4.0211 Offline 29 (1/330)

Elaine Brennan & Allen Renear (EDITORS@BROWNVM.BITNET)
Wed, 20 Jun 90 18:19:06 EDT

Humanist Discussion Group, Vol. 4, No. 0211. Wednesday, 20 Jun 1990.

Date: Wednesday, 20 June 1990 1114-EST
From: Robert Kraft <KRAFT@PENNDRLS>
Subject: OFFLINE 29

<<O F F L I N E 2 9>>
by guest columnist James O'Donnell, for Robert Kraft
[HUMANIST 20 June 1990]
[Religious Studies News 5.4 (Aug)]
[CSSR Bulletin 19.3 (September)]

<Accessing Remote Libraries>

One of the topics of special interest to a number of readers
who responded to the invitation in OFFLINE 28 is how to access
remote libraries. Since everything I know about that topic has
come from my colleague in Classical Studies, Jim O'Donnell, it
seemed sensible to ask him if he would write a guest column for
OFFLINE on that subject. He agreed, to the benefit of us all, and
the results follow. The directions work for me, but then, I am on
the same computer system that Jim uses. You may need to put a bit
more energy into making them work, but keep at it -- the results
are well worth it! Thanks, Jim, for sharing your expertise.

<La bibliothe/que imaginaire d'INTERNET>

"Ad hoc, ad loc, quid pro quo, so little time, so much to
know" was the complaint of the Nowhere Man in the animated film
_Yellow Submarine_. The paradox of the computer age is that it
makes it possible to learn more things in less time, while at the
same time making many more things for us to know. At Bob Kraft's
request, I am going to sketch here one advance that has been
revolutionary for me and for others. Any working scholar with
experience in e-mail will find it all rather transparent; the
scholar who has not yet gone on-line with the world will find it
rather less transparent at first, but should be assured that in
fact it is all a piece of cake. If you can do _anything_ on an
IBM or Macintosh, you can summon the wisdom of the world down the
wires into your machine: it's easier than learning to use a word

The first principle is that the great research libraries of
the world have been and are continuing to make their collections
more accessible through computer cataloguing. Most major
university collections now have at least part of their collection
in an on-line catalogue and most users are now accustomed to
looking not only in a card catalogue but also in one of a row of
terminals usually found standing in the entrance hall of the

The second principle, specific to the world of computers, is
that any information in any computer anywhere in the world is
theoretically available to any other computer anywhere in the
world, including the one on your desk. In practice, there are
often obstacles, but happily librarians genuinely enjoy
minimizing those obstacles. Many exciting developments remain,
but much has already been done.

What can you do? From any modem equipped telephone anywhere,
you can now reach a huge variety of library catalogues. Now
catalogues are not the same thing as books, and it is certainly
frustrating to learn that a book is on a shelf someplace where
you can't go; but it may be useful to know that anyway. Among the
uses of the kind of library searching that I will describe below
are these: browsing specialized collections in remote libraries,
confirming the existence of and locating relatively uncommon
volumes, searching in catalogues better equipped than that of
your home institution (this can be useful in several ways), and
the exhilarating sense of intellectual play that comes from
nosing through any collection of books anywhere.

Some examples. I grew up in New Mexico and Texas. The
University of Pennsylvania library is not specially strong in
southwest regional history and sociology; but the collections of
the Universities of New Mexico, Colorado, and California (to name
the ones I've had access to) are much stronger. I can learn of
the existence of materials, get confirmed bibliographical
records, and (if it came to that) decide which collection(s)
might be so strong as to be worth a visit sometime.

Often I find myself in possession of a defective and obscure
bibliographical reference: title, author, date, with perhaps the
title slightly garbled. The Penn library doesn't have it on-line.
A little intelligent snooping, and I find it in the University of
California system: I get a confirmed title/author/place/date
record and take that the next day to our Interlibrary Loan
office, where they do a much better job of getting the book
quickly than they ever could have with the defective record with
which I started. But, you might wonder, wouldn't the National
Union Catalogues have the same information? Probably, but: (1) I
just finished rearranging all my books at home and I've confirmed
that I _don't_ have room on my shelves for all the NUC volumes,
not even merely (sic) the pre-'56 imprints; and (2) the computer
databases can be searched in ways more cunning than the printed
volumes or a card catalogue. If your reference is really
defective as regards author or title, the computer lets you do
searches by parts of words, keywords, subject, and in some cases
even call-number: it's much easier to turn a bad reference into a
good one from a keyboard than by walking up and down helplessly
in front of a row of NUC volumes. You also need to know less
about library cataloguing and filing conventions than you used
to, and that can be a great time-saver in obscure cases.

Perhaps the most important use, however, is for gaining
access to catalogues better than that which your home institution
can offer. At Penn, for example, the full computerized catalogue
covers items received and catalogued since about 1968/70. That
means there's an awful lot of older material just not on line;
some recent additions have put defective and partial records of a
lot of older stuff on-line, but those additions, while they may
help me locate a book that is in the collection, are no
substitute for the full catalogue information that a regular
catalogue record can allow.

My most idiosyncratic use of the catalogues is not for
everyone: I use them as a grand intellectual toy. I have always
found it an important part of the life of the mind to browse,
rummage, snoop, and generally prowl the libraries. On a leave a
few years ago, I devoted some time to reading through the shelves
of the Bryn Mawr College library collection: I got through all
the philosophy/religion, history, and literature shelves over the
course of a year. Just walking along looking at things, pulling
off whatever struck my fancy, and sitting down to sort them as
often as my arms were full. A richly useful intellectual
experience. With the computer, it is possible to do half of that:
you can't pull books out and look at them, but you can browse and
snoop much more widely, in much bigger collections.

But no description of possibilities can be prescriptive, only
suggestive. Whatever you can do with a library catalogue, you can
do better and faster from your computer: if that means something
to you, read on for directions.


_First_, you need a PC (Mac or IBM are identical for these
purposes) with modem and modem software. _Second_, you need a
connection to the great world. Characteristically, this will be
furnished by your home academic institution. You will probably
have gotten that connection in order to do e-mail of some kind,
or perhaps to have access to on-line student records for
registration or the like. Some institutional connection is
apparently essential: inquiries have failed to find a commercial
service that offers the right kind of interactive link to
INTERNET. There may be an easier way: snoop around academic
institutions close to you. See if any of them allow outsiders to
reach the level of access necessary to get to INTERNET without a
formal account or password; institutional policies will vary
widely. _Third_, you will need to find out from your local
computer gurus how to get on to INTERNET, the nationwide computer
network that links the libraries (and many other facilities).
Usually this is easy. For me, it means giving a single command on
first logging on to the local computer (I merely type "TELNET"
and hit a carriage return <CR>), then I get a new prompt at which
I type the letter T and the "address" (either names or numbers:
example below) I seek. In a matter of seconds, I am linked to the
computer I seek and can begin logging on there. BUT YOU MUST FIND

Now _fourth_, you need more information. The easiest way to
get the information you need is through e-mail. Ask you local e-
mail gurus how to acquire files from remote list-servers. On many
systems, this is as simple as issuing a one-line command from the
basic system prompt; others require you to send a short mail
message. On the common VM/VMS systems, the message you need to
send is:


The crucial elements are the address (UNMVM: a computer at the
University of New Mexico) and the filename (INTERNET LIBRARY).
This file is a collection of very specific and explicit
instructions for gaining access to libraries all over the
country. (If you have had a copy of this file around for some
time and done nothing with it, now would be a good time to get a
new copy: it is updated regularly, with new facilities being
added all the time.) Read this file. (Dr. Art St. George of UNM
deserves at least a medal for his patient work in gathering and
updating this material: it is really the browser's bible.)

At this point, you want to follow your nose and your
inclinations. Which libraries are of most interest to you will be
a matter of taste, and trial and error will confirm them.

For each library discussed in the file, there will be good
instructions how to log on to the individual facility. Here's
where an example helps, so I'll walk you through the University
of Maryland, which is a very easy system to approach.

First on my machine, I have given the link-to-internet command:


Then the call-Maryland command


The number I learned from the INTERNET LIBRARY file.
Some places have also alphabetical addresses like those familiar
from e-mail addresses (e.g., PENNLIB.UPENN.EDU); if you have
addresses in both forms and one doesn't work, try the other; for
Maryland, you would use UMCAT.UMD.EDU. Anyway, after you connect
you get a rather austere prompt, but from INTERNET LIBRARY, you
know to answer:


At this point, you can stop reading this article and start looking
at the help screens you get on the computer. These will be the
same screens you would get if you were in the catalogue department
of the library itself.

You will want to experiment with what you can get. One thing I like
(and don't have at home at Penn) is the capacity to do keyword searches,

k=water buffalo

That is a command that will get you many more "hits" than a
subject search: subject searches are restricted to the kinds of
things that librarians have explicitly selected and ratified, the
kinds of things that were formally listed on the card in the old
card file; but keyword searches look at the whole record, and so
if there is a book with "water buffalo" in the title, you will
get it (even if it's a novel); and if somebody wrote a book under
the pseudonym "Clem Water Buffalo," by golly, you'll hit it. You
may well get therefore more dross with such a search, but also
more gold. (I've also noticed that SUBJECT headings in these
catalogues are the ones most prone to typographical errors: in
the old days, the subject heading typed on the original author
card was not a crucial piece of information: it would be retyped
correctly on the actual subject card. But in the computer record,
_that's_ the place the computer looks for the subject heading.
Curtailing your search target is a good idea here.)

And always remember, a search that doesn't work out takes a
second; you can usually think of a way to improve your question
to get more effective results. A=JONES is going to get thousands
of "hits"; think about how to reduce the range. When in doubt, of
course, always use the shortest possible search target: don't ask
for "water buffaloes" because you'll _only_ get the plural (and
even then somebody might have spelled it "buffalos" and you'd
miss that); ask for the singular and the plural will come along
at no extra charge; and in fact "water buf" is likely to get you
everything you want, and the fewer characters you type, the fewer
chances to make a typing mistake and have to start over.

So, go ahead, play. Browse, snoop, take notes. When you're
done, hang up the modem or disconnect (sometimes systems will
have explicit logoff instructions; if you know them it's polite
to use them, but in an emergency or if you simply don't know
them, just breaking the connection will suffice). Start again. Go
back to INTERNET LIBRARY and look for someplace else to call.


Note-taking. Depending on your software, your hardware, and
the characteristics of the individual library you are calling,
you may be able to "LOG" your call and keep a record of your
results on disk. Sometimes this takes a little practice (at Penn,
for example, there's an undocumented alternate "terminal type"
you need to tell it to emulate in order for most communications
software packages to be able to log successfully). Once you do
that, whatever you see on screen will be recorded on disk for
later editing and manipulation. But almost every computer, modem
software, and printer combination will allow you to PRINT SCREEN,
and that is in fact the way I usually handle it. Get a screen
with interesting information, hit the PRINT SCREEN button, and
nudge my laser printer to eject the paper and I have a printed
record. 99% of the time, this is fine, even wonderful.

Other facilities: There are two main proprietary systems that
enrich the possibilities of on-line library work, OCLC and RLIN.
Both of these systems, which report the holdings of many
libraries, require special access codes and, at least indirectly,
payment of fees. Practices vary sharply from institution to
institution, but at Penn, we may get access to RLIN and our own
account number on request, but the library administration
periodically reviews the costs and benefits and reserves the
right at some point to pass the charges on to users. We do not
have access to OCLC. Consult your local institution (usually
somebody in the library) for information about these systems.
These systems have their own special strengths, special
databases, etc., and RLIN, for example, is the best source I have
for information about very new books --sometimes even finding out
about books before they are actually published, as the Library of
Congress posts information registered by the publishers.


As you play around with INTERNET LIBRARY, you will find that
some institutions have already begun putting other services on-
line: I would like to have a good on-line encyclopedia, and would
be happy to have as many reference databases as possible handy.
At some institutions, you can already check circulation status of
a book and leave, by computer, a request that a book be recalled
or merely paged from the stacks and held for you at the
circulation desk. Similarly, at Penn there is talk of allowing
faculty to initiate their own Interlibrary Loan requests by
machine from home, with the existing ILL staff freed up to
concentrate on the really tough cases and on the management of
the flow of books in and out of the building.

Somewhere beyond INTERNET lies a Borgesian fantasy library,
where all the texts are themselves on-line, and where you may
flit from text to text without budging from your desk in your
study, perhaps miles away from the library. A pretty fantasy, but
not all _that_ unrealistic and worth keeping in mind as the goal
towards which all the interim developments reach.


In the first instance, consult your local computer gurus
about INTERNET access and your local library people about things
like RLIN and OCLC. The INTERNET LIBRARY file gives addresses for
queries directed to other libraries. For queries about this
article, the author may be reached as JODONNEL@PENNSAS.UPENN.EDU
(and would be particularly glad to hear of any corrections or
improvements that might be suggested).


Please send information, suggestions or queries concerning
OFFLINE to Robert A. Kraft, Box 36 College Hall, University of
Pennsylvania, Philadelphia PA 19104-6303. Telephone (215) 898-
5827. BITNET address: KRAFT@PENNDRLS. To request printed
information or materials from OFFLINE, please supply an
appropriately sized, self-addressed envelope or an address
label. A complete electronic file of OFFLINE columns is
available upon request (for IBM/DOS, Mac, or IBYCUS), or from
the HUMANIST discussion group FileServer (BROWNVM.BITNET).