On knowing one's own resistant beliefs

Abstract

Influential views on self-knowledge presuppose that we cannot come to know a resistant belief in a first-personal way. Two theses support this supposition: (i) if a belief self-ascription is grounded in the evidence of the person holding the belief, it is third-personal and (ii) we cannot have first-personal knowledge of beliefs we do not control. I object to both of these theses and argue that we can introspect on beliefs of which we lack control even though we cannot assent to their content.

Keywords:

• resistant beliefs
• self-knowledge
• introspection
• authorship
• ownership

Disclosure statement

No potential conflict of interest was reported by the author.

Funding

This research was partly funded by the research projects “Information and the Fragmented Mind” (granted by the Province of Styria), “The Fragmented Mind” (granted by the FWF Austrian Science Fund) and FFI2013-44836-P (granted by the Spanish Ministry of Economy and Competitiveness).

Notes on contributor

Cristina Borgoni is an assistant professor at the Department of Philosophy of the University of Graz. Her research covers various issues related to the nature of beliefs, self-knowledge, rationality and mental agency. She has published on cognitive dissonance cases, Moore's paradox and epistemic akrasia.

Notes

1. The notion of “rational control over beliefs” used in the text is not the one of voluntary control. For criticisms of the notion of voluntary control over beliefs, see Alston (1988) and Hieronymi (2006). For an alternative view of the type of control we might exercise over our beliefs, see Hieronymi (2009).

2. Thesis (i) is often assumed by most approaches on self-knowledge. Examples are the approaches defended by Wrigth ([1989] 2003, 1998), Bar-On (2004), Finkelstein (2008) and Moran (2001). Ryle (1949) and more recently Carruthers (2011) are the few exceptions that do not hold (i). Although this paper aims at denying (i), it does not share Ryle's or Carruthers's general views on self-knowledge. Thesis (i) relates to the immediacy thesis of self-knowledge criticized by Cassam (2011).

3. Thesis (ii) is held more explicitly by approaches that tie self-knowledge to agency, for example, Moran's (2001).

4. A version of this example can be found in Borgoni (2015). Some of the material in this section draws on the discussion in that essay, in particular the interpretation of the case.

5. The fact that Emilia cannot easily eliminate her belief does not mean that she will not eventually eliminate it.

6. According to the anti-judgment view, Emilia believes that not-P although she judges that P. The shifting view holds that when she is explicitly defending the truth of P, she believes that P. But when she is sexist when assessing female politicians, she believes that not-P. The contradictory-belief view holds that she has two mutually contradictory beliefs: she believes that P and believes that not-P. She does not, however, believe the contradiction P and not-P.

7. Gendler's view can be criticized for her assumption that automatic and instinctive reactions are insensitive to evidence, and therefore cannot express beliefs. Research on implicit associations and automatic responses has shown that the supposition is at least disputable (see Dasgupta and Greenwald 2001; Blair 2002). On the other hand, very often, our judgments related to our endorsed assertions are also unresponsive to evidence (Schwitzgebel 2010, 539).

8. For an analysis and criticism of the in-between believing approach, see Borgoni (2014).

9. The occurrence of resistant beliefs has also been supposed in other philosophical contexts. The recent but already much discussed view on delusions held by Bortolotti (2010) presupposes the existence of resistant beliefs. In her argument against anti-doxastic views of delusions, she discusses numerous examples of beliefs that fail different criteria for rationality. One such criterion relates to the notion of epistemic rationality, according to which beliefs ought to be well supported by and responsive to the evidence available to their subjects. This constraint is, however, a normative rather than a constitutive trait of beliefs. A resistant belief is a belief that fails to meet the epistemic rationality constraint. The classic discussion of irrationality (e.g. Davidson [1985] 2004) also accepts the existence of beliefs that resist the individual's epistemic reasons against the belief. A further philosophical context where resistant beliefs appear is the one on self-knowledge. Moran (2001), for example, refers to this class of beliefs (although he does not use the “resistant belief” terminology) to explain the difference between the first- and the third-person perspectives on our beliefs. As I explain throughout the paper, resistant beliefs are assumed to be beliefs that are only known from the third-person perspective on oneself; the subject is alienated from the belief. Gertler (2011) also refers to resistant beliefs (she does not use this terminology either) to discuss self-knowledge. However, in contrast to Moran's project, she refers to this class of beliefs to argue against transparency accounts of self-knowledge. Other authors like Cavell (2006) also refer to resistant beliefs to investigate the special way by which we know our beliefs.

10. It has been argued that some situations like Emilia's allow the subject to affirm a justified and meaningful Moorean sentence (see Borgoni 2015). Emilia currently endorses the idea that men and women are equally competent in politics. Her case is such that despite her conscious endorsements, she cannot eliminate her prejudicial belief; such a belief sticks in her psychology. If Emilia analyzes herself, she is able to discover that she has a resistant belief. If Emilia knows that she has the belief that not-P, she is justified in attributing such a belief to herself. That fact does not conflict with her knowing that the content of such a belief is false. Thus, she is in a position to affirm “P, but I believe that not-P”. She is able to affirm this type of Moorean sentence precisely because her false belief resists her reasons even though she is aware of the belief. In normal cases, identifying that one has a false belief automatically leads to its elimination. But cases of resistant beliefs are not normal cases.

11. See Section 5 (i.e. detectivist models) for the use of this notion of introspection to explain self-knowledge.

12. A stronger notion of first-person authority holds that the subject's belief self-ascriptions are themselves a criterion for the correctness of the corresponding third-personal claim made by someone else (Wright 1998).

13. Moran's account (2001) is one example of such a view. Views based on the transparency principle (see Evans 1982) would also say that Emilia's self-ascription is not first-personal. I will return to this issue in the next section.

14. This line of reasoning grounds the idea that even if Emilia succeeds in affirming a sentence which looks like a Moorean sentence (of the type “P but I believe that not-P”), she would not be affirming a real Moorean sentence. A real Moorean sentence is one in which the belief self-ascription is made from the first-person perspective. I consider such a position misleading for the reasons I offer in this section. Emilia not only uses the first-person pronoun “I” correctly when she attributes a resistant belief to herself, she can also introspect on her resistant belief.

15. This notion of alienation is narrower than the notions used by Moran (2001) and by Hunter (2011). According to the narrower notion, alienation only requires lack of ownership, whereas the alternate notions of alienation require lack of both ownership and authorship over our beliefs. My proposed notion is grounded in the argument that we should not conflate the notions of ownership and authorship over our beliefs. The robust view of alienation seems to conflate both notions.

16. Hunter correctly avoids this description of the case.

17. Finkelstein (2008) attributes this model to Russell, in The Problem of Philosophy (1912). One might feel inclined to attribute such a model to Descartes. However, since it is unclear whether Descartes actually offers a model of self-knowledge in his works (cf. Broughton 2008), we should be more cautious in making such an identification.

18. Armstrong ([1968] 1994) is one of the defenders of this model.

19. This is an unfortunate result, however, since complete failures of self-knowledge seem to give rise to irrationalities rather than to mere mistakes (see Burge [2007] 2011).

20. The perceptual model has received a lot of criticisms in the last decades (see Shoemaker 1994; Burge [2007] 2011).

21. Constitutivism has received many criticisms concerning its accuracy in explaining self-knowledge. As suggested by Finkelstein, we do not take people to be responsible for their mental states in the way that constitutivism characterizes such a relationship. Normally, we do not blame someone who complains of being in pain. According to the model, since the attribution of mental states constitutes them, somebody who claims to be in pain somehow participates in its constitution. But surely the avowal of a headache is not what makes it awful (Finkelstein 2008, 47, 52).

22. Nor can all resistant beliefs be introspected.