Hostname: page-component-586b7cd67f-gb8f7 Total loading time: 0 Render date: 2024-11-30T16:07:15.419Z Has data issue: false hasContentIssue false
  • English
  • Français

Practical Steps and Reasons for Action

Published online by Cambridge University Press:  01 January 2020

Philip Clark
Philip Clark*
Affiliation:
University of California - Davis, Davis, CA95616-8673, USA
Get access Rights & Permissions [Opens in a new window]

Extract

There is an idea, going back to Aristotle, that reasons for action can be understood on a parallel with reasons for belief. Not surprisingly, the idea has almost always led to some form of inferentialism about reasons for action. In this paper I argue that reasons for action can be understood on a parallel with reasons for belief, but that this requires abandoning inferentialism about reasons for action. This result will be thought paradoxical. It is generally assumed that if there is to be a useful parallel, there must be some such thing as a practical inference. As we shall see, that assumption tends to block the fruitful exploration of the real parallel. On the view I shall defend, the practical analogue of an ordinary inference is not an inference, but something I shall call a practical step. Nevertheless, the practical step will do, for a theory of reasons for action, what ordinary inference does for an inferentialist theory of reasons for belief. The result is a general characterization of reasons, practical and theoretical, in terms of the correctness conditions of the relevant sorts of step.

Type
Research Article
Information
Canadian Journal of Philosophy , Volume 27 , Issue 1 , March 1997 , pp. 17 - 45
Copyright
Copyright © The Authors 1997

Access options

Get access to the full version of this content by using one of the access options below. (Log in options will check for institutional or personal access. Content may require purchase if you do not have access.)

References

1 For a critique of inferentialism about reasons for action, see Milo, Ronald D. ‘The Notion of a Practical Inference,’ American Philosophical Quarterly 15 (1976) 13–21.Google Scholar

2 For the objection, see Audi, Robert Practical Reasoning (London: Routledge 1989), 94Google Scholar; and Kenny, Anthony Aristotle's Theory of the Will (New Haven: Yale University Press 1979) 142 ff.Google Scholar

3 Anscombe, G.E.M. Intention (Ithaca, NY: Cornell University Press 1958), 65Google Scholar

4 See Kenny, Anthony ‘Practical Inference,’ Analysis 26 (1966) 65–73CrossRefGoogle Scholar; and ‘Practical Reasoning and Rational Appetite,’ in Will, Freedom and Power (Oxford: Basil Blackwell 1975) 70-96. See also Ross, Alf ‘Imperatives and Logic,’ Philosophy of Science 11 (1944) 30–46CrossRefGoogle Scholar, and Wright, G.H. von ‘On So-called Practical Inference,’ Acta Sociologica 15 (1972).Google Scholar

5 On content maps, see Stampe, Dennis W. ‘The Authority of Desire,’ The Philosophical Review 96 (1987) 379.CrossRefGoogle Scholar

6 See Carroll, Lewis ‘What the Tortoise said to Achilles,’ Mind (1895), 278CrossRefGoogle Scholar; and Stroud, Barry ‘Inference, Belief and Understanding,’ Mind 88 (1979) 179–96.CrossRefGoogle Scholar

7 On this point it may help to imagine someone claiming to infer a conclusion from a premise, but disavowing any thought that the premise supports the conclusion. For instance, suppose someone predicts a hard winter. You ask why she expects a hard winter, and she says she infers it from the extra woolliness of the woolly worms. You ask what makes her think the extra woolliness is a sign of a hard winter. ‘Sign?’ she says, ‘Why, it never occurred to me that it might be a sign of anything. Still, I do make this inference ….’ The complete disavowal of any thought that the extra woolliness indicates a hard winter, together with the claim to be making an inference, is nonsense. For a different argument to the same effect, see Thomson, Judith Jarvis ‘Reasons and Reasoning,’ in Black, M. ed., Philosophy in America (Ithaca, NY: Cornell University Press 1965) 296–8.Google Scholar

8 Perhaps it will be said that this picture of premises ignores recent developments in epistemology, specifically the rise of reliabilist and proper-function accounts of justification. (See, e.g., Goldman, A.I. Epistemology and Cognition (Cambridge, MA and London, England: Harvard University Press 1986)Google Scholar; Millikan, Ruth Language, Thought, and Other Biological Categories (Cambridge, MA and London, England: The MIT Press 1984)Google Scholar; and Plantinga, Alvin Warrant and Proper Function (New York and Oxford: Oxford University Press 1993.)Google Scholar But a theory of inference is not a theory of justification. It is one thing to say what psychological states are presupposed by the drawing of a conclusion from premises, and quite another to give the conditions under which a belief is justified. Not all beliefs formed via the drawing of a conclusion from premises are justified. Nor need we assume that the drawing of conclusions is the only process by which a justified belief can be formed. Nevertheless, there is nothing in the present view to prevent us from saying that the drawing of a conclusion from premises sometimes forms part of a ‘reliable mechanism’ or a ‘proper functioning’ whereby a justified belief is produced. On the face of it, then, these views of justification are not in competition with our account.

Similarly, it is a mistake to oppose the present view to a causal theory of inference. Goldman, for instance, is concerned to show that inference, insofar as it confers justification, is a causal process. But the view that premises essentially work through connecting thoughts is compatible with the idea that inference is a causal process. In its causal version, our view is just that not all of the beliefs that together do the causing should be thought of as premises.

Goldman does argue that one can make an inference without believing that the inference is valid. If he were to generalize this, and say one can make an inference without thinking that the premises in any way support the conclusion, this would put him at odds with our view. The argument is, roughly, that ‘children and adults untrained in formal logic’ lack the concepts required to form the connecting thoughts, even though they can make the inferences. But there is no question that children can have beliefs about support, e.g., that a certain person's arrival means Mom and Dad are going out, without any formal introduction to concepts like evidence and validity. (See Goldman, Epistemology and Cognition 86Google Scholar. For a less intellectualized view of belief, see Pettit, and Jackson, ‘Moral Functionalism and Moral Motivation,’ Philosophical Quarterly 45 (1995) 20–40.)Google Scholar

9 It may be asked why E is not a connecting proposition. After all, your wanting E is a missing link between your thinking X will lead to E and your doing the action. So why not think of the end as a connecting proposition, and conclude that the content of the thought has every claim to be a premise? Were we to accept the suggestion, consistency would bind us to say the same of theoretical inference. Believing the premises is, in much the same way, a missing link between believing the connecting proposition and believing the conclusion. If you take red night skies to indicate good weather tomorrow, you still won't draw the conclusion unless you think the sky is red tonight. So to be even-handed about it, we would have to allow that theoretical premises are connecting propositions themselves, and that what I've been calling the connecting proposition has as much claim to be called a premise as the others. But this really goes too far. For it misses the point that we just do not call everything behind the conclusion a premise. What we do not call a premise is precisely the element that is about how the others are related to the conclusion. That is the feature that disqualifies the theoretical connecting proposition as a premise. The same feature disqualifies the practical connecting proposition.

10 For the idea of such an objection I thank Warren Quinn.

11 Here I follow Anscombe, Intention, 65Google Scholar, and ‘von Wright on Practical Inference,’ in Schilpp, ed., The Philosophy of Georg Henrik von Wright (La Salle, IL: Open Court 1989), 381Google Scholar. See also Snare, Francis Morals, Motivation and Convention (Cambridge: Cambridge University Press 1991), 140CrossRefGoogle Scholar; Darwall, Stephen Impartial Reason (Ithaca, NY: Cornell University Press 1983), 37Google Scholar; and Pettit, Philip and Smith, Michael ‘Backgrounding Desire,’ Philosophical Review 99 (1990) 565–92.CrossRefGoogle Scholar

12 One could be a non-cognitivist about the judgment that the truth of P supports the conclusion that Q, though not without the usual sorts of sacrifices. See, e.g. Gibbard, Alan Wise Choices, Apt Feelings (Cambridge, MA: Harvard 1990)Google Scholar. The point is just that the regress argument lends no support to such a view.

13 One might, I suppose, use the word ‘premise’ more broadly. One might just choose to call everything believed (except for the conclusion) a premise. In that case, I am arguing for a distinction among ‘premises,’ between those from which the conclusion is inferred, and that which connects them to the conclusion. But this is clearly not what we ordinarily mean by ‘premise.’ The premises of an inference, according to the ordinary notion, constitute that from which the conclusion is inferred.

14 Other formulations for a practical rule of inference have been suggested. For instance, one respondent suggested the following:

If I am to E, and X is necessary for me to E, then I am to X

This would join with the premises

I am to E

and X is necessary for me to E

giving the conclusion

I am to X

On this view, the believed element would come in as a premise that works through the acceptance of the principle. One problem with this suggestion is that the sentence ‘If I am to E, and X is necessary for me to E, then I am to X’ is not obviously grammatical. (We say things like, ‘If I am to survive then I must eat.’ But what would it mean to say ‘If I am to survive, then I am to eat?’) But there is a more serious problem, namely that it is entirely unclear what it would be to accept the proposed principle. What would be the difference between someone who accepted it and someone who didn't? This question will have to be answered. But it isn't clear that there is any way of answering it that makes trouble for my view. For instance, one possible answer is that the one who accepts the principle is caused, by her acceptance of the other two elements, to accept the conclusion. But if this is all accepting amounts to, then the belief that X is necessary for me to E isn't working through the acceptance of the principle after all. It is simply joining with the desire to produce action. So accepting the principle has to be construed as something more.

Several possibilities present themselves.

First, we could take accepting the principle to be believing ‘If I am to E, and X is necessary for me to E, then I must X.’ This is grammatical, but unhelpful. Sentences of the form ‘If I am to E then I must X’ just express the claim that X is necessary for me to E. So the proposed principle expresses the unsurprising claim that I cannot E, given that X is necessary for me to E, unless I X. It is hard to see how the acceptance of this claim would add anything to the explanation of the action that was not already supplied by the acceptance of ‘X is necessary for me to E.’

Second, we could take ‘I am to E’ to express a normative claim such as ‘I should E.’ In that case accepting the principle is believing ‘If I should E, and X is necessary for me to E, then I should X.’ And the resulting inference is an ordinary inference to a believed conclusion about what I should do. As I have argued, the appeal to such inferences vitiates the inferentialist strategy, since ‘I should X’ is too close to ‘X is what I have most reason to do.’

Finally, one might say accepting the principle is having some sort of intention or aim. But what aim? If it is the aim of making certain sorts of practical steps, then the suggestion collapses into an appeal to a rule licensing certain steps. As we have seen, my view is quite compatible with an appeal to rules of that sort. On the other hand, if it is the aim, not of making certain types of steps, but of doing certain types of actions, then we will wind up misrepresenting the agent's reasons, in the way we saw for desires. That is, we will wind up treating actions that are done simply for the sake of E as if they were done for the sake of conforming to the principle. Indeed, we will wind up analyzing acting for the sake of E in terms of acting for the sake of conforming to the principle, a most unpromising strategy, since it is this acting-for-the-sake-of that we need to understand.

Perhaps there is some way of fleshing out the proposal on which it will be both plausible and incompatible with the present view. If so, we have not yet seen it.

15 Suppose that, wanting to lose weight, and believing that if I lose weight I will buy new clothes, I head for the mall. Plausibly, the relation among what I want, believe and do is flawed in this case: the truth of the belief bears no implication that the action is rightly related to the end. But this is no counterexample, because we cannot yet understand it as a case of going to the mall in order to lose weight: we do not yet have an example of a practical step from the desired end to the action.

16 There is a class of what we might call deductive practical steps, some of which are formally valid. Suppose I want at least one of my parents to know where I'm spending my vacation, so I tell my father. Here we might say the connecting proposition has the form ‘Informing my father is a way of informing my father or my mother,’ which has the look of a logical truth. But no logic developed along these lines will capture the correctness of steps whose connecting propositions express contingent truths, e.g., the step from wanting to control one's appetite to drinking coffee.

17 For the neo-Humean view, see Williams, Bernard ‘Internal and External Reasons,’ Harrington, Ross ed., Rational Action (Cambridge: Cambridge University Press 1980)Google Scholar, reprinted in Williams, Moral Luck (Cambridge: Cambridge University Press 1981)CrossRefGoogle Scholar; and ‘Internal Reasons and the Obscurity of Blame,’ Logos 10 (1989) 1-11. For responses in a neo-Kantian spirit, see Korsgaard, Christine ‘Skepticism About Practical Reason,’ Journal of Philosophy 83 (1986), 20–21CrossRefGoogle Scholar; Cohon, Rachel ‘Are External Reasons Impossible?’ Ethics 96 (1986), 555CrossRefGoogle Scholar; and Hooker, Brad ‘Williams's Argument against External Reasons,’ Analysis 47 (1987), 44.CrossRefGoogle Scholar

18 See McDowell, John ‘Might There be External Reasons?’ in Altham, J.E.J. and Harrington, Ross eds., World, Mind and Ethics (Cambridge: Cambridge University Press 1995)Google Scholar; and Smith, Michael The Moral Problem (Oxford: Blackwell 1994).Google Scholar

19 Mackie, J.L. Ethics: Inventing Right and Wrong (London: Penguin 1990), 43Google Scholar. For the full Hobbes, passage, see Leviathan (London: Cambridge University Press 1935), 30.Google Scholar

20 Hurley, Susan combines a form of cognitivism with a view of value judgments as expressing preferences in her Natural Reasons (Oxford: Oxford University Press 1989)Google Scholar. My point is just that we should not abandon large portions of evaluative discourse to a subjectivist analysis, merely because those uses of ‘good’ march with facts about the speakers’ wants.

21 Here I concur with Darwall, Impartial Reason, 36ffGoogle Scholar. See also Locke, Don ‘Reasons, Wants, and Causes,’ American Philosophical Quarterly 2 (1974), 169ff.Google Scholar

22 This paper has benefited from comments by Robert M. Adams, Rogers Albritton, Michael Bratman, David Copp, Michael Jubien, Jeff King, Paul Kjellberg, Warren Quinn, Marleen Rozemond, Samuel Scheffler, Houston Smit, and several referees for this journal.