Humanist Discussion Group, Vol. 16, No. 384.
Centre for Computing in the Humanities, King's College London
www.kcl.ac.uk/humanities/cch/humanist/
Submit to: humanist@princeton.edu
[1] From: Norman Hinton <hinton@springnet1.com> (16)
Subject: Re: 16.380 thinking with the technologies
[2] From: "C. M. Sperberg-McQueen" <cmsmcq@acm.org> (96)
Subject: Re: 16.380 thinking with the technologies
[3] From: Leo Robert Klein <leo@leoklein.com> (14)
Subject: Re: 16.380 thinking with the technologies
--[1]------------------------------------------------------------------
Date: Fri, 13 Dec 2002 07:54:53 +0000
From: Norman Hinton <hinton@springnet1.com>
Subject: Re: 16.380 thinking with the technologies
Willard, this is the best possible way of looking at the problem: it's
what I would expect from you, and if things work that way, it's
marvelous.
Over the past 30+ years, though, l I've seen, over and over again,
people doing types of investigations because the technology leant itself
to the particular kind of approach -- apparently without any attempt to
make the technology work for them instead of the other way around. A
good example is the proliferation of KWIK concordances, whether we need
the or not. I truly prefer deciding what I want to do and then forcing
the technology to do it, or failing that, gong ahead and doing it
without the technology,even though that may slow my work down by factors
of ten and prevent me from getting the job done for a year or so. (Or in
one case, for a decade.) I guess it's what Humpty Dumpty said -- "the
question is who's to be master, that's all."
And of course it's also fun to learn the technology well enough to see
ways to make it do things it isn't 'supposed' to do.
--[2]------------------------------------------------------------------
Date: Fri, 13 Dec 2002 07:53:36 +0000
From: "C. M. Sperberg-McQueen" <cmsmcq@acm.org>
Subject: Re: 16.380 thinking with the technologies
At 2002-12-12 00:27, in 16.380, Willard McCarty wrote:
So how do we not get trapped within the scope defined by any
particular tool? I think this must be *very* difficult. A kind of
controlled two-(or more-)mindedness, a detached engagement, seems
the only answer.
It's good to hear you mention two-mindedness; I had been tempted to
reply to your earlier query (in Humanist 16.372):
A technical question: what might be reliable criteria for
determining when a given research problem involving textual data
is approached with relational database technology, when with text
encoding? A related non-technical question: how does one cultivate
the ability to keep such questions always in mind? As has been
said, we tend to view each thing as something to be hammered with
the hammer in hand.
by observing that the solution is to cultivate a knowledge of a larger
toolkit, in order to have something more than a hammer ready to hand
when work needs doing. Concretely, that means one cultivates the
ability to keep such questions always in mind by doing enough work
both with encoded texts and with relational or post-relational
database management systems to be comfortable with either tool.
The technical indicia I use are fairly simple though
non-deterministic: think about it each way and do it the way that
seems likely to be simpler. If the data fit naturally into relational
form, use SQL. If reducing the data to third normal form, on the
other hand, produces more than ten or twenty tables with real data,
and another ten or twenty which serve solely to identify many-to-many
links among other tables, and there is a reasonably simple textual
form for the data, then using XML is almost surely the way to go. If
the kinds of summary reports at which dbms excel are important and
numerous, a dbms is probably preferable to XML (although if there are
only a small number of reports, hard-coding them in XSLT may be
preferable; I do my time-logs in XML + XSLT, because at the time it
was more convenient to do it that way).
In the future, as others have pointed out, the availability of XQuery
implementations from various sources, including (I expect) most
vendors of commercial SQL systems, will make XML more usable for data
manipulation which now seems most easily handled with SQL.
That said, the problem does not seem to me "*very* difficult". For
any given (finite, delimited) set of tools / ways of thinking, one can
avoid unsuitable bias by working enough with each of them to become
comfortable using it -- the reason our tools risk biasing us is,
surely, that using other tools would be uncomfortable. And for an
undefined and undelimited set of tools, the problem is not soluble.
Nothing we can do can prevent us by being surprised by something we
did not know.
Let me put that again in different words: if you are worried about
succumbing to an inappropriate and unconscious bias toward either XML
or SQL, you can avoid it by acquiring facility in each. This won't
always protect you from bias, but it will help make it conscious and
help reduce the chance that it's inappropriate.
If on the other hand you are worried about an inappropriate and
unconscious bias toward XML or SQL, and against something you do not
know about and cannot identify or characterize, then I don't think
there is any way to guarantee that what you worry about will not come
true. And the only way to try to keep from it is, as Patrick Durusau
suggests, to listen carefully to your data:
At least in the early stages, I think projects should be
formulated without regard to available technology (either locally
or read about) so that researchers can state fully what they would
like to do, without regard to whether that is actually possible
with current technology. A very precise formulation of the
research problem and goals of the project would provide a basis
for evaluating available technologies for the one that most
closely meets the needs of the project.
There is, of course, no guarantee that the goals thus formulated are
achievable: you may discover that your goals include fully automated
translation from one natural language to another, or the solution of
the traveling salesman problem in time linear to the number of cities
visited.
You (Willard) respond to Patrick by saying
I think, however, that the problem we face is much harder than
that would suggest. Are not these technologies (text-encoding,
relational database &al.) imaginative forms that give you a way of
thinking not available otherwise? Do they not tend to lead the
mind in directions it would not otherwise go? How, in fact, can
one formulate a research problem completely independently of the
tool?
By formulating the research problem in terms of the entities in the
research domain, and without reference to the entities of the tool
domain. (I grow confused: what is the mystery here?)
I think you are confusing the formulation of the research problem with
the formulation of a plan of attack, or of a solution to the problem.
The bias introduced by tools is, at least in part, the bias of
thinking about problems which that tool will help us solve and not
about other problems. One corrective is to know about more than one
tool -- that helps ensure we are not unduly biased toward a single
tool. This is the answer to your original question about encoded texts
and dbms.
Another is to cultivate the ability to ignore, in formulating
the problem you really want to solve, anything you know about how it
might be solved; this would help ensure that we are not unduly biased
toward the soluble. The fact that technologies give us ways of
thinking which otherwise would not be available seems to suggest that
our tendency to focus on the soluble is not a function of our tools
but is somewhat more deeply rooted.
Michael
--[3]------------------------------------------------------------------
Date: Fri, 13 Dec 2002 07:54:06 +0000
From: Leo Robert Klein <leo@leoklein.com>
Subject: Re: 16.380 thinking with the technologies
At 07:27 AM 12/12/2002 +0000, you wrote:
>So how do we not get trapped within the scope defined by any particular
>tool? I think this must be *very* difficult. A kind of controlled two-(or
>more-)mindedness, a detached engagement, seems the only answer.
Willard, like anything, the best option would be expertise at both ends of
the question -- both theoretical/visionary and technical. There are not
many individuals like that but where they exist (and a few exist here for
example), you can see the quality of their work.
LEO
--------------------------------------------------------------------
Leo Robert Klein Library Web Coordinator
home :::::::::::::::::::::::::::::::::::::::::: http://leoklein.com
office :::::::::::::::::::::::::::::: http://newman.baruch.cuny.edu
--------------------------------------------------------------------
This archive was generated by hypermail 2b30 : Fri Dec 13 2002 - 03:22:52 EST