neither rare nor pathological cases. They are rife in biologically realistic settings and in domains ranging from simple animat behaviors to language acquisition. Not only are such problems rife – they are standardly solved! This presents a puzzle. How, given the statistical intractability of these type-2 cases, does nature turn the trick? One answer, which we do not pursue, is to suppose that evolution gifts us with exactly the right set of recoding biases so as to reduce specific type-2 problems to type-1 mappings. Such a heavy-duty nativism is no doubt sometimes plausible. But we believe there are other, more general mechanisms also at work. Such mechanisms provide general strategies for managing problems of type-2 complexity. Several such mechanisms are investigated. At the heart of each is a fundamental ploy – namely, the maximal exploitation of states of representation already achieved by prior, simpler learning so as to reduce the amount of subsequent computational search. Such exploitation both characterizes and helps make unitary sense of a diverse range of mechanisms. These include simple incremental learning, modular connectionism, and the developmental hypothesis of “representational redescription”. In addition, the most distinctive features of human cognition – language and culture – may themselves be viewed as adaptations enabling this representation/computation trade-off to be pursued on an even grander scale. (shrink)

Morphological content (MC) is content that is implicit in the standing structure of the cognitive system. Henderson and Horgan claim that MC plays a distinctive epistemological role unrecognized by traditional epistemic theories. I consider the possibilities that MC plays this role either in central cognition or in peripheral modules. I argue that the peripheral MC does not play an interesting epistemological role and that the central MC is already recognized by traditional theories.

The notion of a ‘symbol’ plays an important role in the disciplines of Philosophy, Psychology, Computer Science, and Cognitive Science. However, there is comparatively little agreement on how this notion is to be understood, either between disciplines, or even within particular disciplines. This paper does not attempt to defend some putatively ‘correct’ version of the concept of a ‘symbol.’ Rather, some terminological conventions are suggested, some constraints are proposed and a taxonomy of the kinds of issue that give rise to (...) disagreement is articulated. The goal here is to provide something like a ‘geography’ of the various notions of ‘symbol’ that have appeared in the various literatures, so as to highlight the key issues and to permit the focusing of attention upon the important dimensions. In particular, the relationship between ‘tokens’ and ‘symbols’ is addressed. The issue of designation is discussed in some detail. The distinction between simple and complex symbols is clarified and an apparently necessary condition for a system to be potentially symbol, or token bearing, is introduced. (shrink)

Connectionist attention to variables has been too restricted in two ways. First, it has not exploited certain ways of doing without variables in the symbolic arena. One variable-avoidance method, that of logical combinators, is particularly well established there. Secondly, the attention has been largely restricted to variables in long-term rules embodied in connection weight patterns. However, short-lived bodies of information, such as sentence interpretations or inference products, may involve quantification. Therefore short-lived activation patterns may need to achieve the effect of (...) variables. The paper is mainly a theoretical analysis of some benefits and drawbacks of using logical combinators to avoid variables in short-lived connectionist encodings without loss of expressive power. The paper also includes a brief survey of some possible methods for avoiding variables other than by using combinators. (shrink)

The relation between logic and thought has long been controversial, but has recently influenced theorizing about the nature of mental processes in cognitive science. One prominent tradition argues that to explain the systematicity of thought we must posit syntactically structured representations inside the cognitive system which can be operated upon by structure sensitive rules similar to those employed in systems of natural deduction. I have argued elsewhere that the systematicity of human thought might better be explained as resulting from the (...) fact that we have learned natural languages which are themselves syntactically structured. According to this view, symbols of natural language are external to the cognitive processing system and what the cognitive system must learn to do is produce and comprehend such symbols. In this paper I pursue that idea by arguing that ability in natural deduction itself may rely on pattern recognition abilities that enable us to operate on external symbols rather than encodings of rules that might be applied to internal representations. To support this suggestion, I present a series of experiments with connectionist networks that have been trained to construct simple natural deductions in sentential logic. These networks not only succeed in reconstructing the derivations on which they have been trained, but in constructing new derivations that are only similar to the ones on which they have been trained. (shrink)

Connectionism provides hope for unifying work in neuroscience, computer science, and cognitive psychology. This promise has met with some resistance from Classical Computionalists, which may have inspired Connectionists to retaliate with bold, inflationary claims on behalf of Connectionist models. This paper demonstrates, by examining three intimately connected issues, that these inflationary claims made on behalf of Connectionism are wrong. This should not be construed as an attack on Connectionism, however, since the inflated claims made on its behalf have the look (...) of cures for which there are no ailments. There is nothing wrong with Connectionism for its failure to solve illusory problems. (shrink)

Harnad's main argument can be roughly summarised as follows: due to Searle's Chinese Room argument, symbol systems by themselves are insufficient to exhibit cognition, because the symbols are not grounded in the real world, hence without meaning. However, a symbol system that is connected to the real world through transducers receiving sensory data, with neural nets translating these data into sensory categories, would not be subject to the Chinese Room argument. Harnad's article is not only the starting point for the (...) present debate, but is also a contribution to a longlasting discussion about such questions as: Can a computer think? If yes, would this be solely by virtue of its program? Is the Turing Test appropriate for deciding whether a computer thinks? (shrink)

Minds are said to be systematic: the capacity to entertain certain thoughts confers to other related thoughts. Although an important property of human cognition, its implication for cognitive architecture has been less than clear. In part, the uncertainty is due to lack of precise accounts on the degree to which cognition is systematic. However, a recent study on learning transfer provides one clear example. This study is used here to compare transfer in humans and feedforward networks. Simulations and analysis show, (...) that while feedforward networks with shared weights are capable of exhibiting transfer, they cannot support the same degree of transfer as humans. One interpretation of these results is that common connectionist models lack explicit internal representations permitting rapid learning. (shrink)

This paper examines the use of connectionism (neural networks) in modelling legal reasoning. I discuss how the implementations of neural networks have failed to account for legal theoretical perspectives on adjudication. I criticise the use of neural networks in law, not because connectionism is inherently unsuitable in law, but rather because it has been done so poorly to date. The paper reviews a number of legal theories which provide a grounding for the use of neural networks in law. It then (...) examines some implementations undertaken in law and criticises their legal theoretical naïvete. It then presents a lessons from the implementations which researchers must bear in mind if they wish to build neural networks which are justified by legal theories. (shrink)

Analogy making from examples is a central task in intelligent system behavior. A lot of real world problems involve analogy making and generalization. Research investigates these questions by building computer models of human thinking concepts. These concepts can be divided into high level approaches as used in cognitive science and low level models as used in neural networks. Applications range over the spectrum of recognition, categorization and analogy reasoning. A major part of legal reasoning could be formally interpreted as an (...) analogy making process. Because it is not the same as reasoning in mathematics or the physical sciences, it is necessary to use a method, which incorporates first the ability to specify likelihood and second the opportunity of including known court decisions. We use for modelling the analogy making process in legal reasoning neural networks and fuzzy systems. In the first part of the paper a neural network is described to identify precedents of immaterial damages. The second application presents a fuzzy system for determining the required waiting period after traffic accidents. Both examples demonstrate how to model reasoning in legal applications analogous to recent decisions: first, by learning a system with court decisions, and second, by analyzing, modelling and testing the decision making with a fuzzy system. (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)

In this paper I discuss one of the key issuesin the philosophy of neuroscience:neurosemantics. The project of neurosemanticsinvolves explaining what it means for states ofneurons and neural systems to haverepresentational contents. Neurosemantics thusinvolves issues of common concern between thephilosophy of neuroscience and philosophy ofmind. I discuss a problem that arises foraccounts of representational content that Icall ``the economy problem'': the problem ofshowing that a candidate theory of mentalrepresentation can bear the work requiredwithin in the causal economy of a mind and (...) anorganism. My approach in the current paper isto explore this and other key themes inneurosemantics through the use of computermodels of neural networks embodied and evolvedin virtual organisms. The models allow for thelaying bare of the causal economies of entireyet simple artificial organisms so that therelations between the neural bases of, forinstance, representation in perception andmemory can be regarded in the context of anentire organism. On the basis of thesesimulations, I argue for an account ofneurosemantics adequate for the solution of theeconomy problem. (shrink)

It has been argued that neural networks and other forms of analog computation may transcend the limits of Turing-machine computation; proofs have been offered on both sides, subject to differing assumptions. In this article I argue that the important comparisons between the two models of computation are not so much mathematical as epistemological. The Turing-machine model makes assumptions about information representation and processing that are badly matched to the realities of natural computation (information representation and processing in or inspired by (...) natural systems). This points to the need for new models of computation addressing issues orthogonal to those that have occupied the traditional theory of computation. (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)

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)

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 nonconnectionist 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)

_Two notions of autonomy are distinguished. The respective_ _denials that psychology is autonomous from neurobiology are neuron_ _doctrines, moderate and radical. According to the moderate neuron_ _doctrine, inter-disciplinary interaction need not aim at reduction. It is_ _proposed that it is more plausible that there is slippage from the_ _moderate to the radical neuron doctrine than that there is confusion_ _between the radical neuron doctrine and the trivial version._.

Various systems that learn are examined to show how content is carried in connections installed by a learning history. Agents do not explicitly use the content of such states in practical reasoning, yet the content plays an important role in explaining behavior, and the physical state carrying that content plays a role in causing behavior, given other occurrent beliefs and desires. This leads to an understanding of the environmental reasons which are the determinate content of these states, and leads to (...) a better grasp of how representational content can be carried by systems without an explicit representation. (shrink)

Although visual long-term memory (VLTM) and visual short-term memory (VSTM) can be distinguished from each other (and from visual sensory storage [SS]), they are embodied within the same modality-specific brain regions, but in very different ways: VLTM as patterns of connectivity and VSTM as patterns of activity. Perception and VSTM do not “activate” VLTM. They use VLTM to create novel patterns of activity relevant to novel circumstances.

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)