¹ G. E. M Anscombe, Intention (Harvard University Press, 1963), 11. I will also refer to some supplementary remarks in Anscombe, G. E. M, ‘On Promising and Its Justice, and Whether It Needs be Respected in Foro Interno’, Crítica: Revista Hispanoamericana de Filosofía 3 (1969), 61–83Google Scholar.

² For in the following discussion we are told that doing something intentionally requires awareness which is non-observational – hence somehow more directly acquired than the knowledge which must be gained by looking at how things are (Intention, 49), and also practical as opposed to theoretical (Intention, 57), whatever precisely that comes to. In the final pages of Intention, it seems that we are supposed to realise a way of understanding both intending and doing things intentionally through their association with this non-observational and practical awareness (Intention, 90).

³ Setiya, Kieran, ‘Practical Knowledge’, Ethics 118 (2008), 388–409CrossRefGoogle Scholar. I will also refer to Setiya, Kieran, ‘Practical Knowledge Revisited’, Ethics 120 (2009), 128–137CrossRefGoogle Scholar, and to Kieran Setiya, Reasons Without Rationalism. (Princeton University Press, 2007).

⁴ Donald Davidson, ‘Intending’, repr. in Essays on Actions and Events: Philosophical Essays (Oxford: OUP, 2001).

⁵ This is not merely for lack of space, since I doubt that a direct argument, flowing from independent premises, is possible. If such an argument is impossible, then it is to be expected that Setiya, who is sympathetic to Anscombe's project, gets his inquiries into the matter going by simply noting that ‘[o]n the face of it, […]what we do intentionally, we do knowingly’ ‘Practical Knowledge Revisited’, 389. It is true that Anscombe and Setiya provide various examples which can seem to reinforce such claims. But these examples seem to be illustrations of the general claim, rather than reasons for endorsing it.

⁶ Intention, 11.

⁷ Some are tempted to question Anscombe's claim on the ground that someone can be doing something intentionally without thinking about it, and without consciously entertaining it. I agree with Anscombe and Setiya that these worries should dissolve when we realise that one can be aware of something – such as one's name or the size of one's feet – without consciously entertaining or thinking about it. ‘[W]e [are] not concerned with a thought occurring to someone, but with what he believes’ (Anscombe, ‘On Promising and Its Justice’, 65), and equally not (if this distinction is at all separate) with what an agent is consciously aware of, but just with what they're aware of (Setiya, ‘Practical Knowledge’, 389).

⁸ See ‘Practical Knowledge’ and ‘Practical Knowledge Revisited’.

⁹ ‘Intending’, 5 – emphasis added.

¹⁰ It seems clear that Setiya is in this situation. For after a number of attempts to qualify and reform Anscombe's claim in response to various counterexamples and objections, he still suspects it could be expressed ‘more accurately’ (‘Practical Knowledge Revisited’, 131). Davidson does not make such an admission, and it would take much time and space to formulate a decisive objection to his alternative proposal. But one concern which may be noted in the space of a footnote is that Davidson's entire proposal rests on his controversial view of intentional doings as particulars which can come under many descriptions, and be known under some but not others. Part of what makes this view controversial is that we lack a clear account of when a description that is known (like, for example, ‘I'm doing my best to walk’ or ‘I'm doing this funny thing with my legs here’) applies to the very same particular event as another description (like, for example, ‘this person is walking intentionally’). In connection with this, see Davidson's ‘Reply to Quine on Events’, in his Essays on Actions and Events (Oxford: OUP, 2001), which contains a withdrawal of his early view of event identity, and does not contain any replacement proposal.

¹¹ For a longer presentation of the debate about Anscombe's claim, which describes some differences between these and other theorists who all agree that Anscombe is wrong, see Will Small, ‘Practical Knowledge and the Structure of Action’ in Rethinking Epistemology Volume 2 (Berlin, Boston: De Gruyter, 2012).

¹² ‘Intending’, 91–92.

¹³ ‘Intending’, 92.

¹⁴ At one point, Small (‘Practical Knowledge and the Structure of Action’, 197–199) responds to Davidson by testing our judgements against fleshed-out versions of Davidson's carbon copier example. For part of Small's response seems to be that, once the example is thoroughly filled in, we will realise that the copying is either unknown and unintentional, because the copier isn't sufficiently response to failures in copying, or known and intentional, because the copier is sufficiently responsive. I do not disagree with Small on this point. But I think Small's discussion does little to take the force out of Davidson's example. For the non-fleshed out example exemplifies an abstract argument, hinging on apparently plausible general premises, which seems to undermine Anscombe's claim. I want to respond to Davidson not by discussing the fleshed-out examples, but by undermining the argument which the non-fleshed out example seems to exemplify.

¹⁵ Notice that I say ‘might be thought to be’ and not ‘is’. Those who think that Gettier cases don't exemplify justification will not think that these cases are straightforward counterexamples to the mentioned accounts of knowledge. I do not want to take a stand on such issues here. The case of Gettier cases is merely meant to illustrate the general point that there is a difference between straightforward appeals to counterexample and more theoretically loaded appeals to counterexample. It leaves undone the task of explaining when an example is a clear counterexample to some philosophical claim, as opposed to an example which can counter a claim together with certain extra theoretical commitments. But I hope to have shown it unsatisfying to put Davidson's cases firmly in the first category, thereby ignoring the further commitments which seem to underpin his argument.

¹⁶ If someone who is building a helicopter is very confused about whether they are building a helicopter or some other type of machine, they might not know that they are building a helicopter. But then there seems to be no reason for claiming they are building a helicopter intentionally.

¹⁷ Like many philosophers, including Austin and other presumed ‘ordinary language philosophers’, I am not here arguing from how we would describe things, but am using facts about how we would describe things as clues to crucial distinctions. Once we see the distinctions, there is little point in grinding axes about which phrases we ordinarily use and assent to. Hence of course I grant that people may sometimes say something meaningful and true by using strings of words like ‘this runner was winning, but they didn't end up having won’. But this is only possible in a strained sense where ‘winning’ means being close to winning, or anyway doing something which falls short of being such that one will have won. Once we see what this notion of winning comes to, it becomes totally unthreatening to the point I am making in the main text. For if someone doesn't know they're winning even in this close-to-winning sense, then once again they can't be winning intentionally in this close-to-winning sense, and therefore such a conception of winning, properly distinguished, does not support a Davidsonian challenge.

¹⁸ Gilbert Ryle, The Concept of Mind (London: Hutchinson, 1949).

¹⁹ Vendler, Zeno ‘Verbs and Times’, The Philosophical Review 66 (1957), 143–160CrossRefGoogle Scholar.

²⁰ Mourelatos, Alexander ‘Events, processes, and states’ Linguistics and Philosophy 2 (1978), 415–434CrossRefGoogle Scholar.

²¹ Michael Thompson, Life and Action: Elementary Structures of Practice and Practical Thought (Harvard University Press, 2008).

²² Hornsby, Jennifer ‘Actions and Activity’ Philosophical Issues 22 (2012), 233–245CrossRefGoogle Scholar.

²³ Hornsby, ‘Actions and Activity’, 241.

²⁴ Thompson, Life and Action, note 1, 106.

²⁵ The Concept of Mind, 302.

²⁶ ‘Actions and Activity’, note 16, 244.

²⁷ ‘Events, processes, and states’, 417.

²⁸ ‘Events, processes, and states’, 416.

²⁹ Reasons Without Rationalism, 50.

³⁰ I would like to acknowledge that my understanding of the present topics has been improved tremendously by many discussions with Adrian Haddock. Adrian was the one to draw my attention to the question of whether predictive knowledge need be involved in intentional action, and to the relevance this question has to the present concerns with cognitivism. On the financial side, I would like to acknowledge that Erik & Gurli Hultengren's foundation and the Royal Institute of Philosophy have funded much of my previous research on the present topics and on some connected topics.