Constitutivism and normativity: a qualified defence

Abstract

In this article, I defend a meta-normative account of constitutivism by specifically addressing what I take to be a fundamental criticism of the constitutivist stance, namely, the objection that constitutive standards have conceptual, not normative, force, and so that no practical normativity can be extracted from them as constitutive of agency. In reply to this objection, I argue that the conceptual role of the standards constitutive of agency – their applying to us by virtue of our being the kinds of creatures we are – does not exclude, but rather combines with, the normative role the same standards play in our practical life.

Keywords:

• normativity
• practical reason
• agency
• constitutivism
• metaethics

Acknowledgements

My indebtedness goes to those who have offered helpful comments and critical remarks on previous versions of this essay. I should like to thank in particular David Enoch, Stanley L. Paulson, Corrado Roversi, Filippo Valente and the anonymous reviewers for this journal. Needless to say, responsibility for the views I express here, as well as for any errors, rests solely with me.

Notes

These expressions are used in Ferrero (2009, 304), Setiya (2003, 340), and Enoch (2006, 177), respectively.

I will not argue my claim that agency is somehow more fundamental than action. Instead, I will ask the reader to take this as an intuitive understanding of the relationship between the two, bearing in mind, however, that some authors take agency and others action as the basic starting point of their constitutivist construction. Thus, Gewirth (1978, 21–42) and Velleman (2000, 123–43 and 170–99) lay emphasis on action, while Korsgaard (1996, 90–130, 2009), Schapiro (1999), and Rosati (2003) proceed from agency instead. Needless to say, the two notions are deeply interlinked, and it is therefore somewhat arbitrary and elusive to attempt any clear-cut distinction between them. This is especially evident in Korsgaard (2009) and Schapiro (2001), whose discussions of agency bear directly on action, and vice versa.

Paradigmatic in this sense is the attitude taken by Korsgaard (2009, 29–32), who does not explicitly take up the question as to whether and how normative standards can legitimately derive their normativity from the constitutive standards they embody, such that the normative can be justified simply by pointing out the constitutive standards from which it issues. This attitude is most apparent in Korsgaard's claim that “constitutive standards meet skeptical challenges to their authority with ease … once you've answered the technical questions there is no further room for doubting that the constitutive standard has normative force” (Korsgaard 2009, 29). She (2009, 32) later remarks, in the same spirit, that

constitutive standards have unquestionable authority, while external standards give rise to further questions, and leave space for skeptical doubts. How then can we ever give authority to an external standard, except by tracing its authority back to a constitutive one?

 Statements such as these suggest that Korsgaard takes it to be a position in need of no sustained argument that the force of constitutive standards is at once conceptual and normative. Accordingly, while certain of Korsgaard's considerations may be constructed as addressing the question of whether and how constitutive standards can be normative, one cannot find in her work any systematic attempt to deal with these problems head on. The other constitutivist writers have not approached the question any differently.

This criticism can be found in Roversi (2011, 288–91), among other places, observing that constitutivism seems to locate normativity beyond the sphere of practical choice: if practical standards owe their normativity to their being constitutive of agency, and if agency is just who we essentially are, then we do not properly choose to follow those standards, for that is simply what we would do anyway, on account of our agency.

The objection here – for which see O'Day (1998, 62–9) and Gibbard (1999, 161–4), among others – is that constitutivism may be able to explain why the practical standards grounded in agency should bind me, as an agent, but not why they should be binding on other agents, too.

Those who make this criticism consider agency a construct too thin and formal for us to be able to extract substantive practical standards on that basis alone. This objection is prefigured in Gibbard (1999, 144–9) and fully developed in Setiya (2003, 371–84, 2007, 14–18).

This objection – framed in Roversi (2011, 301–4) – is that agency may not be the primitive notion it is made out to be: if that is so, then agency must owe its normativity to something even more primal, which in turn faces the same problem of showing it is itself a primal notion.

This objection – known in the literature as the “Why be an agent?” objection – can be found in O'Day (1998, 69–82), Skorupski (1998, 345), Railton (2003, 299–321), FitzPatrick (2005, 663–77) and Enoch (2006), among others.

But see, for example, Korsgaard (1996, 134–45), Ferrero (2009), and Velleman (2009, 135–44), where some of these criticisms are addressed.

This concern emerges from, among other places, Railton (2003, 309–13), Fitzpatrick (2005, 663–70), and Enoch (2006, 187–92).

A clarification is in order here to prevent possible misunderstandings. The objection, as I interpret it, does not state that the constitutive features of agency are by themselves not normative and yet they can end up generating normative requirements – at least, hypothetical normative requirements – when they are combined with some further elements (such as a personal desire, interest, goal or commitment). In my reading of it, the objection is more radical than that: it should be taken to mean that the constitutive does not partake of the normative at all and, hence, any normative dimension a constitutive standard may happen to be associated with derives entirely from outside, and is completely independent of, that standard, to which it owes nothing. In other words, what is constitutive of agency by itself neither is nor can become either hypothetically or categorically normative since constitutive standards are normatively inert and so through none of their combinations can they provide access to the normative sphere. It is this reading of the objection that justifies the two-step argumentative strategy I resort to in my reply, where first (in Section 3), I seek to show that the discontinuity between the constitutive and the normative (in any of its forms – hypothetical normativity and categorical normativity), as this discontinuity is asserted by the objection under consideration, is not real, and (only) then (in Section 4), I move to provide the outline of an argument showing that the constitutive features of agency have the specifically categorical kind of normative force.

A discussion of the distinction can be found in Rawls (1955), which has deeply influenced the debate ever since. For an assessment of some of the issues at stake, see also Searle (1969, 33–42).

Apart from the constitutivity they both share, the difference between a constitutive rule and a constitutive standard, or norm – in this article, I regard the terms “standard” and “norm” as synonyms – can best be understood by reference to a distinction, which has now become familiar in legal philosophy: the distinction between rules and principles, a distinction to which no one has made as great a contribution as Alexy. In Alexy's legal philosophy legal requirements can take either of two forms: that of “optimisation commands”, as is the case with general principles of law, or that of “definitive commands”, as is the case with legal rules. The idea of an optimisation command (originally set out in Alexy 2000 and 2002, 44–69) is that of a norm, or standard, “commanding that something be realised to the highest degree that is actually and legally possible” (Alexy 2000, 294–5, cf. Alexy 2002, 47–8). A definitive command, by contrast, is a rule, or directive, which will not admit of any gradation: you either obey it or disobey it; there is no middle ground, for the directive requires its addressees to do exactly as it says.

This thought is briefly stated in von Wright (1963, 5–6). But when he returns to the same topic (1971, 151–2) he does not develop his earlier observation, and indeed may even be seen as withdrawing it.

von Wright explains these two roles thus: “From the point of view of the game itself, the rules determine which are the correct moves, and when viewed from the point of view of the activity of playing, the rules determine which are the permitted moves” (von Wright 1963, 6, emphasis added). This twofold nature of constitutive rules – with their primary conceptual role and their concurrent normative role – has also been pointed out by Searle, observing that constitutive rules “constitute (and also regulate) forms of activity whose existence is logically dependent on the rules” (Searle 1969, 55, emphasis added, cf. Searle 1995, 48). Searle, however, does not develop the point or attempt to explain the twofold role so observed.

This statement should not be interpreted as implying that one can never be responsible for acting either in accordance with or against the standards constitutive of agency. In fact, whilst some remarks I make may lead one to think otherwise, here I do not intend to engage with questions concerning responsibility, the discussion of which is well beyond the scope of this contribution (I would like to thank one of the journal's anonymous reviewers for bringing to my attention the consequences that my argument may have for issues concerning the responsibility for our conduct).

The study (Roversi 2010) is in essence a critical analysis of Searle's account of constitutive rules, and its main thesis is that a proper understanding of constitutive rules requires one to look at the context within which a constitutive rule is framed. Although Roversi does consider whether and how a constitutive rule can also perform a regulative function, that is not his main concern. Indeed, his study does not frontally address the relation between the constitutive and the normative; nor does he engage with the debate on constitutivism or with the idea of agency. For, he is specifically concerned with staking out a position in the discussion of Searle's theory of constitutive rules, without also taking other theories into account. And so, what I offer here is really an extension of Roversi's framework to a broader set of questions: the normativity of practical reason, the constitutivist stance, and the role that agency plays in shaping how we ought to behave. How this extension coheres with Roversi's broader theoretical stance is a matter that cannot be taken up here.

Consider, for example, the technical question of how a chess piece moves on the chessboard: this is a question we can answer simply by pointing out the constitutive rules of chess pertaining to that piece.

Consider in this case the question: Why, in playing a chess match, should one try to win playing by the rules rather than by cheating? As intuitive as the answer to this question may be from a broader perspective, it not a question that can be answered by looking up any of the rules of chess as stated in the official rulebook.

This account is offered in Roversi (2010, 228–33). The introduction to Roversi's account I provide is clearly condensed, but it should be enough to appreciate that his main argument is framed through a number of oppositions: those between (i) the technical vs. the teleological component, (ii) the syntax of an enterprise vs. its dynamics, (iii) individual rules vs. regulative framework and (iv) the close-up vs. the bird's-eye view. The first three oppositions overlap so as to form a single, overarching opposition (and are thus largely equivalent). The fourth one stands apart as an “interpretive” or “meta-descriptive” device by which to assign the other terms a function depending on which side of the underlying opposition they belong to. And the “optical” metaphor helps us to bring out the basic conceptual relations. Thus, the close-up view brings into focus the technical component, or syntax, of an enterprise (showing that they essentially play the same conceptual role), while the bird's-eye view enables us to see the teleological component, or dynamics, of an enterprise (likewise showing a conceptual overlap between these two components).

I thank the journal's anonymous reviewers for pointing out the need to block out the basic shape of the broader argument I have in mind, clarifying in what way I understand the standards constitutive of agency to be non-hypothetically normative.

Or, as Ferrero (2009, 310) puts it, even “the agent who contemplates the possibility of opting out of agency is not challenging the binding force of agency's standards. […] She defers to and abides by the standards of agency in determining the fate of her future participation in it”. See also Velleman (2009, 135–44).

This statement indicates that the argument deployed here is limited in scope. The argument is not meant to demonstrate that one's departure from agency is an unintelligible move period and thus an absolute impossibility. Instead, the argument I outline in this section is of a distinctively practical character and therefore is exclusively concerned with the position someone who inhabits the practical domain necessarily finds herself in, whilst leaving wide open the question as to whether, for beings like us, living outside the practical realm and destroying our agency are viable options and, if so, sustainable and durable ones.

Clearly, there are many questions to be answered before this view can come into focus as a crisp, fully fleshed-out conception. But these questions cannot be taken up in the short space of an article. It suffices, for the moment, to have sketched out in broad strokes a conception enabling us to see how practical standards can be at once non-hypothetical and normative.

Stroud (1968), for example, offers a forceful criticism of transcendental arguments.

In fact, the use of transcendental arguments as a way to provide the foundation of the basic standards of practical rationality has its supporters: witness Skidmore (2002, 124), arguing that transcendental arguments “may be the only way to ground the most fundamental principles of practical reason”. Skidmore, to be sure, restricts the scope of this argumentative strategy to its use as a foundation for instrumental rationality. But as Korsgaard (1997, 243–54) illustrates, instrumental rationality cannot be meaningfully dissociated from moral rationality, and in that sense a constitutivist can rely on transcendental arguments in showing that the standards constitutive of agency ground the non-hypothetical principles of practical rationality – instrumental as well as moral. I should note, though, that the idea of grounding the standards of moral rationality in a transcendental foundation finds an objection in Skidmore (2002, 133–9).