Publish or perish?

April 17, 2009 1:41 PM

How often are academics expected to publish?

I realize that the discipline and the quality/type of publication are major factors. But I'm guessing that publishing in a peer-reviewed journal annually would be expected of most academics, two such publications annually would not be unusual, three would be considered prolific. In addition, a book every 5 years? conference presentation/paper once or twice a year? Just looking for some very broad estimates.

posted by angiep to education (21 comments total)

From my experiences in science I would say at least 1 paper a year with the number being much higher than that depending on how far up the ladder you climb (For example, a grad student should try to aim for 1/year at least in years 3-5, post-docs 1-2 a year, and the principal investigator of a lab many more than that, depending on how big the lab is). I would say grad students/post-docs aim for at least one conference presentation a year, and again the PI either being completely over that stage in life, or expected to do many.

I can't comment on the book writing. Maybe a chapter in a book occasionally, but I don't know many people that have entire books out there.
posted by sickinthehead at 1:45 PM on April 17

That sounds about right, at least for an academic who hopes to attain/has already attained tenure. One paper this year, three next year, two the following, etc. is pretty typical, I think.

It does depend a bit on your field though. I would imagine that for academics in engineering and the hard sciences, where papers frequently have dozens or hundreds of co-authors, things would be a little different.
posted by valkyryn at 1:47 PM on April 17

I think you're estimates are reasonably accurate, however universities are going to have their own "soft" requirements for faculty. This will vary wildly from school to school, department to department according to all types of factors such as school rankings and whether the school considers itself a research institute or not.
posted by wfrgms at 1:47 PM on April 17

You really can't generalize, because it not only depends on the field, but also on the expectations of a given institution. Schools with larger teaching loads expect less publication than research universities.

Here is one example of expectations for an interdisciplinary field in the humanities:

By three-year review: at least one article in one of the principal peer-reviewed journals of the field, plus at least a few other publications.

By tenure review (during the sixth year): at least one book published that is a monograph based on original research, not an edited volume. It should be significantly different than one's dissertation.
posted by umbú at 1:50 PM on April 17

Depending on the size of the lab, in biomedical research three or more publications a year is certainly not unusual, and by no means prolific. Additionally, your name can be listed as a co-author in many, many publications (one of my advisors is listed as a contributer on over 600 publications after a 20-year career).
posted by halogen at 1:51 PM on April 17

Eh, contributor.
posted by halogen at 1:52 PM on April 17

In the sciences, three peer-reviewed papers a year, on which one is primary author (i.e. did much of the work), would be prodigious. A co-author though, or as a head of lab doing an edit, for example, that number can easily rise to the double digits. Including conference papers could double that number.

Books are much less common. Writing a book chapter is like writing a really, really big paper and one could happen every year, at the expense of a paper. Writing a monograph could take years. An active author will update a monograph every decade or so. Editing a book takes about a year, during which time you'll not have the opportunity to do much else. Writing a textbook is also a big endeavour, but publishers will ask to to "update" it frequently.
posted by bonehead at 1:53 PM on April 17

Thank you so much, all! This is exactly the info I wanted.
posted by angiep at 1:54 PM on April 17

The things that factor into this are primarily discipline, single authorship, and type of institution. In general, the following pertains:

- Physical, biological, and medical sciences expect a lot of articles, value books and book chapters much less, and do not value sole authorship any more than multiple authorship. Essentially, they assume that you will be running large research projects, with multiple collaborators, which generate lots of papers on a consistent basis.

- Humanities highly value papers on which you are the sole author; value books higher than articles; and expect fewer articles. Essentially, they assume that you are a lone researcher who may take months or years simply conducting research for a book-length project, which may or may not generate an article or two as a by-product.

- Social sciences value both multiple- and single-authored publications, also value books, and expect a moderate amount of articles. This is somewhere in between the hard sciences and the humanities.

- Finally, smaller, less-prestigious universities and liberal arts colleges will expect a lower output than larger, more prestigious institutions or research universities.

What this means is that its tough to generalize in any way. I know physical scientists at liberal arts colleges who are expected to produce one article per year, and humanities professors at research universities who need to publish at least one book and multiple article every two years to get tenure.
posted by googly at 1:56 PM on April 17 [4 favorites]

It depends on the field. Even related fields can be worlds apart. I recall a high ranking NSA official complaining publicly that CS publishes too many things, and that what they do publish is basically the same thing over and over. Progress reports, conference proceedings, and so on. He complained that the rapid pace of publication was interfering with basic research; instead of further research, people spend time preparing old manuscripts for the next conference deadline.

I wish I could find it now, but he compared all that with the math field, where you publish infrequently, when you have findings. Nobodys interested in a progress report on proving the Riemann Hypothesis, I think was his central point.
posted by pwnguin at 2:08 PM on April 17

As others have said, this varies highly by field, university (as well as type of university), and department, and also changes over time to a certain extent. In some fields, books are unusual/dispreferred but in others they are required, in some fields conference proceedings are as prestigious as journal articles and journal articles are less common, in some fields co-authorship is high (meaning publication standards are also very high in terms of quantity) and in some fields co-authorship is rare, and advisors are typically not co-authors. The best way to figure this out for a particular field is to find a bunch of people who've recently gotten tenure and see what their CV looks like. Also, if by perish you mean not get tenure, at some schools/departments it doesn't matter what you do, you just won't get tenure.

In linguistics, 1-2 journal articles/book chapters a year is approximately typical (though more is not bad, and since lag to publication is high and somewhat unpredictable you have to shoot for more), high amounts of co-authorship (esp. between advisor and student) is not common, books are not required (though it isn't uncommon to revise your dissertation into a book), and some conference proceedings are very prestigious. Also, it would be extremely uncommon for a grad student to have more than 2 peer-reviewed publications by the time of graduating, and it is relatively common for them to have none, or one that is accepted but hasn't appeared.
posted by advil at 2:11 PM on April 17

Just one more thing to add - which is that the approach to publishing also differs according to your country's academic tradition. In the world of biology (not my field; I'm a social science - but I hear things), in the States, it would be acceptable to publish a couple of very big Nature-level papers; in Europe, to publish so few papers would significantly weaken one's job prospects.
posted by archofatlas at 3:04 PM on April 17

I recall a high ranking NSA official complaining publicly that CS publishes too many things, and that what they do publish is basically the same thing over and over. Progress reports, conference proceedings, and so on.

This is so true.

I've watched computer science professors write a paper for the preliminary results of an experimental algorithm, then write a paper describing the implementation, then another paper describing an improvement, and another paper describing another improvement after the first improvement failed to improve things, then a paper on the experimental results with a well-known dataset, then another paper for another dataset, then another paper where it's shown that it works with "arbitrary input", and finally another paper showing the proof of correctness (or statistical derivation, or whatever).

All in the course of, say, two years. Frequently with a rotating cast of co-authors as you swap a clever graduate student hacker for a colleague in the math department or whatever.

And those papers will have significant pieces of text shared between them. It'll be rewritten somewhat, but it started out as copy & paste from the last paper.
posted by Netzapper at 4:22 PM on April 17 [1 favorite]

Publishing is really about becoming well known and respected in your field, though both quantity and quality can accomplish this. In computer science alone, publishing expectations can vary from four or five conference papers a year to one conference paper in five years depending on what area one is in.
posted by pbh at 4:47 PM on April 17

My undergrad advisor published one book-- from his dissertation-- in 1967. The sum total of his scholarship for the entire rest of his career, which ended about five years ago, was a single book review.

Now I have colleagues who publish three and four refereed articles a year (this is in sociology). I have a colleague who last published any refereed product in 1998. I have a colleague who has written 35 books (though "book" is defined loosely). I have two who were just promoted to full professor having never written a book, but having written a good number (averaging, oh, three every two years) of refereed articles. I'm now 16 years post-PhD, have 15 refereed articles and one book. I went to grad school with people who are tenured at major research universities who have, say, four books but only, say, two refereed articles; I have one who is department chair at a teaching-intensive liberal arts college, and he has published a grand total of two refereed articles, and that's it.

THERE IS NO TYPICAL. Expectations vary more by the institution in which one works than in one's discipline. If you have a 4-4 teaching load then even publishing one article every OTHER year is going to look damn impressive. If you're 2-2 and you go up for tenure with two articles to show for 5 years of that luxurious teaching schedule then you can kiss tenure goodbye.
posted by ethnomethodologist at 5:34 PM on April 17

Something that didn't get expounded upon much above is first-authorship, especially for students.

In lab sciences, middle authorships are an indicator that you can do something collaborative or useful to a lab. For example, I might help many people do modeling / a statistical technique. Your first authorship papers identify where you are thinking on your own, doing the dominant part of the work, and probably most of the writing. Those are extremely important, and not exchangeable with really any number of middle authorships. I've heard faculty denigrate job candidates by saying that as grad students or post docs they had failed to distinguish themselves from their labs; that they looked like their PI's tech more than an independent scientist in development.
posted by a robot made out of meat at 5:42 PM on April 17

Googly may be right about most of the humanities--from what I know it's true about English--but at least in philosophy, an article in one of the top-ranked journals is generally going to do more for you than a book.

Then again, most books in the discipline tend to be patch jobs of previously published articles, so maybe that's why.
posted by Beardman at 6:11 PM on April 17

One of the professors (of human genetics) in my Ivy League graduate Anthropology department published, between 2001 (hired by the university) and 2004 (halfway to tenure, which he got in 2007): 12 research articles, 2 review articles, 12 book chapters, 5 book reviews, 2 popular publication articles, and gave or had his name on 29 conference papers. All in 4 years. He's also the first (and only, as of mid-2008) person in the department to make tenure in a decade, since the mid-90's.
posted by The Michael The at 6:13 PM on April 17

In my field, this depends a lot on what level you're at. A graduate or postdoc will be publishing papers as a first author, and will be doing well to have 1-2/year. 3-5 a year would be considered quite prolific; more than 5 would be an incredible year.

Principle investigators who have just begun running a lab won't be publishing first author papers anymore (except for the occasional review), and their per-year publication rate will ramp up as their lab gets started, from 1-2/year in the beginning to on the order of 5-10/year once everything is running smoothly. Some of those will be from collaborative projects in which members of the lab made relatively small contributions.

A tenured PI running a sizable research group can easily produce upwards of 20 papers/year, including collaborative work.
posted by mr_roboto at 6:18 PM on April 17

Another data point: for a junior scholar in law, one or two law review articles a year is fairly typical. Books typically wait until after tenure (which is fairly easy to obtain in law).
posted by raf at 8:24 PM on April 17

Totally depends on lab size, institution, field, as stated above several times. I recommend looking at peer departments and seeing what newly tenured professors have put out in the last 6 years or so.
posted by mezamashii at 6:29 AM on April 19