Saturday, August 22, 2009

What is Procedural Knowledge?

Ordinary thought distinguishes two kinds of incapacity: I might lack (say) the brute strength to accomplish some task, or I might just not know how to do it. Is this an important (deep, principled) distinction? If so, what is its basis? (What justifies counting some abilities, but not others, as constituting a kind of knowledge?)

It doesn't just seem to be the difference between physical vs. mental incapacities, since even the latter (e.g. working memory limitations) may be intuitively classified as brute incapacities. Nor does the distinction seem to rest on differences in how the capacity may be acquired: practicing the piano does not seem obviously different in kind from lifting weights, though only the former is thought to yield procedural knowledge. Perhaps we need to combine several conditions, but then the resulting classification starts to look a bit ad hoc.

I feel like I must be missing something obvious here. Any thoughts?


  1. It might be more helpful to have three categories: procedural knowledge, trainable brute capacity, and untrainable brute capacity.

    It seems to me that one can have procedural knowledge about playing the piano without practicing it, and that what practice builds is the trainable capacity to carry out the tasks that one's procedural knowledge directs (i.e. strike the keys at the right time without slipping, etc.). Of course, if one has no innate musical talent whatsoever, one is unlikely even to be trainable at the piano.

    Trying to articulate the distinction: the key feature of procedural knowledge seems to be that if you have the (trainable and untrainable) brute capacity to T but not the procedural knowledge, then all that you need to T is for someone to give you information. If you have the procedural knowledge but not the trainable brute capacity, then you need training. And training isn't simply a matter of information acquisition, because you can have information without training (the piano player who tries to learn how to play a Thelonious Monk tune from a book) and training without information (the piano player who is simply drilled in specific key presses without understanding).

    Of course, it's possible to lack both the trainable brute capacity and the procedural knowledge. And sometimes they must be gained in a specific order. That's the case of the piano player -- he must gain some procedural knowledge before he can train his brute capacity by practicing.

  2. What about technique? If I lack procedural knowledge, I lack mastery of a technique; whereas, if I'm not able to do something because it's just not possible for me to do it, that's because there's no technique I could learn (given how I'm constituted) that would enable me to do it.

    (Incidentally, lifting weights also involves technique, e.g. so that one doesn't get injured, uses the appropriate muscles to maximize lifting capacity, etc...)

  3. "all that you need to T is for someone to give you information."

    This would seem to collapse the distinction between procedural and propositional knowledge. As the standard example goes, you can talk all you like about the mechanics of riding bicycles, but you won't know how to ride a bike until you practice (internalizing the information in such a way as to yield practical fluency).

  4. Matthew - that sounds promising! (And you're right that there's know-how involved in lifting weights safely and efficiently, etc. But hopefully it was clear enough from context what I meant.)

    I wonder about certain kinds of mental training. On the one hand, it initially seems odd to think that increasing your memory capacity could involve a kind of knowledge-how. But maybe that's because we implicitly assume that there is no such technique. When I focus on a particular technique, e.g. "chunking", then it no longer seems so strange to say that the chunker "knows how" to remember lots of numbers, whereas the ordinary man on the street does not. If they could come to learn this technique, then presumably this would indeed be an expansion of their procedural knowledge.

    We might then push the problem back a step: what makes something a technique, rather than just a kind of training? Perhaps it is that a technique is used in the performance of the action, whereas training is more indirect: it merely enables us to (later) perform the action.

  5. It seems like you might want to distinguish between, say, the use of a capacity, and the application of a technique to a capacity. And here again, distinguish between the employment of a technique that involves exercising a capacity, and the employment of a technique intended to improve, or enhance, a capacity.

    E.g. there are certain techniques that I can use on the piano to improve my dexterity although, in a certain way, those techniques are different from the techniques I would employ in playing a song. (Practically, there's overlap: perhaps playing a certain song helps improve my dexterity.)

    Re: chunking. Some techniques seem to be opaque, even to the learner. E.g.: chicken-sexing. But clearly, chicken-sexing pupils are learning how to do something, even though they aren't aware explicitly of how they do what they've learned to do...

  6. I suspect that there just isn't a sharp distinction here, so you shouldn't be so worried if your results look a bit ad hoc.

    I think the same about the procedural/propositional knowledge line. I'm persuaded by the teleofunctionalists that having a belief is pretty much a matter of being able to do things with the belief. What the belief is/can be used for determines the content of the belief. But I think it would be hasty to collapse propositional and procedural knowledge on that basis; that they're closely related and probably on a continuum doesn't mean that it isn't possible and probably useful to make what fuzzy distinctions we can.

  7. Richard, I would combine Paul's and Matthew's answers.

    There is

    1. Brute untrainable capacity. e.g. some people are just naturally more athletic than others

    2. Brute trainable capacity. If I train hard, I can improve my stamina and speed.

    3. Technique. I need to know how to pace myself, how long I should stride etc. I could have the first 1 and 2 but lacking 3, I may run too fast in the beginning and be winded, or start off too slowly and not make it in time even though I had the capacity to run 2.4km under 12 min. This, like 2 also needs training, but of a different kind

    4. Procedural knowledge. I know what is involved in running 2.4km. This is distinct from 3 as 3 involves getting a feel for it etc. i.e 3 is about how to apply a techinque while 4 is about when to apply a technique.

    The distiction between 3 and 4 becomes sharper with if we are talking about taking a broad jump. I know that I should swing my arms forward just as I push off from the ground, tuck my legs in when I'm in the air, and keep my arms and head forward as I land so that I dont fall back (4). This is distinct from getting the hang of doing this. That is also distinct from having the power to push off in order to jump far enough to pass. This is divided into what capacities I can improve thorugh training (2) and what I cannot (1).

    I think I've exhausted everything necessary in order to successfully complete a task. Some of these components are rather negligible in some tasks depending on what the task is and who the person doing it is.


Visitors: check my comments policy first.
Non-Blogger users: If the comment form isn't working for you, email me your comment and I can post it on your behalf. (If your comment is too long, first try breaking it into two parts.)