LINGUIST List 6.1423

Sat Oct 14 1995

Disc: Response to suggestions

Editor for this issue: Anthony M. Aristar <aristartam2000.tamu.edu>

Directory

1. Karen Ward, Re: 6.1393, Disc: Self-censorship
2. "The LINGUIST List", Moderators' Note

Message 1: Re: 6.1393, Disc: Self-censorship

I must respectfully disagree with several of the well-meant and
thoughtful suggestions that Lloyd Anderson (ecolingaol.com) has made
for improving the list. Personally, I find the Linguist list as
currently managed to be the most cost-effective mailing list I
receive. However, I am willing to acknowledge that other folks'
experience may differ.

My specific concerns with Dr. Anderson's suggestions:
>Another possibility is that the QQ category be subdivided into elementary and
>advanced QQ. Access to the elementary QQ list could be unrestricted, and
>subscribers would not be compelled to receive these messages.

I'm concerned about the effect of ghettoizing a class of queries; the
term "elementary" may be perceived as somewhat dismissive. These will be
less likely to be answered, and less likely to be answered by someone
with the knowledge to answer correctly. In my opinion, if the question
is really too elementary to bother the main list with, then it shouldn't
be accepted at all. Instead, the questioner can be politely redirected to
another forum.

Of course, either rejecting the query or relegating it to a secondary
list will place an additional, possibly uncomfortable, burden on our
moderators. I don't see a good way out of this, however.

>(2) Topic-subdivided subscriptions
> If a number of core interest areas could be identified, subscribers could
>then tailor their subscriptions to the range of topics which suit them.
	[...]
>the AZTLAN network (using the Listserv software to do this). Subscriptions
>thus can gradually approximate a boolean search on keywords. That surely is
>a goal? How far can we move in that direction?

This can also be approximated by simply dividing the list into many
separate, more specialized, lists. Although I can understand others'
impatience with the range of topics discussed on this list, I enjoy
the diversity and variety and I would hate to see it fragmented.

If you wish to consider other models, though, you might consult with Anita
Borg (borgpa.dec.com). Dr. Borg founded and managed another highly
successful (read: large, at times unwieldy) mailing list for women in
computer science. She recently bent her research interests to the
problems of maintaining the benefits of a mailing list without
inundating folks' mailboxes and without going the Linguist route
of employing a team of human moderators. The result is the Mecca
system, which makes use keywords and registration similar to that
proposed by Dr. Anderson. It is still in early deployment, so it is
difficult for me as a subscriber to comment on its ultimate success
in meeting its goals. However, Dr. Borg may have some early results
and insights to share; I don't know what has been published to date
on this project.

>(a) Include fewer distinct topics per message, so that even the truncated
>subject titles are more revealing.

To offer another data point, I find the current compilation policies to
be appropriate. My perception is that it is generally only short
summaries and questions that have multiple topics per message, and those
are generally easy to scan quickly in any case. I would prefer *not* to
see more, shorter messages from Linguist; the sheer numbers of messages
would cause an unacceptable level of mailbox clutter for me.

>(c) Drastically shorten the general intro material which prefaces each
>message. Only the parts should be included which are necessary for
>graciousness and politeness towards fellow list users. Following two
>examples:

Well, I like knowing who our intrepid moderators are, and I think that
it's only appropriate to give an explicit nod to the individual who put
the issue together. Still, I suppose that we could consider posting the
administrative information only periodically, say once per month or once
per quarter. But please don't remove this information altogether.

> (This point could be promoted by another software device: The Return
>Address or Reply Address for Queries could be set to the individual
>originator only, rather than as the LIST address. Then "Reply to sender"
>buttons in email software will not cause messages to be sent to the entire
>list. )

I believe that on many (most?) mailers, this will only work if you assume one
query per message.


Karen Ward (wardkcse.ogi.edu) 			http://www.cse.ogi.edu/~wardk/
Department of Computer Science and Engineering
Oregon Graduate Institute of Science & Technology

 Moderators' Note

Once again we've received a list of sane and helpful suggestions
about how LINGUIST could be improved. Many thanks to you all.
We think that it's perhaps time for us to respond publicly
to these, since many of the ideas are ones we are already intending
to incorporate into the "new" version of LINGUIST--i.e., the
Worl Wide Web direct access site we are in the process of setting up at
Eastern Michigan U.

Some background info may be in order here: in 1993, we got a $4000
grant from NSF for the purpose of developing faster, more reliable
editing software. Like all software projects, this one has
taken a lot longer than we anticipated. But, in fact, the
delay has worked in our favor--as has our choice of programmer.
This summer John Remmers, a computer science prof at Eastern
Michigan (whose name now appears on that overlong header!) finished
an editing interface that has cut our issue-making time in half and
is simple and failsafe enough to greatly decrease the time it takes
us to train new student editors each year. We've been testing
this "makeissue" program for the last few months.

Now John (still working at, we figure, the payrate of $0.0001 an hour)
is developing software that will create hypertext issues of LINGUIST
and allow us to mount a WWW direct-access site at Eastern Michigan U.
later this year. Our involvement in this software development, and our
knowledge that a new hypertext version of LINGUIST was shortly to appear,
has made us postpone many planned revisions in the e-mail issues. Some
of these we have, however, incorporated into the design of the hypertext
issues and WWW site.

For example, simplifying the "how-to's" and shortening the head notes is
already part of the hypertext design, and we have discussed the idea of
making different reply options available in different kinds of issues,
e.g., reply to poster only in queries issues. This is implementable
in hypertext issues--but not, as far as I know, with e-mail.

Most important, however, is that the new Web version of the list will,
we hope, offer a genuinely hypertext environment. It will, when
completed, link issues into the appropriate threads, topics
and summaries, and allow you, the readers of LINGUIST, to jump
between related issues, to follow the thread of a discussion, and
to go directly to earlier issues of the list to which references
have been made. This means you'll be able to bypass much of the
auxiliary material which now clutters LINGUIST.

About the header: we are aware that seeing the same material over
and over again at the top of an issue may be irritating and it certainly
slows reading. But this is one area where the needs of subscribers
may have to be subordinated to LINGUIST's need to secure support.
As long as LINGUIST remains crucially dependent on resources donated by
the English Depts of EMU and Texas A&M, we need to give these institutions
prominent credit. And as long as LINGUIST continues to be staffed by
graduate students and faculty volunteers (see above), we are very loath to
leave any of their names off the header. Name recognition within the
discipline is almost the only thing that LINGUIST has to offer the people
that work so hard for us.

About topics: we have finally got the "set topics" function to work for
our e-mail subscribers. (The first time we did this, the listserv
protested by shutting down virtually all the mailing lists in the
eastern U.S. Neither we, nor the sysop, could ever figure out why; but
it made us wary of setting topics again.) Now, however, this listserv
function seems to be working well. However, the topics are issue types
(e.g., Books, Jobs, Confs, Calls, Qs, FYI, TOCs, and Disc - for discussion),
not linguistic fields. We have always resisted streaming LINGUIST into
different linguistic topics (phonetics, phonology, syntax, morphology,
etc.), because there are already specialized lists treating almost all
of these topics, whereas LINGUIST remains one of the few places
where people of all linguistic persuasions are encouraged
to confront each other and to consider issues outside their own
sub-specialties. We think such cross-disciplinary interchange has
value.

Thanks again,

Helen & Anthony