More than 50 years ago, Bongard introduced 100 visual concept learning problems as a challenge for artificial vision systems. These problems are now known as Bongard problems. Although they are well known in cognitive science and artificial intelligence, only very little progress has been made toward building systems that can solve a substantial subset of them. In the system presented here, visual features are extracted through image processing and then translated into a symbolic visual vocabulary. We introduce a formal language (...) that allows representing compositional visual concepts based on this vocabulary. Using this language and Bayesian inference, concepts can be induced from the examples that are provided in each problem. We find a reasonable agreement between the concepts with high posterior probability and the solutions formulated by Bongard himself for a subset of 35 problems. While this approach is far from solving Bongard problems like humans, it does considerably better than previous approaches. We discuss the issues we encountered while developing this system and their continuing relevance for understanding visual cognition. For instance, contrary to other concept learning problems, the examples are not random in Bongard problems; instead they are carefully chosen to ensure that the concept can be induced, and we found it helpful to take the resulting pragmatic constraints into account. (shrink)

In a previous paper, I argued that expert nursing intuition is a form of what James J. Gibson termed ‘direct perception’ and, as such, is information‐based and can be accepted as part of nursing science. In this paper, I explore the philosophical basis for these claims. I begin by describing analogous problems in philosophy and psychology related to how we know the world. After describing the various solutions proposed and the problems they engender, I summarize Gibson’s theoretical solution together with (...) some of the supporting empirical evidence, but emphasizing the ecological realism on which it relies. I then use these insights to reconsider nursing intuition and the implications for its further exploration. (shrink)

Experts in Artificial Intelligence (AI) development predict that advances in the dvelopment of intelligent systems and agents will reshape vital areas in our society. Nevertheless, if such an advance isn't done with prudence, it can result in negative outcomes for humanity. For this reason, several researchers in the area are trying to develop a robust, beneficial, and safe concept of artificial intelligence. Currently, several of the open problems in the field of AI research arise from the difficulty of avoiding unwanted (...) behaviors of intelligent agents, and at the same time specifying what we want such systems to do. It is of utmost importance that artificial intelligent agents have their values aligned with human values, given the fact that we cannot expect an AI to develop our moral preferences simply because of its intelligence, as discussed in the Orthogonality Thesis. Perhaps this difficulty comes from the way we are addressing the problem of expressing objectives, values, and ends, using representational cognitive methods. A solution to this problem would be the dynamic cognitive approach proposed by Dreyfus, whose phenomenological philosophy defends that the human experience of being-in-the-world cannot be represented by the symbolic or connectionist cognitive methods. A possible approach to this problem would be to use theoretical models such as SED (situated embodied dynamics) to address the values learning problem in AI. (shrink)

A common kind of explanation in cognitive neuroscience might be called functiontheoretic: with some target cognitive capacity in view, the theorist hypothesizes that the system computes a well-defined function (in the mathematical sense) and explains how computing this function constitutes (in the system’s normal environment) the exercise of the cognitive capacity. Recently, proponents of the so-called ‘new mechanist’ approach in philosophy of science have argued that a model of a cognitive capacity is explanatory only to the extent that it reveals (...) the causal structure of the mechanism underlying the capacity. If they are right, then a cognitive model that resists a transparent mapping to known neural mechanisms fails to be explanatory. I argue that a functiontheoretic characterization of a cognitive capacity can be genuinely explanatory even absent an account of how the capacity is realized in neural hardware. (shrink)

The central aim of this paper is to shed light on the nature of explanation in computational neuroscience. I argue that computational models in this domain possess explanatory force to the extent that they describe the mechanisms responsible for producing a given phenomenon—paralleling how other mechanistic models explain. Conceiving computational explanation as a species of mechanistic explanation affords an important distinction between computational models that play genuine explanatory roles and those that merely provide accurate descriptions or predictions of phenomena. It (...) also serves to clarify the pattern of model refinement and elaboration undertaken by computational neuroscientists. (shrink)

Tae Wan Kim, John Hooker, and Thomas Donaldson make an attempt, in recent articles, to solve the alignment problem. As they define the alignment problem, it is the issue of how to give AI systems moral intelligence. They contend that one might program machines with a version of Kantian ethics cast in deontic modal logic. On their view, machines can be aligned with human values if such machines obey principles of universalization and autonomy, as well as a deontic utilitarian principle. (...) Programming machines to do so might be useful, in their view, for applications such as future autonomous vehicles. Their proposal draws both on traditional logic-based and contemporary connectionist approaches, to fuse factual information with normative principles. I will argue that this approach makes demands of machines that go beyond what is currently feasible, and may extend past the limits of the possible for AI. I also argue that a deontological ethics for machines should place greater stress on the formula of humanity of the Kantian categorical imperative. On this principle, one ought never treat a person as a mere means. Recognition of what makes a person a person requires ethical insight. Similar insight is needed to tell treatment as a means from treatment as a mere means. The resources in Kim, Hooker, and Donaldson’s approach is insufficient for this reason. Hesitation regarding deployment of autonomous machines is warranted in light of these alignment concerns. (shrink)

After presenting Kripke’s criticism to Frege’s ideas on context dependence of thoughts, I present two recent attempts of considering cognitive aspects of context dependent expressions inside a truth conditional pragmatics or semantics: Recanati’s non-descriptive modes of presentation (MOPs) and Kaplan’s ways of having in mind (WHIMs). After analysing the two attempts and verifying which answers they should give to the problem discussed by Kripke, I suggest a possible interpretation of these attempts: to insert a procedural or algorithmic level in semantic (...) representations of indexicals. That a function may be computed by different procedures might suggest new possibilities of integrating contextual cognitive aspects in model theoretic semantic. (shrink)

Cognitive scientists used to deem reasoning either as a higher cognitive process based on the manipulation of abstract rules or as a higher cognitive process that is stochastic rather than involving abstract rules. I maintain that these different perspectives are closely intertwined with a theoretical and methodological endorsement of either cognitivism or connectionism. Cognitivism and connectionism represent two prevailing and opposed paradigms in cognitive science. I aim to extoll the virtues of connectionist models of enthymematic reasoning by following means: via (...) the phenomenon of creative enthymeme, viz. the inference where one cannot even articulate the missing premise, I introduce a connectionist mechanism of pattern recognition as underlying expertise; via Gestalt switch or Gestalt click, I demonstrate how differences in pattern recognition of an expert and a novice can be construed as qualitatively different, and not merely a matter of faster reasoning. (shrink)

A number of researchers today make an appeal to quantum physics when trying to develop a satisfactory account of the mind, an appeal still felt to be controversial by many. Often these "quantum approaches" try to explain some well-known features of conscious experience (or mental processes more generally), thus using quantum physics to enrich the explanatory framework or explanans used in consciousness studies and cognitive science. This paper considers the less studied question of whether quantum physical intuitions could help us (...) to draw attention to new or neglected aspects of the mind in introspection, and in this way change our view about what needs explanation in the rst place. Although prima facie implausible, it is suggested that this could happen, for example, if there were analogies between quantum processes and mental processes (e.g., the process of thinking). The naive idea is that such analogies would help us to see mental processes and conscious experience in a new way. It has indeed been proposed long ago that such analogies exist, and this paper rst focuses at some length on David Bohm's formulation of them from 1951. It then briefly considers these analogies in relation to Smolensky's more recent analogies between cognitive science and physics, and Pylkko's aconceptual view of the mind. Finally, Bohm's early analogies will be briefly considered in relation to the analogies between quantum processes and the mind he proposed in his later work. (shrink)

A number of researchers today make an appeal to quantum physics when trying to develop a satisfactory account of the mind, an appeal still felt to be controversial by many. Often these "quantum approaches" try to explain some well-known features of conscious experience (or mental processes more generally), thus using quantum physics to enrich the explanatory framework or explanans used in consciousness studies and cognitive science. This paper considers the less studied question of whether quantum physical intuitions could help us (...) to draw attention to new or neglected aspects of the mind in introspection, and in this way change our view about what needs explanation in the first place. Although prima facie implausible, it is suggested that this could happen, for example, if there were analogies between quantum processes and mental processes (e.g., the process of thinking). The naive idea is that such analogies would help us to see mental processes and conscious experience in a new way. It has indeed been proposed long ago that such analogies exist, and this paper first focuses at some length on David Bohm's formulation of them from 1951. It then briefly considers these analogies in relation to Smolensky's more recent analogies between cognitive science and physics, and Pylkko's aconceptual view of the mind. Finally, Bohm's early analogies will be briefly considered in relation to the analogies between quantum processes and the mind he proposed in his later work. -/- [This article is a modified version of an article that was first published in the anthology Being and Brain: At the Boundary between Science, Philosophy, Language and Arts, ed. by G. Globus, K. Pribram and G. Vitiello, Advances in Consciousness Research 58, John Benjamins, Amsterdam 2004, pp. 165-195.]. (shrink)

This book contains a representationalist theory of self-consciousness and of the phenomenal first-person perspective. It draws on empirical data from the cognitive and neurosciences.

This is an amended version of material that first appeared in A. Clark, Microcognition: Philosophy, Cognitive Science, and Parallel Distributed Processing (MIT Press, Cambridge, MA, 1989), Ch. 1, 2, and 6. It appears in German translation in Metzinger,T (Ed) DAS LEIB-SEELE-PROBLEM IN DER ZWEITEN HELFTE DES 20 JAHRHUNDERTS (Frankfurt am Main: Suhrkamp. 1999).

The cell assembly is an important concept for cognitive psychology. Cognitive processing will to a large extent depend on the relations that can exist between different assemblies. A potential relation between assemblies can already be seen in the occurrence of (classical) conditioning. However, the resulting associations between assemblies only produce behavioristic processing or so-called regular computation. Higher-level cognitive abilities most likely result from nonregular computation. I discuss the possibility of this form of computation in terms of cell assemblies.

Consciousness and experience, William G. Lycan. Cambridge, MA: MIT Press, 1996. ISBN 0–262–12197–2 (hc)Mind as motion: explorations in the dynamics of cognition, Robert Port & Timothy Van Gelder (Eds). Cambridge, MA: MIT/Bradford, 1995. ISBN 0–262–16150–8Perception, Kathleen Akins (Ed.). New York: Oxford University Press, 1996. ISBN 0–19–508461 (hbk); 0–19–508462–4 (pbk)Context and consciousness, B. Nardi (Ed.). Cambridge, MA: MIT Press, 1996. ISBN: 0–262–14058–6Catching ourselves in the act, Horst Hendriks‐Jansen. Cambridge, MA: MIT Press, 1996. ISBN 0–262–08246–2The search for mind. A new foundation for (...) cognitive science, Sean O'Nuallain. Norwood, NJ: Ablex, 1995Mindblindness: an essay on autism and theory of mind, Simon Baron‐Cohen. Cambridge, MA: MIT Press, 1995. ISBN 0–262–02384–9. (shrink)

This paper examines whether a classical model could be translated into a PDP network using a standard connectionist training technique called extra output learning. In Study 1, standard machine learning techniques were used to create a decision tree that could be used to classify 8124 different mushrooms as being edible or poisonous on the basis of 21 different Features (Schlimmer, 1987). In Study 2, extra output learning was used to insert this decision tree into a PDP network being trained on (...) the identical problem. An interpretation of the trained network revealed a perfect mapping from its internal structure to the decision tree, representing a precise translation of the classical theory to the connectionist model. In Study 3, a second network was trained on the mushroom problem without using extra output learning. An interpretation of this second network revealed a different algorithm for solving the mushroom problem, demonstrating that the Study 2 network was indeed a proper theory translation. (shrink)

There is widespread belief that connectionist networks are dramatically different from classical or symbolic models. However, connectionists rarely test this belief by interpreting the internal structure of their nets. A new approach to interpreting networks was recently introduced by Berkeley et al. (1995). The current paper examines two implications of applying this method: (1) that the internal structure of a connectionist network can have a very classical appearance, and (2) that this interpretation can provide a cognitive theory that cannot be (...) dismissed as a mere implementation. (shrink)

PDP networks that use nonmonotonic activation functions often produce hidden unit regularities that permit the internal structure of these networks to be interpreted (Berkeley et al., 1995; McCaughan, 1997; Dawson, 1998). In particular, when the responses of hidden units to a set of patterns are graphed using jittered density plots, these plots organize themselves into a set of discrete stripes or bands. In some cases, each band is associated with a local interpretation. On the basis of these observations, Berkeley (2000) (...) has suggested that these bands are both subsymbolic and symbolic in nature, and has used the analysis of one network to support the claim that there are fewer differences between symbols and subsymbols than one might expect. We suggest below that this conclusion is premature. First, in many cases the local interpretation of each band is difficult to relate to the interpretation of a network's response; a more appropriate relationship only emerges when a band associated with one hidden unit is considered in the context of other bands associated with other hidden units (i.e., interpretations of distributed representations are more useful than interpretations of local representations). Second, the content that a band designates to an external observer (i.e., the interpretation assigned to a band by the researcher) can be quite different from the content that a band designates to the output units of the network itself.. We use two different network simulations – including the one described by Berkeley (2000) – to illustrate these points. We conclude that current evidence involving interpretations of nonmonotonic PDP networks actually illustrates the differences between symbolic and subsymbolic processing. (shrink)

One new tradition that has emerged from early research on autonomous robots is embodied cognitive science. This paper describes the relationship between embodied cognitive science and a related tradition, synthetic psychology. It is argued that while both are synthetic, embodied cognitive science is antirepresentational while synthetic psychology still appeals to representations. It is further argued that modern connectionism offers a medium for conducting synthetic psychology, provided that researchers analyze the internal representations that their networks develop. The paper then provides a (...) detailed example of the synthetic approach by showing how the construction (and subsequent analysis) of a connectionist network can be used to contribute to a theory of how humans solve Piaget's classic balance scale task. (shrink)

This paper critically examines the claim that parallel distributed processing (PDP) networks are autonomous learning systems. A PDP model of a simple distributed associative memory is considered. It is shown that the 'generic' PDP architecture cannot implement the computations required by this memory system without the aid of external control. In other words, the model is not autonomous. Two specific problems are highlighted: (i) simultaneous learning and recall are not permitted to occur as would be required of an autonomous system; (...) (ii) connections between processing units cannot simultaneously represent current and previous network activation as would be required if learning is to occur. Similar problems exist for more sophisticated networks constructed from the generic PDP architecture. We argue that this is because these models are not adequately constrained by the properties of the functional architecture assumed by PDP modelers. It is also argued that without such constraints, PDP researchers cannot claim to have developed an architecture radically different from that proposed by the Classical approach in cognitive science. (shrink)

Cognitive psychology and cognitive science are concerned with a domain of cognition that is much broader than the realm of judgement, belief, and inference. The idea of states with semantic content is extended far beyond the space of reasons and justification. Within this broad class of states we should, however, differentiate between the states distinctive of thinking persons — centrally, beliefs, desires, and intentions — and other states. The idea of consciousness does not furnish a principle of demarcation. But the (...) distinction between states whose content is conceptualized by the person whose states they are and states for which this is not so is more promising. This principle of demarcation contains the seeds of a problem for distributed connectionism. The article ends with some more general reflections about cognitive science. (shrink)

In this paper, I define tacit knowledge as a kind of causal-explanatory structure, mirroring the derivational structure in the theory that is tacitly known. On this definition, tacit knowledge does not have to be explicitly represented. I then take the notion of a modular theory, and project the idea of modularity to several different levels of description: in particular, to the processing level and the neurophysiological level. The fundamental description of a connectionist network lies at a level between the processing (...) level and the physiological level. At this level, connectionism involves a characteristic departure from modularity, and a correlative absence of syntactic structure. This is linked to the fact that tacit knowledge descriptions of networks are only approximately true. A consequence is that strict causal systematicity in cognitive processes poses a problem for the connectionist programme. (shrink)

The cell assembly as a simple attractor cannot explain many cognitive phenomena. It must be a highly structured network that can sustain highly structured excitation patterns. Moreover, a cell assembly must be more widely distributed in space than on a square millimeter.

I challenge the dominant understanding of what it means to say two thoughts are associated. The two views that dominate the current literature treat association as a kind of mechanism that drives sequences of thought. The first, which I call reductive associationism, treats association as a kind of neural mechanism. The second treats association as a feature of the kind of psychological mechanism associative processing. Both of these views are inadequate. I argue that association should instead be seen as a (...) highly abstract filler term, standing in for causal relations between representational states in a system. Associations, so viewed, could be implemented by many different mechanisms. I outline the role that this view gives associative models as part of a top-down characterization of psychological processes of any kind and of any complexity. (shrink)

The Problems of Mental Causation. Functionalism in the philosophy of mind identifies mental states with their dispositional connections with other mental states, perceptions and actions. Many theories of the mind have sailed under the Functionalist flag. But what I take to be essential to Functionalism is that mental states are individuated causally: the reality of mental states depends essentially on their causal efficacy.

In their enthusiasm for programming, computational linguists have tended to lose sight of what humansdo. They have conceived of conversations as independent of sound and the bodies that produce it. Thus, implicit in their simulations is the assumption that the text is the essence of talk. In fact, unlike electronic mail, conversations are acoustic events. During everyday talk, human understanding depends both on the words spoken and on fine interpersonal vocal coordination. When utterances are analysed into sequences of word-based forms, (...) however, these prosodic aspects of language disappear. Therefore, to investigate the possibility that machines might talk, we propose acommunion game that includes this interpersonal patterning. Humans and machines would talk together and, based on recordings of them, a panel would appraise the relevant merit of each machine's simulation by how true to life it sounded. Unlike Turing's imitation game, the communion game overtly focuses attention, not on intelligence, but on language. It is designed to facilitate the development of social groups of adaptive robots that exploit complex acoustic signals in real time. We consider how the development of such machines might be approached. (shrink)

In his recent book, Concepts: Where Cognitive Science Went Wrong, Jerry Fodor retracts the radical concept-nativism he once defended. Yet that postion stood, virtually unchallenged, for more than twenty years. This neglect is puzzling, as Fodor's arguments against concepts being learnable from experience remain unanswered, and nativism has historically been taken very seriously as a response to empiricism's perceived shortcomings. In this paper, I urge that Fodorean nativism should indeed be rejected. I argue, however, that its deficiencies are not so (...) obvious that they can simply be taken for granted. Fodor can counter extant objections by stressing two distinctions: between historicist and counterfactual semantic theories and between explaining reference and explaining concept-acquisition. But, I argue, this victory is pyrrhic. Reformulated as objections to his account qua theory of concept-acquisition, and not qua theory of reference, analogous difficulties are fatal to the Fodorean position. (shrink)

The study of signed languages has inspired scientific' speculation regarding foundations of human language. Relationships between the acquisition of sign language in apes and man are discounted on logical grounds. Evidence from the differential hreakdown of sign language and manual pantomime places limits on the degree of overlap between language and nonlanguage motor systems. Evidence from functional magnetic resonance imaging reveals neural areas of convergence and divergence underlying signed and spoken languages.