Tracking Scientific Information

(This page records a process of thinking out loud, and goes through 1 Dec 2000. Most of its importance is in the entries after 11 October 2000, but the earlier ones connect the discoveries and decisions to the earlier articulations of problems and possible solutions.)
13 August 1999
The landscape of scientific information is tectonically active: tremors here, subductions there, whole continents adrift... and I need to do something systematic about trying to keep track of what's happening and what the implications are for undergraduates and science in the liberal arts. And so of course another weblet, which may eventually be useful for considering the issue as a problematic in courses. The topic continues on from or links to other weblets I've created, including one for a VFIC workshop, another on envisioning for technology in teaching, one on instructional technology, one on ecology of information, one on "information literacy", and ruminations on CA, ISI, etc. There's one from Dec 1997 on literature access in the sciences that's somewhat outdated now.

The 6 August 1999 issue of Science reports the construction by DOE of PubSCIENCE, a parallel to PubMED and thus an essential finding and navigation tool. The text suggests that the service may be available by October.

26 Aug
A question that recurs about every 3 months: What possible solutions are there to the ChemAbs problem (all that paper, no easy access, very little use)? Is SciFinder Scholar any sort of possibility? ARE there any viable alternatives to paper, where 'viable' includes $$ as a factor? See Oberlin Science Librarians (Mar '99) on the current range of possibilities.

11 October 2000
I need to do the basic thinkwork of figuring out what our options are for supplying literature access for the sciences, considering even the unthinkable (viz: getting everything by document delivery, instead of paper subscriptions) and laying out just what the big choices are. This is precipitated by the latest vendor outrage, in the form of a listserv rumor that Chemical Abstracts is due to be withdrawn from DIALOG, and that users are expected to shift to the ACS's Science Finder, at a cost of something like $10,000 per year.

So once again we must address just what it is that our users want, need, use... and how best to get it into their ken. It's very tempting to grumble about the scanty use of journals, the grasping habits of the likes of ISI and the publishers... but what's really needed is constructive fact-gathering, to support what I anticipate will be a bigtime budget proposal for rationalization of information access in the sciences.

This may be our opportunity to get Web of Science, and to set W&L's users on a different and more productive path. But the costs will be extra-ordinary, and probably entail shedding some of the ordinary. That will have to be done, if at all, with enthusiastic faculty support.

12 Oct
The early morning maunderings produced this beginning of summary view of where we are:

Providing access to scientific literatures is the raison d'etre of the Science Library. It's not a full summary of what we do, but it's why we're there. We order books and shelve periodicals and instruct people in the niceties of the hunt for information, and we consult on specific retrieval problems. We maintain the suite of tools and portals that faculty and students can use to search on their own, and we mediate more complex searching and retrieval episodes.

We preside over an exceedingly lopsided system: vast oceans of print flow in through the Science Library's doors, but only a minuscule fraction is ever retrieved, and even less is actually read. Books and periodicals roost on the shelves, awaiting the unlikely someday that they will be wanted. Most of our efforts in collection development are inspired by optimistic visions of what we think our users might want, or should want. Occasionally it turns out that we were right, and a delighted user finds the item sought.

The systems of indexing that are the access points to our collections --the online catalog for books and the array of electronic databases that serve periodical literatures-- are vastly powerful but scarcely understood and little-used by faculty and students. The fact is that scientific literatures have little to do with the teaching and learning of undergraduates, in part because the level of most material is simply beyond the capacities of undergraduates, buut also because these literatures are seldom employed in undergraduate courses. Teaching in most courses is based in textbooks, and not in the current literatures of scientific disciplines. Only in the last two years or so of undergraduate study is the student expected to attempt research, and most students who take introductory science courses for general education requirements never get to upper level courses in the sciences.

Much of what we do as a Library is constrained by hallowed tradition. Current periodicals are arrayed for browsing because, well, because that's what a library is supposed and expected to do --and not because many library users actually browse. Expensive encyclopedias of this and that are purchased for the Reference collection (and are soon replaced with still more expensive new editions) because they seem to distill knowledge... but most are consulted only by Reference Librarians, and that infrequently. We maintain standing orders for serials because, well, we don't want to break the set... And we order monographs from publishers' catalog descriptions because they seem like they should be in the collection. But the fact is, if we stopped doing those things, hardly anybody would notice.

So we have a problem here, a problem that sprawls across pedagogy and intellectual activity and budgeting and self image and accumulated tradition. Our users rely more and more upon electronic access and we need to provide them with the appropriate tools. Those tools continue to evolve and become more expensive as they become more powerful. Sometimes electronic versions supplant traditional paper forms of the same tool, but there is often a long period of overlap, when the library has to supply both the new and the old form, often at considerable cost --and rarely is the electronic form less expensive than the paper it supercedes.

Electronic journals are one case in point: the great advantages of desktop delivery-on-demand and searchable online archives make electronic versions much more effective. At present about half of our periodical subscriptions have electronic editions (and more than 10% are in online-only form). Cross-linking of online versions through the online catalog directs some users to electronic editions, and Web pages listing periodical holdings provide pointers --but we do not participate in any of the retrieval schemes that link directly from search databases to online versions.

Converting to integrated search and retrieval would (really the word is will, since it's inevitable in the long run) be expensive, and fraught with teaching challenges, but the gains over the tatterdemalion present of literature access are very clear: our users will be able to find and get what they seek much more efficiently than they are now able to.

There are information fluency issues here as well as questions of ease and efficiency of access. The case is especially clear in the sciences, where the linkage of a primary research report to its intellectual progenitors is essential: the reader of an article must know on whose shoulders the writers have stood. And, in order to understand a line of research, the reader must also know who else has cited the article. Our students now have no direct way to make this last --and essential-- step in the process of understanding the moving frontier of science. The tool certainly exists (ISI's Web of Science) , and is in widespread use throughout higher education, but its expense (on the order of $15,000 per year, and about $25,000 for the first year) has put it beyond our grasp without extraordinary funding and/or a wholesale reorganization of our approach to access and retrieval.

Such a reorganization implies reallocation of the present funds for information access in the sciences, which I don't think we have ever analyzed in their myriad forms. To put our budgetary situation in perspective: in 1999-2000 "Science Books" amounted to some $49,000, and "Science Serials" to about $240,000 (these figures do not include the cost of electronic access to Academic Press IDEAL journals or the JSTOR archive). Search engine costs include $5500 for Cambridge Scientific Abstracts (mostly used by Biology and Environmental Studies), $2751 for GeoRef (Geology and Environmental Studies) , and $3634 for MathSciNet (Mathematics and Computer Science). FirstSearch science databases and PsychInfo reach us through VIVA and VICULA arrangements which are difficult to parcel out. The IDEAL access to Academic Press journals (most in the sciences) costs $17,307, and access to the JSTOR archive is $5125 (though about 1/3 of that is for non-science titles). One might also add in the salaries and other expenses of staffing the Science Library, to the tune of another $100,000 or so a year... and perhaps the costs of computing machinery and support and building maintenance as well... The point of this exercise is that W&L has a lot of money invested in information for the sciences, and so ought to be gimlet-eyed about getting the best value for the investment. In my considered opinion, the weakest link is in search and access --our users don't use the tools they have very well, and they lack one especially vital tool because of the limited access to citation indexing.

What are the prospects for a one-stop information integrator, or (more realistically) which of the various vendors comes closest, or (still more realistically) what combination of services can best fill the needs of the science faculty and students of small liberal arts colleges? We must begin to answer this question by laying out the present-day landscape of search and access, and recognizing that ISI's Web of Science is the only candidate for a cross-disciplinary integrator --but also recognizing that Web of Science cannot do everything and is not a standalone solution to every need for information access in the sciences. Other elements of the search-and-access landscape:

The various document delivery services which offer 24-hour (or sometimes faster) FAX delivery of articles seem to be much the same in terms of price charged: the basic charge for an article is in the $12-15 range, plus a 'copyright fee' that varies from publisher to publisher but often pushes the cost of an article to $30-35. UnCover, ISI's Document Solution, and the new Scientific World are all in this ballpark.

InterLibrary Loan has carried much of the weight of article access, the bulk of requests for which (university-wide) come from Psychology and Biology. It is difficult to calculate the cost of an ILL transaction, or to imagine an external service competent to replace it.

Web of Science

We need this. The question is how to get it.

ISI's Web of Science is potentially the cornerstone for a fundamental reorganization of access to and use of information for the sciences, and could be the place most searches in the sciences would start (since the service indexes most of the journals we subscribe to, and many more) . The greatest single value-added feature of Web of Science is the ability to place an article in context by tracing by whom it has been cited, and examining its bibliography of sources cited. The advantages of this approach to literature are abundantly clear to librarians, but without a concerted effort educate faculty and student users, the glories would be largely unrecognized and underappreciated. Consequently, a campaign to develop and publicize the utility of Web of Science would need to accompany its addition to our searching resources. If we gain access to Web of Science, I propose to set such a campaign in motion by way of a University Scholars course in The Moving Frontiers of Science, aimed at developing the skill sets and analytical frameworks and teaching methods to extend the use of Web of Science to broader audiences. I would then integrate Web of Science into Biology 182, which reaches some 40 students each year, and into other science courses for which I provide information access components. The course could well be developed into a Gen Ed offering in science literacy, for which Web of Science would be an important (but not by any means the only) tool.

So the problem is: how to come up with about $25,000 to set up Web of Science, and then approximately $10,000 per year in annual update fees? What do we need to forego to liberate funds on this order?

Chemical Abstracts

Chemical Abstracts presents the problem of expensive, little-used, arcane --but indispensable information. American Chemical Society acreditation requirements say that "Chemical Abstracts [hard copy or online] must be a part of the collection" (, and we must find a means to shed the print version. The terms of the Small College Program that brings us STNEasy at a reasonable price requires that "schools must maintain a subscription to the printed CA and indexes or CA on CD". The disadvantages of CD-ROM (no easy way to circulate or distribute in electronic form, multiple CDs to search, necessity for librarian-mediated searching ) are probably outweighed by the fact that CA on CD does save shelf space, and a purchase of previous years on CD-ROM would allow us to recover still more shelf space when necessary, while still preserving the completeness that makes our present holdings remarkable. But the price is astronomical.

The ideal solution for our needs would have these features:

CA's SciFinder Scholar appears to be what we should have, in terms of the range of information it makes available.

18 October
A blitz campaign to find big bucks for Web of Science by cancelling several high-priced but little-used journals seems to have met with about $30K of success.

An additional advantage of Web of Science as a main entrée is that it includes links to full text for some providers --notably Academic Press, Elsevier , and soon Cambridge, HighWire, Project Muse, and Wiley. Details can be seen at The links to Web of Science in Science's online version are an added inducement.

Now on to Chemical Abstracts:

It appears that CA on CD is slightly more expensive than the print form (the full price for 2001 is $22,600 for print or microform, and $25,200 for CD --a four-simultaneous license, since they assume that the material will be mounted on a server), BUT (as I've just learned from a call to CAS) theres's no small college discount for CD-ROM. With the paltry discount they do offer, it would be $21,690 next year for CD-ROM. Forget THAT.

SciFinder Scholar's price (one user , 24-hour access) is $13,125. I am given to understand that The Committee on Professional Traini ng (CPT) of CAS "approves Scholar as a substitute for Chemical Abstracts on a case-by-case basis". I also asked about the rumor that CA is being withdrawn from DIALOG, but the customer rep knew nothing about that.

So it would seem that our only choices are

  1. stay with paper, and assume that DIALOG will continue (alternative would be a "standard" account with STN for occasional daytime use, which has a one-time $50 fee and high search charges)
  2. find an extra $6000 per year for SciFinder Scholar, and fight the battle with the CPT so that we can drop the paper . We would then cease to use STN's after-5 service.
SciFinder Scholar sounds like it's an excellent product, in that it actually makes Chemical Abstracts usable. Finding that money is another question, unless (and it's most unlikely) the Chem folks would agree to give up Tetrahedron Letters... In any case, I can't find $6000 more in the serials budget.
The general bottom line is starting to look like this to me:
Working toward a set of courses and course modules that attack science information from an integrated perspective seems a worthwhile goal, in the contexts of departmental needs for pretty specific information skills (along the lines of what exists for Biology) and in the realms of 'fluency' from the General Education perspective. I don't know of any courses like that, but I'm imagining one that I'd be the right person to teach, as a hands-on empirical exploration.

19 October
The solution to funding SciFinder Scholar is to wait a year --and use the continuing savings from the cancellations. What it comes down to is: a subvention of $5000 plus my savings of $30,000 is enough to get immediate access to Web of Science. In 2001-2002 about $10,000 of that $30,000 will go toward Web of Science; the rest can be allocated toward desirable electronic resources, including SciFinder Scholar if it turns out to be as useful as anticipated.

So it seems that we're working toward a pretty comprehensive reorganization of literature access for the sciences. In some cases this will involve substituting electronic access for print (overall, a good step from the perspectives of enhanced accessibility --desktop and linking via databases-- and saved shelf space), acquisition of other needed resources and titles as they come up, and some changes in how these resources are taught (to faculty and students alike). (to be continued...)

Links to the biggies:

ISI on Web of Science and ISI Links to full text, and ISI Document Solution

CAS on SciFinder Scholar, and CAS home page and STN on the Web

Elsevier's ScienceDirect

...and to what Science (6 October) identifies as "New Kid in E-Journals Block": The Scientific World

Another Internet venture that aims to shake up traditional scientific publishing made its debut last week. The Scientific World promises a slew of online resources for researchers, including an articles database, chat rooms, lab-supply shopping, and its own peer-reviewed paper publishing arm.

There is no charge to search the site's database of over 11 million article titles drawn from the British Library, or to sign up for e-mail alerts to new papers. But full-text articles cost $12 and up. The 20,000 journals are said to span physics, chemistry, life sciences, and environmental science, although there's no list of titles. The site also intends to publish papers in all disciplines for free, making money by charging for full-text downloads. Publishing chief Graham Lees says he's sent e-mails to over 200 potential editors and the response has been "very positive."

The Scientific World has $8 million in financing and a slate of prominent scientists as advisers. But it's got competition--such as BioMed Central, another e-publishing site--and it's not yet clear whether many scientists will send their manuscripts to unknown, electronic-only journals.

Others who are exploring these perilous swamps:

University of Ottawa IDENTIFICATION OF CONTENT & DATABASES DRAFT (Last revised: February 12, 1999) --see also their array of Issues and Projects

SPARC Notes, and the SPARC home page

Academic Press

U Md Electronic Journals page, and a similar compendium from Yale

Morning Becomes Electric : Post-Modern Scholarly Information Access, Organization, and Navigation (Gerry McKiernan) --a comprehensive look at past and present, much more complete than anything I'd be able to do myself. The issues are laid out clearly and links to relevant sources abound. Bravo.

The ISI. Web of Science. - Links and Electronic Journals: How links work today in the Web of Science, and the challenges posed by electronic journals (Helen Atkins) from D-Lib Magazine

Accessing Electronic Journals by Sharon Cline McKay DATABASE, April 1999

Electronic Journals Project: Issues & Implications (Peter McDonald) --a White Paper from Syracuse University, April 2000


SCI: why teach? from Stanford Chem Info page (see also

Eugene Garfield his own very very self... (links to papers and so on, a real trove)

1 December
So... in the last month, success in finding the means to get Web of Science, and a whole raft of thoughts about Moving Frontiers of Science --whether as a course or just a campaign on my own isn't clear yet.

And the preliminaries completed for trial of SciFinder Scholar in January.