The book presents the case that cognitive science should turn its attention to developing theories of human cognition that cover the full range of human perceptual, cognitive, and action phenomena. Cognitive science has now produced a massive number of high-quality regularities with many microtheories that reveal important mechanisms. The need for integration is pressing and will continue to increase. Equally important, cognitive science now has the theoretical concepts and tools to support serious attempts at unified theories. The argument is made (...) entirely by presenting an exemplar unified theory of cognition both to show what a real unified theory would be like and to provide convincing evidence that such theories are feasible. The exemplar is SOAR, a cognitive architecture, which is realized as a software system. After a detailed discussion of the architecture and its properties, with its relation to the constraints on cognition in the real world and to existing ideas in cognitive science, SOAR is used as theory for a wide range of cognitive phenomena: immediate responses ; discrete motor skills ; memory and learning ; problem solving ; language ; and development. The treatments vary in depth and adequacy, but they clearly reveal a single, highly specific, operational theory that works over the entire range of human cognition, SOAR is presented as an exemplar unified theory, not as the sole candidate. Cognitive science is not ready yet for a single theory – there must be multiple attempts. But cognitive science must begin to work toward such unified theories. (shrink)

Because Müiller fails to distinguish between two senses of the term “autonomy,” there is a danger that his results will be misinterpreted by both linguists and neuroscientists. Although he may very well have been successful in refuting one sense of autonomy, he may actually have helped to provide an explanation for the correctness of autonomy in its other sense.

This chapter contains sections titled: Highlights of Past Literature Current Work Future Work.

This article surveys the philosophical literature on theoretical linguistics. The focus of the paper is centred around the major debates in the philosophy of linguistics, past and present, with specific relation to how they connect to the philosophy of science. Specific issues such as scientific realism in linguistics, the scientific status of grammars, the methodological underpinnings of formal semantics, and the integration of linguistics into the larger cognitive sciences form the crux of the discussion.

In this article, I describe a novel position on the semantics of artificial intelligence. I present a problem for the current artificial neural networks used in machine learning, specifically with relation to natural language tasks. I then propose that from a metasemantic level, meaning in machines can best be interpreted as radically contextualist. Finally, I consider what this might mean for human‐level semantic competence from a comparative perspective.

This paper attempts to describe and address a specific puzzle related to compositionality in artificial networks such as Deep Neural Networks and machine learning in general. The puzzle identified here touches on a larger debate in Artificial Intelligence related to epistemic opacity but specifically focuses on computational applications of human level linguistic abilities or properties and a special difficulty with relation to these. Thus, the resulting issue is both general and unique. A partial solution is suggested.

I argue that there are no mental representations, in the sense of “representation” used in standard computational theories of the mind. I take Cummins' Meaning and Mental Representation as my stalking-horse, and argue that his view, once properly developed, is self-defeating. The argument implicitly undermines Fodor's view of the mind; I draw that conclusion out explicitly. The idea of mental representations can then only be saved by appeal to a Dennett-like instrumentalism; so I argue against that too. Finally, I argue (...) that there is no good metaphysical reason in favour of believing in mental representations and that cognitive science can manage perfectly well without them. (shrink)

O'Brien & Opie do not address the question of the psychotherapeutic role of unconscious representational states such as beliefs. A dilemma is proposed: if they accept the legitimacy of such states then they should modify what they say about dissociation, and if they do not, they owe us an account of why.

It is widely held in contemporary philosophy of mind that states with underived representational content are ipso facto psychological states. This view—the Content View—underlies a number of interesting philosophical projects, such as the attempt to pick out a psychological level of explanation, to demarcate genuinely psychological from non-psychological states, and to limn the class of states with phenomenal character. The most detailed and influential theories of underived representation in philosophy are the tracking theories developed by Fodor, Dretske, Millikan and others. (...) Tracking theorists initially hoped to ‘naturalize’ underived representation by showing that although it is distinctively psychological it is not irreducibly so, yet they ended up developing theories of representation that by their own lights don’t pick out a distinctively psychological phenomenon at all. Burge sets out to develop a theory of underived representation that does pick out a distinctively psychological phenomenon. His theory promises to vindicate the Content View and the various philosophical projects that depend on it. In this paper I argue that Burge’s theory dementalizes representation for the same reason tracking theories do: These theories hold that representations are states with underived accuracy conditions, yet such states are found in all sorts of mindless systems, like plants. (shrink)

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

Natural language meaning has properties of both (embodied) cognitive representations and formal/mathematical structures. But it is not clear how they actually relate to one another. This article argues that how properties of cognitive representations and formal/mathematical structures of natural language meaning can be united remains one of the puzzles in cognitive science. That is primarily because formal/mathematical structures of natural language meaning are abstract, logical, and truth‐conditional properties, whereas cognitive/conceptual representations are embodied and grounded in sensory‐motor systems. After reviewing the (...) current progress, this work offers, in outline, the general formulations that show how these two different kinds of representations for semantic structures can (potentially) be unified and also proposes three desiderata for testing, in brain dynamics, the mathematical equivalence between formal symbolic representations (and their transitions), and neuronal population codes (and their transitions). (shrink)

This article will provide a unifying perspective on perception and cognition via the route of linguistic representations of emotion. Linguistic representations of emotions provide a fertile ground for explorations into the nature and form of integration of perception and cognition because emotion has facets of both perceptual and cognitive processes. In particular, this article shows that certain types of linguistic representations of emotion allow for the integration of perception and cognition through a series of steps and operations in cognitive systems, (...) whereas certain other linguistic representations of emotion are not so representationally structured as to permit the unity of perception and cognition. It turns out that the types of linguistic representations of emotion that readily permit the desired unity of perception and cognition are exactly those that are linguistically encoded emotive representations of everyday objects, events, and things around us. It is these ordinary objects, events and things that provide the scaffolding for task-dependent or goal-oriented activities of cognitive systems including autonomous systems. In this way, cognitive systems can be saliently tuned to the outer world by being motivated and also subtly governed by emotion-driven representations. This helps not only tie together perceptual and cognitive processes via the interface between language and emotive representations, but also reveal the limits of emotive representations in amalgamating perceptual and cognitive processes in cognitive systems. (shrink)

I propose a version of inferential role theory which says that having a concept is having the disposition to draw most of the inferences based on the stereotypical features associated with this concept. I defend this view against Fodor and Lepore.

Dead reckoning is a feature of the navigation behaviour shown by several creatures, including the desert ant. Recent work by C. Randy Gallistel shows that some connectionist models of dead reckoning face important challenges. These challenges are thought to arise from essential features of the connectionist approach, and have therefore been taken to show that connectionist models are unable to explain even the most primitive of psychological phenomena. I show that Gallistel’s challenges are successfully met by one recent connectionist model, (...) proposed by Ulysses Bernardet, Sergi Bermúdez i Badia, and Paul F.M.J. Verschure. The success of this model suggests that there are ways to implement dead reckoning with neural circuits that fall within the bounds of what many people regard as neurobiologically plausible, and so that the wholesale dismissal of the connectionist modelling project remains premature. (shrink)

This paper proposes a model for implementation of intrinsic natural language sentence meaning in a physical language understanding system, where 'intrinsic' is understood as 'independent of meaning ascription by system-external observers'. The proposal is that intrinsic meaning can be implemented as a point attractor in the state space of a nonlinear dynamical system with feedback which is generated by temporally sequenced inputs. It is motivated by John Searle's well known (Behavioral and Brain Sciences, 3: 417–57, 1980) critique of the then-standard (...) and currently still influential computational theory of mind (CTM), the essence of which was that CTM representations lack intrinsic meaning because that meaning is dependent on ascription by an observer_._ The proposed dynamical model comprises a collection of interacting artificial neural networks, and constitutes a radical simplification of the principle of compositional phrase structure which is at the heart of the current standard view of sentence semantics because it is computationally interpretable as a finite state machine. (shrink)

Chomskyian claims of a genetically hard-wired and cognitively autonomous “universal grammar” are being promoted by generative linguistics as facts about language to the present day. The related doctrine of an evolutionary discontinuity in language emergence, however, is based on misconceptions about the notions of homology and preadaptation. The obvious lack of equivalence between symbolic communicative capacities in existing nonhuman primates and human language does not preclude common roots. Normal and disordered language development is strongly influenced by the genome, but there (...) is no evidence for the existence of specific genes underlying “universal grammar.” In the mature brain, stages of language processing can be distinguished and “first-pass” syntactic analyses appear to precede semantic decoding. However, this partial seriality – as well as behavioral and clinical dissociations between lexical and functional categories – can be best described, not in terms of serially activated and discrete modules, but in terms of classes of cell assemblies that differ in their distributional properties. Whereas cell assemblies involved in semantic interpretation (“content word” assemblies) are widely distributed and are generally less vulnerable to focal lesion, those involved in structural decoding (functor assemblies) are primarily distributed within the left perisylvian cortices and are selectively vulnerable to left perisylvian lesion. These distributional differences are explained in terms of the perceptuomotor components involved in the acquisition of relevant representations. The emerging “motivated” or “toposemantic” brain regional specificity can only be accommodated with “soft” and maturational versions of modularity. The failure to reproduce double dissociations in current connectionist models is due to overly simple neuroscientific assumptions, notably that of overall equipotentiality. Linguistic models should not be expected to be “implemented” in the brain, but need to be constrained by neuroscientific evidence on how biological brains function. (shrink)

The concepts of the innateness, universality, species-specificity, and autonomy of the human language capacity have had an extreme impact on the psycholinguistic debate for over thirty years. These concepts are evaluated from several neurobiological perspectives, with an emphasis on the emergence of language and its decay due to brain lesion and progressive brain disease.Evidence of perceptuomotor homologies and preadaptations for human language in nonhuman primates suggests a gradual emergence of language during hominid evolution. Regarding ontogeny, the innate component of language (...) capacity is likely to be polygenic and shared with other developmental domains. Dissociations between verbal and nonverbal development are probably rooted in the perceptuomotor specializations of neural substrates rather than the autonomy of a grammar module. Aphasiologicaldata often assumed to suggest modular linguistic subsystems can be accounted for in terms of a neurofunctional model incorporating perceptuomotor-based regional specializationsand distributivity of representations. Thus, dissociations between grammatical functors and content words are due to different conditions of acquisition and resulting differences in neural representation. Human brains are characterized by multifactorial interindividual variability, and strict universality of functional organization is biologically unrealistic.A theoretical alternative is proposed according to which (1) linguistic specialization of brain areas is due to epigenetic and probabilistic maturational events, not to genetic ”hard-wiring,” and (2) linguistic knowledge is neurally represented in distributed cell assemblies whose topography reflects the perceptuomotor modalities involved in the acquisition and use of a given item of knowledge. (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 distributed neural network model called SPEC for processing sentences with recursive relative clauses is described. The model is based on separating the tasks of segmenting the input word sequence into clauses, forming the case‐role representations, and keeping track of the recursive embeddings into different modules. The system needs to be trained only with the basic sentence constructs, and it generalizes not only to new instances of familiar relative clause structures but to novel structures as well. SPEC exhibits plausible memory (...) degradation as the depth of the center embeddings increases, its memory is primed by earlier constituents, and its performance is aided by semantic constraints between the constituents. The ability to process structure is largely due to a central executive network that monitors and controls the execution of the entire system. This way, in contrast to earlier subsymbolic systems, parsing is modeled as a controlled high‐level process rather than one based on automatic reflex responses. (shrink)

Symbols enable people to organize and communicate about the world. However, the ways in which symbolic knowledge is learned and then represented in the mind are poorly understood. We present a formal analysis of symbolic learning—in particular, word learning—in terms of prediction and cue competition, and we consider two possible ways in which symbols might be learned: by learning to predict a label from the features of objects and events in the world, and by learning to predict features from a (...) label. This analysis predicts significant differences in symbolic learning depending on the sequencing of objects and labels. We report a computational simulation and two human experiments that confirm these differences, revealing the existence of Feature-Label-Ordering effects in learning. Discrimination learning is facilitated when objects predict labels, but not when labels predict objects. Our results and analysis suggest that the semantic categories people use to understand and communicate about the world can only be learned if labels are predicted from objects. We discuss the implications of this for our understanding of the nature of language and symbolic thought, and in particular, for theories of reference. (shrink)

Predictive Processing (PP) is an increasingly influential neurocognitive-computational framework. PP research has so far focused predominantly on lower level perceptual, motor, and various psychological phenomena. But PP seems to face a “scale-up challenge”: How can it be extended to conceptual thought, language, and other higher cognitive competencies? Compositionality, arguably a central feature of conceptual thought, cannot easily be accounted for in PP because it is not couched in terms of classical symbol processing. I argue, using the example of language, that (...) there is no strong reason to think that PP cannot be scaled up to higher cognition. I suggest that the tacitly assumed common-sense conception of language as Generative Grammar (“folk linguistics”) and its notion of composition leads to the scale-up concerns. Fodor’s Language of Thought Hypothesis (LOTH) plays the role of a cognitive computational paradigm for folk linguistics. Therefore, we do not take LOTH as facing problems with higher cognition, at least with regard to compositionality. But PP can plausibly play the role of a cognitive-computational paradigm for an alternative conception of language, namely Construction Grammar. If Construction Grammar is a plausible alternative to folk linguistics, then PP is not in a worse position than LOTH. (shrink)

Several arguments suggest that motor planning may share embodied neural mechanisms with mental rotation (MR). However, it is not well established whether this overlap occurs regardless of the type of stimulus that is manipulated, in particular manipulable or non-manipulable objects and body parts. We here used high-density electroencephalography (EEG) to examine the cognitive similarity between MR of objects that do not afford specific hand actions (chairs) and bodily stimuli (hands). Participants had identical response options for both types of stimuli, and (...) they gave responses orally in order to prevent possible interference with motor imagery. MR of hands and chairs generated very similar behavioral responses, time-courses and neural sources of evoked-response potentials (ERPs). ERP segmentation analysis revealed distinct time windows during which differential effects of stimulus type and angular disparity were observed. An early period (90–160 ms) differentiated only between stimulus types, and was associated with occipito-temporal activity. A later period (290–330 ms) revealed strong effects of angular disparity, associated with electrical sources in the right angular gyrus and primary motor/somatosensory cortex. These data suggest that spatial transformation processes and motor planning are recruited simultaneously, supporting the involvement of motor emulation processes in MR. (shrink)

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

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)

Smolensky's (1995) proposal for a connectionist explanation of systematicity doesn't work.

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