Victorian splendour brought up-to-date

VICTORIAN DATABASE Compiled and edited by Brahma Chaudhuri, James Mulvihill and Fred Radford LITIRDatabase, c/o Department of English, University of Alberta, Canada. Email: litir@gpu.srv. ualberta.ca

Institutional licence: US$995/Can$1,350. Annual update. Renewal: US$245/Can$335

MS-DOS/Windows CD

Victorian Database, unlike some electronic publications, fulfils its promises to the user. It is a unique cross-disciplinary bibliography, compiled from 60,000 books and 500 journals over a 25-year span from 1970-95. To be annually updated with 2,500 titles, it will prove invaluable to researchers, whether they are mature scholars or graduate students. Under such broad categories as fine arts, religion, politics and technology, it assembles and cross references entries from different disciplinary fields, from music, law and medicine, women's studies and colonial history, poetry, fiction and economics.

Having said this, a few acerbic questions need to be raised. What can Victorian Database do, speed of access apart, that printed resources cannot? The odyssey of discovery which all bibliographies provide derives partly from discovering an exact solution to an enquiry and partly from the ludic pleasure of chance, or at least unexpected juxtaposition, where classification genuinely surprises, and extends one's sense of the subject under review. Theoretically, electronic media should increase such potential. They can produce multiple collocations, a sorting of taxonomies which, because of digital technology's capacity to reconfigure groupings and classifications instantaneously under different heads, should vastly extend the possibility of interdisciplinary relationships. Interdisciplinarity has no content - it is not a predictable cocktail of disciplinary components - but is a method and a practice. Thus everything depends on the discriminating and flexible classification of the subjects and topics generated by particular scholarly works. This database terms these subject categories the "descriptor", and these can be configured in many ways: one work, for example, generating the descriptors "clothes", "politics", "religion", "technology". This is where the human element in selecting keywords and matching them to descriptors is paramount. To see how successful the Victorian Database could be in configuring information I tested it with a selection of different doctoral topics I am supervising: feminism in the Middle East, confessional lyric, "new woman" poets, George Meredith, and Margaret Oliphant and Providential thought.

The first two topics produced encouraging results. "Middle Eastern women" elicited only two entries, but one, Maria Frawley's article of 1991, "Desert Places/Gendered Spaces: Victorian Women in the Middle East", was spot on. The keyword "confession" yielded among other items a dissertation of 1982 on Church and Confession (Walter H. Conser Jr), and Nigel Yates's account of the controversy around ritualist confession in the Journal of Ecclesiastical History for 1988. The crossing of theology with literary text these references make possible is exactly what my student wants. The other topics were more disappointing. The keyword "new woman" elicited a range of material on fiction and drama and nothing on poetry. Entering the names of poets produced obvious information. "George Meredith" produced six items, all re-edited texts. Over 25 years, articles and chapters on this extraordinary poet and novelist must be lurking somewhere. "Margaret Oliphant" was disappointing for different reasons. An adequate listing of 21 items is produced but one item is misreferenced. Ester Schor's 1993 article on the supernatural fiction was not published in Women: A Cultural Review. And when I tried for "predestination" (no matches) and "providence" (11 matches), the results were feeble.

It became clear that two opposite problems occur. The more abstract the topic pursued, as in the case of "providence", the more difficult it is to discover a bibliography that matches it. Conversely, the more specific an enquiry the harder it is to find information. What such a bibliography has chosen to know is not necessarily what you want to know. The taxonomical categories are its, not yours. This is partly because the keyword index apparently has been compiled mechanistically titles and rarely, it seems, from a close assessment of the intellectual content of a work. The descriptors are correspondingly weak. Two examples suffice. Mrs Hemans's The Image in Lava suggests that there was an early 19th-century discourse around the excavations of Pompeii. "Pompeii", "excavation", "archaeology", elicited nothing. Some other way of accessing work on the Victorians and the ancient world will have to be found. Again, economic domination of India, in particular the point when colonial rule imposed the sale of land on the country, was an enquiry of the moment. India (1,056 items) had to be tried under at least five cumbersome categories before the database yielded relevant material.

The rigidity of this resource is its great limitation. The keywords "Crystal Palace" and "Great Exhibition" yield fascinating information, such as Patricia Mainardi's architectural historical article (1986) of the unbuilt picture gallery of the Great Exhibition. But neither listing includes Thomas Richards's The Commodity Culture of Victorian England: Advertising and Spectacle 1851-1914 (1990), an important interdisciplinary study of the Crystal Palace and culture. What will happen to Andrew Miller's Novels Behind Glass (1996)? You have to know about the books before you can find out about them - looking glass bibliography.

The keyword principle occasionally produces wildly random juxtapositions which are almost a celebration of the irrational fallibility of human taxonomies. "Rome" produces references to Tractarianism, foreign policy, and a bizarre article on Gissing's witnessing of a man hurling himself from a house in Rome, tracked by the descriptors "Gissing", "Suicide" and "Italy". Taxonomies are often conducive of levity and the Victorian Database is no exception. Perhaps this is no bad thing, despite its massive ambition. It is as well to remember the intellectual work digital technology cannot perform. On the other hand, this database does perform, perhaps inadvertently, a somewhat unexpected but welcome task. Its range is so wide that it is hardly possible now to bring its various materials under the homogenising term, "Victorian". This is all to the good. Another cultural descriptor is required. It is time the name of a stodgy queen ceased to stand guard over the infinite variety of the 19th century, a variety wonderfully disclosed through this work.

Isobel Armstrong is professor of English, Birkbeck College.