This paper exploits recent work on the normative and constitutive roles of knowledge in practical rationality, to put pressure on the idea that speakers could communicate without exploiting linguistic knowledge. I defend cognitivism about meaning, the view that speakers have rationally accessible (i.e., implicit rather than tacit) knowledge of semantic facts and principles, and that this knowledge is constitutive of their linguistic competence.

This paper critically examines the argument structure of Fodor's theory of modularity. Fodor claims computational autonomy as the essential properly of modular processing. This property has profound consequences, burdening modularity theory with corollaries of rigidity, non-plasticity, nativism, and the old Cartesian dualism of sensing and thinking. However, it is argued that Fodor's argument for computational autonomy is crucially dependent on yet another postulate of Fodor's theory, viz. his thesis of strong modularity, ie. the view that functionally distinct modules must also (...) have physical counterparts in the neural architecture of the brain. Yet, Fodor offers little or no independent support for this neurological speculation. Moreover, due to the cognitivist underpinnings of Fodor's theory his view of modules as 'mental organs'faces an untenable dilemma that is to be traced back to the earliest history of modem cognitive science, viz. to the rationalist-computationalist research program initiated by Descartes and Male-branche. The tension characteristic for the Cartesian program was one that arose between information correlation and information processing accounts of the transactions between body and mind. Similarly, the tension characteristic for Fodor's theory of modularity is one between a causal account of modules on the model of simple detection mechanisms, and an information processing account of modules on the model of vast and elaborate cognitive systems. It is argued that the resulting concept of a cognitive module Fodorian style constitutes an amalgam of incompatible desiderata that fails to stake out a natural kind for cognitive science. As an alternative account, the final section shows connectionism to be capable of encompassing both Gibsonian and 'new look' accounts of cognitive achievements within one theoretical perspective, thus providing a fruitful interfield theory capable of combining the theoretical resources of the ecological approach with the indispensable theoretical complement provided by psychological processing accounts. This change of perspective would ultimately involve recasting the symbo-functionalist notion of cognitive function along bio-psychological lines. (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.

As AI systems become increasingly complex it may become unclear, even to the designer of a system, why exactly a system does what it does. This leads to a lack of trust in AI systems. To solve this, the field of explainable AI has been working on ways to produce explanations of these systems’ behaviors. Many methods in explainable AI, such as LIME, offer only a statistical argument for the validity of their explanations. However, some methods instead study the internal (...) structure of the system and try to find components which can be assigned an interpretation. I believe that these methods provide more valuable explanations than those statistical in nature. I will try to identify which explanations can be considered internal to the system using the Chomskyan notion of tacit knowledge. I argue that each explanation expresses a rule, and through the localization of this rule in the system internals, we can take a system to have tacit knowledge of the rule. I conclude that the only methods which are able to sufficiently establish this tacit knowledge are those along the lines of Olah : 4901–4911, 2017), and therefore they provide explanations with unique strengths. (shrink)

In "Representations without Rules, Connectionism and the Syntactic Argument'', Kenneth Aizawa argues against the view that connectionist nets can be understood as processing representations without the use of representation-level rules, and he provides a positive characterization of how to interpret connectionist nets as following representation-level rules. He takes Terry Horgan and John Tienson to be the targets of his critique. The present paper marshals functional and methodological considerations, gleaned from the practice of cognitive modelling, to argue against Aizawa's characterization of (...) how connectionist nets may be understood as making use of representation-level rules. (shrink)

Visual modules can be viewed as expressions of a marked analytic attitude in the study of vision. In vision psychology, this attitude is accompanied by hypotheses that characterize how visual modules are thought to operate in perceptual processes. Our thesis here is that there are what we call “intrinsic reasons” for the presence of such hypotheses in a vision theory, that is, reasons of a deductive kind, which are imposed by the partiality of the basic terms in the definition of (...) a module, and by peculiar characteristics of those terms. Specifically, we discuss three hypotheses of functional attributes: successive stages in the action of modules, residual indeterminacy of their effects, and the role of prior constraints. For each of the three, we indicate its occurrence in perceptual psychology, explain corresponding intrinsic reasons, and illustrate such reasons with examples. (shrink)

[This is an earlier (1997), much longer and more detailed version of my entry on LOTH in the _Stanford Encyclopedia of Philosophy_] The Language of Thought Hypothesis (LOTH) is an empirical thesis about thought and thinking. For their explication, it postulates a physically realized system of representations that have a combinatorial syntax (and semantics) such that operations on representations are causally sensitive only to the syntactic properties of representations. According to LOTH, thought is, roughly, the tokening of a representation that (...) has a syntactic (constituent) structure with an appropriate semantics. Thinking thus consists in syntactic operations defined over representations. Most of the arguments for LOTH derive their strength from their ability to explain certain empirical phenomena like productivity, systematicity of thought and thinking. (shrink)

Modular approaches to the architecture of the mind claim that some mental mechanisms, such as sensory input processes, operate in special-purpose subsystems that are functionally independent from the rest of the mind. This assumption of modularity seems to be in tension with recent claims that the mind has a predictive architecture. Predictive approaches propose that both sensory processing and higher-level processing are part of the same Bayesian information-processing hierarchy, with no clear boundary between perception and cognition. Furthermore, it is not (...) clear how any part of the predictive architecture could be functionally independent, given that each level of the hierarchy is influenced by the level above. Both the assumption of continuity across the predictive architecture and the seeming non-isolability of parts of the predictive architecture seem to be at odds with the modular approach. I explore and ultimately reject the predictive approach’s apparent commitments to continuity and non-isolation. I argue that predictive architectures can be modular architectures, and that we should in fact expect predictive architectures to exhibit some form of modularity. (shrink)

Management faces replacement by automated processes. Workflow automation in the information processing sectors of the economy is changing the way information and knowledge workers do their jobs. I consider the changing nature of the information supply chain from the creation of knowledge in firms to the supply of information to consumers. The changing nature of data and the development of data science and machine learning methods that enable the analysis of unstructured data have meant that what was once viewed as (...) tacit knowledge is now just a problem in information processing. The rise of workflow automation in the IT industries and the emergence of the reproducible research movement in the sciences is leading to increased automation of the production of information in both industry and academia and this has profoundly changed the nature of the information supply chain. (shrink)

What is the best way to understand 'applies to' when it is said of a linguistic theory that it applies to a particular language-user? We can answer by saying that a linguistic theory is applicable to an individual language-user just in case that individual tacitly-knows the theory. But this is an uninformative answer until we are told how to understand 'tacit-knowledge'. The end goal of this thesis is to defend the claim that we should take tacit-knowledge to be, simply, knowledge. (...) Towards this end I argue against the satisfactoriness of competing ways of understanding 'tacit-knowledge'. For example, the instrumentalist position is neutral on whether linguistic theories are actually known by the ordinary language-users who tacitly-know them; instead, linguistic theories are to be such that knowing them would enable someone to do whatever it is that the tacit-knower can do. Other competing positions hold that, though tacit-knowledge is a psychological relation of some sort, it is not genuine knowledge. I also attempt to meet specific objections to the claim that a typical language-user could plausibly be said to know a linguistic theory. An objection on which I focus is based on the claim that typical language-users do not possess the requisite concepts for having genuine knowledge of a linguistic theory. The aim in attempting to meet these objections is to open up the way for the linguistic theorist to exploit a paradigm of explanation: explanation of behaviour by knowledge attribution. Attributing knowledge of linguistic theories would be potentially explanatory of linguistic behaviour in exactly the same way that attributions of knowledge in non-linguistic spheres are potentially explanatory of behaviour. Finally, because my emphasis is specifically on semantic theories, I attempt to explicate and defend the claim that a semantic theory could and should have the form of a theory of truth. (shrink)

If Australasian philosophers constitute the kind of group to which a collective identity or broadly shared self-image can plausibly be ascribed, the celebrated history of Australian materialism rightly lies close to its heart. Jack Smart’s chapter in this volume, along with an outstanding series of briefer essays in A Companion to Philosophy in Australia and New Zealand (Forrest 2010; Gold 2010; Koksvik 2010; Lycan 2010; Matthews 2010; Nagasawa 2010; Opie 2010; Stoljar 2010a), effectively describe the naturalistic realism of Australian philosophy (...) of mind. In occasional semi-serious psychogeographic speculation, this long-standing and strongly-felt intellectual attitude has been traced back to the influences of our light, land, or lifestyle (Devitt 1996, x; compare comments by Chalmers and O’Brien in Mitchell, 2006). Australasian work in philosophy of mind and cognition has become more diverse in the last 40 years, but is almost all still marked, in one way or another, by the history of these debates on materialism. (shrink)

Thinking is special. There is nothing quite like it. Thinking.