Google's new scholarly search service (http:///scholar.google.com) may deliver to the scientific and scholarly worlds some of the facilities that the OA community have been labouring over - most specifically a search service which is limited to scholarly and scientific resources but which unites material from all web sites within that domain across the complete spectrum of scholarly enquiry. It also provides a basic (but highly welcome) level of citation analysis so that individual papers are listed with an entry which indicates how many citations they have received. As far as gross coverage goes, a search for the definite article indicates that 315 million articles (or at least, citable scholarly artefacts) have been indexed - a figure that appears to correspond to 13 years of the world's scientific and scholarly peer-reviewed research journal output. (Doubtless further analyses will soon show a more accurate interpretation of this figure and the relative coverage of each discipline.)
It may be tempting to suggest that in the face of such an overwhelming resource, there is little point in putting continued effort into Institutional Archiving, with its administrative overheads and individual metadata entry requirements for each paper. After all, with Google on the job, all we apparently need to do is to encourage individual researchers to leave a copy of their papers on the web! This is, after all, in line with Harnad's original "Subversive Proposal" from ten years ago. http://www.arl.org/sc/subversive/ However, we need to bear in mind some crucial factors (a) Google is indexing data from commercial publishers as well as the open web - and so it may still be impossible for any individual to read the vast majority of articles that are returned as the result of a query (again, more thorough analysis will give a better indication). (b) Most researchers aren't even putting their articles on the web let alone in a managed institutional repository (and for the same reasons - they don't see the need, they are worried about copyright etc etc). So Google is not offering increased Open Access, just improved resource discovery of current ad-hoc OA. To advance we still need to offer carrot and stick, policies and mandates, and that is one of the places where a managed Institutional Archive comes into its own. An institutional archive offers many advantages to the institution in terms of research management, reporting, statistics gathering, publicity and research assessment. These institutional benefits mean that the institutional support, policies and mandates are more likely to be forthcoming, which in turn WILL DIRECTLY increase the amount of open access material available. (Of course it's not all stick! The carrot comes in the form of improved impace, automatically generated CVs, up-to-date bibliographies on personal web pages and the enormous relief of automatic data collection for institutional bean-counters - an enter-once system which caters for the many uses to which researchers need to apply their output.) Google, of course, indexes entries in Institutional Archives! Whether EPrints or DSpace, these OA repositories are providing grist to Google's mill. So the Google scholarly search engine is a welcome addition to the arsenal of services that researchers use to mine the literature - but it is still OA Institutional Repositories that provide the best chance of getting readable copies of those papers into Google! --- Les Carr EPrints Development Manager & University of Southampton Institutional Repository Manager
