Externalism and Authoritative Self‐Knowledge

In this chapter, we first present the characteristic features of what may be regarded as truly first-personal self-knowledge—namely, groundlessness, transparency and authority (§1). The key idea defended in the following is that they are not contingent but necessary and a priori aspects of what goes by the name of " (first-personal) self-knowledge ". For massive failures at self-knowledge would display either the lack of the relevant psychological concepts or failures at rationality, understood in a " thick " sense, which is accordingly specified. These characteristic traits of first-personal self-knowledge are then defended against possible objections stemming from scepticism about knowledge of the content of our own propositional attitudes deriving from the endorsement of semantic externalism and from recent findings in cognitive sciences (§2). It is argued that none of this shows that we never have essentially first-personal self-knowledge. Rather, it shows that its scope is limited and does not extend to our deep-seated and future dispositions, to the dispositional elements of our feelings and emotions, and to the causal relations among our various mental states, which are known, if and when they are, in a third-personal way. The various modes of third-personal self-knowledge are then presented and discussed (§3). Yet all this is compatible with the fact that we have essentially first-personal knowledge of a wide range of mental states and at least of their " narrow " content, if and when they have it, such as our ongoing sensations, perceptions, basic emotions and propositional attitudes as commitments.

2002

This article develops a constitutive account of self-knowledgethat is able to avoid certain shortcomings of the standard response to the perceived prima facieincompatibility between privileged self-knowledge and externalism. It argues that ifone conceives of linguistic action as voluntary behavior in a minimal sense, one cannot conceive ofbelief content to be externalistically constituted without simultaneously assuming that the agent hasknowledge of his beliefs.

Synthese, 1995

Externalism in the philosophy of mind is threatened by the view that subjects are authoritative with regard to the contents of their own intentional states. If extemalism is to be reconciled with first-person authority, two issues need to be addressed: (a) how the non-evidence-based character of knowledge of one's own intentional states is compatible with ignorance of the empirical factors that individuate the contents of those states, and (b) how, given extemalism, the non-evidence-based character of such knowledge could place its subject in an authoritative position. This paper endorses a standard strategy for dealing with (a). The bulk of the paper is devoted to (b). The aim is to develop an account of first-person authority for a certain class of intentional states that is capable of explaining (1) why knowledge of one's own intentional states is peculiarly authoritative, and (2) why such authority is compatible with externalism.

Philosophy and Phenomenological Research, 1999

What is introspective knowledge of one's own intentional states like? This paper aims to make plausible the view that certain cases of self-knowledge, namely the cogito-type ones, are enough like perception to count as cases of quasi-observation. To this end it considers the highly influential arguments developed by Sydney Shoemaker in his recent Royce Lectures. These present the most formidable challenge to the view that certain cases of self-knowledge are quasi-observational and so deserve detailed examination. Shoemaker's arguments are directed against two models of ordinary perception, the "object perception model" and the "broad perceptual model". I argue that the core theses that Shoemaker associated with them are either dubious in their own right or applicable to certain cases of self-knowledge. Overall the aim is to show that there is such a variety of patterns in each case that simple analogies or disanalogies are unhelpful. What is introspective knowledge of one's own intentional states like? Is it in any way like observation or perception? If so, how like? In this paper I hope to make plausible the view that certain cases of self-knowledge, ones in which subjects are authoritative with respect to knowledge of the contents of their own intentional states, are enough like perception to qualify as cases of quasi-observation, and so to justify the claim that authoritative selfknowledge is a form of inner observation. In doing this I am going against the current trend in philosophical literature, which is decidedly hostile to observational models of self-knowledge (Shoemaker 1994, McDowell 1994, Peacocke 1996, 1998, Wright 1998). However, I think that much of this hostility is voiced against the background of assumptions concerning the socalled observation model itself, and specifically, assumptions about what ordinary perception involves, which themselves are questionable. If the perceptual model upon which the analogy with self-knowledge is based is itself a faulty paradigm for ordinary perception, the fact that cases of selfknowledge do not conform to it is neither here nor there. One cannot establish that self-knowledge is not a form of inner observation or perception on the basis of a paradigm that is too crude or inaccurate to fit perception itself. SYMPOSIUM 711 My strategy for making plausible the claim that certain cases of selfknowledge are, broadly speaking, observational, is to focus on various models of ordinary perception and argue that the ones that cause trouble for an observational account of self-knowledge are either unsatisfactory paradigms for ordinary perception, or, where they are not, are applicable to certain cases of self-knowledge. I shall proceed by examining two models that Sydney Shoemaker (1994) discusses extensively, the "object perception model" and the "broad perceptual model", and the reasons he advances for thinking that introspective self-knowledge does not conform to either of these. In the case of the first, I shall argue that the core theses that Shoemaker associates with the paradigm are either dubious in their own right with regard to ordinary perception, or applicable also to certain cases of self-knowledge. In the case of the second, I shall argue that Shoemaker's claim that first-order and secondorder intentional states do not, whereas perceptual states and their objects do, meet the condition of existing, or being capable of existing, independently of one another, is false. Certain cases of self-knowledge, viz. the so-called cogito-type ones, I argue, can be construed as meeting the independence condition. My overall aim is to show that there is such a variety of patterns in each case that simple analogies or disanalogies are unhelpful. If this is right, then the contrast between perception and introspection distorts both modes of awareness. Before embarking on this enterprise, I should make clear what the point of the exercise is and why it is significant. The class of intentional states with which I am concerned is, in the first instance at least, a very restricted but important one. I am interested in what Tyler Burge (1985, 1988, 1996) has referred to as the cogito-type states; states in which one is currently consciously thinking about a thought (with a given propositional content p) while thinking that thought. Such cases of self-knowledge are thought by many to be ones where the subject is authoritative with regard to the nature and contents of the first-order thought. Suppose, for example, that it is true that I am currently, consciously thinking that I am thinking that water is transparent, and that this reflective thought constitutes knowledge. This knowledge seems to be peculiarly immediate or direct (Burge 1985, 1988, 1996, Heil 1988). Descartes (1969) thought that the special status of such knowledge that derives from this immediate relation between subject and thought, or thinker and thought reviewed, makes for a kind of transparency of thought reviewed to the reviewing subject. He thought that it is this difference between a subject's access to the contents of her own intentional states and others' access to those contents that confers an epistemic advantage on the subject in the sense of better placing her to know what the contents of those states are, and so in giving her authority over others with regard to knowledge of such contents (see Burge 1985).

Language, Knowledge, and Representation, 2004

SELF-KNOWLEDGE AND EXTERNALISM BILL BREWER I want to discuss the possibility of combining a so-called 'externalist' theory of empirical content, on which the contents of a person's beliefs are determined in part by the nature of his extra-bodily environment, with a plausible account of self-knowledge, in particular, of a person's knowledge of the contents of his own beliefs. A difficulty for this combination is thought to be that it provides a wholly non-empirical source of knowledge about the mind-independent physical world which is intuitively intolerable. 1 The inference which is held to create this difficulty can be put like this. (E1) I believe that p. (E2) If x believes that p, then x's environment is thus-and-so. ∴ (E3) My environment is thus-and-so. An example of this might be the following. (w1) I believe that water is wet. (w2) If x believes that water is wet, then x's environment contains (or did contain) water. ∴ (w3) My environment contains (or did contain) water.

Rationalität, Realismus, Revision / Rationality, Realism, Revision

The question whether direct, authoritative knowledge of one's own thought contents and an externalist individuation of those contents are compatible has been, and still is, the object of a wide discussion. In this paper I shall present the main lines of the discussion and put forward the makings of an affirmative answer to the compatibility question. Owing to space restrictions, my presentation is bound to be rather sketchy, though I will try to bring out the central points of my perspective as clearly as possible. The simplest way in which incompatibilism could be established would be to start from the premise according to which if content depends on external factors, knowledge of content must depend on knowledge of those factors (cf. Bonjour 1992, p. 136). Given externalism, this premise would entail that in order to know what we think we should first investigate our surroundings, which in turn leads to the conclusion that we do not have direct, authoritative knowledge of our own thoughts. This premise, however, does not seem to be true. Think, for example, that though my existence (metaphysically) depends on my parents' existence, I can know that I exist even if I do not know about my parents' existence. This holds not only in cases of metaphysical dependence, but also in cases of conceptual dependence: someone can know that a certain figure is a triangle and not know that its internal angles measure 180 degress even though this figure's being a triangle depends upon its internal angles' measuring 180 degrees. Some philosophers (Burge 1988, Heil 1988, Davidson, unpubl. ms) have tried to defend compatibilism by noting that reflexive self-ascriptions of thoughts include the content of the ascribed thought itself, whatever the way this content is determined. On the inclusion model of self-knowledge, as this proposal might be called (Bernecker 1996), Cogito-like judgments are reliably true in that they are contextually self-verifying, as Burge insists. A subject need not know what the individuation conditions of his thoughts are in order to correctly ascribe these thoughts, with their right contents, to himself. One major objection to the inclusion model has been put forward by Boghossian (1989, 1992), on the basis of thought experiments in which a subject is unwittingly switched between distinct but observationally undistinguishable environments, say between Earth and Twin Earth. Let's baptize our inter-world traveller 'Peter'. Suppose that Peter is unwittingly 1 Research for this paper has been funded by the Spanish Government's DGES as part of the project PB96-1091-C03-02. My thanks to this institution for its help and encouragement. I want also to express my gratitude to Carlos Ulises Moulines, Julian Nida-Rümelin and Wilhelm Vossenkuhl for inviting me to present a version of this paper to the 3rd Congress of the Gesellschaft für Analytische Philosophie. I am also grateful to Sven Bernecker, Andreas Kemmerling and Nenad Miscevic for their useful comments and criticism.

In this chapter, we consider epistemically robust accounts of self-knowledge. We start with recent inner sense theories (§1), which have been proposed by David Armstrong and William Lycan. Their differences notwithstanding, they hold that self-knowledge is the product of a reliable cognitive mechanism that tracks first-order propositional attitudes and produces the corresponding self-ascriptions. The mechanism is physically realised in our brains and is operative at the subpersonal level. Indeed, according to Lycan, there is a real inner sense faculty, dedicated to keeping track of first-order mental states through the operation of attention. Although the model accounts for groundlessness, transparency and authority, it severs the connection between self-knowledge and rationality and concepts' possession. Hence, it is found to be problematical.

Synthese, 1999

Pacific Philosophical Quarterly, 1999