PHIL2606: Knowledge, Reason and Action
Week 14: Breckenridge and Kearns on Knowing

    The idea

  1. Here's the idea:

    OneK: In the following three sentences we mean the very same 2-place relation by the word 'know':

  2. When John knows Mary he stands in a certain relation to the person Mary; when John knows that Mary likes to laugh he stands in a certain relation to the proposition that Mary likes to laugh; when John knows how to make Mary laugh he stands in a certain relation to the action of making Mary laugh. The idea is that it is the same relation in each case.
  3. Note: the claim is not that there are three distinct relations here, one of which is more basic, the other two of which can be explained in terms of it.
  4. Nor is it the claim that for John to know Mary is for John to stand in a certain relation to a proposition (although there might be a reading of 'John knows Mary' on which it is).
  5. So what?

  6. If our interest is in knowledge of propositions, then what does it matter whether OneK is true?
  7. Well, if OneK is true then we would expect that many claims about knowledge of propositions can be generalized to knowledge of other things. If a certain claim cannot be generalized, then that is some reason to be suspicious of it; if it can, that is a mark in its favour. Moreover, in trying to make such generalizations we might expose some interesting phenomena. (More below.)
  8. A reason for

  9. Here is a reason to think that OneK is true: The following sentences are not zeugmatic:
  10. If OneK were not true then they would be zeugmatic. Compare:
  11. Three reasons against

  12. First reason. In 'John knows Mary', 'John knows that Mary likes to laugh', and 'John knows how to make Mary laugh', 'know' takes complements of different grammatical types, and so must be used with a different meaning in each case.
  13. But if this were a good reason then we would have a good reason to think that we use 'live' with a variety of meanings too:

    Same too for 'see', 'explain', 'guess', 'believe', and so on.

  14. Second reason. In 'John knows Mary', 'John knows that Mary likes to laugh', and 'John knows how to make Mary laugh', the second relatum is a different kind of thing in each case (person, proposition, action), so 'know' must be used with a different meaning in each case.
  15. But it is possible for a single relation to obtain between relata of very different kinds:
  16. Third reason. When translating 'John knows Mary', 'John knows that Mary likes to laugh', and 'John knows how to make Mary laugh' into certain languages, we have to translate 'know' into a different word in each case (French: 'savoire' and 'connaitre'), which suggests that we use it with a different meaning in each case.
  17. But rather than conclude that we use the one word for different things, why not conclude that they use different words for the one thing?
  18. Or perhaps this is what is going on: we have a word for something; they do not have a word for that thing but have words for several different types of that thing (consider: 'consume', 'eat', 'drink'). If so, then it is strictly speaking wrong to translate our word 'know' as either of their two words.
  19. JTBX accounts

  20. If OneK is true, then this might be another nail in the coffin for JTBX accounts of propositional knowing. What is it to know p? It is to stand in the 'know' relation to p. What is it to stand in the 'know' relation to something? It is implausible that this has anything to do with that thing's being true (for we can stand in the 'know' relation to people and other things that do not bear truth values); and it is implausible that it has anything to do with believing that thing, let alone doing so justifiably.
  21. Unless we can generalize truth, belief, and justification ...
  22. Truth

  23. It is commonly said that propositional knowing is factive - you cannot know a proposition that is false. What would be the corresponding claim for knowing people, or actions? What would be the general claim about knowing?
  24. We propose: knowing is actualitive. That is, you can only know things that are actual:
  25. Other relations seem to be actualitive: seeing, causing. Perhaps it is surprising that there can be relations that are not actualitive (such as believing).
  26. Belief

  27. It is commonly said that you cannot know a proposition unless you believe it. How does this claim generalize?
  28. In the case of knowing people, is there a relation R such that if x knows y then x R y, a relation that is the analogue of believing?
  29. Is there a more general relation that subsumes both believing and R.
  30. Williamson suggests that to believe a proposition p is to treat p as if you know it. If this is right, then maybe this carries over directly to knowing people, and knowing actions.
  31. Justification

  32. What about justification - does this generalize?
  33. Gettier cases

  34. Can these be generalized? Perhaps cases involving identical twins can be used to gettierize knowing people.
  35. Skeptical arguments

  36. Can we generalize the BIV argument? I know stephen. Or do I? We need an argument whose conclusion is that I do not. Perhaps cases involving identical twins can be used here too.
  37. Closure

  38. It is commonly said that propositional knowing is closed under known entailment:

    Necessarily, for all s, p, and q, if s knows p, and s knows that p entails q, then s knows q.

  39. Is there a similar closure principle for knowing people or actions? What about a single principle that covers them all?
  40. Consider a simple case. Is knowledge of pluralities distributive? That is, suppose I know (A and B). Does it follow that I know A and that I know B? Possible counterexample: I know John and Mary (husband and wife), but I don't know John and I don't know Mary, because they act quite differently when together.
  41. Suppose that for x to entail y is for x to be a determinate of the determinable y. Then here is an idea:

    Necessarily: for all s, x, and y: if s knows x, and x is a determinate of y, then s knows y.

  42. The KK principle

  43. The KK principle:

    Necessarily: for all s and p: if s knows p then s knows that s knows p.

  44. Can we generalize this?
  45. Here is a similar principle:

    Necessarily: for all s and x: if s knows x then s knows s's knowing x.

  46. Degrees

  47. It is very plausible that knowing a person comes in degrees. If OneK is true, then we would expect that knowing a proposition does too - it would be odd for the 'know' relation to come in degrees when its second relatum is a thing of one kind, but not when its second relatum is a thing of some other kind. Does knowing a proposition come in degrees?
  48. Knowing how

  49. According to Stanley and Williamson, to know how to ride a bicycle is to know, under a practical mode of presentation, for some way to ride a bicycle w, that w is a way to ride a bicycle. I.e. it is to stand in the know relation to a proposition (under a certain mode of presentation of that proposition). Do we like this? What we might say: it is to stand in the know relation to a way of riding a bicycle. There are two readings of 'John knows where Mary lives.' On one reading, it is true iff for some x such that Mary lives at x, John knows that Mary lives at x. On the other reading, it is true iff John knows x, where x is the place that Mary lives. So too, there are two readings of 'John knows how to ride a bicycle'. On one reading, it is true iff for some w that is a way to ride a bicycle, John knows that w is a way to ride a bicycle. On the other, it is true iff John knows w. This knowledge is not propositional (so we claim).
  50. Other

  51. A concern: does OneK wrongly allow that we can know false propositions? No. It allows that John can know the proposition that grass is black - he can know of it. But it does not allow that John can know it, in the relevant sense. Compare: knowing the president.