Abstract
Argues for a conception of reasons as premises of practical reasoning. This conception is applied to questions about ignorance, advice, enabling conditions, “ought,” and evidence.
This is a preview of subscription content, log in via an institution to check access.
Similar content being viewed by others
Notes
I assume that normative reasons are facts in that they are true propositions, not worldly states that correspond to them: normative reasons must be as finely individuated as the premises of practical reasoning.
Raz treats the conclusion of practical inference as a deontic proposition—that one ought to ϕ relative to these considerations, or that there is reason to ϕ—not as action, intention, or desire (Raz 1978, pp. 5ff). The formulation in the text abstracts from this dispute.
Like Raz, Williams takes practical inference to involve the belief that one has reason to ϕ (Williams 1979, pp. 104, 107–108), though his principal concern is motivation.
It is thus immune to the objections of Broome (2007, pp. 368–369, 371–373), who notes that practical rationality is consistent with failing to act or intend to act on the ground that p, when the fact that p is a reason for me to ϕ. This reason may be outweighed, or I may have misleading evidence to that effect. Despite this, rationality does require a correct response to reasons, in the form of motivation or desire.
On internalism about reasons, see the essays collected in Setiya and Paakkunainen (2012).
Throughout this essay, “practical rationality” means the excellence of practical reason, not just freedom from “practical irrationality” in a narrow sense. On the restricted use of “irrationality” as a constraint on reasons, see “Against Internalism” (Setiya 2004).
For an example of this kind, attributed to Frank Jackson, see Dancy (2000, pp. 65–66).
If degrees of belief are correct when they match epistemic probabilities, beliefs may be correct only when they are both true and justified, or epistemically sound. For instance, we might restrict the premises of sound reasoning to what the agent knows. For a view in this spirit, see Hawthorne and Stanley (2008).
Smith counts full information as a condition of full rationality, along with correct deliberation (Smith 1994, p. 156). My terminology separates the two.
Compare Smith 2006, in which “should” is indexed to contextually relevant expectations and values. This move could be adapted to reasons. In each case, the effect is to build disputed claims about the rational treatment of risk and uncertainty into the nature of normative properties. Reasons is less tendentious.
Smith (1994, pp. 151–152, 164–174).
The strength of Smith’s condition can be hard to make out. After all, if the circumstance includes psychological facts about the agent, we are bound to want the same things when our circumstances are the same. This is, however, irrelevant to the advice model, which asks for the desires we would have about our behaviour in C, not what desires we would have in C itself. Suppose, for instance, that practical rationality is purely instrumental, a matter of putting means to ends; and suppose that A is altruistic, desiring happiness for all, while B is utterly selfish. If A were fully rational and fully informed, what desires would he have about his behaviour in the unfortunate circumstance in which he becomes like B? Being altruistic, instrumentally rational, and well-informed, he wants even those without altruistic desires to act in ways that benefit others: that is what he wants himself to do in the circumstance described. In contrast, B would want himself to act only in ways that promote his own interests. Even in conditions of full rationality and full information, A and B would differ in the relevant desires. These issues are further discussed in Sobel (1999).
In addition to theoretical arguments for the advice model, there are cases in which it seems to give a better verdict than Reasons, as when it is impossible to act on a putative reason without defeating its purpose. Is the fact that I have forgotten a meeting at noon a reason for me to check my calendar? It is not a reason by which it would be rational to be moved. As I have argued elsewhere (in Setiya 2009, p. 538), facts like this are not reasons to act but reasons to respond to action in related ways: the fact that I have forgotten my appointment is a reason to be glad if I check my calendar and dismayed if I do not.
See Pettit and Smith (1990).
In this way, Reasons restricts the phenomenon of “silencing.” If the fact that doing something will cause pleasure is a reason for me to do it, but not when the pleasure is guilty, the possibility of guilt must be excluded by the starting points of sound practical thought. My reasoning must include beliefs whose truth implies that the pleasure is innocent. In other words: reasons can be silenced by a change in circumstance only if sound reasoning anticipates the relevance of that change. This may seem to count against Reasons, but it does not. We can see this by reflecting on practical thought in conditions of ignorance. If it is good practical thought to be moved to ϕ by C-and-the-belief-that-doing-ϕ-will-cause-pleasure, even though C does not contain beliefs whose truth is incompatible with guilt, it is, in effect, practically rational to bet on the pleasure’s being innocent. That reason survives even if the gamble fails, like my reason to bet on the losing horse. Note that, even in this case, C had better include the absence of beliefs which imply that the pleasure is guilty, or there would be reason to pursue it even when I believe it is.
Note that holism about reasons does not entail, or even support, the more radical “particularism” of Dancy 2004. As Dancy (2004, pp. 81–82) concedes, holism is consistent with the codifiability of practical reason; and Dancy’s challenge to codifiability—how does it follow from the nature or “logic” of reasons to act?—does not depend on it.
This formula ignores sound reasoning that does not involve belief, since it does not correspond to reasons. In my view, this is harmless, since there is no such thing. When it is valid, practical reasoning always involves beliefs about the circumstance of action, even if it also involves non-cognitive states.
It follows that even complete or inclusive reasons may be incomparable and so not quasi-additive.
For this version of the example, see Kolodny and MacFarlane (2010); they take the case from Derek Parfit, who credits it in turn to Donald Regan.
Kolodny and MacFarlane (2010, §2.2). For what it is worth, I am not convinced by their arguments. If we think of “ought” as being evaluated with respect to a set of reasons fixed by a body of information, which is fixed in turn by the discursive context, and we assume that the deliberator intends to focus on the information that will be available before he acts, we can explain how advisors and deliberators disagree. The advisor’s information, however unexpected, is relevant to the deliberator’s claim. Nor is that claim excessively precarious or speculative, since the deliberator may have good reason to doubt that new information will come to light. What about distant observers? If they mean to rely on inaccessible information, their claims do not conflict with those we make in the context of deliberation. On the other hand, an observer may want to evaluate an action from our point of view. It will then be natural to say, of the rescuers in the mineshaft case, “They did what they should in blocking neither shaft,” and in doing so, one speaks the truth. The position of the deliberator is equally flexible, as we can see by looking at conditionals like these:
-
If the miners are in shaft A, we ought to block shaft A.
-
If the miners are in shaft B, we ought to block shaft B.
Such conditionals may seem true. But ask yourself: suppose the miners are in shaft A, but you don’t know it; should you block shaft A? Surely not! Even if the miners are in shaft A, we ought to block neither shaft. What is happening here is a shift in the contextually salient body of information. On one reading, we should block neither shaft, not knowing where the miners are. The mere fact that the miners are in shaft A, or B, does nothing to change this, and the conditionals come out false. (This point is more vivid the more obscure the relevant facts.) On another reading, we accommodate the assertion of the conditionals by evaluating “ought”-claims with respect to a different body of information, including facts about the location of the miners. Then the conditionals come out true; and it is a mistake to say, as we originally did, “We ought to block neither shaft.”
-
As on the theory of evidence as raising epistemic probability; Kearns and Star (2009, pp. 231–232).
I defend these possibilities in Setiya (2010).
Nothing I have said here provides a test for when evidence that one ought to ϕ is or is not a reason to ϕ. That is a question about the content of practical rationality. I am trying to state the relationship between its standards, whatever they are, and facts about what there is reason to do.
They also claim that R best explains how reasons can be weighed (Kearns and Star 2009, §2.6). But they do not consider Reasons and the corresponding theory of relative strength, developed in the text above. Their complaint is more persuasive against the theory of reasons as explaining oughts in Broome 2004; see Kearns and Star (2008, pp. 42–5).
Here I agree with Parfit (2001), among others.
References
Alvarez, M. (2010). Kinds of reasons. Oxford: Oxford University Press.
Audi, R. (2004). The right in the good. Princeton, NJ: Princeton University Press.
Berker, S. (2007). Particular reasons. Ethics, 108(1), 109–139.
Broome, J. (2004). Reasons. In R. J. Wallace, M. Smith, S. Scheffler, & P. Pettit (Eds.), Reason and value: Themes from the moral philosophy of Joseph Raz (pp. 28–55). Oxford: Oxford University Press.
Broome, J. (2007). Does rationality consist in responding correctly to reasons? Journal of Moral Philosophy, 4(3), 349–374.
Chang, R. (1997). Incommensurability, incomparability, and practical reason. Cambridge, MA: Harvard University Press.
Dancy, J. (2000). Practical reality. Oxford: Oxford University Press.
Dancy, J. (2004). Ethics without principles. Oxford: Oxford University Press.
Davidson, D. (1963). Actions, reasons, and causes. Reprinted in Essays on actions and events (pp. 3–19). Oxford: Oxford University Press, 1980.
Hawthorne, J., & Stanley, J. (2008). Knowledge and action. Journal of Philosophy, 105(10), 571–590.
Kearns, S., & Star, D. (2008). Reasons: Explanations or evidence? Ethics, 119(1), 31–56.
Kearns, S., & Star, D. (2009). Reasons as evidence. Oxford Studies in Metaethics, 4, 215–242.
Kolodny, N., & MacFarlane, J. (2010). Ifs and oughts. Journal of Philosophy, 107(3), 115–143.
Parfit, D. (2001). Reasons and rationality. In D. Egonsson, J. Josefsson, B. Petersson, & T. Rønnow-Rasmussen (Eds.), Exploring practical philosophy (pp. 17–39). Aldershot: Ashgate.
Pettit, P., & Smith, M. (1990). Backgrounding desire. Philosophical Review, 99(4), 565–592.
Raz, J. (1978). Introduction. In J. Raz (Ed.), Practical reasoning (pp. 1–17). Oxford: Oxford University Press.
Setiya, K. (2004). Against internalism. Noûs, 38(2), 266–298.
Setiya, K. (2007). Reasons without rationalism. Princeton, NJ: Princeton University Press.
Setiya, K. (2009). Reply to Bratman and Smith. Analysis, 69(3), 531–540.
Setiya, K. (2010). Sympathy for the devil. In S. Tenenbaum (Ed.), Desire, practical reason, and the good (pp. 82–110). Oxford: Oxford University Press.
Setiya, K., & Paakkunaninen, H. (2012). Internal reasons: Contemporary readings. Cambridge, MA: MIT Press.
Shafer-Landau, R. (2003). Moral realism. Oxford: Oxford University Press.
Smith, M. (1987). The Humean theory of motivation. Mind, 96(381), 36–61.
Smith, M. (1994). The moral problem. Oxford: Blackwell.
Smith, M. (1995). Internal reasons. Philosophy and Phenomenological Research, 55(1), 109–131.
Smith, M. (2006). Moore on the right, the good, and uncertainty. In T. Horgan & M. Timmons (Eds.), Metaethics after Moore (pp. 133–148). Oxford: Oxford University Press.
Sobel, D. (1999). Do the desires of rational agents converge? Analysis, 59(3), 137–147.
Watson, G. (1975). Free agency. Journal of Philosophy, 72(8), 205–220.
Williams, B. (1979). Internal and external reasons. Reprinted in Moral luck (pp. 101–113). Cambridge, England: Cambridge University Press, 1981.
Acknowledgments
For discussion of these topics, I am grateful to Samuel Asarnow, Jonathan Dancy, Cian Dorr, Alida Liberman, Richard Moran, Joseph Raz, Michael Smith, Jonathan Way, and Ralph Wedgwood; a reader for the journal gave very helpful comments on earlier drafts.