Just for fun* I decided to look up how often various classic ecology papers have been cited over the years. I restricted attention to papers that are old enough to be considered classics, but are young enough to still be cited (Darwin’s Origin is a classic, but it’s so old that it’s now considered part of scientists’ shared background knowledge and so isn’t ordinarily cited).
Just off the top of my head, these are the papers I decided to look at:
- May 1972 (stability-complexity)
- May 1975 (chaos)
- Connell 1978 (intermediate disturbance hypothesis)
- Paine 1969 (keystone predation)
- MacArthur and Pianka 1966 (optimal foraging)
- MacArthur and Levins 1967 (limiting similarity)
- Likens et al. 1970 (deforestation and nutrient cycling)
- Ehrlich and Raven 1964 (coevolution)
- Hairston, Smith, and Slobodkin 1960 (trophic cascades, though they didn’t use that term)
- Price 1970 (the Price equation; couldn’t resist)
I also tried to look at Levins (1969), the paper introducing the idea of “metapopulations”, but it’s from the Bulletin of the Entomological Society of America and doesn’t show up in a Web of Science search. And Cole (1954), a classic paper on life history evolution, seems to be too old to show up in a Web of Science search. Obviously, there are many other papers I could’ve looked up, but this seemed like a nice mix.
It was actually really interesting to see just how much variation there is among these papers in terms of how much they’ve been cited, and in their patterns of citation over time. They’re all unquestionably classic papers, but apparently not all classics are created equal, at least when it comes to accumulating citations! And no, the differences among them in citation patterns are not primarily a matter of some papers having had more time to accumulate citations. I’ll have more to say about their citation patterns in a future post (although right now let me emphasize that I do not think that citation counts are a good measure of how “classic” a paper is! I’m just having a bit of fun here.)
One thing in particular that surprised me is that one of these papers has been cited far more often than any of the others, and one has been cited far less. So here’s a little quiz. Without cheating and looking it up, can you guess which paper is which?
Answers to the questions, and more discussion of the data, in a future post.
*For “for fun” read “to procrastinate”
Dynamic Ecology 
A wrinkle here is that most people cite Paine (1966) for keystone predation because it is the key experimental study, rather than Paine (1969) which coins “keystone” predation. Or, because they think that Paine (1966) is the keystone piece because they have not read it. Similarly, HSS is often cited as originating “trophic cascades” but that term was not coined until Paine (1980). The point is only that citation trails can be misleading over time – people cite a paper because everyone else has, and there is a positive feedback cycle to reinforce the emergence of a few “classics” while leaving many others behind. I wonder how many ecologists have gone back and read all of these “classics” and their kin before they cited them?
Good point re: papers originating a concept vs. papers coining the now-accepted term for it.
Re: “citation bandwagons”, I share your curiosity. My future post on the data will shed some (dim) light here.
Along the same lines, many people cite May’s 1973 book rather than his 1972 paper for his stability-complexity work. So I suppose one approach to the quiz is to guess that papers that “compete” for citations with other papers or books I didn’t list will be cited less. I don’t say that that’s the right approach, though. 😉
Just going by memory here, but I have read HSS 1960, MacArthur and Levins 1967, Paine 1969, Price 1970, May 1972, May 1975, and Connell 1978. I may have once skimmed MacArthur and Pianka 1966, not sure. And I’ve had occasion to cite most but not all of the ones I’ve read. I’ve never read nor cited the others.
Levins 1969 does show up in a cited reference search. It’s got a nice cloud of citation mutants around it (a reference quasispecies?):
Journal Yr Volume Page Cites
B ENTOMOL SOC 1969 15 27 1
B ENTOMOL SOC AM 1969 15 236 6
B ENTOMOL SOC AM 1969 15 273 1
B ENTOMOL SOC AM 1969 15 445 1
B ENTOMOLOG SOC AM 1969 15 131 1
B ENTOMOL SOC AM 1969 15 227 3
Bulletin of the Entomological Society of America 1969 15 3 237 1316
BTW, nice cliffhanger — now I’ll have to tune in next time to see if I guessed well!
Hmm, not sure why it didn’t show up for me…I did find it weird that it wasn’t showing up. Same for Cole 1954, actually.
I agree that classic papers often don’t say what is attributed to them. They often generate new ideas, if not new subfields, but the specific concepts attributed to them when cited are not to be found inside them. Another key paper that is inappropriately cited is Fittkau and Klinge 1973. (I do think I’m as guilty of this as anybody else.)
This is why, as a gift to students leaving my lab for grad school, I give them the “Foundations in Ecology” book as a parting gift, or “Foundations of Tropical Forest Biology.”
Yes, the Foundations books are a great resource for students.
MacArthur and Levins 1967 is infamous for being misinterpreted. Joan Roughgarden has a classic paper pointing out that many of the “predictions” that ecologists of the 70s and early ’80s claimed to draw from MacArthur and Levins (and other papers of MacArthur’s) were actually assumptions that MacArthur made purely for the sake of mathematical tractability.
May 1972 is another that’s widely misinterpreted, but for different reasons–basically, people substituting other definitions of “stability” and “complexity” for the very specific definitions May 1972 used. Which is unfortunate, because May’s results are specific to his definitions, and don’t carry over to other definitions of those key terms. I have an old post on this.
There’s an old line about Karl Marx, something about how his curse is that he’s been interpreted. Probably something like that could be said of MacArthur and Levins 1967, May 1972, and perhaps other papers on this list. 😉
Terry,
Great call on Fittkau and Klinge – an all time classic demonstrating the massive abundance of tropical social insects, that is, until you actually read it. Other than the plants, all their other biomass estimates were based on the OPINIONS of experts that were consulted by the authors. .
Wow! Are you serious? In terms of miscitation or misinterpretation, that’s FAR worse, and FAR less understandable/semi-excusable, than the examples I gave.
Maybe I should do a whole post asking people for examples of “most egregiously mis-cited classics”…
Well, I was being a little dramatic, but it is not too far off to say “opinion.” Here is the text describing how they arrived at faunal estimates (from FittKau and Klinge 1973, Biotropica,On Biomass and Trophic Structure of the Central Amazonian Rain Forest Ecosystem):
During this time many
collections were made of large and conspicuous arthropods,
amphibians, and reptiles. The vertebrates
collected during this time were submitted to Dr. P.
Muller, SaarbriickenG, ermany,a nd the invertebrates
went to H. Schubart and E. J. Fittkau, Plon, Germany
for identification. When evaluating this animal
sampling and our observations on animals in
central Amazonia made over the last 10 years, we
also made use of any information gathered from
Indians, settlers, hunters, and professional biologists.
In order to obtain weight data of fauna we listed
first those animal taxa (orders, families, etc.) which
were observed to have some bearing upon biomass
because of their individual weight and/or frequency.
Then we estimated density taking into account all
observations gathered by ourselves and the abovenamed
sources. Finally, density was converted into
weight of the respective group by multiplying weight
by density, using average weight of individuals contained
in our collections or otherwise determined
weight. None of the values thus obtained was adjusted
with respect to data found in literature for
similar groups in tropical regions elsewhere.
This sounds like two good cases for #overlyhonestmethods
(1) We asked some guys we met down the pub how many bugs there were around.
(2) We didn’t read the original citation, but someone else cited it and that was good enough for us.
Yeah, it is mighty bad how that one gets cited. I think we might blame that string on citations on Wilson, but I’m not sure who started it.
I think that about half of the time I’m reviewing or editing a paper which cites me, the citation is wrong, and either inflates my finding or actually states the opposite of what I found. (I don’t see this as often in papers that have gone to press. This does suggest that peer review works to some extent.)
Jeremy, great idea re, ‘I should do a whole post asking people for examples of “most egregiously mis-cited classics”…’. I’d suggest anything written by Frederick Clements. Routinely cited by vegetation ecologists keen to highlight that each round of new ecosystem models brings something new to the table, without having read anything by Clements beyond the textbook view we all know. (I’m not in any way promoting traditional Clementsianism, am just critical of lazy ways of professing novelty). If the Clements estate could sue the world’s ecologists for mis-representation I reckon they’d make a fortune from generations who quote him without reading anything he wrote. If you run such a post you’re bound to generate a wonderfully eclectic comments stream.
Pingback: Citation patterns of classic ecology papers: the most-cited classic is a zombie | Dynamic Ecology