15.313 language, habit and computation

From: by way of Willard McCarty (willard@lists.village.Virginia.EDU)
Date: Sat Oct 13 2001 - 12:24:00 EDT

  • Next message: by way of Willard McCarty: "15.315 conferences: computer games, computational linguistics"

                   Humanist Discussion Group, Vol. 15, No. 313.
           Centre for Computing in the Humanities, King's College London
                   <http://www.princeton.edu/~mccarty/humanist/>
                  <http://www.kcl.ac.uk/humanities/cch/humanist/>

             Date: Sat, 13 Oct 2001 09:16:12 -0700
             From: Willard McCarty <willard.mccarty@kcl.ac.uk>
             Subject: a thought about language and habit

    This is to solicit help in thinking through what at the moment seems to me
    a productive question about language with respect to computing. This
    question is as follows. Please forgive my relatively untutored speculations
    about at least one academic field in which I have no training whatever.

    In language production we tend to be very conservative if not hide-bound
    when we imagine what it is possible to say or what a set of rules will
    generate; thus, as I understand a frequent complaint, the unreliability of
    intuition, including a linguist's. We tend to be radical, however, in our
    actual production of new forms, taking delight in breaking rules, doing the
    supposedly impossible. (I recall once John Sowa declaring that his system
    would of course admit the sentence "the man is eager to eat the food" but
    of course exclude as nonsensical the sentence "the food is eager to be
    eaten by the man". No one would ever say that -- except, prominently, in
    the Hitchhiker's Guide and by many, many people subsequently.) The question
    I have comes from reflecting that a language-generation program is thus
    neither conservative nor radical with respect to language generation. Does
    this mean that we "follow rules" in a very different sense than the
    computer, or not really at all, even or especially when we are trying our
    hardest to do so? One of course can argue that we follow rules but make
    exceptions and/or are subject to unpredictable deviations. But is this not
    a *model* of human behaviour rather than the form of it? If so, then our
    relationship to the machine is very interesting indeed.

    This morning, before sunrise, I made cappucino as usual, while still half
    in dreamland. I would have said (and would have said about myself, had I
    been observing me) that I was operating according to a set of
    self-established rules, then as close to a mechanical being as I ever get.
    Granted the rules are complex and do admit some variation, but (again) I
    would have said that these are not only finite but also small in number.
    One of these exceptions is that when I empty a milk-bottle I rinse it out,
    then take it to the front door, placing it just outside for the milkman to
    collect (yes, this is a very civilised country). I have done this hundreds
    of times. Yet this morning I looked up into the sky and saw the sharp,
    brilliant crescent moon, which in a sense changed everything. Was I
    following a set of rules? Ordinarily one would call my behaviour habit,
    which is a very different kind of thing, admitting within its strictures
    perfect freedom, perhaps even establishing the conditions for freedom.
    Every way I have for thinking about this, in language, surfaces as another
    model -- or I should say, metaphor. Which prompts me to ask, does "model"
    model metaphor?

    Not to get too fuzzed out, I return to the question of rules. If we are not
    in fact rule-governed but habitual, then is not the silent assumption that
    we are rule-governed fatal to clear thinking about computing?

    Yours,
    WM

    -----
    Dr Willard McCarty / Senior Lecturer /
    Centre for Computing in the Humanities / King's College London /
    Strand / London WC2R 2LS / U.K. /
    +44 (0)20 7848-2784 / ilex.cc.kcl.ac.uk/wlm/



    This archive was generated by hypermail 2b30 : Sat Oct 13 2001 - 04:32:33 EDT