Citation circus ignites explosion

Two thousand new research journals spin off the presses every year. Why? reports

Ever heard of Bradford's law? It is the rule that of the cornucopia of journals produced by academics to publish the latest research in any given subject, only a choice few will ever be of any use. In the 1950s, when the rule was formulated by C. F. Bradford, most people were able to keep track of developments in their academic field through the relevant journals. Since then journal numbers have exploded dramatically. Educated guesses put the global total of reputable journals at between 40,000 and 80,000, with over 2,000 new journals appearing every year. So, how can libraries and academics keep up?

In a desperate bid to stay on top of the latest findings many academics turn to formal computerised filters such as BIDS and the use of annals, digests and the like. One response to the publishing explosion has been the rise of bibliometric analysis, whereby journals are ranked in a league table according to the frequency with which their articles are cited in other publications. The American firm ISI produces the most comprehensive citation indices and journals such as The Lancet now quote their ISI citation rating in their publicity.

But there are worries that the vagaries of such computer-generated analysis and the importance attached to the lists it produces are distorting research. In the humanities, says English literature professor Catherine Belsey, "CD-Rom digests are having a bad effect. If something is not in the Modern Languages Association bibliography, for example, it drops off the map. The technology is beginning to determine the frame of reference."

Similarly in the sciences. ISI tables are top-heavy with American journals, and while the country does publish most of the world's journals, the fact that American authors tend to cite other Americans creates a vicious citation cycle. "The problem is that, in psychology at least, the high-status journals tend to be North American and rather conservative," according to Bristol psychology professor Alan Baddeley.

The use of citation records to establish which journals are the most prestigious worries many, particularly since some assessment panels in the last research assessment exercise announced that the standing of the journal in which an article was published would be taken into account when assessing that article's value.

"Bibliometric analysis does have a role to play," concedes the Royal Society's John Rawlinson. "It is perfectly reasonable for (the government's chief scientist) Sir Robert May to look at such analysis to see whether British authors are being cited more than French or German authors to get a measure of the health of British science." But, he goes on, "one of the problems is that people judge people's merit not on the quality of the work but by saying, did they publish it in Phys Rev Letters, well, it must be first-class work, or did they publish it in the Romanian Journal of Physics, in which case we can forget about it."

But any computerised system will have its limits. As Rawlinson recounts, "The Cambridge chemistry department looked up which of its members was most cited in the past five years and found that one of its stars, Alan Hirsch, one of its best-known people, wasn't on the list because he publishes in biochemical journals and ISI doesn't recognise those."

The citation culture encourages "salami slicing" of research, too - the division of a single piece of research into several articles. "I've had papers to referee and when I've looked into the antecedent works of the same author, I've found that one is in a journal in the States, and another is in a German journal and now he's trying to publish (what is fundamentally the same piece of work) in an English journal. I tend to be unsympathetic to that, " says Rawlinson.

Moreover, the culture of citing references is very different in different parts of science. Biologists cite more papers than mathematicians, for instance. Mathematicians, when they publish a paper, only have three or four references in it, the paper's one or two immediate predecessors. A biological paper will typically have 30 to 40 references. "This certainly distorts citation records," says Rawlinson.

In response to such problems some academics, such as Peter Brandon of Salford University, are bypassing traditional journals altogether. "If it's news that you want about what's going on electronic information is probably the way in which things are moving," he says. "The advantage of a journal is that it consolidates knowledge in a field and takes a snapshot of it at a particular point in time. The trouble is that journals sometimes take two years to turn around. In a fast-moving field like mine, IT, it's just too long."

Two years is just long enough for a spot of plagiarism to take place though. Liverpool John Moores engineering academic Brian Rowe avoids one US journal for this reason. Like most journals it uses a panel of referees to assess the worth of an article before taking a decision about whether to publish it. "This particular journal is appalling. It will allow the referees to hang on to a paper, dither about it, make excuses and then a bit later a paper will come out from one of the referees that's fairly similar. If referees see a good idea they can, if not directly copy it, take advantage of it, and publish somewhere quicker."

With rumours flying that false footnotes are being appended to online journals and meta-tagged for optimum citation effect, perhaps journals have not changed so much after all.

THE WORLD'S MOST CITED SCIENCE JOURNALS

(source for all ISI's JCRs for Science and Social Science, 1996)

1 Journal of Biochemistry

2 Proceedings of the National Academy of Sciences of the USA

3 Nature

4 Science

5 Journal of the American Chemical Society

LEAST CITED: Bangladesh Journal of Botany Artificial Intelligence Review (joint no. 4,535)

MOST CITED SOCIAL SCIENCE JOURNALS

1 Archives of General Psychiatry

2 American Journal of Psychiatry

3 Journal of Personality & Social Psychology

4 British Journal of Psychiatry

5 Psychological Bulletin

Least cited:

Russian Politics & Law

Nouvelles Questions Feministes

Australian Feminist Studies (joint no. 1,510)

FACT FILE

Academic journals published in the UK

UK market:

Domestic Pounds 107m.

Export Pounds 5m.

Total Pounds 382m

(Source UK LIb Statistics, SCONUL, LISU) Number of U Kpublished journals:

Academic/professional 1,966

Scientific/technical/medical 2,245

(Source PA statistic collation scheme)

ACADEMIC JOURNALS

JOURNAL ONE:

Past & Present. Est. 1952

Published by: Oxford University Press

How much? Indiv. Pounds 30, Inst. Pounds 71 a year

How often? quarterly

Circulation: 3,500

Subject: Global historical, social and cultural change

Editors: Joanna Innes and Chris Wickham

Referee process: Unusual in beingrefereed mainly by members of 25-strong editorial board

Lag: one to two years

Best scoop: 1971 article by Edward P. Thompson, "The Moral Economy of the English Crowd in the 18th century", coining the subsequently widely adopted term "moral economy".

Reader comment: "It is very much a flagship journal, in some senses more highly regarded than the English Historical Association journal. Its articles tend to be more exciting. It has a left-wing background and lively articles. It deliberately decided to have no reviews, and that gives it a particular character." Michael Prestwich, Durham.

Style: As olde worlde as its subject matter. 200-page A5 book with quaint grey & scarlet handbill cover. Typical article: "Gossip and Litigation in a Suffolk village".

JOURNAL TWO:

Modern Language Review Published by: W. S. Maney & Sons for the Modern Humanities Research Association

General Editor: Professor M.C. Cook, Exeter University

How much? Pounds 78 trade, Pounds 25 MHRA members

How often? quarterly

How many? 1,300

Aimed at: English, German, Slavonic and French literature scholars Refereed by?: four section editors send out articles to referees.

Lag: at least one year

"One of its virtues is that it is international. It also has a good, wide-ranging book review section. There's a nice account in the January 1998 issue of 'Dread in the Victorian Body'. It's about the fear of penetration - by vampires, for instance. Catherine Belsey Style: Sober Penguin classic.

JOURNAL THREE

Nature: Est. 1869 Editor: Philip Campbell Published by: Macmillan Magazines How much? Indiv. Pounds 175 Instit. Pounds 395. Student Pounds 99 How often? Weekly How many? over 58,000 Subject: Life, physical, chemical, earthsciences Refereed by: Editorial board of 19-20 sends papers out to 2-4 anonymous reviewers Lag: varies according to importance of paper Style: Loves its psychedelic cover images and teasing cover lines such as "what keeps sandcastles standing?" and "Tyrannosaurus suffered from gout."

JOURNAL FOUR

Electronic Journal of Information Technology in Construction (ITcon)

Editor Bo-Christer Bjork

Published by: Royal Institute of Technology, Stockholm

How much? Internet access. Printed copies $60

How often: Once a year. Articles published when ready.

First two of second volume already on net, further four undergoing review. Printed version every two years

Subject: the use of IT in construction

"Electronic journals such as this are the way forward for fast-moving fields, especially in science." Peter Brandon, Salford University Style: hard copy looks like some 1970s A4 computer textbook.