Abstract
The paper is a critical examination of the metaethical position taken up recently by Terence Horgan and Mark Timmons, called ‘cognitivist expressivism’. The key component of the position is their insistence that some beliefs are nondescriptive. The paper argues against this thesis in two ways: First by sketching an independently plausible account of belief, on which belief is essentially a certain kind of descriptive representational state; and second by rebutting Horgan and Timmons’ positive arguments in favor of their account. The final section argues that Horgan and Timmons’ view cannot survive abandonment of the thesis that moral beliefs are nondescriptive in character.
This is a preview of subscription content, log in via an institution to check access.
Similar content being viewed by others
Notes
Classic works of the new naturalism include Railton (1986), Boyd (1988), and Brink (1989). Contemporary forms of nonnaturalism are developed by Shafer-Landau (2003) and Audi (2004). Cognitivist error-theories are set out in Mackie (1977) and Joyce (2001). Kalderon (2005) argues for a noncognitivist sort of error-theory. Recent versions of expressivism are featured in Blackburn (1993, 1998) and Gibbard (1990, 2003).
These points are not uncontroversial. For example, it is widely believed that if perceptual experience is to play a role in warranting perceptual beliefs then the former must be able inferentially to support the latter, which would seem to require that the content of perceptual experience be conceptual or propositional; see for example Brewer (1999). This form of argument overlooks noninferential, nonpropositional forms of epistemic warrant; cf. Burge (2003, section V). As Burge notes, there is also strong empirical reason for allowing that perceptual experience is nonconceptual. For a great many kinds of animals perceptually represent their surroundings; but in most such cases it is not plausible that the animals have concepts or propositional attitudes.
Horgan and Timmons have an ambiguous relationship with the use of the direction of fit metaphor to distinguish different sorts of mental states. They claim at one point (2000, p. 124) that it presupposes the thesis of semantic unity, the doctrine that all beliefs, or belief-like states, function to describe the world. If this were so then they would of course be committed to rejecting it. But later in the same paper (146) they treat the idea that belief aims at truth as a platitude, and attempt to make room for it in their theory. Though I will not go into detail on the point, I believe that this tension is indicative of a more general lack of comportment between Horgan and Timmons’ rejection of the idea that beliefs are descriptive, on the one hand, and their endorsement of a deflationary conception of truth and reference, on the other. If reference, representation, truth, and description are all understood in a deflationary way, it is a question whether an expressivist should—or even could—deny that moral beliefs are descriptive in character. If, on the other hand, description is held to a different standard than are the other semantic notions mentioned, some explanation or justification is surely owed.
Horgan and Timmons use ‘representational’ interchangeably with ‘descriptive’. This causes a problem, inasmuch as they presumably do not wish to deny that nondescriptive beliefs, like all propositional attitudes, relate a thinker to an intentional content. (I think it would be obviously incoherent to deny this, at least in the absence of an adequate alternative characterization of what a belief is, which they do not provide.) It is a problem because ‘intentional’ and ‘representational’ are themselves typically used interchangeably in the literature, and Horgan and Timmons explicitly and repeatedly deny that evaluative beliefs are representational. As a consequence of this, I am in this paper somewhat awkwardly reserving the term ‘descriptive’ for the kind of content characteristic of mental states with mind-to-world direction of fit. And I will sometimes employ the ugly phrase ‘descriptively representational’ to mark the same sort of distinction. Thus I am taking it that Horgan and Timmons allow that all beliefs are representational, in the sense of intentional – which is just to say that they are about something – but yet deny that evaluative beliefs are descriptive.
Earlier in the same paper Horgan and Timmons put the point more directly, calling attention to “those phenomenological features that [moral judgments] share with nonmoral descriptive beliefs … which (together with considerations of semantic assessability and functional role) we claim qualify them as genuine beliefs” (263).
Note that I am ignoring their extensive treatment of the semantic assessability of moral judgments here because a proper consideration of it would require discussing deflationary views of truth and reference. As indicated in the introduction, the present paper aims to discuss only one of the two central strategies for conjoining expressivism with cognitivism (but cf. note 5).
Such a thesis was suggested to me by an anonymous referee.
Horgan and Timmons claim, in the quotation given earlier, that the what-is-likeness involved in the phenomenology of belief is experienced, ‘perhaps peripherally in consciousness’, as being a sufficient reason for categorizing as one does, and as a judgment that is apt for assertion. They may hold that all of their claims about the way in which occurrent belief is typically experienced should be qualified in this manner. Two points in response: First, it is difficult to make sense of the notion of experiencing something, as being a given way, peripherally in consciousness. Certainly there may be degrees of conscious awareness, and like most concepts that of consciousness may be vague. But if it is true that one experiences X as being F, even peripherally, then it must be true that one was in some measure aware of X as being F. And it is not plausible that in all cases of occurrent belief one is so aware of the referent of one’s thought (if it has one). Being peripherally aware of something in consciousness, like being directly aware of it, takes time. And as I will argue one can form more beliefs in a short temporal period than there is time to consider in one’s space of conscious awareness. Second, the point made in the last paragraph of the section does not turn on the degree of awareness present in occurrent belief. For if one is aware, in any sense at all, of X’s being F then one must possess and exercise the concept F.
Notice that acceptance of this analysis-sketch does not commit one to functionalism. For it does not require that the relevant forms of behavior be nonintentionally characterized. Functionalist analyses, if they are to have any materialist teeth, must disallow use of representational notions in their specification of the dispositions to behavior crucial to possession of the relevant propositional attitudes. I think that no such analyses can be adequate to the phenomena, and so do not build-in such a requirement into my characterization of belief.
That phenomenology is largely irrelevant to the nature of propositional thought is a central lesson of Wittgenstein’s Philosophical Investigations. The point was later exhibited, in different ways, in work on the nature of translation and interpretation by Quine and Davidson, and in functionalist treatments of representational mental concepts and states.
Smith (1987, pp. 45ff.) argues that even desires have no essential connection to phenomenology. There are large points in his paper that I do not accept. But Smith seems to me fundamentally correct to want to separate any qualitative content that propositional attitudes such as desires may possess, on the one hand, from their propositional content, on the other; and to suggest that a phenomenological conception of the essence of the latter is not plausible.
Thanks to an anonymous referee for suggesting this line of response to the argument presented in the text.
I am not certain how Horgan and Timmons understand the relevant notion of experiencing. They claim at one point that moral judgments “are experienced as psychologically involuntary, and as grounded in reasons: given one’s evidence, one cannot help but make certain moral judgments” (2000, p. 128). There is no standard notion of experiencing on which the fact that one’s evidence compels one to believe that p implies that one experiences anything as involuntary, or as grounded in reasons. Perhaps they mean simply that if one cannot help but believe that p, in the face of evidence, then one’s belief is involuntary, and is grounded in reasons. One problem with this claim is that not all epistemically warranted beliefs are supported by reasons – some are grounded rather in nonconceptual perceptual experience. Cf. note 3. More important, though, is the fact that this ostensible conception of experience has nothing to do with phenomenology, and so does nothing to further the view that moral judgments express beliefs, despite their nondescriptive nature. Many mental states are involuntary, and are grounded in reasons, without so much as being present to consciousness. The formation of certain dispositional beliefs serve as examples.
At (269), and also in their (2000, pp. 129, 134, 141).
Darwall (1983, p. 54); Brink (1989, p. 42); and Smith (1994, p. 61). Strictly speaking, Smith is here characterizing a view he rejects. He thinks that if one sincerely judges that one ought to , then one will be so motivated—unless one is practically irrational. But I am understanding ‘motivation’ in such a way that one could be motivated to , and one’s motivation could yet be overridden, perhaps by weakness of will. In this sense Smith accepts the view as stated.
Also note that the issue between internalism and externalism here is often put in terms of the possibility of the amoralist. This is equivalent to the characterization I have offered, inasmuch as the amoralist is precisely one who sincerely judges that she ought to φ, and yet is not motivated to any extent to do so. Since Horgan and Timmons accept the possibility of the amoralist, they are externalists about moral judgment and motivation.
For an example of an externalist who allows and even emphasizes this close sort of connection between moral judgment and motivation, see Shafer-Landau (2003, chapter 7).
References
Audi R (2004) The good in the right. Princeton University Press, Princeton
Blackburn S (1993) Essays in quasi-realism. Oxford University Press, Oxford
Blackburn S (1998) Ruling Passions. Oxford University Press, Oxford
Boyd R (1988) How to be a moral realist. In: Sayre-McCord G (ed) Essays on moral realism. Cornell University Press, Ithaca, NY
Brewer B (1999) Perception and reason. Oxford University Press, Oxford
Brink D (1989) Moral realism and the foundations of ethics. Cambridge University Press, Cambridge
Burge T (2003) Perceptual entitlement. Philos Phenomenol Res 67:503–548
Darwall S (1983) Impartial reason. Cornell University Press, Ithaca, NY
Gibbard A (1990) Wise choices, apt feelings. Harvard University Press, Cambridge, MA
Gibbard A (2003) Thinking how to live. Harvard University Press, Cambridge, MA
Horgan T, Timmons M (2000) Nondescriptivist cognitivism: framework for a new metaethic. Philos Pap 29:121–153
Horgan T, Timmons M (2006) Cognitivist expressivism. In: Horgan T, Timmons M (eds) Metaethics after Moore. Oxford University Press, Oxford
Humberstone L (1992) Direction of fit. Mind 101:59–83
Joyce R (2001) The myth of morality. Cambridge University Press, Cambridge
Kalderon M (2005) Moral fictionalism. Oxford University Press, Oxford
Mackie JL (1977) Ethics: inventing right and wrong. Penguin, New York
Railton P (1986) Moral realism. Philos Rev 95:163–207
Shafer-Landau R (2003) Moral realism: a defence. Oxford University Press, Oxford
Smith M (1987) The Humean theory of motivation. Mind 96:36–61
Smith M (1994) The moral problem. Blackwell, Oxford
Tenenbaum S (2006) Direction of fit and motivational cognitivism. In: Shafer-Landau R (ed) Oxford studies in mathematics, volume 1. Oxford University Press, Oxford
Acknowledgment
I am grateful to Russ Shafer-Landau, and to a pair of anonymous referees, for good advice.
Author information
Authors and Affiliations
Corresponding author
Rights and permissions
About this article
Cite this article
Majors, B. Cognitivist Expressivism and the Nature of Belief. Ethic Theory Moral Prac 11, 279–293 (2008). https://doi.org/10.1007/s10677-007-9100-8
Received:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s10677-007-9100-8