Maurice B. Line
Alexandria. The Journal of National a n d International Library and Information Issues, London
1
7
Research ' carried out under my supervision 15-20 years ago showed that bib-liographic control in the social sciences was poor. There was a multiplicity of bibliographies, large and small, with a huge overlap between them but with significant gaps. There was no attempt at coordination. To obtain good coverage in somé subjects, four or five bibliographies had to be used - and even then coverage was by no means complete. Articles often had to wait several months, sometimes many months, before appearing in indexes. Subject access was hap-hazard; access to the contents of books, a major source in most of the social sciences, was extremely poor. This last was and is true in science and technology as well, but it matters far less there; books account for a very small proportion of citations in those fields, whereas in the social sciences they account for as many as journals.
Somé of these deficiencies are attributable in part to characteristics of the social sciences. In the "softer" areas, such as sociology and political science, terminology is unstable over time and variable between different ideologies; Marxist countries had different concepts from capitalist countries, and often different terms for similar concepts. Few of the social sciences have fixed boundaries, and there is a low level of self-suffíciency; for example, over two thirds of references made in sociology journals were to other disciplines such as anthropology, management and political science. One explanation that has been put forward for this is that the social sciences are "immature", but it is more likely that cross-disciplinary
interdependence is an intrinsic and permanent feature; it would be surprising, and probably not healthy, if sociology was largely self-sufficient.
Another characteristic of the social sciences is that, by contrast with science and technology, potential users of bibliographic tools are neither numerous nor wealthy. The market for indexing services may therefore not be large or rich enough to support publications on the scale of, say, Biological Abstracts.
Moreover, again in contrast with scientists, there has been little demand from social scientists for better services; the penalties for ignorance are far less than in science, largely because absolute replication of research is rare - material objects do not change between one project and another or from one country to another, but human beings do, and the same study at different times or in different places is unlikely to yield the same results.
Some of the above problems can be alleviated by use of the Social Sciences Citation Index, since this cuts across subject boundaries, and also partially com-pensates for the deficiencies of subject indexing by taking a different approach to access. But it is best regarded as an additional tool, not as a substitute for more conventional ones.
What has changed in the last 20 years? Not a great deal, it must be said. In the 1970s I expressed the hope that some entrepreneur would produce a better service or services, covering wider areas more comprehensively than most existing bibliographic tools, taking their markets and driving them out of business. From the large databases subsets could be produced, either on demand or as published bibliographies if there was a market for them; these published tools would be of manageable size and therefore moderately priced. This has not happened. All that has really changed is that nearly all bibliographic services now construct databases by computer and use these as the basis for printed and online services. This has made them available more quickly and more widely, and also perhaps made searching easier, but the actual contents are no better.
The need for better access is more acute than it was 20 years ago. There is more literature, and at the same time libraries are able to acquire less and less of it. Browsing on the shelves is therefore less likely to yield useful results, and the ability to find out with some precision what there is on different subjects is more important. It will become still more important.
There is now hope of improvement with the recent transformation of Unesco's International Bibliography of the Social Sciences (IBSS), a set of four volumes relating respectively to Anthropology, Economics, Political science and Sociology.
In its old form this tool, published since 1952, was of very limited value, since it was published only once a year and was thus out of date by the time it appeared;
it was highly selective in its coverage; and its subject arrangement was not very helpful. Where it did perform a useful job was in its coverage of many journals
168 Thoughts in the library"
in less familiar languages and from less developed countries. It thus usefiilly supplemented other bibliographies, but the utility was marginal, and it is doubtful if much use was made of it by social scientists.
This has now changed. In 1990 the British Library of Political and Economic Science (BLPES - the library of the London School of Economics) took over editorial responsibility for the IBSS, and this provided the opportunity to merge it with the London Bibliography of the Social Sciences. This last bibliography was confined to books, largely the acquisitions of the BLPES. There is thus now a much more comprehensive database, in machine-readable form, which is used to produce not only annual volumes but a monthly bibliography called Inter-national Current Awareness Services (ICAS). This is a collection of contents pages of somé 13,000 serials and of collections of papers published in monograph form, with full indexing. IBSS, which covers over 1,500 serials as well as books, and ICAS, with its much wider coverage of serials, complement one another. In due course the database will be accessible online, and this will greatly increase its value - as well, one hopes, as its income, for upon this will ultimately depend the success of the venture.
One of the greatest hindrances to good access to the literature is the very inadequate subject indexing of books in bibliographies. It has been interesting to observe that as OPACs (Online Public Access Catalogues) have come into exist-ence the ability that most of them offer to search by keywords in titles has been heavily exploited by users. A few experiments have been conducted that involve putting the contents pages of books into machine-readable form and providing keyword access to them. This is undoubtedly crude, since many chapter headings are not descriptive of their contents, any more than all book titles are. But it is very cheap, and it offers something that was not available before.
It is however possible to do much better than this. Detailed subject analysis of books is expensive and time-consuming, but the effort and cost are, page for page, no greater than for the indexing that is already carried out for periodicals.
If this were done for all books of scholarly interest, a first rate database should result that would attract a big enough markét to pay for itself, partly by sale of printed products, partly by income from online use. Most academic and research libraries would buy the printed products, just as they buy now many expensive bibliographic tools, somé of them of dubious usefulness.
Such a database could form the foundation of a still better system. Contents pages could be digitally scanned and put on to microfilm or optical disc and coded for retrieval, so that they could be called up by searchers as required. A searcher in a bibliography would then be able to consult the contents page of any book he thought might be of interest to him. Ideally, there would be a direct link between the bibliography and the film or disc, but the system would work well enough without that.
169
It would be possible to go one better, and put the actual indexes of books on to film or disc. The user could then call up the indexes of likely books and see whether they had material of relevance to him. (This would incidentally reduce the number of books requested unnecessarily on interlibrary loan in the hope that they might contain something of interest.) This procedure would be better than entering the indexes of different books into one huge database for direct searching of entries. For one thing, the file would be almost unusable because of incon-sistencies of indexing and vast numbers of entries under certain headings; for another, the keyboarding that would be necessary for character encoding would be far more costly than scanning the pages digitally.
In this ideal future system the user's first stop would be the local catalogue, supported and supplemented in turn by (1) the comprehensive database, (2) the file of contents pages, and (3) the file of indexes. No publisher or database producer could be expected to take the risk of initiating a comprehensive project on the above lines. However, it is not unreasonable to hope that an experiment could be undertaken in a particular field, say books in criminology published in the last five years.3
Access to documents has been improving faster than bibliographic access. This has been due entirely to the application of electronic technology. The automation of union catalogues and the catalogues of major research libraries, and online access to them, has given wider and wider exposure to the contents of a larger and larger number of libraries. Requests can be sent to them instantaneously by computer, and responses can be sent back equally quickly in the case of unavaila-bility or delay.
What happens after that is still open to improvement. If items are sent by mail, supply can be slow and unreliable. It must also be said that many libraries do not handle requests from users in other libraries with anything approaching the ur-gency of requests from their own users. Unless they do, we shall not see really fast supply times under any system, and the acceptance by libraries of this re-sponsibility must be a prime aim. One way of helping to achieve this is to introduce a scheme whereby the full direct costs of supply are charged to requesting libraries, so that they are under an obligation to give a good service, and moreover cannot make the excuse of financial inability to provide it. In any case, there is no reason why supplying libraries should pay, any more than booksellers pay for the books they supply to libraries.
For books, or any items that cannot be copied (that is, anything over 20 or 30 pages), it is hard to see what more can be done unless mail systems improve.
However, for journal articles electronic technology is coming to our aid again, in the form of telefacsimile. The CCTV Group 3 machines that are in general use at present are too slow, and the copy they produce is of too poor quality, to make
170
them satisfactory for any but urgent items, but Group 4 machines, which will come into use when ISDN (Integrated Services Digital Network) is widely in-stalled, should be at least as cheap as making photocopies and sending them by mail, and a great deal faster. Moreover, the copies are of much higher quality.
Many have suggested that the ultimate ideal is the online journal, the text of which can be searched word by word and read on screen. Reading on screen without the ability to search is possible already for journals that have been digitally scanned, as is the case for quite a few in business and management. They can also be printed out at the receiving end; users will find on-screen scanning adequate for some articles, but most people will want to read some articles more intensively or will wish to have a copy to keep.
Whether direct searching of text will prove very useful is difficult to say. My own guess is that, given the problems of terminology in the social sciences, it will not, so long as bibliographies give adequate subject access to the contents of journal articles.
It seems then that we can look forward to significantly improved access to social science literature in the next few years: better bibliographic control through the new International Bibliography of the Social Sciences and the International Current Awareness Services, greater ability to locate books in libraries through online catalogues, and faster supply of journal articles through fax and online journals.
What needs attention is, in ascending order of difficulty, the speed of supply of items requested on interlibrary loan, subject access to the contents of books, and the mail service. The former is wholly within the control of libraries; and the second would have more chance of happening if libraries pressed for it. As for the third, many bodies other than libraries have an interest in seeing better mail services, and one can only hope.
References
1. Bath University Library. Towards the improvement of social science information systems:
overview of research carried out 1971-1975. (Design of Information Systems in the Social Sciences, Research Reports, Series A, no. 1). Bath, University Library, 1980.
2. Line, Maurice B.: Secondary services in the social sciences: the need for improvement and the role of libraries.
In: Behavioral and Social Sciences Librarian, 1(4), Summer 1980, pp. 263-273.
3. These ideas are explored more fully in: Line, Maurice B.: Bibliographic records for users:
from disordered superabundance to cost-effective satisfaction.
In: Aslib Proceedings, 42(2), February 1990, pp. 41-49.