11.0481 perl and programming

Humanist Discussion Group (humanist@kcl.ac.uk)
Wed, 24 Dec 1997 14:32:20 +0000 (GMT)

Humanist Discussion Group, Vol. 11, No. 481.
Centre for Computing in the Humanities, King's College London

[1] From: David Halsted <halsted@h-net.msu.edu> (12)
Subject: Re: 11.0478 perl is what you want

[2] From: Leo Robert Klein <kleinl@is2.nyu.edu> (29)
Subject: Re: 11.0478 perl is what you want

[3] From: Don Wilkins <dwilkins@ucr.campus.mci.net> (30)
Subject: Re: 11.0478 perl is what you want

[4] From: Eric Johnson <johnsone@jupiter.dsu.edu> (12)
Subject: Programming for the Humanities

Date: Fri, 19 Dec 1997 18:01:58 -0500
From: David Halsted <halsted@h-net.msu.edu>
Subject: Re: 11.0478 perl is what you want

Well, the reason I asked about the Perl book was that I suspect many
humanists won't see immediately how even (the excellent) _Learning Perl_
applies to their work, or could apply to their work. Seems to me that
there's rather a dearth of computer books aimed at people in the Humanities
and social sciences. Is this impression shared by others?

Dave Halsted

Date: Fri, 19 Dec 1997 18:31:48 -0500 (EST)
From: Leo Robert Klein <kleinl@is2.nyu.edu>
Subject: Re: 11.0478 perl is what you want

On Fri, 19 Dec 1997, Humanist Discussion Group wrote:

> you need to know is basic Perl - so get a basic book. I highly recommend
> anything published by O'Reilly; if you are new to programming, get
> _Learning Perl_.

Having just been through the above work, I would hardly recommend it to
anyone without a pretty good programming background already under one's
belt. The whole is peppered with comments like:
"and for those already familiar with C++, you'll be happy to know
that also in Perl..."

For those not already familiar with C++, the book is not particularly
helpful. It is definitely not written with someone with a background in
the humanities in mind. The introductory "Tour of Perl" is gratuitiously
intense and overwhelming. This is a wonderful book if you feel like
reading it over maybe fifteen times before you can make head or tail of

While I haven't read the Larry Wall book ("Programming Perl"--or
therabouts), which is another Oreilly publication, I have heard it is
rather dense--a quality shared in large measure, by the way, by the
work, "Introduction to Perl", mentioned above.

I tried desparately to find a manual on Perl which was written is
something approximating English and was none too successful. Hopefully
others have had more luck.

Leo Robert Klein

Leo Robert Klein 70 Washington Square South
Web Coordinator & Reference Associate New York, NY. 10012
General & Humanities Reference Tel.: (212) 998-2666
Elmer Holmes Bobst Library Fax: (212) 995-4383
New York University Email: leo.klein@nyu.edu

Date: Sun, 21 Dec 1997 07:54:11 -0800
From: Don Wilkins <dwilkins@ucr.campus.mci.net>
Subject: Re: 11.0478 perl is what you want

Though I don't work with Perl, I am familiar with similar languages (e.g.
Tcl/Tk) and would agree with Gregory Murphy in principle. I was tempted to
respond to the original question about C/C++ and humanities because I am
heavily involved with C/C++ in my own humanities research (Greek, which is
blessed with huge databases). C/C++ has an overwhelming learning curve, and
unless you have the full moral support of your department--which I
unfortunately have not had--you will get little else done and pay dearly for
it in terms of job security or advancement. However, there are at least two
drawbacks to Perl: (1) if I am understanding Gregory's closing comment about
speed and efficiency correctly (I assume that Perl is an interpreter, like
Tcl/Tk, HyperCard, etc.), it may be too slow to do a good job on any large
database, and (2) it may have a learning curve that is still too steep for
the typical humanities prof (meaning that most will not have the patience).
Gregory may very well object to my first comment, and perhaps he will or
would say that Perl is adequately fast. In my own work, however, I find
myself trying to eek our every tiny microsecond of improvement in speed even
in my C++ programming. While I can, in effect, live with a massive search
that perhaps takes a couple of hours, it nevertheless makes a significant
difference if I can cut the time in half or so. Therefore, if you are asking
yourself--to parody Microsoft-- "Where do I want to go in terms of
programming," you should also ask, "How long do I want to take to get
there?" Also, Gregory may have an objection to my comment about the learning
curve for Perl, but I have looked a Perl files before at internet sources
and can only say that I lost patience too quickly to learn it, which is
saying something for someone who has taken on C/C++. Perhaps my problem was
a (false?) perception that Perl is not sufficiently user-friendly. In that
case, I'll conclude by encouraging others to at least take a look at Perl.
You might also take a look at HyperCard, which I think is about as
user-friendly as it gets, but far too slow to do any serious text processing
or searching.

Don Wilkins

Date: Sun, 21 Dec 1997 07:10:35 -0600 (CST)
From: Eric Johnson <johnsone@jupiter.dsu.edu>
Subject: Programming for the Humanities

A recent posting on HUMANIST asked,
> Or doesn't anybody use Snobol/Spitbol anymore?
The answer is that some of us do indeed continue to use modern versions
of SNOBOL4 and SPITBOL. Descriptions of programs for text analysis are
contained in my article "The World Wide Web, Computers, and Teaching
Literature." The full text of my article can be found on the Web at


I would be interested in any comments.

--Eric Johnson

Humanist Discussion Group
Information at <http://www.kcl.ac.uk/humanities/cch/humanist/>