Skip to main content Accessibility help

The Scope of Intention: Action, Conduct, and Responsibility

  • Robert Audi (a1)


Intention takes various forms. Must its objects be acts or activities? How much can be encompassed in the content of a single intention? Can intentions can have the content: to A for R, where ‘A’ ranges over act-types and ‘R’ over reasons for action, for instance to keep my promise? The question is particularly important on the widely accepted assumption that, for concrete actions (act-tokens) that are rational and have moral worth, both their rationality and their moral worth depend on the reason(s) for which they are performed. If intentions can have content of the form of ‘to A for R’, should we conclude that (contrary to the position of many philosophers) we have direct voluntary control of the reason(s) for which we act? If intentions cannot have such content, how can we intend to do, not just what we ought to do, but to do it with ‘moral worth’? This question is also raised by the idea that we can be commanded to treat others as ends in themselves – which presumably has moral worth. If the commandable is intendable, then, to understand commands and other directives, we need a theory of the scope of intention. This paper explores kinds and objects of intention, outlines an account of its scope, and brings out some implications of the account for moral responsibility.


Corresponding author


Hide All

1 Detailed discussions of intention have been profuse since Anscombe's, G.E.M. Intention (Oxford: Blackwell, 1957; 2nd ed., Oxford: Blackwell, 1963), but the usual focus is on their role in explaining action and, especially in this century, in providing reasons for action or determining moral responsibility.

2 I have heard it said that forgiveness is an action, but I do not think so. I can say I intend to forgive, at least where I see why someone failed me and resolve to forgive the broken promise.  But we can also say ‘I intend to love’, and surely ‘love’ here is not an action-reporting term (what intending to love one's neighbor comes to will be discussed below). Perhaps ‘I intend to forgive’ is typically a way of saying either that one will express forgiveness, which is readily understandable, or that one will try to achieve forgiveness, in which case ‘try to’ leaves open a number of possible forgiveness-related acts but does not designate forgiveness itself. An account of forgiveness is not possible here, but there is now much philosophical literature on the topic (including books by Charles Griswold and Glen Pettigrove).

3 One can intend to find something out by consulting reference works, or intend to speak to someone who is visibly occupied with driving.  Neither all that one believes is involved in consulting those works nor, of course, someone else's driving, is in the content of these intentions – of what one intends.

4 For discussion of how adverbs figure in action theory see, e.g. Davidson, Donald, ‘The Logical form of Action Sentences’, in Essays on Actions and Events (Oxford: Oxford University Press, 1980).  Detailed discussion is found in ch. 3 of my Means, Ends, and Persons:  The Meaning and Psychological Dimensions of Kant's Humanity Formula (New York and OxfordOxford University Press, 2016).

5 The variety of scripts is greater than here indicated. One possibility is a de re grasp of a complex kind of behavior, as where someone who hears and well remembers a tune forms the intention to sing that.

6 Suppose, however, that ‘What did S do?’ could be properly answered by ‘S A-ed for R’, say A-ed in order to appear kind (where ‘to appear kind’ expresses the reason).  If A-ing-for-R can itself be performed for a reason, we must apparently countenance the idea that, for reason R1, S can-A-for-R.  But that higher-order act can presumably also be for a reason, R2. This is not to imply a vicious regress.  That would arise if intentionally A-ing entailed intention to A-for-a-reason and the higher-order act this requires must itself be intentional. Still, this picture forces us to posit, for agents with finite comprehensional capacities, a kind of intentional action that – contrary to the most plausible conception of intentional action – cannot be for a reason. Call A-ing-for-R a double-barreled intentional act – double-barreled because there is both a report of a first-order intentional action and a specification of a higher-order explanatory condition, for R. A double-barreled intentional act, as intentional, is still for a reason; we thus need to posit further intentional acts involving R1, R2 … etc. But, for every agent with finite capacity, there would be a kind of brute double-barreled intentional action, A-ing for Rn, where n is too large for the agent's comprehension in the relevant way, to which the idea of intentional action as action for a reason would not apply. This is an implication action theory can readily avoid. See ch. 2 of Ross, W.D., The Right and the Good (Oxford: Oxford University Press, 1930), on this point.

7 Strictly speaking, without an appropriate conative attitude, since strong hopes can yield action toward the relevant end even when the agent thinks it likely unreachable.  Granted, even for such hopes it is arguable that the agent must intend some act, such as trying to A (where A is roughly achieving the hope).

8 See Kant, Immanuel, Groundwork of the Metaphysics of Morals (1785), Wood, Allen W., trans. (New Haven: Yale University Press, 2002), Sec. 422.

9 Even if I should be mistaken in denying that ‘treating as an end’ expresses an act-type, the three kinds of intentions described in this section can be taken to indicate how such an “act” can be intended.

10 A conjunction of intentions might also fit the psychology of the agent, sometimes better, and it may be that in different situations one or the other case better fits the agent. Certainly if what is, say, resolved, is a course of action that is constituted by doing dozens of discrete things that the agent in some sense foresees, it may be more plausible to take the agent to intend to A, and to B, and to C, etc., as opposed to having a single intention to A and B and C, etc. With certain possible exceptions, we would expect the same behavior with either of the corresponding ascriptions.

11 W.D. Ross may have so interpreted him.  See, e.g. The Categorical Imperative (Oxford: Clarendon, 1954). Ross read Kant as thinking we can perform acts that harness actions to our ends.  For discussion of this issue see my Practical Reasoning and Ethical Decision (London: Routledge, 2006), e.g. 59–60.

12 Detailed discussion of the traceability thesis in given in my Responsible Action and Virtuous character’, Ethics 101 (2)(1991), 304321 .

13 This paper has benefited from discussions following its presentation as a Royal Institute for Philosophy Lecture in October, 2015, and later at the Universities of Adelaide, California – Riverside, and Pennsylvania.  Section 5 represents refinements and extensions of Sections 2 and 3 of Chapter 8 in my Means, Ends, and Persons. For helpful comments I thank Maria Alvarez, Claire Finkelstein, Peter J. Graham, the late Hugh J. McCann, Anthony O'Hear, Eric Schwitzgebel, and, especially, Garrett Cullity.

Related content

Powered by UNSILO

The Scope of Intention: Action, Conduct, and Responsibility

  • Robert Audi (a1)


Full text views

Total number of HTML views: 0
Total number of PDF views: 0 *
Loading metrics...

Abstract views

Total abstract views: 0 *
Loading metrics...

* Views captured on Cambridge Core between <date>. This data will be updated every 24 hours.

Usage data cannot currently be displayed.