Tuesday, July 24, 2007

Recently Cited

The topic o' the day is Citations, as in one's personal citation record. The topic was inspired by a conversation with a colleague earlier today about citations, a topic of great interest/obsession in the corridors of academe, at least among scientists and engineers -- perhaps humanities people are not so afflicted by this(?).

When we talked, my colleague was en route to the library -- the actual physical library on campus -- to read an ancient (1993) article. He mentioned a hypothesis that was discussed at a recent conference he attended: namely, that the likelihood that pre-mid-1990's papers will be cited is lowered because many journals do not have their entire archive available online. Perhaps online publishing has contributed to the precipitous decline of citations of papers published before about 1996?

There are of course 'classic' papers from days of yore that are cited again and again, in some cases because they are great papers and in some cases because it's easiest to cite the same papers (snowball effect). Some 'old' papers are no longer relevant, and that just shows the exciting pace of science, but it would be tragic if the wisdom of pre-1990's papers is lost just because the articles aren't available online.

I checked my own citation record just now, making one of those zippy graphs that you can do in Web of Science, and there is an abrupt upward spike that corresponds to the year after which electronic articles are widely available. No, I'm not going to conclude anything from this one anecdotal record that doesn't contain nearly enough data to evaluate. There are other explanations for the spike exactly at that time. Nevertheless, it's intriguing how neatly it matches the hypothesis my colleague described.

Some of my colleagues could tell you their citation statistics very exactly and give weekly updates. I seldom check my citations, though when I do, it is typically for one of two reasons: (1) to see if there are new papers that would be of interest to me on certain topics and that I want to be sure I haven't missed in my perusal of journals; and (2) to check on a certain paper of which I am particularly fond.

I know the latter reason doesn't sound too sane, but there is a paper of which I am particularly fond because it was so much fun to write. It's a fairly short paper, so it doesn't have all that emotional/psychic baggage that some longer papers have. Fortunately, it is doing fine, so I will have to put off for now any brooding about how misunderstood my work is.


Quantum Moxie said...

I completely understand being fond of a particular paper. I do the same thing with one of mine that I am particularly fond of. It's not that I have any illusions about its importance, simply that it was a gratifying paper to write.

Anonymous said...

On a half-related note, I imagine also if you were the young man who had written the (infrequently cited but frequently mentioned and referenced) "Electron Banding Structure in Germanium, My Ass" you would have similar emotional attachment to a short paper. (I highly recommend searching for the paper online, it's worth the laugh and is widely published now--not too surprisingly, the author is now very successful.)

Anonymous said...

I'm afraid that there is adeeper issue behind the fact that old papers are rarely cited: People just think that the old stuff must be out of date and often don't bother to read papers older than, say, 10 years. It's a common saying that you can publish the same stuff every 15 years without anyone noticing.

Quantum Moxie said...

Oh how true that is. My PhD thesis was partly historical and I sometimes do some research in the history of physics and it is amazing the number of times I have witnessed the exact same thing "re-discovered" a few decades later. On the other hand, keeping up-to-date is an increasingly difficult task. In a book published over a decade ago, David Wick mentions there bing over 70,000 scientific journal titles. Can you imagine the number now?

Mr. B. said...


At the risk of being considered cynical, I propose a bifurcated strategy.

When you write papers, cite the old literature, as well as the new. It is good for younger readers, or even students, to realize that scientific things actually happened before 2000.

When you write grant proposals, don't cite the old literature as this will be held against you. Find the latest (relevant) references that you can.

Mr. Bonzo

Anonymous said...

Perhaps you could increase your citation index by making sure that *your* early papers turn up in Google Scholar, etc.