NetNews Usenet Archive 1992 #31/NN_1992_31.iso/spool/soc/feminism/3063

Internet Message Format | 1993-01-01 | 6.0 KB

Path: sparky!uunet!cs.utexas.edu!uwm.edu!spool.mu.edu!agate!remarque.berkeley.edu!muffy
From: [e-mail] (Russell Turpin)
Newsgroups: soc.feminism
Subject: The numbers problem
Date: 1 Jan 1993 21:05:27 GMT
Organization: CS Dept, University of Texas at Austin
Lines: 116
Sender: [e-mail] (Muffy Barkocy)
Approved: [e-mail]
Distribution: world
Message-ID: <[e-mail]>
NNTP-Posting-Host: remarque.berkeley.edu
Summary: Now for something really different.
Originator: [e-mail]

-*----
A few years ago there was a discussion on the net started (I
think) by a woman who was a graduate student in physics.  The
problem that she recounted was that she studied and worked in an
environment where she was one of very few women, and she received
too much unwanted male attention.  A respondent asked whether she
wanted her male colleagues to treat her as out of bounds
socially.  No, she replied, she just wanted "normal" treatment,
including the possibility of dating, but not so much unwanted
attention.

I think if she had applied her analytic acumen to the problem,
she would have realized that what she wanted was -- sadly --
impossible to achieve in the situation at hand.  If *each* of her
male colleagues treated her normally (which might well have been
the case), then by virtue of their number, she would receive more
attention than desired, much of it unwanted.  There is no
practical behavioral rule that her colleagues could *individually*
follow that would alleviate the problem.

This is one instance of what I now see as a general problem.
I call this the numbers problem.  It is characterized by:

  1) Having a group of people whose behavior is viewed as
     undesirable because of its *cumulative* effect.

  2) Requiring a solution that ameliorates the cumulative
     effect but that does not totally eliminate the individual
     behavior.

  3) Requiring a non-institutional, deterministic, and
     actor independent solution, i.e., a behavioral rule that
     each member of the group can follow that is independent
     of other members' behavior and that does not require
     randomizing mechanisms nor centralized control.

Another instance of this problem is the predominance of men in
most Internet newsgroups, including those having to do with
women's issues and feminism.  Assuming that all male posters are
cooperative, what rule should they follow to solve this?

At the extreme, one could propose the rule "don't post."  This
is fine if the desire is a newsgroup in which *only* women
participate.  I do not oppose such a newsgroup, but it falls
outside the problem I describe here.  (It fails the second
criterion above.)  As with the physics student, if she had desired
*no* male attention, the rule that solves the problem is obvious.

Let's assume cases where *some* male participation is desireable,
but not so much.  Then what rule does one propose?  "Only some of
you post?"  How does a male poster know whether he is one of the
ones who should or shouldn't post?  "Post only when you have
written something really good."  Of course, every male poster
thinks that he only writes great stuff.  (Or at least, enough of
us do to break this solution.  Similarly, in the case of the
physics student, it is hard to know beforehand if one's initial
social attention is wanted or not.)

I do not think that the problem as stated has a solution.  The
solutions I know weaken the third criterion above in various
ways.  They fall into three categories.  (I suspect if I wanted
to formalize the problem, I could prove this.)

  Institutional:
     For example, the moderators of soc.feminism could impose
     tighter criteria for men's posts to achieve the desired
     balance.

  Actor dependence:
     For example, before posting, each man counts the number
     of men's posts in the last hundred in his news spool and
     then does not post if it is above a certain ratio.

  Probabilistic:
     Before posting, each man roles a dice and rejects the
     post he would have written if the number is too low.
     (The moderators vary the number from month to month
     in order to achieve the desired balance.)

I suspect that solutions in the last two categories are
impractical, because they violate the principle of poster
vehemence, i.e., the well-known fact that posters who are very
agitated about an issue do not want their voice squelched
mechanically or randomly and are unlikely to follow a rule that
requires this.  They want at least a chance to speak their say,
if only they are eloquent and insightful enough for the
moderators to pass on their post.

In short, this is a hard problem.  I would end with three
observations.  First, it is not the men's fault.  The problem
stems from their number, not their individual behavior.
(Obviously, there are also problems with some individuals'
behavior, but solving that will not solve the numbers problem.)
Second, if this is perceived as a serious problem in
soc.feminism, then the only solution is likely to stem from
action on the part of the moderators.  I know they don't want to
hear this, but I call them as I see them.  Third, this can be a
serious problem in places such as physics departments, where
education and career are on the line.  There, none of the three
qualified solutions is broadly useful.  The institutional solution
is out because few universities these dayswant to impose social
coordinators who control dating, and few graduate students would
accept such an arrangement.

A general solution would be useful, but I doubt it exists.  I was
tempted to cross-post this to sci.math, but that would just
invite more male voices into soc.feminism.  Hee, hee.  Sigh.

Russell


--
Post articles to soc.feminism, or send email to [e-mail].
Questions and comments should be sent to [e-mail].  This
newsgroup is moderated by several people, so please use the mail aliases. Your
article should be posted within several days.  Rejections notified by email.