17.813 what's needed; pedagogical use of text-analysis

From: Humanist Discussion Group (by way of Willard McCarty willard.mccarty@kcl.ac.uk)
Date: Fri May 07 2004 - 16:56:23 EDT

  • Next message: Humanist Discussion Group (by way of Willard McCarty

                   Humanist Discussion Group, Vol. 17, No. 813.
           Centre for Computing in the Humanities, King's College London
                       www.kcl.ac.uk/humanities/cch/humanist/
                            www.princeton.edu/humanist/
                         Submit to: humanist@princeton.edu

       [1] From: "Amsler, Robert" <Robert.Amsler@hq.doe.gov> (44)
             Subject: RE: 17.802 what's needed?

       [2] From: "Paul Edward Oppenheimer" (80)
                     <paul.oppenheimer@cox.net>
             Subject: Re: 17.804 pedagogical use of text-analysis tools

    --[1]------------------------------------------------------------------
             Date: Wed, 21 Apr 2004 06:40:04 +0100
             From: "Amsler, Robert" <Robert.Amsler@hq.doe.gov>
             Subject: RE: 17.802 what's needed?

    Three things come to mind as having served well in other fields and which
    could be applied to humanities computing.

    The first is from mathematics, namely the creation of a series of "grand
    challenges" to the members of the discipline.
    A Grand Challenge is a problem which typically has a simple description and
    which has eluded a satisfactory and definitive answer for generations of
    practitioners of a discipline. The quest in humanities computing would be
    to state what could become grand challenges, perhaps to computationally
    determine authorship of famous unknown works, or to translate an unknown
    language using computational tools, etc.

    The second is from computational linguistics and information retrieval,
    namely to host "competitions" in which software competitors attempt to
    solve a hard problem in which evaluation of competing systems has proven
    difficult to assess. Computational Linguistics and information retrieval
    were handed this paradigm by their funding agencies, which were faced with
    numerous competing software systems which all claimed to address problems
    such as the extraction of facts from text, the disambiguation of word
    meanings, etc. These tasks are ones in which there exists a number of
    software systems each of which show promise, but which have never be
    applied under similar circumstances to the same body of data so they could
    be compared for the quality of their output. The competitions are typically
    administared by an impartial (non-competing agency) such as the National
    Bureau of Standards or an international committee of representatives from
    academia, who devise the competition's rules, develop a special set of data
    to be used by the competitors and associate with this data the answer
    protocols which will be used to judge the results. Typically, the data is
    judged by humans and those judgements are used to evaluate the output of
    the participant's computer programs. There is usually a training period for
    competitors, in which data is released upon which the participants can
    practice,, and then a release of the test corpus with a deadline for
    responses. At the end, a report/proceedings is prepared by the organizers
    with submissions/papers by the participants. The event is then held again
    annually for a few years in the hopes that improvements in the
    methodologies will result in improved software.

    The third method is from computer science and computational linguistics,
    namely to produce standards for the field. Standards include things such as
    recommendations for the field's academic course content and for sets of
    courses to be included in Masters and Doctoral Degree programs. They also
    have included the publishing of algorithms for standard techniques in the
    field and the establishment of a dictionary of terminology for the field.
    In the humanities there are some standards, such as those the Text Encoding
    Initiative provided, but these apply to text markup and not software. Any
    field which emphasizes computers could presumably also have standards for
    computer software.

    --[2]------------------------------------------------------------------
             Date: Wed, 21 Apr 2004 06:43:12 +0100
             From: "Paul Edward Oppenheimer" <paul.oppenheimer@cox.net>
             Subject: Re: 17.804 pedagogical use of text-analysis tools

    Francois,

    Please expand your cryptic analogy. Thank you!

    Paul
    ----- Original Message -----
    From: "Humanist Discussion Group
    <willard.mccarty@kcl.ac.uk>)" <willard@lists.village.virginia.edu>
    To: <humanist@PRINCETON.EDU>
    Sent: Sunday, April 18, 2004 11:39 PM

    > Humanist Discussion Group, Vol. 17, No. 804.
    > Centre for Computing in the Humanities, King's College London
    > www.kcl.ac.uk/humanities/cch/humanist/
    > www.princeton.edu/humanist/
    > Submit to: humanist@princeton.edu
    >
    >
    >
    > Date: Mon, 19 Apr 2004 07:19:15 +0100
    > From: lachance@chass.utoronto.ca (Francois Lachance)
    > Subject: Re: 17.744 MonoConc (Pro), with thoughts on teaching
    >
    > Willard,
    >
    > I want to pick up the thread of learning objectives and the
    > pedagogical use of tools for textual analysis. I think your comments on
    > MonoConc relate to a blog entry by Matt Kirschenbaum about the exercise
    > set by Douglas Hofstadter in the early pages of Godel, Escher, Bach.
    > See
    > http://www.otal.umd.edu/~mgk/blog/archives/000339.html
    >
    > Matt invites readers to consider why Hofstadter introduces his
    > discussion of the Mechanical, Intelligent, and Unmode with what can turn
    > out to be a frustrataing exercise. That invitation raises similar
    > questions about the value of learning by doing that your MonoConc example
    > embodies.
    >
    > For some reason, the example put forward by Matt and your example have me
    > wondering if certain teachers complement the exploration of the
    > application with the exploration of the objects of study. Does anyone
    > teaching humanities computing set up exercises along the following lines?
    >
    > Present a class with a given distribution and then invite students to
    > discover if the given distribution is replicated in an analysis of
    > different versions of a text. Repeat the exercise with one set of students
    > introducing a typo in one version (or altering it in some form); another
    > group of students is assigned the task of determining if the comparative
    > analysis picks up the change. Repeat with the student groups switching
    > tasks.
    >
    > > Date: Sun, 28 Mar 2004 08:59:01 +0100
    > > From: Willard McCarty <willard.mccarty@kcl.ac.uk>
    > >
    > <snip/>
    >
    > > MonoConc is very easy to learn -- as I said, 5 to 10 minutes is all
    that's
    > > required. The students I've had tend to come to humanities computing
    > > believing that it's about pushing buttons. So I've tried to rush them
    past
    > > the push-button interface to problems of interpretation. The more
    > > sophisticated-in-features this interface is, the harder that is to do,
    the
    > > more they take what they see as a harder problem of the kind they've
    mostly
    > > already mastered rather than a new sort of problem entirely.
    > >
    > <snip/>
    >
    >
    >
    >
    >
    > --
    > Francois Lachance, Scholar-at-large
    > http://www.chass.utoronto.ca/~lachance
    >
    > Wondering if...
    >
    > mnemonic is to analytic
    > as
    > mimetic is to synthetic



    This archive was generated by hypermail 2b30 : Fri May 07 2004 - 16:56:44 EDT