Normative roles, conceptual variance, and ardent realism about normativity

ABSTRACT

In Choosing Normative Concepts, Eklund considers a “variance thesis” about our most fundamental (and seemingly most “authoritative”) normative concepts. This thesis raises the threat of an alarming symmetry between different sets of normative concepts. If this symmetry holds, it would be incompatible with “ardent realism” about normativity. Eklund argues that the ardent realist should appeal to the idea of “referential normativity” in response to this challenge. I argue that, even if Eklund is right in his core arguments on this front, many other important challenges for ardent realism remain that also stem from the issues about possible variance in normative concepts that he considers. Following this, I introduce further issues about conceptual variance. These are issues that arise within the context of the framework that Eklund proposes the ardent realist use to confront the variance theses he considers. In particular, the issues concern what normative role as such is, as well as, relatedly, which roles associated with a concept (or predicate) get to count as part of its normative role. The upshot is that issues about conceptual variance in normative domains might be even more challenging for the ardent realist to deal with than Eklund argues.

KEYWORDS:

• Ardent realism
• normative role
• Matti Eklund

Acknowledgements

Thanks to David Braddon-Mitchell, Mark Budolfson, Alexis Burgess, Herman Cappelen, Billy Dunaway, Matti Eklund, Tristram McPherson, Knut Skarsaune, Rachel Sterken, Tim Sundell, and Daniel Wodak for helpful discussion and feedback.

Disclosure statement

No potential conflict of interest was reported by the author.

Correction Statement

This article has been republished with minor changes. These changes do not impact the academic content of the article.

Notes

1 My use of the phrase “alarming symmetry” to describe the situation here draws on (McPherson 2018b, 2019).

2 See (Eklund 2017), chapter 1. See Eklund's discussion chapter 8, section 3, for some thoughts on the possibility of non-realist views that still make good on key motivations behind “ardent realism”.

3 On the idea of “authoritative” (or “robust”) vs. “generic” (or “merely formal”) normativity, see (Copp 2005; McPherson 2011, 2018a; McPherson and Plunkett 2017; Plunkett and Shapiro 2017).

4 On this point, see (McPherson 2018a).

5 In turn, if one's focus with authoritative normativity is on the “all-things-considered” normatively authoritative “should”, one can then think of there being a range of properties that play a contributory role relative to it; e.g. normative reasons that contribute to determining what one should (authoritatively) do. Parallel points apply to the relevant “authoritative” kind of value. See (McPherson 2018a; McPherson and Plunkett Forthcoming).

6 This is arguably true for a wide range of work in the area, including (Korsgaard 1996; Gibbard 2003; Street 2006; Schroeder 2007; Enoch 2011; Parfit 2011).

7 For defenses of the coherence of the idea, as well as theories of what it amounts to, see (Wodak 2018; McPherson 2018a). For critical discussion, see (Copp 2005; Tiffany 2007; Baker 2018).

8 As I read him, Eklund assumes that the thinnest concepts are the most authoritatively normative ones. However, I think it is conceptually possible that this is not so, and so one would need to provide (which Eklund does not) an account of why “thin” concepts are more authoritative than “thick” ones. For discussion, see (McPherson and Plunkett Forthcoming).

9 For Eklund's own discussion of authoritative (or “robust”) vs. generic (or “merely formal”) normativity, see (Eklund 2017, 184–186). As suggested by my frontloading this cut here in setting up Alternative, in a way that Eklund does not, I think that Eklund underappreciates the import of that distinction for his own project. For connected thoughts, see (McPherson 2019).

10 In this paper, I use smallcaps caps (e.g. cat) to pick out concepts. Single quotation marks (e.g. ‘cat’) are used strictly to mention linguistic items. Double quotation marks (e.g. “cat”) are used for a variety of tasks including quoting others' words, scare quotes, and mixes of use and mention.

11 Following Eklund, I am putting things here in terms of extensions. It should be noted, however, that this isn't crucial to spelling out the core idea in Alternative. The core idea in Alternative is that we and the alternative community employ normative concepts with the same normative role, but that these concepts are alternatives to each other in that their application has different upshots for thought. That is most straightforwardly captured in terms of thinking of them delivering non-identical extensions, or in terms of picking out alternative normative properties. But the core idea could here could be translated without discussion of extensions (or properties), if one were attracted to views of the meaning of these terms in which they didn't yield extensions (or denote properties). However, given the focus on ardent realism, these issues are not crucial for my discussion.

12 For a representative example of his use of “normatively privileged” in this way, see (Eklund 2017, 44).

13 For connected discussion about difficulties that arise in using concepts in forming normative questions about whether to use those very concepts, see (Burgess and Plunkett 2013b; Burgess Forthcoming).

14 Note that another possible take on the Further Question is that, in the Further Question, we aren't interested in which normative concepts we should use, but rather just which ones to use. If this “what to do” question isn't framed using normative concepts at all, then that might get around some of the issues involved here in the first horn of Eklund's dilemma for Alternative-friendly theories. (For a defense of this kind of idea in a related context, also concerning the selection of alternative normative concepts, see (Skarsaune Manuscript)). There are two main issues with this proposal in the present context. First, it is not clear that “what to do” questions are fully distinct from normative questions about what one should do. On this front, consider Allan Gibbard's view in (Gibbard 2003), according to which the topic of what to do just is the topic an agent is thinking about when she thinks about what she should do. On other views, these questions come apart much further: such that one can think one should do something but still ask oneself the practical reasoning question “what to do?” in light of that. For example, see (Silverstein 2017). In the dialectical context we are considering, someone who pursues this line thus takes on the burden of first showing why the Gibbard-style view is wrong, and then, second, defending a view on which the further practical reasoning question that Silverstein highlights really is devoid of all normative concepts. Second, and perhaps more importantly in this context, it is hard to see how the ardent realist is going to be happy thinking that the choice between alternative sets of normative concepts (e.g. between ours and those of the alternative community) is just going to come down to something we choose in a way that isn't fundamentally regulated by, or something that (put intuitively) should be responsive to, fully objective, mind-independent normative facts about how we should choose. So, at the end of the day, at best, this line of saying the Further Question is really a “what to do” question puts us squarely on the second horn of the dilemma.

15 I say “put roughly” because there are obviously thorny issues here about using “normatively privileged” to describe what's at issue here, as Eklund emphasizes. For example: why not be concerned with what is normatively* privileged (where that is a property picked out by the alternative normative concepts employed by Bad Guy). The issues here in using the language of “normatively privileged” as a gloss for what's intuitively at issue here are tied to the challenges for Alternative-friendly views that Eklund discusses (and which I glossed above in describing his dilemma for such views). For my purposes here, I am just sticking with “normatively privileged” as the gloss here, following Eklund himself (as on (Eklund 2017, 44)).

16 For example, consider here the sort of constitutivist view in (Korsgaard 1996).

17 See (Enoch 2011, ch. 3), drawing on his arguments in (Enoch 2006). It should be noted that Enoch does think that facts about rational inescapability play an important role – namely as a key part of the argument for why his preferred form of Robust Realism is the correct metanormative theory. See (Enoch 2011, ch. 3). So maybe the ardent realist should take a page from Enoch's book here and go for “rational inescapability” here in this context as well, rather than psychological inescapability. This doesn't seem that promising in the current context, however. This is because ‘rational’ here is a normative term, and would thus be something where there is (at least potentially) an alternative concept used to play the same normative role associated with Enoch's ‘rational’. So it doesn't seem that ardent realists will make much progress in appealing to rational inescapability in responding to Alternative.

18 See (Lewis 1983). For connected discussion, developing this idea further, see (Sider 2011).

19 See (Eklund 2017, 202–203).

20 See (Eklund 2017, 9, 28–32).

21 Note that this response might be used independently of the normative sparseness reply, or in combination with it.

22 Notice that these issues are distinct from skepticism about the very coherence of the idea of “normative role” in the first place, or skepticism about whether concepts/words in fact have normative roles. Those are important issues too, as Eklund emphasizes. But they are different than the ones I focus on.

23 A further question (which I am not pushing on here) is what exactly it means for a concept to be “associated” with a normative role. See (Eklund 2017, 45–49) for discussion of that issue.

24 For some work that emphasizes this role of normative concepts, see (Korsgaard 1996; Gibbard 2003).

25 For example, compare (Gibbard 2003) to (Smith 1994).

26 For work that emphasizes this role of normative concepts, see (Railton 2003; Darwall 2006).

27 For work that emphasizes this role of normative concepts, see (Railton 1986; Smith 1994; Manne 2014).

28 For work that emphasizes this role of normative concepts, see (Gibbard 1990; Blackburn 1998).

29 This sort of “upstream” role of normative concepts (emphasized, for example, in (Smith 1994; Enoch 2011)) is largely absent from Eklund's discussion, as McPherson discusses in (McPherson 2019).

30 For an overview of the issues involved in this debate, see (Faraci and McPherson 2017).

31 See (McPherson and Plunkett 2017; Plunkett and Shapiro 2017).

32 See (Eklund 2017, ch. 10). Eklund describes this methodological approach in terms of “conceptual engineering”, drawing on (Scharp 2013). For a similar use of the term ‘conceptual engineering’, see (Cappelen 2018). For connected discussion, see (Burgess and Plunkett 2013a, 2013b) on the topic of “conceptual ethics”.

33 See (Cappelen 2018, ch. 6). It should be underscored that, on Cappelen's view, a change in metasemantics here need not result in change in topic. So, for example, we could still be talking about the topic of marriage, even if the metasemantics (and, indeed, the semantics) for ‘marriage’ has changed. See (Cappelen 2018, chs. 9–11).