Knowing how to

Semantics

Consider again one of these sentences in which ‘S knows’ is followed by a clause with an embedded question. This question will have a number of correct answers (‘such-and-such is a place where Fred gets his beer’, ‘so-and-so is a place where Fred gets his beer’, etc.). We begin by saying that for it to be true that S knows where Fred gets his beer S must know all these propositions, in other words S must know all the correct answers to the embedded question.

A number of qualifications are then needed. What counts as all these propositions is relative to context. In certain circumstances (perhaps when we are trying to get something done quickly) it may be important only that S know one of the correct answers. In many cases the embedded question has deontic or normative significance: it is more a matter of where/when/how etc. one ought to, or can best, do something than of just being able to. But none of these qualifications affect the point that we are dealing with a cognitive relation between the subject S and one or more propositions, not a relation between S and an action or action-type. Semantically, as well as syntactically, knowing how to is a form of knowing that.

This line of thought isn’t without its difficulties. When we apply it to ‘I know how to Φ’ it gives:

A problem appears when we consider (for example) the case where W is a demonstrative:

The trouble is that it seems a long way from

The latter strongly suggests that I am competent to promote Φ-ing (whether that be by doing it myself or by advising others) whereas (1) doesn’t. In Stanley and Williamson (2001) the authors respond by saying that the subject needs to know that W is a way to Φ ‘under a practical mode of presentation’.

The notion of a mode of presentation is well illustrated by the following (see for example Perry 1979). Suppose I see in a mirror a man standing next to a firework and holding a glowing fuse-paper. I think ‘That man is in danger’. Now the man is in fact, although I have not realized it, me. If I did realize it, I would be thinking ‘I am in danger’, and this would be to think of the (same) situation in a first-person mode of presentation. It is obvious that mode of presentation makes a big difference: in the first case (‘That man is in danger’) I will stay where I am and shout a warning, in the second (‘I am in danger’) I will do neither, simply run. Similarly, there is a present-tense mode of presentation: the belief that the meeting begins at twelve o’clock does not, even when in fact it is twelve o’clock, give rise to the same behaviour as the belief that the meeting is beginning now. We are to understand a practical mode of presentation analogously: knowledge under a practical mode of presentation that W is a way of Φ-ing is linked to a certain complex of behavioural dispositions to which a demonstrative mode of presentation (among others) is not. To know how to Φ, then, is to know, under a practical mode of presentation, that some W is a way to Φ.