How often are you cited correctly?


Sometimes, I get cited incorrectly. I have some feelings about this.

When I started doing science, the Science Citation Index was a thing. But it was a big volume of books, in the reference section of the library. Looking up citations of authors and papers was a rare occasion. I guess we would find out what the “citation classics” were, but otherwise, who would even know where or how often papers got cited, except when you’d see it in review or in print?

This changed when the citation index became available via Web of Science, and even more so when Google Scholar came on the scene. Now, folks are acutely aware when they get cited — or at least, you can be acutely aware. (Google was sending me notifications, but I managed to figure how to turn that off.)

Instead of operating in a vacuum of citation information, citations now are getting too much attention, as career currency. Anybody could tell you precisely which papers you’d need to cite to bump up their h-score a few notches. (ahem, I just looked, maybe these three?)

My spambox is laden with desperate missives from researchgate. The last time I looked in there, I saw that they give you a little snippet of text from the sentences where you were cited. At closer inspection, some of these sentences just weren’t quite right. They weren’t necessarily factually inaccurate, but in at least some of the cases, they clearly implied that certain things were in my papers that were just not there. Does this bug me? Well, yes and no. It bugs me in the sense that I see what looks like a flaw in the scientific record — that a myth could end up getting propagated by a game of citation telephone. I’m not bugged in the sense that I feel wronged in anyway. I don’t think it’s about me, as I feel the same way when I see other incorrect citations of other papers too (I’m just more situated to detect them if I wrote the paper).

I suppose incorrect citations to my work are inconvenient when I’m reviewing a manuscript, because this puts me in the position of having to be that person — the one who who tells you to cite this paper this way, and that paper that way, blah blah blah. My personal guideline is that if I feel a need to mention citations to my own work, then I will always sign those reviews.

Why do incorrect citations happen? I think they fall into a few categories.

Sometimes, I think folks just remember the wrong paper. Someone could just remember a fact or finding from a particular lab, and then they cite what they think is the right paper, but get it wrong.

Some folks just want to be able to say something because they think it to be true (or want it to be), and then try to hunt down a citation that validates their statement. This is, of course, poor writing practice, but it surely happens. For example, folks will make some factual claim about ants, citing Hölldobler and Wilson, but sometimes, the thing they wrote just isn’t in that book. It’s a standard joke among ant people: got a thing you want to say, don’t have a citation? It must be in Hölldobler and Wilson! (okay, we’re normally funnier than that.)

Other times, papers become a standard citation for a particular idea or set of ideas, even if it’s not in the paper. For example, a long time ago, I was writing a grant with a pal and was looking for a citation for a general statement about the spatial heterogeneity of nutrients in tropical soils. He said I should cite Fittkau and Klinge. I was like, that’s not in that paper, is it? And he said, well, that’s the paper that people cite when you stay stuff like this. And indeed, that’s what happens. Once a paper gets cited for something, it’s more likely to get cited for that thing again. Even if it doesn’t say what people say it says.

Another way bad citations can happen is when authors upsell their work in their abstracts. Just because an abstract says a paper says something, if you read the whole paper, there’s a chance that the abstract isn’t quite representative of what happens in the paper.

So, bad citations are bad, right? So of course, when I cite a paper, am I 100% sure that this paper says what I’m claiming it says? Well, hmm. That’s a rough question. Because at the time, I feel 100% sure. But that’s based on my memory of what the paper says. And as anybody familiar with the criminal justice system can tell you, memory is mighty poor evidence. And I often do the thing I mentioned before, where I know that a certain person had a certain finding — and then I hunt down the paper that I can cite for this. I might get wrong, sometimes, I guess.

I doubt that citation error is steering science in the wrong direction. I think the biggest risk cost of this practice is that it gives us the impression that there aren’t gaps in our knowledge. A lot of citation errors, I suspect, are when a sweeping general statement is made even though the cited papers are very clear that the finding doesn’t constitute a sweeping generalization. So we use citations to gloss over what we don’t know, which ultimately gets in the way of discovery.