Forgetting older science

By Grant Jacobs 24/10/2009

Are the electronic literature databases making us forget about the older literature?

A short blog by Isis, a self-proclaimed domestic and laboratory goddess (and who are we to disagree?), reminded me that I have been meaning to write a blog about this issue for some time and that I should get on with it.

Isis mopes that articles not available electronically are just not worth the effort tracking down. At least not when it’s raining. Or something like that.

A few years ago, I noted an annoying trend in my research field, bioinformatics. The odd research paper–one or two too many for my tastes–re-invented or re-discovered old results and methods. These authors, presumably innocently, seemed unfamiliar with the earlier work. These papers seemed mostly from younger research groups, bioinformatics scientists who joined the field after the genome era. I put this down to older group leaders knowing the earlier literature. But then, again, maybe it’s the confirmation bias of a slightly older scientist…

After some ruminating, I wondered if the cause was PubMed and laziness.

PubMed is an on-line database of research paper abstracts and links to their original copies hosted by the NCBI. PubMed is a–if not the–main source of access to the scientific literature for many biologists.

There were two main things that I put these “re-inventions” down to. One I want to touch on in a later article as it bothers me: some researchers seem to incorrectly think that computational biology (aka bioinformatics) started with the genome projects, or at least that literature prior to that time can be ignored. This is peculiar to this field, but the general concept may apply to other fields that chugged along relatively unnoticed for years then suddenly become very popular and in demand.

The other was that the lack of presence of the older literature in the on-line databases: those who only used on-line databases and electronic copies of research articles would miss the older works.

It was quite a disturbing thought. Science relies on the continuity of the record. If that’s broken, a lot is lost.

PubMed, in trying to make the literature more accessible, may have–at least temporarily–encouraged “lazy” researchers to break the chain of knowledge.

This has since been alleviated by efforts to bring the back catalogues to electronic form. This will no doubt continue and in time we’ll all be in a stronger position. (Assuming some of the those grotty scans of early manuscripts are tidied up!)

Nevertheless, I can’t help but worry that the combination of easy access to the newer literature, the work involved locating the older literature and the slow, but sure, mothballing of the paper (card) and microfiche index systems encouraged researchers to consider old(er) papers irrelevant and this has persisted to establish a new norm.

Certainly it is tempting to ignore them. Research fields move a such a clip that it’s hard to find time to keep up, let alone dig up leads fifty or more years old. Hardly anyone seems to rummage the dusty corners of the library to locate their treasures anymore. OK, it’s time-consuming and carrying stacks of bound tomes works the arm muscles, but it suggests it’s thought that if it can’t be gotten on-line, it’s not worth the while.

Am I alone on this one? Or am I just getting older and fussier…?!


This of course justifies those review papers that give long views of a field, mostly written by those starting to sport white hairs 🙂

0 Responses to “Forgetting older science”

  • Where to start. One problem with older literature is indeed that we have become lazy about going to the library. But there are other problems as well. PubMed does not go far enough in time (in my opinion), and if you are to go to the library I would doubt you would find the older issues that are not listed in PubMed anyway. I have as a result moved to google scholar, where I find citations dating back to the 1800’s. (getting them is a different story). Regarding the ‘rediscovery by lack of citation’ I don’t think that laziness is the only explanation. I have seen ‘rediscoveries’ two years after the original discovery, but if the original is not cited, then that is that. Another way in which ‘original’ research gets lost is through ‘renaming’ (and I have seen this a lot too). When an author changes the name of a structure or of a phenomenon, then there is no google scholar or pubmed that will take you to the original literature anyway (’cause you are not using the original term on the search). And those who practice this will forever be known as the ‘first’ to report on it. Even worse, I keep getting annoyed at reviewers and editors that request I remove ‘old’ references from my work. You are not alone, but we may just be a handful.

  • Hi Fabiana,

    I agree, there is much more to it.

    I can remember that the hunt for the papers was part of the fun, for me anyway. There may be an element of people not caring in the sense that it probably will have no effect on them. (Unless, perhaps, they strike a referee who insists that they include an old citation the referee “just happens” to know well!)

    Your example of “rediscovery” only two years later, makes me think that some people aren’t searching by keywords enough, or just not reading enough.

    Renaming is a common issue in bioinformatics, e.g. with gene and protein names. It’s less of an issue than it used to be. Once when proteins were mainly described by activity it was relatively common to discover later that two sets of suspiciously similar activities were in fact the same protein (or pathway). Now people tend to work back to the gene or protein sequence, it’s less of an issue. Also, bioinformatics people have been maintaining aliases of gene and protein names for a long time now.

    I’m surprised at reviewers & editors asking you to remove old citations, aren’t they supposed to insist on the earliest reports being presented?! There was something about this relatively recently by an editor in one of the top-end journals (I forget the details now), saying something to the effect that the habit of only citing recent references wasn’t right.

    Misattributing credit can have some bad side-effects on people’s careers to, but that’s another ball park.