14.0658 black-box vs glass-box methods

From: by way of Willard McCarty (willard@lists.village.Virginia.EDU)
Date: Sat Feb 10 2001 - 02:55:49 EST

  • Next message: by way of Willard McCarty: "14.0660 the Luhn papers?"

                   Humanist Discussion Group, Vol. 14, No. 658.
           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, 10 Feb 2001 07:54:24 +0000
             From: Willard McCarty <willard.mccarty@kcl.ac.uk>
             Subject: Re: 14.0653 black-box vs glass-box methods

    David Hoover, in Humanist 14.0653, has I think moved us toward the core (or
    at least a core) of the matter I was blundering toward in my question about
    black- and glass-box methods. His two additional scenarios do this by
    setting out extremes at the crossover-point of which, it seems to me,
    humanities computing is most itself -- and most interesting to me. To
    refresh our memories,

    >1. Suppose researcher A gets access to an important medieval literary
    >manuscript, digitizes it, creates a web site at which a searchable PDF
    version
    >of the document resides, along with links to related materials, and
    >creates and
    >moderates a listserve devoted to discussion of the manuscript's
    importance and
    >implications. There is nothing particularly revolutionary here, and the
    >researcher is not writing primarily about the computer techniques or
    software,
    >but rather using them to create a significant scholarly literary resource
    that
    >would be impossible without the technology. Is the project valuable? To
    >whom? Is it humanities computing? If not, what is it?

    We need many, many researchers like A, of course. A's method may not be
    interesting from a technical point of view, but it does exemplify the large
    intellectual difference simple tools can make. It seems to me that a study
    of that difference is among those activities that we wish to include under
    the rubric of humanities computing, in a subdivision of our interdiscipline
    corresponding to the history & sociology of science. The fact of that
    difference is prime, exportable news for the other disciplines with which
    we interrelate. Even if we ungenerously say that A is not doing humanities
    computing him- or herself, someone else can by looking at what A has done.

    Let me try for another example of the above. Suppose a mathematician
    proposed and proved theorem Z in number theory. Subsequently a computer
    scientist applied theorem Z and produced important results in CS with it.
    Those results would not become part of number theory, as Z certainly would,
    but would certainly become part of the story of number theory and might
    well be included in a broader course situating mathematics within related
    disciplines. If mathematics were pursued as an interdiscipline, then its
    number theoreticians would want to include the application of Z among the
    connecting points that helped to define their practice. (Would it not be
    interesting to enquire into why mathematics differs from humanities
    computing in respect of its relation to other fields? Can someone help here?)

    >2. Suppose researcher B selects a cutting-edge computer technique normally
    >used
    >in the business world and applies it to a set of well-known texts.
    >Collaborating with one of the programmers who wrote the software, B writes
    >an article in
    >which he explains how the technique works and the modifications that were
    >required to
    >apply it to literary texts. By using the new techniques, B also is able to
    >show
    >that Faulkner couldn't have written _The Sun Also Rises_. Here B is writing
    >about the technology primarily, and even making a contribution that might
    >be of
    >interest in other fields that use the technique, but the results might not be
    >very interesting to literary scholars. Is this project valuable? To whom?
    >Is it
    >humanities computing? If not, what is it?

    A real example of good humanities computing that also borrows a technique
    from elsewhere is Peter Robinson's use of cladistics (from evolutionary
    biology) in the analysis of mss interrelations. His results for Chaucerian
    studies are most impressive. In the spirit of Dr Hoover's example let us
    suppose, however, that Peter had not done his work; rather a very different
    researcher like B had, applying cladistics to mss study, demonstrating with
    it that later, derivative ms. Q is the authorial copy and that oldest,
    historically closest ms R has been copied from it! Ridicule everywhere, of
    course; the researcher relocates to a remote island in the Black Sea and
    becomes a goatherd. Nevertheless, one of us is not so easily fooled,
    investigates and finds that there are no methodological errors -- the
    application of cladistics should have worked, but didn't. A very
    interesting research problem for us thus comes to light. The disgraced
    researcher's work per se is worthless, but he or she has accidentally
    stumbled upon something valuable.

    >Willard seems to be suggesting that achieving disciplinarity for Humanities
    >Computing may require that work recognized as belonging to the discipline be
    >focused on the implications or theory behind the computer application.
    I'm not
    >sure what I think about this, but I do wonder whether it doesn't create a
    >precariously narrow definition for a discipline. If the work moves too far
    >toward the computer techniques, it risks becoming marginal computer
    >science. If
    >it moves too far toward one of the established humanities disciplines, it
    >risks
    >becoming marginal history, musicology, literary study, or whatever.

    I certainly agree that we inhabit the middle ground and would argue that we
    must do everything we can not to be intellectually or socially exclusionary
    -- but at the same time we need to be able to say what is interesting in
    terms of the interdiscipline. We'd be making a huge error to strive toward
    a narrow definition of the field. A de-finition is, however, an act of
    drawing boundaries. Perhaps one implication of Jonathan Culler's point
    about the myth of foundationalism -- the historically incorrect notion that
    a discipline, such as English studies, once enjoyed Edenic unanimity -- is
    that defining a scholarly practice is something that its practitioners must
    attempt but can only achieve by killing the practice. The Exodus story all
    over again, not surprisingly.

    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 Feb 10 2001 - 03:01:06 EST