Graduate survey papers published?

I Find Karma (adam@cs.caltech.edu)
Tue, 17 Jun 97 21:07:51 PDT


[from a Red Rock Eater post]

> From rre-request@weber.ucsd.edu Tue Jun 17 20:01:50 1997
> Subject: notes
> ...
> Graduate school is a good thing, and I think we should keep it regardless
> of what all the fashionable think-tank anti-intellectualists say. Going
> through graduate school means getting a new identity: shifting from
> the anonymous identity of undergraduate life to the unique professional
> identity that one acquires by presenting one's research in public.

Losing between 3 and 12 years of one's youth in the process.

> That's a painful transition, and good graduate schools try to smooth it.
> One common part of this transition is making the students write a survey
> paper: a paper that summarizes and interrelates "the literature" -- all
> of the work that has previously been written on some topic. It's a great
> exercise because you're both contributing to scholarship and mapping out
> the intellectual territory in which you're planning to locate yourself.

Wait, *everyone* cannot write survey papers. There aren't enough
sub-disciplines.

> Some of these graduate student survey papers are very good.

Most are not.

> Unforunately, in the paper-based system we have now they mostly just
> get filed away. I think we should put them all on the Internet.

I think he's nuts. Most survey papers I've read have been awful.

> Let's ask some librarians to fashion a standard for these documents
> and their cataloguing, indexing, and management. At one level we'd
> want to evolve a genre for these survey papers, with somewhat
> standardized kinds of language and notation. Then at the next level
> we'd want formatting templates so that the papers look alike and
> afford the same kinds of automatic processing. For example, we
> could use SGML (assuming that SGML doesn't croak) to give each field its
> own templates. Or we could follow the technical fields and use LaTeX and
> postscript, with automatic (and hopefully much improved) HTML generators
> for online viewing.

This idea is really batty. Why should we trust graduate students to
write authoritative surveys on anything???

> Each survey paper would go through a formalized refereeing process. For
> example, the paper could be made available in draft form to the author's
> local research group first, then to a specified broader circle of friends,
> and then finally to public comment.

Yo buddy, I have enough trouble getting comments out of people under the
current circumstances. If people are continually having to read survey
papers too I will never get a free moment of their time.

> Each field would have its own editor of surveys, just as journals have
> editors now, and this editor would ask for anonymous referees'
> comments in addition to the public comments. The paper might be
> revised several times before being officially published, and the
> publication date would mark a new graduate student's official
> debut into the research community.

Again, the number of subdisciplines just doesn't make this feasible.
Unless I'm willing to put up with 500 surveys of computer security, for
example.

> Once online, the survey papers would be searchable by automatic tools,
> starting from the topic or authors, or working backwards from particular
> citations. They would be given their own catalog entries and included in
> online library catalogs. The catalog entries would be hyperlinked to the
> papers themselves, and the papers' citations would be hyperlinked to the
> catalog entries for the publications they represent.

Hyperlinking is a pain. Until better tools exist, this idea isn't
feasible, either.

> Taken together, the technology and institution of the graduate student
> survey papers would become a pivotal resource for the world of research.

But why let the lifeblood of fields depend on the newbies?

> The papers would be extremely numerous, and they would be easy to find.

Does anyone else see the contradiction in this sentence.

> Barriers to entry into a research community would be greatly lowered,
> since someone wishing to conduct research in a new area would simply
> be able to fetch the appropriate surveys from a well-known location.

And you think we want this?? Once we start shpreading the bits and clue
around, we invite all kinds of mechanical cockroaches into the mix.

> The surveys themselves would be extremely numerous, given that thousands
> of graduate students write these things every year, and students would
> be challenged to define their survey areas in new ways.

And he doesn't see this as a problem?

> No longer would it be necessary for a hundred graduate students to
> write survey papers on the sociology of organizations, although it
> wouldn't hurt to have ten such papers, written from different
> conceptual or national perspectives.

Jiminy crickets.

> Best of all, the graduate students themselves would acquire well-defined
> identities fairly early in their graduate careers.

Is this really so hard to do? Everyone knows Rohit, and he's not even a
graduate student yet.

> It would then be much less painful to submit one's first conference
> paper, currently a source of tremendous anxiety for many students.

Who the heck has anxiety about this?

> Students would have a standardized mechanism for getting feedback
> before their work was released to the whole world, thus alleviating
> the equally extreme anxiety of publishing work that -- surely, one
> always thinks -- will get brutally shot down in public by some
> powerful researcher.

I guarantee that the potential for this to happen will still exist.

> Of course, everyone will look back and wish that their public survey
> paper was stronger. But that's okay; once you've started your career
> and obtained your unique professional identity you can focus on the
> future, building professional relationships and presenting new and
> better work to bigger and friendlier audiences.

I really dislike the whole schmoozing aspect of academia nuts.

> Many people never make it through that transition because it's just
> too daunting. I want to change that.

Like Rohit would say, if they're too daunted, they don't deserve to be
there in the first place. *Justify* *your* *existence*.

----
adam@cs.caltech.edu

Hey, I eat healthy. If I have to take out an eye, that's the breaks.
-- Jerry to George in the diner after Jerry squirted grapefruit juice in
George's eye on Seinfeld