The nature of cognition is being re-considered. Instead of emphasizing formal operations on abstract symbols, the new approach foregrounds the fact that cognition is, rather, a situated activity, and suggests that thinking beings ought therefore be considered first and foremost as acting beings. The essay reviews recent work in Embodied Cognition, provides a concise guide to its principles, attitudes and goals, and identifies the physical grounding project as its central research focus.

Hauser argues that his pocket calculator (Cal) has certain arithmetical abilities: it seems Cal calculates. That calculating is thinking seems equally untendentious. Yet these two claims together provide premises for a seemingly valid syllogism whose conclusion - Cal thinks - most would deny. He considers several ways to avoid this conclusion, and finds them mostly wanting. Either we ourselves can't be said to think or calculate if our calculation-like performances are judged by the standards proposed to rule out Cal; or (...) the standards- e.g., autonomy and self-consciousness- make it impossible to verify whether anything or anyone (save oneself) meets them. While appeals to the intentionality of thought or the unity of minds provide more credible lines of resistance, available accounts of intentionality and mental unity are insufficiently clear and warranted to provide very substantial arguments against Cal's title to be called a thinking thing. Indeed, considerations favoring granting that title are more formidable than is generally appreciated. Rapaport's comments suggest that, on a strong view of thinking, mere calculating is not thinking (and pocket calculators don't think), but on a weak, but unexciting, sense of thinking, pocket calculators do think. He closes with some observations on the implications of this conclusion. (shrink)

This dissertation examines aspects of the interplay between computing and scientific practice. The appropriate foundational framework for such an endeavour is rather real computability than the classical computability theory. This is so because physical sciences, engineering, and applied mathematics mostly employ functions defined in continuous domains. But, contrary to the case of computation over natural numbers, there is no universally accepted framework for real computation; rather, there are two incompatible approaches --computable analysis and BSS model--, both claiming to formalise algorithmic (...) computation and to offer foundations for scientific computing. -/- The dissertation consists of three parts. In the first part, we examine what notion of 'algorithmic computation' underlies each approach and how it is respectively formalised. It is argued that the very existence of the two rival frameworks indicates that 'algorithm' is not one unique concept in mathematics, but it is used in more than one way. We test this hypothesis for consistency with mathematical practice as well as with key foundational works that aim to define the term. As a result, new connections between certain subfields of mathematics and computer science are drawn, and a distinction between 'algorithms' and 'effective procedures' is proposed. -/- In the second part, we focus on the second goal of the two rival approaches to real computation; namely, to provide foundations for scientific computing. We examine both frameworks in detail, what idealisations they employ, and how they relate to floating-point arithmetic systems used in real computers. We explore limitations and advantages of both frameworks, and answer questions about which one is preferable for computational modelling and which one for addressing general computability issues. -/- In the third part, analog computing and its relation to analogue (physical) modelling in science are investigated. Based on some paradigmatic cases of the former, a certain view about the nature of computation is defended, and the indispensable role of representation in it is emphasized and accounted for. We also propose a novel account of the distinction between analog and digital computation and, based on it, we compare analog computational modelling to physical modelling. It is concluded that the two practices, despite their apparent similarities, are orthogonal. (shrink)

This essay continues my investigation of `syntactic semantics': the theory that, pace Searle's Chinese-Room Argument, syntax does suffice for semantics (in particular, for the semantics needed for a computational cognitive theory of natural-language understanding). Here, I argue that syntactic semantics (which is internal and first-person) is what has been called a conceptual-role semantics: The meaning of any expression is the role that it plays in the complete system of expressions. Such a `narrow', conceptual-role semantics is the appropriate sort of semantics (...) to account (from an `internal', or first-person perspective) for how a cognitive agent understands language. Some have argued for the primacy of external, or `wide', semantics, while others have argued for a two-factor analysis. But, although two factors can be specified–-one internal and first-person, the other only specifiable in an external, third-person way–-only the internal, first-person one is needed for understanding how someone understands. A truth-conditional semantics can still be provided, but only from a third-person perspective. (shrink)

This dissertation examines aspects of the interplay between computing and scientific practice. The appropriate foundational framework for such an endeavour is rather real computability than the classical computability theory. This is so because physical sciences, engineering, and applied mathematics mostly employ functions defined in continuous domains. But, contrary to the case of computation over natural numbers, there is no universally accepted framework for real computation; rather, there are two incompatible approaches --computable analysis and BSS model--, both claiming to formalise algorithmic (...) computation and to offer foundations for scientific computing. The dissertation consists of three parts. In the first part, we examine what notion of 'algorithmic computation' underlies each approach and how it is respectively formalised. It is argued that the very existence of the two rival frameworks indicates that 'algorithm' is not one unique concept in mathematics, but it is used in more than one way. We test this hypothesis for consistency with mathematical practice as well as with key foundational works that aim to define the term. As a result, new connections between certain subfields of mathematics and computer science are drawn, and a distinction between 'algorithms' and 'effective procedures' is proposed. In the second part, we focus on the second goal of the two rival approaches to real computation; namely, to provide foundations for scientific computing. We examine both frameworks in detail, what idealisations they employ, and how they relate to floating-point arithmetic systems used in real computers. We explore limitations and advantages of both frameworks, and answer questions about which one is preferable for computational modelling and which one for addressing general computability issues. In the third part, analog computing and its relation to analogue (physical) modelling in science are investigated. Based on some paradigmatic cases of the former, a certain view about the nature of computation is defended, and the indispensable role of representation in it is emphasized and accounted for. We also propose a novel account of the distinction between analog and digital computation and, based on it, we compare analog computational modelling to physical modelling. It is concluded that the two practices, despite their apparent similarities, are orthogonal. (shrink)

This paper presents a framework for analysing perceptual Twin Earth thought experiments. Visual content normally has an analogue character, and it is argued in this paper that this sets certain constraints on the extent to which Twin Earth thought experiments can be successful. The argumentation in the paper is developed by using examples from visual spatial content. It is argued that visual spatial content can only be “twin-earthed” in a very limited way. Whereas the metrics of space can be twin-earthed, (...) visual experience has a structure that means that it can only be the vehicle for representing entities with geometrical structures. (shrink)

A myth has arisen concerning Turing's paper of 1936, namely that Turing set forth a fundamental principle concerning the limits of what can be computed by machine - a myth that has passed into cognitive science and the philosophy of mind, to wide and pernicious effect. This supposed principle, sometimes incorrectly termed the 'Church-Turing thesis', is the claim that the class of functions that can be computed by machines is identical to the class of functions that can be computed by (...) Turing machines. In point of fact Turing himself nowhere endorses, nor even states, this claim (nor does Church). I describe a number of notional machines, both analogue and digital, that can compute more than a universal Turing machine. These machines are exemplars of the class of _nonclassical_ computing machines. Nothing known at present rules out the possibility that machines in this class will one day be built, nor that the brain itself is such a machine. These theoretical considerations undercut a number of foundational arguments that are commonly rehearsed in cognitive science, and gesture towards a new class of cognitive models. (shrink)

The objective of research in the foundations of Al is to explore such basic questions as: What is a theory in Al? What are the most abstract assumptions underlying the competing visions of intelligence? What are the basic arguments for and against each assumption? In this essay I discuss five foundational issues: (1) Core Al is the study of conceptualization and should begin with knowledge level theories. (2) Cognition can be studied as a disembodied process without solving the symbol grounding (...) problem. (3) Cognition is nicely described in propositional terms. (4) We can study cognition separately from learning. (5) There is a single architecture underlying virtually all cognition. I explain what each of these implies and present arguments from both outside and inside Al why each has been seen as right or wrong. (shrink)

Some have suggested that there is no fact to the matter as to whether or not a particular physical system relaizes a particular computational description. This suggestion has been taken to imply that computational states are not real, and cannot, for example, provide a foundation for the cognitive sciences. In particular, Putnam has argued that every ordinary open physical system realizes every abstract finite automaton, implying that the fact that a particular computational characterization applies to a physical system does not (...) tell oneanything about the nature of that system. Putnam''s argument is scrutinized, and found inadequate because, among other things, it employs a notion of causation that is too weak. I argue that if one''s view of computation involves embeddedness (inputs and outputs) and full causality, one can avoid the universal realizability results. Therefore, the fact that a particular system realizes a particular automaton is not a vacuous one, and is often explanatory. Furthermore, I claim that computation would not necessarily be an explanatorily vacuous notion even if it were universally realizable. (shrink)

I do not think that the world or the sciences would ever have suggested to me any philosophical problems. What has suggested philosophical problems to me is things which other philosophers have said about the world or the sciences. (G.E. Moore, 1942, p. 14).

Theoretical commentaries on AI often operate as a metadiscourse on the way in which science represents itself to a wider public. The sciences and humanities do the same kind of work but in different fields that encourage them to talk about their work differently: science refers to a natural world that does not talk back, and the humanities refer continually to a world with communicative people in it. This paper suggests that much AI commentary is misconceived because it models itself (...) on the way that science represents itself, rather than on the actual practice of science.AI theorists have become increasingly worried about the lack of evaluation in AI, the lack of reflexivity, and the lack of contact with society. Frequently these writers turn to concepts of tacit knowledge to work through these worries. In doing so they are recognising the problem of AI's second-order representation of science and trying to deal with it. However, this recognition of a problem with the representations of science simply turns back to the legitimation crisis of Western politics where many commentators use science precisely as a ‘model’ for western political institutions. They do so because science is one of the few areas of knowledge where it has been legitimate to use plausible methodology for representation that allows for arbitrary designations of authority as well as parallel systems of different authority. However, the plausible rejects any control on reflexivity, assumes an ethnocentric club culture and does not address social context.It is in this sense that the problems of legitimation in political liberalism are similar to those of legitimation in sciences, both are rooted in their uses of representation. AI's link with the representation of science places it in the heart of this debate about legitimacy. This paper suggests that AI does need to learn about reflexivity and that it might well do so by looking at the recent work on experimentation and representation by historians of science, and by looking to the debates about representation by historians of science, and by looking to the debates about representation within the humanities. However, reflexivity may not be enough. Devising rules of thumb for the appropriate halting of reflexivity, is also needed to address social context and take action. (shrink)

This work analyzes the central role of interpretation in non-routine design. Based on this analysis, a theory of computer support for interpretation in cooperative design is constructed. The theory is grounded in studies of design and interpretation. It is illustrated by mechanisms provided by a software substrate for computer-based design environments, applied to a sample task of lunar habitat design. ;Computer support of innovative design must overcome the problem that designers necessarily make extensive use of situated tacit understanding while computers (...) can only store and display explicit representations of information. The automation techniques used for routine design are not applicable: techniques are needed to support creative, tacit human understanding with explicit computer representations. ;The process by which designers transform their tacit preunderstanding into explicit knowledge is termed "interpretation". Interpretation is necessary for solving design problems and collaborating with other designers. Considerable explicit knowledge is thereby generated in the natural course of designing. Often this knowledge includes the most valuable information that can be presented to designers who revisit these design projects or undertake similar projects in the future. If representations of this knowledge have been defined using computer-based design support systems, then the representations can be captured by these systems for the support of subsequent design work. ;A theory of computer support for interpretation in design is presented in three stages. First, the role of interpretation in design is explored by reviewing descriptions of design by Alexander, Rittel, and Schon; by conducting a protocol analysis of lunar habitat design; and by applying Heidegger's philosophy of situated interpretation. Second, this analysis of interpretation is extended to define a theory of computer support. The features of this theory--support for the situated, perspectival, and linguistic characteristics of interpretation--are used to evaluate previous work on software design rationale systems. Third, design principles are discussed for H scERMES, a prototype hypermedia substrate for building computer-based design environments to support interpretation in tasks like lunar habitat design. The hypermedia integrates a perspectives mechanism and an end-user language to capture and modify representations of the design situation, alternative perspectives on design tasks, and terminology for conceptualizing design issues. (shrink)

By following the arguments developed by Vygotsky and employing the cultural-historical activity theory (CHAT) in addition to dialectical logic, this paper attempts to investigate the interaction between psychology and artificial intelligence (AI) to confront the epistemological and methodological challenges encountered in AI research. The paper proposes that AI is facing an epistemological and methodological crisis inherited from psychology based on dualist ontology. The roots of this crisis lie in the duality between rationalism and objectivism or in the mind-body rupture that (...) has governed the production of scientific thought and the proliferation of approaches. In addition, by highlighting the sociohistorical conditions of AI, this paper investigates the historical characteristics of the shift of the crisis from psychology to AI. Additionally, we examine the epistemological and methodological roots of the main challenges encountered in AI research by noting that empiricism is the dominant tendency in the field. Empiricism gives rise to methodological and practical challenges, including challenges related to the emergence of meaning, abstraction, generalization, the emergence of symbols, concept formation, functional reflection of reality, and the emergence of higher psychological functions. Furthermore, through discussing attempts to formalize dialectical logic, the paper, based on contradiction formation, proposes a qualitative epistemological, methodological, and formal alternative by using a preliminary algorithmic model that grasps the formation of meaning as an essential ability for the qualitative reflection of reality and the emergence of other mental functions. (shrink)

Pervasive digitality reveals us as analogue creatures that are unprepared for a world and a logic generated increasingly through automation. Promulgated by capitalism, digitality has created a new form of alienation, one far more powerful and comprehensive than that envisaged by either Marx or Lukács in the analogue-industrial age. Digital alienation-through-automation is the central process in our digital post-modernity. The effects reach increasing registers and spheres of culture, economy and politics. This essay considers the effects within the production of knowledge (...) within higher education. Through a critique of Clayton Christensen’s concept of ‘disruptive technologies’, it develops the argument that knowledge produced in our higher institutions of learning suffer from two forms of ‘knowledge capture’. The first and most comprehensive is knowledge increasingly oriented toward instrumental outcomes, generated in vocationally oriented and market-pressured contexts, and produced by the general-sector institutions; second is knowledge, much of it profound and potentially world-changing—in economics and climate science in the cases discussed here. These are generated in both prestigious general-sector universities, and in the elite Ivy League and Russell Group universities but are unable to gain traction with an ascendant and alien post-Westphalian, post-modern and post-Enlightenment political context—created by a globalising digitality. I argue that unless we recognise our analogue essence, and that this has been alienated by digital automation, then the aporias of digitality, with knowledge production an important element, then our institution of higher learning will continue to produce much valuable knowledge that has diminishing practical (political) effect. (shrink)

In this paper we consider the concept of a self-aware agent. In cognitive science agents are seen as embodied and interactively situated in worlds. We analyse the meanings attached to these terms in cognitive science and robotics, proposing a set of conditions for situatedness and embodiment, and examine the claim that internal representational schemas are largely unnecessary for intelligent behaviour in animats. We maintain that current situated and embodied animats cannot be ascribed even minimal self-awareness, and offer a six point (...) definition of embeddedness, constituting minimal conditions for the evolution of a sense of self. This leads to further analysis of the nature of embodiment and situatedness, and a consideration of whether virtual animats in virtual worlds could count as situated and embodied. We propose that self-aware agents must possess complex structures of self-directed goals; multi-modal sensory systems and a rich repertoire of interactions with their worlds. Finally, we argue that embedded agents will possess or evolve local co-ordinate systems, or points of view, relative to their current positions in space and time, and have a capacity to develop an egocentric space. None of these capabilities are possible without powerful internal representational capacities. (shrink)