3.597 the analytic Mac (91)

Willard McCarty (MCCARTY@vm.epas.utoronto.ca)
Tue, 17 Oct 89 19:39:36 EDT

Humanist Discussion Group, Vol. 3, No. 597. Tuesday, 17 Oct 1989.

Date: Tue, 17 Oct 89 11:09 ADT
Subject: Stylistic/linguistic analysis on the Macintosh

Re: Stylistic/linguistic analysis on the Mac

Recent postings have brought up the issue of software for the
Macintosh which would do the concordance/text retrieval work many of us
need for our analyses, be it stylistic or -- in my case at least -- linguistic.
Brian Whittaker asks why it isn't there. The answer is simply that that
particular "vertical market" has not yet been developed. One of the reasons
that far more has been done for the IBM environment is probably because
the universities which have developed such software happened to have been
generously equipped with machines by IBM. What I want to add here is a)
Some packages and prototypes do exist b) new software development tools
are coming out, which will make it easier to design Mac software c) even
if existing IBM programs (such as WordCruncher, as Geoff Rockwell hints)
were to be released as Mac versions, there is still a lot of room for im-
provement before a truly satisfactory concordance/text retrieval package
for Humanists is attained. So anyone who sets out to develop appropriate
sofware for the Mac has my full support. Here are some elements.

a) In addition to Sonar and Gofer, The Mac Buyer's guide lists, under the
heading Text Search and Retrieval: ArchiText and Roundup. Sonar was reviewed
by John J. Hughes in Bits and Bytes Review 1.9, Aug. 1987. At that time is
was the only available program of its type, although Hughes also mentions
the existence of MicroDynamics MARS System. That issue of B &B, and the
preceding one focus exclusively on text retrieval programs, and Hughes'
detailed "wish list" of features to be included in an ideal system for humanists
could still be a useful starting point for designing a new program
In the Hypercard world there is of course Tex, (by Mark Zimmerman), which,
although much improved over Texas, is still rudimentary as compared e.g.
with WordCruncher. A beautiful prototype was demonstrated by Etienne
Brunet at the Dynamic Text conference in Toronto, HyperBase. If it were
made generally available, and Brunet said he had no such intention, it
would satisfy most of my needs. HyperBase was made available to the
general public at the Pompidou centre in Paris during the Bicentennial of
the French Revolution to search a set of texts from that period. It con-
tains, however, sophisticated tools for stylistic analysis, including a
statistical component (correspondence analysis). I am sure Brunet (Brunet@
frmop11, Universite de Nice, France) would be a helpful contact for any-
one working in the HyperCard environment.

b) Rather than HyperCard, I would strongly recommend SuperCard, which I have
had for a few weeks. It is much easier to use, more powerful, and allows for
multiple windows on screen. You can import existing HyperCard stacks into it
and refine them to make use of the additional features. You cannot run
SuperCard "projects" on HyperCard, but one important feature of SuperCard is
that the "projects" can be made into standalone programs, which anyone can
use, on a 1MB Mac, without needing either HyperCard or SuperCard. To develop
programs on SuperCard, you do need 2MB or more, especially if you want color.
SuperCard is available from Silicon Beach Software and costs about $200 Can.
Another new product which I have seen advertised is Serius 89 (from
Serius: "Programming for the rest of us" ). According to the ad, the program
"presents a pallet of programming components easy to identify, easy to use and
HyperCard compatible. As fast as you can select icons and link them together
on the screen, Serius 89 Object Interaction Protocol connects the chains of
functions and converts them into a bug-free application."

c) Even though I use the Macintosh for most of my computing needs, I found
it necessary to invest in a MS-Dos system, which, in my case, is completely
dedicated to concordance work, using two IBM programs alternately, since
neither can fulfil all my needs, all the while keeping my eyes open for
developments in the Mac world. One is of course WordCruncher, the other is
a less known program called SATO, developed by the Centre d'ATO, Universite
de Quebec a Montreal, which has a French interface only. It is the best
program I know if you need to tag a text extensivly. It allows you to have
many "layers" of tags, that you can make selectively visible or invisible, and
sort by and search for. These tags (i.e. sets of values within a category)
can be attached to words in the text both at the "type" and the "token" level.
Any Mac program that could take over would have to have this type of feature,
at least for my type of work.

I'd be happy to supply more details on any of the above points.

Karin Flikeid
Dept. of Modern Languages
Saint Mary's University
Halifax, N.S. B3H 3C3