Natika Newton

This paper looks at a central issue with embodiment theories in cognition: the role, if any, they provide for mental representation. Thelen and Smith (1994) hold that the concept of representations is either vacuous or misapplied in such systems. Others maintain a place for representations (e.g. Clark 1996), but are imprecise about their nature and role. It is difficult to understand what those could be if representations are understood in the same sense as that used by computationalists: fixed or long-lasting neural structures that represent the sensory stimuli that caused them (e.g. neural response patterns in the visual cortex), or whose “meaning” is fixed innately or in early development for particular functions (e.g. the body schemas for Meltzoff and Gopnik 1993). The paper proposes a distinctions between, on the one hand, neural patterns, traces of sensory activation that while not in themselves representations are available for representational activity, and on the other the act of representing, which is what gives representational content to neural patterns.

Understanding Quine's analysis puts us in a position to connect imagination and logical possibility. The description of what is to be imagined may induce an imaginative experience which can be related to the description in various ways. The description may be 'observational,' meaning that everything that it describes can be observed; in this case imagining under a description is decisive evidence for the possibility of what is described. If the description is non-observational, or contains more information than can be captured by observation alone, then imagining under the description is not proof of the possibility of what is described, although if the description is not believed to be the denial of an analytic sentence then imagining can be evidence that it describes a logical possibility. When the description is believed to be logically impossible, all experiences induced by it will be rejected as not conforming to it, and hence logically impossible things cannot be imagined. Finally, if there is doubt about the logical possibility of a description, attempts to imagine under it will not resolve the doubt, since one who rejects the description will reject attempts to apply it to any experience. Thus one's beliefs about what is possible cannot be changed by imagining per se, but only by independent argument that the description of what is imagined does not entail the denial of an analytic sentence. ;We then discuss the concept of logical possibility. We consider three approaches to the concepts; metaphysical theories, formal or logical theories, and conventionalist theories. Of these only the conventionalist approach sees possibility as something for which one could have evidence. We adopt a conventionalist account of possibility as describability in a given language, where something is describable if it is not the denial of an analytic sentence. We look in detail at W. V. O. Quine's account of analyticity, which involves his theory of language learning in general. One learns the meanings of certain sentences at the same time that one learns to assent to them; these are the sentences we call 'analytic'. Hearing the denial of such sentences later activates a disposition to dissent coupled with a feeling of bewilderment about the meaning of the sentence. ;This dissertation examines the connection between imagination and logical possibility. Both philosophers and non-philosophers use imagination as evidence for logical possibility, thereby taking for granted that a connection exists. We begin our research by looking at two theories of imagination, those of George Berkeley and David Hume, which show how imagination could be used as evidence for possibility. These philosophers analyze imagination as having both a conceptual and a perceptual component, which enable them to connect it with possibility. We then look at two contemporary theories. Gilbert Ryle offers a dispositional analysis of imagination, in which it is presented as a form of thinking, or more specifically as pretending. Bernard Williams discusses the connection between the description of what is to be imagined and the visualizing which follows. He examines ways in which one can be mistaken about what one is imagining. We use the theories of Ryle and Williams as a framework to develop our own theory. We speak of imagining as creating an imaginative experience under a description. An imaginative experience is defined dispositionally as being similar to actual sense experience. Elements of imaginative experience, which is derived from past experience, are then selected to conform to the description. Three constraints on imagination--the referential, the experiential, and the interpretive constraint--determine the acceptability and unacceptability of claims about imagining

Any physical explanation of consciousness seems to leave unresolved the ‘explanatory gap': Isn't it conceivable that all the elements in that explanation could occur, with the same information processing outcomes as in a conscious process, but in the absence of consciousness? E.g. any digital computational process could occur in the absence of consciousness. To resolve this dilemma, we propose a biological-process-oriented physiological- phenomenological characterization of consciousness that addresses three ‘paradoxical’ qualities seemingly incompatible with the empirical realm: The dual location of phenomenal properties ‘out there’ yet ‘in here’ in consciousness; the mysterious ‘thickness’ of the specious present; the feeling of ‘free agency', that we can voluntarily direct our actions, including the act of conscious attention, while at the same time attention and the emotions that direct it seem responsive to physiological substrates with physical causes. These paradoxes are then resolved by relating three elements of consciousness: organismically interested anticipation; sensory and proprioceptive imagery generated by the interested anticipation rather than by sensory input; resonating of these activities with activity stimulated by sensory data, where the interested anticipation precedes the processing of the input. Each of these elements is bridged to physiological processes such that, if they occur in a certain relation to each other, we can understand why they would inevitably be accompanied by the corresponding elements of conscious experience

The sensorimotor theory of cognition holds that human cognition, along with that of other animals, is determined by sensorimotor structures rather than by uniquely human linguistic structures. The theory has been offered to explain the use of bodily terminology in nonphysical contexts, and to recognize the role of experienced embodiment in cognition. This paper defends a version of the theory which specifies that reasoning makes use of mental models constructed by means of action-planning mechanisms. Evidence is offered from cognitive psychology, the neurosciences, and work in primatology and evolutionary theory. Further scientific and philosophical advantages of the theory are discussed: the theory is more parsimonious than its rival, and it offers better accounts of intrinsic intentionality and consciousness. The paper concludes with observations concerning the role of language in thought.

[opening paragraph]: It is easy to conceptualize a problem in a way that prevents a solution. If the conceptualization is entrenched in one's culture or profession, it may appear unalterable. But there is so much precedent for the discovery of fruitful reconceptualizations that in the case of most philosophical and scientific puzzles it is probably irrational ever to give up trying. The notion of qualia, understood as phenomenal properties of sensations that can exist as objects of experience for a conscious subject, is too recent in origin and too specialized in usage to warrant concluding that qualia cannot be understood in terms of physical processes. Humphrey offers an analysis of qualitative mental states that purportedly renders them commensurate with brain states, allowing them to be described in terms of the same dimensions. If his attempt is successful the conceptual gap between mind and body could be closed. Is it successful?

There is a distinction between phenomenal properties and the "phenomenality" of those properties: e.g. between what red is like and what it is like to experience red. To date, reductive accounts explain the former, but not the latter: Nagel is right that they leave something out. This paper attempts a reductive account of what it is like to have a perceptual experience. Four features of such experience are distinguished: the externality, unity, and self-awareness belonging to the content of conscious experience, and the phenomenon of awareness itself. It is argued that these features are accounted for in the work of recent scientists, including F. Crick, G. Edelman, C. Koch and V. Mountcastle. This account reinforces a common way of treating the "knowledge problem."

In this paper I sketch an account of (a) the origin of the terms and concepts of folk psychology, and (b) the true nature of mental states. I argue that folk psychology is built on metaphors for the functioning physical body, and that mental states are neurological traces which serve as schematic ‘mental images’ of those same functions. Special attention is paid to the folk psychology of self-consciousness. In particular, I argue that the notion of introspection is mistaken, and I criticize recent claims of Patricia and Paul Churchland on this subject. I conclude by discussingrecent empirical evidence in support of my approach.

John Searle has argued that one can imagine embodying a machine running any computer program without understanding the symbols, and hence that purely computational processes do not yield understanding. The disagreement this argument has generated stems, I hold, from ambiguity in talk of 'understanding'. The concept is analysed as a relation between subjects and symbols having two components: a formal and an intentional. The central question, then becomes whether a machine could possess the intentional component with or without the formal component. I argue that the intentional state of a symbol's being meaningful to a subject is a functionally definable relation between the symbol and certain past and present states of the subject, and that a machine could bear this relation to a symbol. I sketch a machine which could be said to possess, in primitive form, the intentional component of understanding. Even if the machine, in lacking consciousness, lacks full understanding, it contributes to a theory of understanding and constitutes a counterexample to the Chinese Room argument

This paper argues that phenomenal consciousness arises from the forced blending of components that are incompatible, or even logically contradictory, when combined by direct methods available to the subject; and that it is, as a result, analytically, ostensively and comparatively indefinable. First, I examine a variety of cases in which unpredictable novelties arise from the forced merging of contradictory elements, or at least elements that are unable in human experience to co-occur. The point is to show that the uniqueness of consciousness is comprehensible in terms of a more general kind of emergence. I then argue that phenomenal consciousness essentially involves synchronous activations of representations of ‘identical’ intentional objects with distinct temporal tags, and is thus a case of the emergence of novelty from forced blending of incompatible components. It follows from the general nature of such emergence that consciousness would be indefinable and hence seem mysterious. This analysis will show why phenomenal consciousness would be impossible to resolve into its constituents by the conscious subject. The result is, I hope, a happy blend of physicalist explanation with respectful acknowledgement of the robustness of subjective experience

The enactivist account of consciousness posits that motivated activation of sensorimotor action imagery anticipates possible action affordances of environmental situations, resulting in representation of the environment with a conscious “feel” associated with the valences motivating the anticipations. This approach makes the mind–body problem and the problem of mental causation easier to resolve, and offers promise for understanding how consciousness results from natural processes. Given a process-oriented understanding of the way many systems in non-conscious nature are “proto-motivated” toward realizing unactualized possibilities, and can use symbolic objects to “proto-represent” unactualized possibilities, it becomes more clear how self-organizing systems can subserve subjective consciousness. If a system executes, in a unified way, both a proto-desire and a proto-representation of the same unactualized possibility — in order to provide a kind of causal power for the unactualized possibility — then the result is the familiar experience of phenomenal consciousness

This paper looks at a central issue with embodiment theories in cognition: the role, if any, they provide for mental representation. Thelen and Smith hold that the concept of representations is either vacuous or misapplied in such systems. Others maintain a place for representations , but are imprecise about their nature and role. It is difficult to understand what those could be if representations are understood in the same sense as that used by computationalists: fixed or long-lasting neural structures that represent the sensory stimuli that caused them , or whose “meaning” is fixed innately or in early development for particular functions . The paper proposes a distinctions between, on the one hand, neural patterns, traces of sensory activation that while not in themselves representations are available for representational activity, and on the other the act of representing, which is what gives representational content to neural patterns

  Sydney Shoemaker argues that introspection, unlike perception, provides no identification information about the self, and that knowledge of one''s mental states should be conceived as arising in a direct and unmediated fashion from one''s being in those states. I argue that while one does not identify aself as the subject of one''s states, one does frequently identify and misidentify thestates, in ways analogous to the identification of objects in perception, and that in discourse about one''s mental states the self plays the role of external reality in discourse about physical objects. Discourse about any sort of entity or property can be viewed as involving a domain or frame of reference which constrains what can be said about the entities; this view is related to Johnson-Laird''s theory of mental models. On my approach evidence, including sensory evidence, may be involved in decisions about one''s mental states. I conclude that while Shoemaker may well be right about different roles for sense impressions in introspection and perception, the exact differences and their significance remain to be established

Nicholas Maxwell takes on the ambitious project of explaining, both epistemologically and metaphysically, the physical universe and human existence within it. His vision is appealing; he unites the physical and the personal by means of the concepts of aim and value, which he sees as the keys to explaining traditional physical puzzles. Given the current popularity of theories of goal-oriented dynamical systems in biology and cognitive science, this approach is timely. But a large vision requires firm and nuanced arguments to support it. Here Maxwell's work is weakest; his arguments for contingent mind-body identity and for free will, on which his larger theory depends, are inadequate. The book is valuable both for its comprehensive view of the human condition and its mysteries, and for its demonstration of the difficulties in making such a view coherent.

John Searle has argued that one can imagine embodying a machine running any computer program without understanding the symbols, and hence that purely computational processes do not yield understanding. The disagreement this argument has generated stems, I hold, from ambiguity in talk of 'understanding'. The concept is analysed as a relation between subjects and symbols having two components: a formal and an intentional. The central question, then becomes whether a machine could possess the intentional component with or without the formal component. I argue that the intentional state of a symbol's being meaningful to a subject is a functionally definable relation between the symbol and certain past and present states of the subject, and that a machine could bear this relation to a symbol. I sketch a machine which could be said to possess, in primitive form, the intentional component of understanding. Even if the machine, in lacking consciousness, lacks full understanding, it contributes to a theory of understanding and constitutes a counterexample to the Chinese Room argument.