The relevance of Wolfgang K hler's psychoneural isomorphism principle to contemporary cognitive neuroscience is explored. K hler's approach to the mind—body problem is interpreted as a response to the foundational crisis of psychology at the beginning of the twentieth century. Some aspects of his isomorphism doctrine are discussed, with a view to reaching an interpretation that is both historically accurate and pertinent to issues currently debated in the philosophy of psychology. The principle was meant to be empirically verifiable. Accordingly, some (...) similarities between K hler's approach and current neural network modeling are pointed out, and it is shown that some recent trends in the neurosciences are broadly compatible with K hler's views on cortical functioning. Isomorphism is interpreted as a form of neuroreductionism constrained by bridging laws relating mental phenomena to macrosocopic parameters of neural function. While isomorphism is probably valid for perceptual phenomena, its applicability to higher mental processes remains doubtful. (shrink)

According to the language of thought (LOT) approach and the related computational theory of mind (CTM), thinking is the processing of symbols in an inner mental language that is distinct from any public language. Herein, I explore a deep problem at the heart of the LOT/CTM program—it has yet to provide a plausible conception of a mental symbol.

Although machine learning has been successful in recent years and is increasingly being deployed in the sciences, enterprises or administrations, it has rarely been discussed in philosophy beyond the philosophy of mathematics and machine learning. The present contribution addresses the resulting lack of conceptual tools for an epistemological discussion of machine learning by conceiving of deep learning networks as ‘judging machines’ and using the Kantian analysis of judgments for specifying the type of judgment they are capable of. At the center (...) of the argument is the fact that the functionality of deep learning networks is established by training and cannot be explained and justified by reference to a predefined rule-based procedure. Instead, the computational process of a deep learning network is barely explainable and needs further justification, as is shown in reference to the current research literature. Thus, it requires a new form of justification, that is to be specified with the help of Kant’s epistemology. (shrink)

This paper articulates what it would take to defend representationalism in the case of emotions – i.e. the claim that emotions attribute evaluative properties to target objects or events. We argue that representationalism faces a significant explanatory challenge that has not yet been adequately recognized. Proponents must establish that a representation relation linking emotions and value is explanatorily necessary. We use the case of perception to bring out the difficulties in meeting this explanatory challenge.

Plant and animal species are information-processing entities of such complexity, integration, and adaptive competence that it may be scientifically fruitful to consider them intelligent. The possibility arises from the analogy between learning and evolution, and from recent developments in evolutionary science, psychology and cognitive science. Species are now described as spatiotemporally localized individuals in an expanded hierarchy of biological entities. Intentional and cognitive abilities are now ascribed to animal, human, and artificial intelligence systems that process information adaptively, and that manifest (...) problem-solving abilities. The structural and functional similarities between such systems and species are extensive, although they “are usually obscured by population-genetic metaphors that have nonetheless contributed much to our understanding of evolution.In this target article, I use Sewall Wright's notion of the “adaptive landscape” to compare the performances of evolving species to those of intelligent organisms. With regard to their adaptive achievements and the kinds of processes by which they are achieved, biological species compare very favorably to intelligent animals by virtue of interactions between populations and their environments, between ontogeny and phylogeny, and between natural, interdemic, organic, and species selection. Addressing the question of whether species are intelligent could help to refine our ideas about species, evolution, and intelligence, and could open new lines of empirical and theoretical inquiry in many disciplines. (shrink)

Plant and animal species are information-processing entities of such complexity, integration, and adaptive competence that it may be scientifically fruitful to consider them intelligent. The possibility arises from the analogy between learning and evolution, and from recent developments in evolutionary science, psychology and cognitive science. Species are now described as spatiotemporally localized individuals in an expanded hierarchy of biological entities. Intentional and cognitive abilities are now ascribed to animal, human, and artificial intelligence systems that process information adaptively, and that manifest (...) problem-solving abilities. The structural and functional similarities between such systems and species are extensive, although they “are usually obscured by population-genetic metaphors that have nonetheless contributed much to our understanding of evolution.In this target article, I use Sewall Wright's notion of the “adaptive landscape” to compare the performances of evolving species to those of intelligent organisms. With regard to their adaptive achievements and the kinds of processes by which they are achieved, biological species compare very favorably to intelligent animals by virtue of interactions between populations and their environments, between ontogeny and phylogeny, and between natural, interdemic, organic, and species selection. Addressing the question of whether species are intelligent could help to refine our ideas about species, evolution, and intelligence, and could open new lines of empirical and theoretical inquiry in many disciplines. (shrink)

Previous neural network models of concept learning were mainly implemented with supervised learning schemes. However, studies of human conceptual memory have shown that concepts may be learned without a teacher who provides the category name to associate with exemplars. A modular neural network architecture that realizes concept acquisition through two functionally distinct operations, categorizing and naming, is proposed as an alternative. An unsupervised algorithm realizes the categorizing module by constructing representations of categories compatible with prototype theory. The naming module associates (...) category names to the output of the categorizing module in a supervised mode. In such a modular architecture, the interface between the modules can be conceived of as an “information relay” that encodes, constrains, and propagates important information. Five experiments were conducted to analyze the relationships among internal conceptual codes and simple conceptual and lexical development. The first two experiments show a prototype effect and illustrate some basic characteristics of the system. The third experiment presents a bottom‐up model of the narrowing down of children's early lexical categories that honors mutual exclusivity. The fourth experiment introduces top‐down constraints on conceptual coding. The fifth experiment exhibits how hierarchical relationships between concepts are learned by the architecture, and also demonstrates how a spectrum of conceptual expertise may gradually emerge as a consequence of experiencing more with certain categories than with others. (shrink)

Previous neural network models of concept learning were mainly implemented with supervised learning schemes. However, studies of human conceptual memory have shown that concepts may be learned without a teacher who provides the category name to associate with exemplars. A modular neural network architecture that realizes concept acquisition through two functionally distinct operations, categorizing and naming, is proposed as an alternative. An unsupervised algorithm realizes the categorizing module by constructing representations of categories compatible with prototype theory. The naming module associates (...) category names to the output of the categorizing module in a supervised mode. In such a modular architecture, the interface between the modules can be conceived of as an “information relay” that encodes, constrains, and propagates important information. Five experiments were conducted to analyze the relationships among internal conceptual codes and simple conceptual and lexical development. The first two experiments show a prototype effect and illustrate some basic characteristics of the system. The third experiment presents a bottom‐up model of the narrowing down of children's early lexical categories that honors mutual exclusivity. The fourth experiment introduces top‐down constraints on conceptual coding. The fifth experiment exhibits how hierarchical relationships between concepts are learned by the architecture, and also demonstrates how a spectrum of conceptual expertise may gradually emerge as a consequence of experiencing more with certain categories than with others. (shrink)

In their article entitled “Ethical Particularism and Patterns”, Frank Jackson, Philip Pettit, and Michael Smith (JPS henceforth) argue that moral particularism is a cognitively implausible theory since it appears to entail the view that one might have a skill that is not grounded in an ability to recognise and represent natural patterns in the world. This charge echoes the complaints of computational theorists of cognition against their embodied cognition counterparts, namely that, theories of cognition that eschew talk of mental representation (...) are implausible qua theories of cognition. In both debates, the cognitive role of generalisation is central to the discussion; however, contrary to the received wisdom, I want to suggest that the dispute is not between generalisation or mental representation on the one hand and no generalisation or mental representation on the other, but rather between what I will call global and local generalisation. Using the dialogue between JPS and Dancy (our paradigm particularist) to frame this discussion, I show that by replacing Dancy's connectionist model for particularist reasoning with a case-based one, we not only vindicate his response to JPS, but we also gain insight into how it is the global/local distinction rather than the generalisation/no generalisation distinction that divides the two views. (shrink)

In an influential critique, Jerry Fodor and Zenon Pylyshyn point to the existence of a potentially devastating dilemma for connectionism (Fodor and Pylyshyn [1988]). Either connectionist models consist in mere associations of unstructured representations, or they consist in processes involving complex representations. If the former, connectionism is mere associationism, and will not be capable of accounting for very much of cognition. If the latter, then connectionist models concern only the implementation of cognitive processes, and are, therefore, not informative at the (...) level of cognition. I shall argue that Fodor and Pylyshyn's argument is based on a crucial misunderstanding, the same misunderstanding which motivates the entire language of thought hypothesis. (shrink)

I begin by tracing some of the confusions regarding levels and reduction to a failure to distinguish two different principles according to which theories can be viewed as hierarchically arranged — epistemic authority and ontological constitution. I then argue that the notion of levels relevant to the debate between symbolic and connectionist paradigms of mental activity answers to neither of these models, but is rather correlative to the hierarchy of functional decompositions of cognitive tasks characteristic of homuncular functionalism. Finally, I (...) suggest that the incommensurability of the intentional and extensional vocabularies constitutes a strongprima facie reason to conclude that there is little likelihood of filling in the story of Bechtel''s missing level in such a way as to bridge the gap between such homuncular functionalism and his own model of mechanistic explanation. (shrink)

This study addresses the dynamical nature of a “representation‐hungry” cognitive task involving an imagined action. In our experiment, participants were handed rods that systematically increased or decreased in length on subsequent trials. Participants were asked to judge whether or not they thought they could reach for a distant object with the hand‐held rod. The results are in agreement with a dynamical model, extended from Tuller, Case, Ding, and Kelso (1994). The dynamical effects observed in this study suggest that predictive judgments (...) regarding the possibility or impossibility of a certain action can be understood in terms of dynamically evolving basins of attraction instead of as depending on representational structures. (shrink)

The commentaries reflect three core themes that pertain not just to our theory, but to the enterprise of connectionist modeling more generally. The first concerns the relationship between a cognitive theory and an implemented computer model. Specifically, how does one determine, when a model departs from the theory it exemplifies, whether the departure is a useful simplification or a critical flaw? We argue that the answer to this question depends partially upon the model's intended function, and we suggest that connectionist (...) models have important functions beyond the commonly accepted goals of fitting data and making predictions. The second theme concerns perceived in-principle limitations of the connectionist approach to cognition, and the specific concerns these perceived limitations raise for our theory. We argue that the approach is not in fact limited in the ways our critics suggest. One common misconception, that connectionist models cannot address abstract or relational structure, is corrected through new simulations showing directly that such structure can be captured. The third theme concerns the relationship between parallel distributed processing (PDP) models and structured probabilistic approaches. In this case we argue that there the difference between the approaches is not merely one of levels. Our PDP approach differs from structured statistical approaches at all of Marr's levels, including the characterization of the goals of cognitive computations, and of the representations and algorithms used. (shrink)

Engineers use neural networks to control systems too complex for conventional engineering solutions. To examine the behavior of individual hidden units would defeat the purpose of this approach because it would be largely uninterpretable. Yet neurophysiologists spend their careers doing just that! Hidden units contain bits and scraps of signals that yield only arcane hints about network function and no information about how its individual units process signals. Most literature on single-unit recordings attests to this grim fact. On the other (...) hand, knowing a system's function and describing it with elegant mathematics tell one very little about what to expect of interneuronal behavior. Examples of simple networks based on neurophysiology are taken from the oculomotor literature to suggest how single-unit interpretability might decrease with increasing task complexity. It is argued that trying to explain how any real neural network works on a cell-by-cell, reductionist basis is futile and we may have to be content with trying to understand the brain at higher levels of organization. (shrink)

We agree with Müller's epigenetic view of evolution and ontogeny and applaud his multilevel perspective. With him, we stress the importance in ontogeny of progressive specialisation rather than prewired structures. However, we argue that he slips from “speech” to “language” and that, in seeking homologies, these two levels need to be kept separate in the analysis of evolution and ontogeny.

This study was aimed at explaining how and under what conditions surface similarity leads to the retrieval of an analogous base problem in LTM. Some elements of a theory of the organisation of knowledge in memory are proposed. Two levels of representation are distinguished. The first level represents directly accessible, local surface properties. The second level represents more abstract information pertaining to the category with which each analogous problem can be associated. Some results will be described showing that access to (...) an analogue in LTM based on surface properties is determined by the existence of a similarity at a higher level of abstraction. This is a search-area effect, in that categorical information seems to delimit a memory search space, within which the connection between local surface properties is more likely to be made. (shrink)

A popular argument form uses general theories of cognitive architecture to motivate conclusions about the nature of moral cognition. This paper highlights the possibility for modus tollens reversal of this argument form. If theories of cognitive architecture generate predictions for moral cognition, then tests of moral thinking provide feedback to cognitive science. In certain circumstances, philosophers' introspective attention to their own moral deliberations can provide unique data for these tests. Recognizing the possibility for this sort of feedback helps to illuminate (...) a deep continuity between the disciplines. (shrink)

Cortical reverberations may induce synaptic changes that underlie developmental plasticity as well as long-term memory. They may be especially important for the consolidation of synaptic changes. Reverberations in cortical networks should have particular significance during development, when large numbers of new representations are formed. This includes the formation of representations across different sensory modalities.

The principal aim of this essay is to discuss some logical features of the so-called Classical model of cognitive architecture as it is advocated by J. Fodor and Z. Pylyshyn in their much discussed article 'Connectionism and Cognitive Architecture: A Critical Analysis'. It is pointed out that their structural assumptions have consequences of a logical kind which call into question the view that the Classical architecture (in their sense) can be employed to model human cognition. It seems that the consequences (...) have escaped Fodor and Pylyshyn's notice, or else they have failed to appreciate them, since some of their claims evidently conflict with them. It is also investigated whether the human mind can be characterized as being logical in some weaker sense of logic. Furthermore, it is argued that J. H. Fetzer's view that it is a semiotic system is more realistic than the Classical model, but the distinction he suggests between human cognition and other kinds may be problematic. (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 (1994) establishing a limiting case correspondence between symbolic (...) and connectionist representations and, on the other hand, study the problem, preliminarily investigated in Rantala (1998), 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)

In this paper, I explore the implications of recent empirical research on concept representation for the philosophical enterprise of conceptual analysis. I argue that conceptual analysis, as it is commonly practiced, is committed to certain assumptions about the nature of our intuitive categorization judgments. I then try to show how these assumptions clash with contemporary accounts of concept representation in cognitive psychology. After entertaining an objection to my argument, I close by considering ways in which conceptual analysis might be altered (...) to accord better with the empirical work. (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.

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.

Considerable debate in philosophy of psychology has recently focussed upon two central themes. One concerns the ontological status of propositional attitudes like beliefs and desires, the other on the proper computational account of cognitive architecture. In the ontological debate, the two most prominent positions are eliminativism, which claims that commonsense psychology is false because there are no such things as beliefs and desires; and versions of intentional realism, which counters that beliefs and desires actually do exist in the mind/brain. In (...) the cognitive architecture debate, there are again two outstanding views: classical cognitivism, which holds that cognition is something closely akin to symbol manipulation; and parallel distributed processing, which roughly maintains that cognition is the distributed activation of several simple, non-symbolic processing units. Furthermore, in spite of their different topics, the two debates have been linked by a number of authors who suggest that where you stand in the cognitive architecture debate should help determine where you stand in the debate over prepositional attitudes. So, for example, writers like Jerry Fodor have used the plausibility of classical cognitivism to defend a realist interpretation of propositional attitudes, while writers like Steve Stich and myself have argued that certain forms of connectionism support eliminativism. (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)

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)

A distinction should be made between the formation of stimulus-driven associations and cognitive concepts. To test the learning mode of a neural network, we propose a simple and classic input-output test: the discrimination shift task. Feed-forward PDP models appear to form stimulus-driven associations. A Hopfield network should be extended to apply the test.

Manipulation of physical models such as tangrams and tiles is a popular approach to teaching early mathematics concepts. This pedagogical approach is extended by new computational media, where mathematical entities such as equations and vectors can be virtually manipulated. The cognitive and neural mechanisms supporting such manipulation-based learning—particularly how actions generate new internal structures that support problem-solving—are not understood. We develop a model of the way manipulations generate internal traces embedding actions, and how these action-traces recombine during problem-solving. This model (...) is based on a study of two groups of sixth-grade students solving area problems. Before problem-solving, one group manipulated a tangram, the other group answered a descriptive test. Eye-movement trajectories during problem-solving were different between the groups. A second study showed that this difference required the tangram's geometrical structure, just manipulation was not enough. We propose a theoretical model accounting for these results, and discuss its implications. (shrink)