Constructivism, Expressivism and Ethical Knowledge

Abstract

In the contemporary metaethical debate, expressivist (Blackburn, Gibbard) and constructivist (Korsgaard, Street) views can be viewed as inspired by irrealist ideas from Hume and Kant respectively. One realist response to these contemporary irrealist views is to argue that they are inconsistent with obvious surface‐level appearances of ordinary ethical thought and discourse, especially the fact that we talk and act as if there is ethical knowledge. In this paper, I explore some constructivist and expressivist options for responding to this objection. My conclusion is that, although both constructivists and expressivists can capture other surface‐level features of ethical thought and discourse, the possibility of ethical knowledge causes special problems for these versions of irrealism. I end with some comments about where I think irrealists should begin to look for a response to these special problems, which points, somewhat surprisingly, towards an alternative inferentialist form of irrealism about epistemic and ethical thought and discourse, which is inspired by Sellars.

Keywords:

• meta‐ethics
• ethical knowledge
• ethical irrealism
• ethical constructivism
• ethical expressivism
• inferentialism

Notes

1 Compare Price (2004: p. 73), who makes a distinction between ‘object naturalism=‘ and ‘subject naturalism’.

2 With expressivism, I have in mind the family of views represented by Ayer (1936/1946), Stevenson (1936), Blackburn (1993, 1998), Gibbard (1990, 2003), Timmons (1999), and Ridge (2006). With constructivism, the view has been less well worked out in the metaethical literature, but it owes inspiration to Rawls (1980), and I mostly have in mind Korsgaard (2003) and Street (2008). There are realist versions of expressivism (Copp, 2001) and constructivism (Lynch, 2009: Ch. 8). These views fall outside the scope of my interest in this paper.

3 Both expressivists and constructivists will often claim to be able to capture the other sense in which there appears to be an ‘internal’ connection between ethics and action. I don’t deny that this is possible; rather my aim here is to sketch the central motivation for the differing views.

4 This is mainly because of the pioneering work of Blackburn’s (1984: Ch. 6 and 1993) development of the quasi‐realist programme, which seeks to recapture for irrealists the sorts of semantically infused talk (for example of truth, fact, property, etc.) that tempt many to realism in various areas. For more elucidation and further citations, see n. 9 below.

5 Compare FitzPatrick, 2005 and Hussain and Shah, 2006, where a similar point is made.

6 This claim requires some qualification. For some kinds of luck are compatible with knowledge. For example, knowledge that a pelican just flew by the car may depend on being lucky enough to have looked out the window at the opportune moment. However, other kinds of luck are not compatible. I’m discussing those other kinds of luck here (without attempting to specify them precisely). See Pritchard, 2005: Chs 5–6 for discussion of permissible and impermissible forms of epistemic luck.

7 There’s a promising footnote in Street’s paper where she hints at a constructivist meta‐epistemology, but she doesn’t develop the view there. It’s something like this which I think a constructivist will need to overcome the realist argument from appearances.

8 For example, one of the views defended by Quine (1970), Field (1986, 1994), Horwich (1990), or Wright (1992).

9 This strategy received its first explicit expression when Blackburn proposed the ‘enterprise of quasi‐realism’, which ‘tries to earn, on the slender basis [of expressivist anti‐realism], the features of moral language … which might tempt people to realism’ (1984: p. 171). It has been subsequently developed by inter alia Blackburn (1993), Stoljar (1993), Horgan and Timmons (1993, 2000), Timmons (1999), and Gibbard (2003).

10 Perhaps ordinary speakers are committed to the in‐principle resolvability of disagreements in ethical belief (Smith, 1994: pp. 5–6). However, I doubt that this is true. It’s a different issue whether expressivism is consistent with a defensible theory of ethical disagreement. On this, Egan (2007) poses a serious challenge to expressivism which is outside the scope of the present paper.

11 This is the basic strategy I pursued in my own previous defence of epistemic expressivism; see my 2007.

12 The sort suggested by Grice (1957), and worked out in different ways by Schiffer (1972) and Davis (2003).

13 No participant to the metaethical debate denies that it is highly debatable whether expressivism about ethical claims can provide a semantics for ethical sentences that is even structurally adequate for the task of distinguishing the semantic values of ethical claims under simple logical operations and propositional attitude reports. Recently Schroeder (2008), picking up some strands from inter alia Unwin (1999) and Dreier (2006), has persuasively argued that most extant versions of expressivism have inadequate explanations of the semantic function of negation and propositional attitude reports.

14 For more on involuntarism and why it’s nevertheless compatible with normative epistemic claims, see my 2008a.

15 See my 2008c for an attempt at an account of propositional knowledge which is inspired by Sellars’ slogan and aims to meet a number of desiderata stemming from current debates in epistemology. See my forthcoming for some more discussion of the relationship between epistemic expressivism and epistemic inferentialism.

16 This is the sort of semantic programme developed by Rosenberg (1974) and Brandom (1994).

17 Compare Craig, 1990, Williams, 1992, and Rosenberg, 2002: Chs 5–6).

18 This does raise the question of how to distinguish realism from irrealism in a given domain, if not in terms of truth, fact, belief, or knowledge. I address this question in my 2008b.

19 Compare Brandom, 2001: Ch. 2 for the beginning of such a view. Things get more complicated when we consider ought‐claims attaching, in the first instance, not to actions but to states of things.

20 For helpful comments on previous versions of this material, I’m grateful to Michael Ridge, Ana Barandalla Ajona, the Epistemology @ Edinburgh research group, and the participants of the Constructivism and Normative Epistemology workshop organized by James Lenman at the University of Sheffield.