A colleague from another university told me over dinner last night about an exchange he’d had with one of his PhD students. The supervisor had asked the student which journal a particular paper he’d quoted had appeared in; the student didn’t understand the question. It seems that the student, who is some way into his thesis, has never used a journal or a journal website: instead he’s worked entirely with preprints and done his bibliographic cross-referencing exclusively through the arXiv. It was a genial sort of evening; somebody else was paying for the wine; and we didn’t really pursue our sense of discomfort about this beyond a vague observation that this was how the new generation of academics would work, and that we’d probably better learn to live with it.
I wonder whether something of this nature lay behind an incident last summer, when a comment in a draft manuscript sent me to the library to run to earth the source of a claim I didn’t particularly believe. The reference was to a paper by the eminent Danish hydraulic engineer Frank Engelund, and I naïvely thought it shouldn’t take me long to check. A little while later, I was little the wiser about the truth or otherwise of the claim, but I felt I’d learned something both about the engineering gray literature and about the reliability of the bloke who drafted the manuscript.
The work in question started life as part of a Progress Report from the Hydraulics Laboratory of the Technical University of Denmark, published in August 1965. It was recorded as part of their regular digest of recent publications by the French engineering journal La Houille Blanche (p. 607, vol. 8, issue 6, October 1965), and two issues later an extract was published in the Notules Hydrauliques section (La Houille Blanche 8(8): 802; December 1965). It may be worth noting that pages in the Notules do not carry the name and volume/issue number as a running head like most pages of LHB, and also that the Notules are not included in LHB’s online archive of articles.
According to Web of Knowledge, the work has been cited in the following ways (presented below, for sanity’s sake, in a standardised format).
- F. Engelund (1965) Criterion Occurrence, p. 802. [1 citation]
- F. Engelund (1965) Houille Blance, vol. 19, p. 607. [1 citation]
- F. Engelund (1965) Houille Blanche, vol. 20, p. 801. [1 citation]
- F. Engelund (1965) Houille Blanche, vol. 8, p. 7. [5 citations]
- F. Engelund (1965) Houille Blanche, vol. 8, p. 802. [1 citation]
- F. Engelund (1965) Houille Blanche, vol. 6. [1 citation]
- F. Engelund (1965) Houille Blanche, vol. 6, p. 607. [1 citation]
- F. Engelund (1965) Houille Blanche. [1 citation]
- F. Engelund (1965) Houille Blanche, p. 7. [1 citation]
- F. Engelund (1965) Houille Blanhe, p. 802. [1 citation]
(There are also nine citations of the original technical report, in a total of eight different formats.)
I make that a total of one entirely correct citation (5); one accurate but slightly incomplete citation (10), at least if you allow for a transcription error in WoK; two citations (1, 8) that are incomplete to the extent of making the work almost impossible to find; and ten that are plain wrong in one way or another. That little block of citations to vol. 8, p. 7 all appear to be from the same school, and I suspect that a photocopy hand-annotated with incorrect information has been doing the rounds.
Does any of this matter, except to a pedantic dominie with a cack-handed co-author? If nothing else, this example illustrates how readily a piece of work can get detached from its bibliographic context, and how few researchers will make the effort to relocate it once it’s been detached. All harmless, of course, until somebody who hasn’t been passed the same copy as everybody else needs to try to track the work down and verify what it actually says.
There’s a popular idea, to which my colleague’s student is far from the only subscriber, that the need for journals is over, and that papers can reasonably circulate through samizdat and preprint servers without the need for cumbersome systems of identification. Maybe my experience of trying to track down the original content of a Danish technical report republished in a French journal and cited inaccurately by an English researcher quoting an Italian paper suggests that we have little to fear from the age of samizdat: it’s hard to imagine the arXiv making a worse hash of this than traditional bibliographical methods have. What I’ll miss in the new age is the context: the paper as part of the conversation recorded by a particular journal, rather than as a node in yet another sprawling and impenetrable social network.
Now that last night’s wine and geniality have both faded a little, I think that maybe I’m glad that my involvement with research is dwindling into nothing. Research, it seems, may soon become an activity carried out in the same spirit, and with the same technology, as Facebook: a conversation conducted in erudite graffiti rather than in books. I’ll be happy, I think, to leave it to the people who can cope with it. What was it Auden’s Prospero said about Ferdinand and Miranda?
Probably I over-estimate their difficulties;
Just the same, I am very glad I shall never
Be twenty and have to go through that business again,
The hours of fuss and fury, the conceit, the expense.
(The Sea and the Mirror, 1944)