Posted by Paul on March 29th, 2007
Scientists have been pretty good about developing standards to ease communication across scientific disciplines, but we have yet to decide on a standard protocol for determining the order of authors on papers. Since papers are widely regarded as the principal metric of scholarly achievement, maybe it’s time we did.
If you asked anyone what position they’d want in an author list, the general consensus would be first. Knowing nothing else about a paper, most of us would assume that the first author made the most important contributions. In the majority of chemistry labs, the first position is usually granted to the researcher who took primary responsibility for conducting experiments, writing the paper, and shepherding the project to completion. The remaining junior authors are usually listed in descending order of perceived scientific contribution, until you get to the principal investigator, who almost always goes last (1 2 3).
But making the above assumption can lead you astray, as that’s not the way all chemists do it. Although it has become decreasingly common for the principal investigator to take first authorship (1 2 3), some young professors are continuing this trend. It seems to be more common in organic synthesis than in other areas. My understanding of this practice was that it simplified finding papers out of a particular research lab back when papers were sorted alphabetically in literature indexes. While the advent of computerized literature searches has rendered this advantage obsolete, the practice of some PIs “going first” persists. Perhaps it is simply a matter of tradition, or maybe a means of senior authors “protecting their brand,” since long author lists are usually shortened to “First Author, et al.”
What do other types of scientists do? Astrophysicists list authors in descending order of contribution (1 2), which means that the advisor is usually the second author listed. While this system seems the most fair to me, I can see how some bosses might lament the prospect of getting lost in long author lists. When I’m scanning through ASAP alerts, after the title, I usually glance quickly at the end of the author list to see where the paper came from. In this respect, the PI going first or last is rather convenient. An astrophysics professor I talked to here said that he usually assumes that the last author on a paper was the guy who drove the rest of the team to the telescope.
While the practice of listing authors in alphabetical order is exceedingly rare in chemistry, it turns out to be standard protocol in the economics community (1 2 3). Does anyone else find it disturbing that, when left to their own devices, economists will revert to academic communism? Isn’t there something to be said for providing incentives for hard work? Oh well.
Aside from author order, there is also the wacky issue of joint authorship. Sometimes author order assignments are so troublesome that people choose to indicate joint first-authorship. While it is occasionally seen in Angewandte Chemie, it has actually been banned from JACS:
JACS does not permit equal or partial authorship designations. JACS assumes that all authors have made substantial contributions to the work.
I’ve got no problem with having these notices of equal contribution; I’m not sure why JACS does. The issue of joint authorship is so important to some authors that they actually published addition/correction notices to set the record straight (1 2).
Finally, I’d be remiss if I didn’t link to Jorge Cham’s Ph.D. comic strip that illustrates what the order of authors really means. There’s more than a bit of truth in it.
So…what style do I advocate? First, any of the above styles is acceptable. It’s a personal decision that is determined at the sole discretion of the PI. So long as the community knows how each PI operates, there won’t be any problems. That said, I like the PI-last format—with a twist. With the PI-last ordering, the people in the trenches have added incentive to get papers out and the PI gets to stand out too. My twist is that I think the first pages of supporting information for any paper should contain a short summary of each author’s contributions. This way, anyone who wants to know who did what can find out. The added detail would be a double-edged sword, because with each credited contribution comes the added burden of being responsible for the integrity of the data. Assigning responsibility in this manner would make it harder for fraudsters and sloppy researchers to hide behind their colleagues. Furthermore, I think this list should be followed by a signed statement from each author asserting that, to the best of his or her knowledge, the results are authentic, are accurate, and were obtained in an academically honest manner. It would be naive to think that this simple change would solve all of our community’s problems, from inflated yields to outright fabrication, but it might make some people think twice about going down those roads.