Intention

G. E. M. Anscombe was Professor of Philosophy at the University of Cambridge.

• 1. The subject introduced under three heads: expression of intention for the future, intentional action, and intention in acting. 2. Intuitive understanding of the difference between 'prediction' and 'expression of intention' rejected as a foundation for a philosophical account of expressions of intention. Prediction defined so as to comprise orders and expressions of intention as well as estimates of the future. The falsity of expressions of intention in the simple future tense (a) as lying and (b) as falsity because the intention is not carried out. 3. Usefulness of considering the verbal expression of intention for the future in order to avoid various dead ends. Uselessness of an introspective explanation of intention. Expressions of intentions distinguished from estimates of the future by the justification, if any, given for them. 4. Are there any statements of the form 'A intends X' which can be made with fair certainty? Descriptions of a man's actions often descriptions truly substitutable for 'X ' in 'A intends X'. Reasons why we suppose a man the sole authority on his own intentions. 5. Intentional actions defined as those to which a certain sense of the question 'Why?' is given application. Difficulty of defining the relevant sense and danger of moving in a circle in our explanations of 'reason for acting' and 'action'. 'I knocked the cup off the table because I was startled' gives an answer to a question 'Why?' about something done. 6. The question 'Why?' is refused application by the answer 'I did not know I was doing that'. The same action can have many descriptions, in some of which the agent knows it and in some not. 7. The question also refused application when the action was involuntary; but this notion cannot be introduced without treating as solved the very kind of problem we are discussing. Difficulties of the notion 'involuntary'. 8. 'Non-observational knowledge' introduced as e.g. the knowledge one has of some of one's own movements. There is also non-observational knowledge of the causation of a movement, as when I say why I gave a start. We can define one class of involuntary movements without begging any questions, as the class of movements known without observation, but where there is no room for non-observational knowledge of causality: e.g. the muscular spasm one sometimes has in dropping off to sleep. 9. In one sense of 'Why?' the answer mentions evidence; but an answer to a question 'Why?' about an action, which does not mention evidence, does not therefore necessarily give a reason for acting. The cases where it was difficult to distinguish a cause from a reason turn out to be ones where there is non-observational knowledge of the causation. 10. This kind of causation labelled 'mental causality'. Mental causes should be distinguished from motives of actions and objects of feelings. 11. And also from intentions with which a person acts, even though these may be expressed in the form 'I wanted...' Mental causality is not important in itself, but it is important to make these distinctions. 12. Motives have been sharply distinguished from intentions by philosophers, and described as causes. Popularly motive and intention are not so distinct; but 'motive' is a wider notion than' intention'. A motive is not a cause at all. 13. Among motives that are not intentions for the future we can distinguish between backward-looking motives like revenge (I killed him because he killed my brother) and motive-in-general (He did it out of friendship). Motive-in-general can also be called 'interpretative' motive. 14. What distinguishes backward-looking motives from mental causes? The notions of good and harm are involved in them. 15. In some cases the distinction between a mental cause and a reason is not sharp-E.g. 'I put it down because he told me to'. 16. Summary of results reached so far. 17. The question 'Why?' is not refused application when the answer is e.g. 'For no particular reason' or 'I don't know why I did it'. Consideration of the latter answer. 18. The fact that 'For no particular reason' is a possible answer to the question 'Why?' about an action does not shew that this answer always makes sense. But when we speak of it as not making sense, we mean that we cannot understand the man who says it, rather than that 'a form of words is excluded from the language'. The question 'Why?' identified as one expecting an answer in the range we have described, which range we use to define the class of intentional actions. 19. We do not mention any extra feature attaching to an action at the time it is done by calling it intentional. Proof of this by supposing there is such a feature. 20. Discussion whether intentional actions could still have the characteristic of being intentional although there were no such thing as expression of intention for the future, or further intention with which one acts. There would be no such thing as our question 'Why?' or intentional action if the only answer were: 'For no particular reason'. 21. Criticism of the Aristotelian proof of a final end for a man's actions. Still, we can now see that some chains of reasons for acting must occur if there is such a thing as intentional action at all. 22. Discussion of intention with which, when this mentions something future. In order for it to be possible to say that an agent does P in order that Q, he must treat an acknowledgement of 'But if P, Q won't happen 'as incompatible with his having that intention in acting. 23. Is there any description which is the description of an intentional action when intentional action occurs? An example is invented in which to examine the question: a man who moves his arm in pumping water to replenish a house water-supply to poison the inhabitants and is also doing other things with the pump handle at the same time. Any true descriptions of what he is doing which satisfy our criteria, are descriptions of intentional actions. Are there as many actions and as many intentions as there are such descriptions? 24. Difficulties. If 'he is poisoning the inhabitants' is one of these descriptions, when does he do this? How is moving his arm up and down an act of poisoning the inhabitants? 25. Supposing the man to know the water will poison the inhabitants, but to say 'I didn't care about that, I was only doing my job of pumping', this answer does not fall within the range of answers to 'Why?' by which we have defined intentional action. Can one determine one's intentions just by what one says they are? The interest of a man's intentions, apart from what he actually did. 26. Answer to the questions of 23. The A-D order: i.e. the order of descriptions of an action as intentional, such that each term of the series can be said to be an intention in the action as described by the previous term, and the last term an intention of the action as described by the first or any intermediate term. 27. Is there ever any place for an interior act of intention, which really determines what is or is not going on under the tide 'such-and-such a kind of action'? 28. Further enquiry into non-observational knowledge. Knowledge of one's own intentional actions-I can say what I am doing without looking to see. 29. But must there not be two objects of knowledge-.--what I am 'doing', i.e. my intention, and what is actually taking place, which can only be given by observation? Philosophical views on will and intention which have arisen from this problem. 30. An example to prove that it is wrong to try and push the real intention, or act of will, back to something initiating the movements that then take place. 31. Attempt at solution by comparing the facts which may falsify a statement of intentional action to the facts which may make an order fall to the ground. Inadequacy of this solution. 32. Example of man with a shopping list: the relation of this list to what he buys, and of what he buys to a list made by a detective following him. The character of a discrepancy between the list and what is bought in the two cases. Is there such a thing as 'practical knowledge' in the sense of ancient and medieval philosophy? 33. This notion can only be understood by first understanding what Aristotle called 'practical reasoning'. The practical syllogism is not a form of demonstration of what I ought to do. It is a different kind of reasoning from that of the proof syllogism, but this has been misunderstood in modern times. 34. Practical syllogisms are not confined to ones that look parallel to proof syllogisms. The starting point for a piece of practical reasoning is something wanted, and the first premise mentions something wanted. 35. Occurrence of evaluative terms in the first premise of practical syllogisms given by Aristotle. Not every statement of a reason for acting shews practical reasoning. 'I want' does not rightly occur in the premises, but the first premise must mention something wanted. 36. In the relevant sense of 'wanting' 'X' in 'A wants X' does not range over all describable objects or states of affairs. Volition and sense-knowledge cannot be described independently of one another. Problem of wanting a wife, and generally of wanting what the agent does not even suppose to exist yet. 37. If a man wants something, he can always be asked what for, or in what aspect it is desirable; until he gives a desirability-characterisation. 38. The question 'What for?' cannot significantly be asked in a continuation of the series of such questions, once a desirability-characterisation has been reached. The point illustrated by an example: 'It befits a Nazi to spend his last hour exterminating Jews'. This does not mean that the practical reasoning cannot be assailed so long as it is not fallacious. 39. The fact that a desirability-characterisation is required does not shew that any is compulsive in relation to wanting. Bonum est multiplex. 40. Comparison of the problem of the relation of 'wanting' to 'good' with that of the relation of 'judging' to 'true'. 41. The mark of practical reasoning is that the thing wanted is at a distance from the particular action. 42. The 'absurdity' of setting practical reasonings out in full. The point is to describe not what (psychologically) goes on, but an order; the same order as I described in discussing what 'the intentional action' was. 43. Contrast between 'the stove is burning' and 'the man is paying his gas bill': enormous apparent complexity of 'doing' in the latter case. 44. Consideration of 'If I do this, this will happen, if that, that' followed by action: cases in which this is, and in which it is not 'practical reasoning'. 45. Practical knowledge considered as the knowledge of what is done in the man who directs a project without seeing it. Problem: how is this knowledge, if his orders do not get carried out? 46. The description of something as e.g. building a house or writing on the blackboard employs the concept of human action, which we have seen to be defined by means of our question 'Why?' 47. The term 'intentional' relates to a form of description of events. Intention in animals. 48. Many descriptions of events effected by humans are formally descriptions of executed intentions. Elucidation of the notion of practical knowledge. 49. Account of 'voluntary' action. 50. Return to expression of intention for the future. What has been said about intention in present action also applies to future intention. A prediction is an expression of intention when our question 'Why?' applies to it. 51. Consideration of 'I just want to, that's all' in regard to an expression of intention for the future. 52. 'I am not going to-' as an expression of intention, and 'I am going to-' as an expression of belief. Cases where they might occur together.

Anscombe's classic work is the font from which all subsequent philosophical thought about agency flows.