The nature of intuitions remains a contested issue in (meta-)philosophy, yet, intuitions are frequently cited in philosophical work, featuring most prominently in conceptual analysis, the philosophical method par excellence. In this paper, we approach the question about the nature of intuitions based on a pragmatist, and particularly Wittgensteinian, account of concepts. To Wittgenstein, intuitions are just immediate ‘reactions’ to certain cognitive tasks. His view provides a distinct alternative to identifying intuitions with either doxastic states or quasi-perceptual experiences. We discuss its implications for intuitions’ role in conceptual analysis and show that a Wittgensteinian account of intuitions is compatible even with ambitious metaphysical projects traditionally associated with this method.
Abstract: Even though much reduced from their original powers as divinely installed organizers of universal knowledge, concepts nevertheless continue to feature prominently in the cognitive and psychological sciences. I will argue that the current discourse around concepts provides neither a clear explanatory target nor a source of explanations for these sciences. Rather, the role is a constitutional one in organizing scientists (and philosophers of cognitive science) around the study of various capacities upon which there has been evolutionary convergence in taxa as distinct as humans and honeybees. On this view, concepts are neither things (e.g., representations) in the brain, nor capacities of the organism (such as categorization, predication, etc.). Rather, the discourse around concepts helps constitute a framework within which these capacities and their underlying mechanisms are investigated empirically and the findings communicated efficiently. I will not argue that the science must be framed in this way, but I will argue that we need stronger reasons for overthrowing the current constitutional arrangements, rejecting any role for talk about concepts, than have been provided thus far.
There seem to be various notions of *concept* at work in cognitive science. Eduoard Machery has argued that the notion is so fragmented that it should be eliminated from our ontology — it serves no useful purpose, at least not in science. I argue that, despite appearances, the notion plays a unified explanatory role in cognitive science, philosophy, and commonsense, and that, whatever the fate of other notions of concept, this unified notion cannot be eliminated without great loss of explanatory power.
Experiences and concepts are traditionally conceived as two distinct types of mental representations. It is also common to portray their relationship in terms of one of the following three paradigms:
I adopt a different ‘capacity approach’: concepts and experience are best conceived of in terms of different yet interconnected abilities.
In developing this un-orthodox approach I argue both against the Fodorian idea that concepts are mental representations and against the pragmatist alternative that they are abilities. Concepts are best conceived as principles of higher mental operations like classification and inference.
I then apply this conception of concepts and their relation to experience to two contrasting ‘myths’: the empiricist myth of ineffable ‘qualia’ and the rationalist myth of ‘spontaneity’.
I consider the many roles that concepts are required to play in cognition, and the different forms they take in different domains, from axiomatically defined functions to vaguely associated collections. I argue that science and science-based professions very commonly use terms that fail the strict test of being well-defined natural kinds, and that there are good reasons for this. The concept of concept is of the same kind as the concept of arthritis, reptile or acid. They usefully capture a range of knowledge and provide the links between different forms of explanation.
What does it take to master abstract terms? More specifically, do we need to invoke abstract representations in order to explain our mastery of abstract terms? Nowadays, many would answer in the negative and claim that embodied representations are all that is required. The idea that embodiment is the key to language comprehension has gathered considerable momentum in recent debates. In the paper, I first outline the basic idea behind embodied accounts of language comprehension and examine some of the evidence that has been put forward in their favor. In a second step, I discuss the neo-empiricist heritage of such accounts and their conception of abstract ideas. In a third step, I raise a novel objection according to which embodied accounts seriously underestimate the cognitive prerequisites of employing abstract terms as well as the cognitive benefits their use affords us. In the remainder of the paper, I argue for the cognitive indispensability of non-embodied, abstract representations by highlighting some of the cognitive benefits they bestow upon language users.
In this talk I defend the view that concepts are accessed in a context insensitive manner in contrast to influential contextualist approaches, according to which contexts are created on the fly. Behavioral and neuroscientific evidence is brought to bear on the issue.
The key question of the talk is: how can be understand the connection between associative cluster of information attached to an object or property and a conceptual representation of an object or a property. The main proposal is that this development can be best explained within the framework of mental files. Furthermore, this framework enables us to develop a new account of concepts, situated between empiricist accounts on which concepts can be fully analyzed in terms of a network of associated perceptual information (Barsalou 1999, Prinz 2004), and rationalist accounts on which concepts are radically different in format from perceptual representations. The latter holds for e.g. Fodor’s (Fodor 1975) language of thought and for Dretke’s theory of digital representations (Dretske 1983). I argue that a theory of concepts should integrate the main aspects of both the empiricist and the rationalist accounts. The alternative view of concepts involves two main claims concerning the organizational structure of concepts: (1) Concepts can be fruitfully understood as consisting of two components, (a) an integrated network of associated information (the empiricist part) used for the categorization of some types of entities (e.g. objects) according to properties, and (b) a handling system that organizes this associative network (the rationalist part); both components are implemented in mental files such that concepts can be characterized as templates based on mental files. This framework can be applied to account for the large variety of different types of concepts including natural kind concepts, descriptive concepts, and phenomenal concepts. If we redescribe phenomenal concepts using mental files we receive a plausible naturalistic understanding of the famous philosophical story of Mary, the omniscient neuroscientist: she already has a concept of red in the black and white room; she only learns to integrate new information about the color appearance into the already existing phenomenal concept.
The explanatory challenge to theories of concepts holds that concepts cannot be a single kind of mental representation storing a single kind of information, because many different representational kinds are required to do the explanatory work in cognitive science. In response to the explanatory challenge, Machery (2009), Weiskopf (2009), and Vicente & Martinez Manrique (2014) have developed eliminativist, pluralist, and hybrid theories of concepts, which each afford the kind CONCEPT a different explanatory role. Here, I argue that we cannot decide between hybrid, pluralist, and eliminativist theories of concepts, because each endorses a different specification of the explananda of cognitive science and, hence, a different interpretation of the relevant cognitive scientific data. It follows that the appeal to cognitive science fails as a methodology for arriving at a consensus about the best theory of CONCEPT.
There is an ongoing debate about whether the philosophers' and the psychologists' notions of concepts are to some extent incommensurable. My talk will briefly touch on this issue, but its focus will be neither philosophy nor psychology, but semantics. I will argue that many systematic patterns of denotation and composition require rich lexical meanings that make such meanings very similar to the conceptual structures that psychologists postulate.
Contextualists hold that concepts are mutable entities that are highly sensitive to aspects of the situations in which they are tokened. Invariantists, by contrast, hold that concepts persistently maintain relatively stable forms. Despite their opposition, in order to achieve empirical adequacy both views have to account somehow for the facts of widespread variability in cognitive performance. The challenge they jointly face centers on the explanatory role that context plays. Either concepts are stable, in which case variability is explained by contextual factors, or else concepts are unstable, in which case their explanatory power is entirely derivative from contextual factors. In either case, context is doing all or most of the causal-explanatory work in capturing variability. Concepts themselves appear to be causally nugatory, and therefore eliminable. I argue that this misconstrues the role that context plays in the explanation of variability, and present an alternative picture of the causal role of context in complex systems. Recognizing that concepts are inextricable from these wider systems should not invite eliminativism, but rather reorient our explanatory priorities towards the system-level structures that generate and sustain this adaptive variation.