politics vs. software

Thu Dec 17 21:54:54 UTC 1998

At 09:47 AM 12/17/98 -1000, Brian MacWhinney wrote:
>It's fine by me to chide people (Bralich or others) for doing "general
>purpose" politics on FunkNet.  However, I feel that Phil Bralich (and Derek
>Bickerton) should be welcome to tell us about their linguistic analysis
>software.  I realize that some readers view these software announcements as
>"ads".  But these are not "general purpose" software programs.  Rather they
>are programs designed for a very small niche which many of us occupy.  On
>the other hand, such announcements would more of a contribution if Bralich
>and/or Bickerton could explain how the analyses produced by their programs
>help illuminate interesting issues in the functional (as opposed to merely
>formal) analysis of language and conversational processes.
>

Thanks for the support and the criticism.  When I sent the political post
I understood there would be some negative feedback, but felt it was risk
I had to take and I had to suffer the consequences whatever they may be
in the face of situation that is seriously threatening a very good
political system and one that is important in the world. However, I
think one can be sure my real interests lie in theoretical syntax and
in computational linguistics as a proving ground of a theory of syntax.
Thus, you can be sure that the political post was as unusual as the
current political sitiuation.  I and Derek Bickerton both are interested
in the functional issues that are triggered by our results, but of
course our attention is regularly drawn to the immeidate needs of
our further development of the theory and the NLP tools.  As you all
know my main point about this particular formalism is that its status
as an effective model of what the brain does is established by its
ability to actually perform grammatical functions and operations
that humans can do and that the particular parser we have does it
far better than others and for that reason should be considered a
very interesting challenge to theoretical linguistics and theoretical
syntax.

I am also aware that my posts do have an advertising air to them, but
I try to be sure they are never purely ads.  I am in the interesting
position of being on the industry side of a partnership between academia
and industry -- one of a growing number of such partnerships that are
becoming crucial to the future of academia and to a number of industries--
high tech being one of the most obvious.  However, I cannot shirk my
responsibility to those on either side of the fence.  There are students
and professors as well as marketers and CEOs and CTO's who need the
full range of the information I present to make decisions about the
futures of their careers as well as of their respective departments.
For example, I cannot walk into a board meeting and then be shocked
or surprised because I am presenting as new something that exists at
an MIT or IBM web site.  This is also true for others in linguistics.
Thus, if I increase awareness of our theory but only tell a few people
that there are working software programs that illustrate them I am
creating a situation where some of your colleagues could be embarassed
at a board or committee meeting.  (Having someone pop at a meeting with
software tools you claim do not exist is as embarassing as having some
one site a major article from a major publication that contradicts
your position--one that you are unaware of.  So while my advertising may
have a somewhat "unclassical" academic edge to it, it does server two
purposes (besides the benefits my company receives) 1) it allows those
who are working within academia and industry to be aware of what is
available and avoid potential embarassments, and 2) it provides evidence
of NLP abilities and of a theory of syntax that is important to
researchers PARTICULARLY because no one else can create similar tools in
this domain.

Thus, while I recognize the political post was problematic, I felt
the situation warranted dramatic means, and while I understand that
advertising per se is not pure academics, it is a reality of modern
life that we all have to be aware of.  And Derek and I are aware that
our particular formalism has interesting implications for functionalists
and that we do not have time to address them directly.  However, I have
not excluded our posts from this list because there are syntacticians
and theoretical and computational linguists on this list who use this
information in their thinking and as part of their work.

Phil Bralich

Philip A. Bralich, Ph.D.
President and CEO
Ergo Linguistic Technologies
2800 Woodlawn Drive, Suite 175
Honolulu, HI 96822

Tel: (808)539-3920
Fax: (808)539-3924
bralich at hawaii.edu
http://www.ergo-ling.com