One of our Ph.D. students, with whom I’m writing a paper, asked me the other day to edit the Acknowledgements section of our paper, and this got us into a conversation about the “rules,” such as they are, for including acknowledgements in a paper. I’m of the opinion that one should acknowledge everyone who will hold still long enough: the funding agency, personal contacts at the funding agency, the study respondents, colleagues who gave you ideas or allowed you to bounce ideas off them or edited early versions… and so on. The one thing I don’t generally do is acknowledge anonymous reviewers, which is something that I see in the literature fairly often. And my bone to pick with reviewing has nothing to do with this. I just figure, it’s the reviewers’ job to review my paper – even though they’re not getting paid, it’s the job they agreed to. I suppose it’s polite to thank someone for doing their job, particularly if they do it well, but not when there are other people who are doing you a favor to be thanked.

The NSF’s Grant General Conditions actually requires that the NSF be acknowledged in any publication, and, helpfully, even gives you the language so all you have to do is copy & paste (section 20, Publications). I always forget which NSF document this is in, so it’s probably good that I’m posting this. Of course now I have to remember that I posted this, the next time I publish something from an NSF grant.

Anyway, this reminded me of this article that I saw in Nature News recently, about a study of the acknowledgements in published papers, using text mining methods. And here’s the original article from PNAS. Citation analysis is a way of representing formal networks in scholarship, but what’s cool about this study is that it’s a way of representing informal networks.

(Can I just say, PNAS is just about the worst acronym? Though equally bad is a developing NISO standard for transferring questions and answers between digital reference services, the Question/Answer Transaction Protocol… the QATP. Though I suppose whether these are awful or not depends on whether you think they’re acronyms or initialisms.)

Anyway… acknowledgement analysis seems to me to be bordering on a semantic web-style issue. The Nature article makes the point that “there are many reasons to acknowledge somebody,” and the PNAS study does not differentiate. Of course the same is true of citation: citation analysis doesn’t distinguish between

“As Pomerantz (2005), in a work of staggering genius, writes…”


“As Pomerantz (2005), that idiot, writes…”

So the situation would be improved if we had typed links.

And of course the PNAS paper has an acknowledgements section in which the authors thank colleagues, the NSF, and anonymous reviewers.