Based on my interest in authorship practices in ecology, I decided to look at papers published in Ecology in each of the past seven decades to see how corresponding authorship changed over that time.* I looked at the first (or second**) issue of Ecology in 1956 and every ten years thereafter.
tl:dr version of the results: Not surprisingly, the number of authors increased over time. For corresponding authorship, I found that, in 1996 and earlier, the corresponding author was almost never indicated. Looking every 5 years from 2001-2016, the first author*** was usually the corresponding author, though expanding the analysis to include AmNat and Evolution**** suggests that some of the changes might be due to some of the more mundane aspects of publication.
The purpose of my analysis wasn’t to look at how the number of authors on Ecology papers changed over time, but that was a striking feature while going through the issues. (This is a feature that has been noted regarding the scientific literature generally. I feel sure someone must have made a similar plot specific to ecology papers, but I can’t find it. If you know of one, please let me know in the comments! As I say in footnote 4, I’ve been told Daphne Fairbairn did a similar analysis for Evolution, but can’t find that online.)
The first paper I saw that had a note related to correspondence was Kalisz & Teeri 1986, which includes a footnote that says “Send reprint requests to first author”. This surprised me, since I would have assumed that most people would send the reprint requests to the first author. I’m not sure what the story is behind that footnote. One question this raises (that I would love to hear thoughts on in the comments!) is what the default was in the pre-email days when author addresses were all listed. If you wanted to find out more about a paper, who did you write? The first author? The last author? (Assume you didn’t know any of the authors personally.) I thought it was the first author, but this footnote makes me wonder.
The next paper that had a note related to correspondence was Murcia & Feinsinger 1996, which gets bonus points for having authorship order determined by a coin toss. (I have a post on fun ways of deciding authorship order.) Their footnote says “Address correspondence and reprint requests to second author,” making it the first footnote I saw in this analysis that had what I would consider a standard correspondence footnote.
My scan of old issues indicates that the email address of at least one author started appearing in Ecology in 2001. Looking at 2001, 2006, 2011, and 2016, it’s clear that the first author is usually the corresponding author for Ecology papers. The proportion of papers with the first author was the corresponding author remained high over the whole period. 2016 had the highest proportion of papers with the last author as the corresponding author, but it’s still a pretty low percentage.
note: “ND” means the corresponding author was not designated
The pattern in Ecology made me wonder if it would hold across journals. I initially started to look at AmNat, but quickly ran into a snag where, in 2006, almost all of the authors had their email addresses listed. Based on an email conversation with Trish Morse at AmNat, it sounds like that was a reflection of their submission/peer review system, so I decided not to include AmNat in my analysis.
Instead, I looked at Evolution. Right as I was about to start the analysis of the Evolution data, an evolutionary biology colleague stopped by my office. When I explained what I was doing, her guess was that the proportion of first authors as corresponding authors would increase, which was the opposite of what I was guessing. We were both kind of right (well, she definitely was definitely right, and I may have been, too):
It seems clear that the number of first author corresponding authors is higher now than it was 15 years ago. It’s a bit less clear for last author corresponding authors, but, as for Ecology, 2016 was the year with the highest percentage of last author corresponding authors. The biggest decrease of time was in the percentage of papers where both/all of the authors had their email addresses listed (lumped together into “all” in the figure). And, in some of the earlier issues, there were some unusual combinations (e.g., first and third authors on a paper with four authors having email addresses, but not the other two). For 2016 (with a reminder that this analysis just includes data from the first issue), the corresponding author was always the first or the last author.
One issue this analysis suggests is that some of these changes over time may reflect changes in submission systems and/or people working out what it meant to have an email address on a paper. AmNat and Evolution both show us that it used to be relatively common to have email addresses for all authors, but that this is now rare. So, while I started this analysis thinking that it would tell me about peoples’ big picture views on corresponding authorship, now I’m wondering if some of it just reflects more mundane aspects of the process of submitting a paper.
Then again, maybe that is exactly the point: prior to having just one or two people with an asterisk next to their name indicating that they should receive the correspondence, corresponding authorship wasn’t a thing that people could note or track and, therefore, wasn’t something people really cared about. Now that we have that asterisk, it is becoming something that people care about and argue over.
One of the things I’m most interested in is whether being corresponding and/or last author is a position of prestige, which is harder to get at with this sort of analysis. But my earlier poll results suggests that most ecologists now view the last author as the senior author, and I am not the only person to have noticed and commented on the apparent increase in the number of last authors who are corresponding author.
The biggest motivation for me continuing to do this analysis is thinking about how early career scientists might be influenced by these views and practices. To me, it’s problematic if some people view last and/or corresponding authorship as something prestigious but others do not. Knowing that at least some people view corresponding authorship as something valuable, I will continue to have my default be for my grad students and postdocs to be corresponding author. What to do about last authorship seems a bit trickier for pre-tenure folks with their own labs. At Georgia Tech, I was told directly that I should be last (and corresponding) author on papers coming out of my lab.***** But, when I got to Michigan, it seemed like some of the senior folks I spoke with expected that I would have been first author on most papers out of my lab while I was an assistant professor. That argues for figuring out the local culture, but, even with doing that, that doesn’t solve the problem of tenure letter writers having different views on this, and of the potential knock on effects on grad students and postdocs in the lab. Those issues seem much less tractable to me.
Do you think it’s a relatively recent thing that people view corresponding authorship as something that matters? Do you think that people care about it now simply because we’ve created a way of indicating it (by having one or two people who get an asterisk next to their name)? What’s the best solution to this problem for early career folks? (Before you say we should just go to authorship contribution statements, read this post of Jeremy’s for more on why those are problematic, too.)
*I had this idea while up all night with a stomach virus over the holidays. In the acknowledgments to his Evolution of Infectious Disease, Paul Ewald notes that, if he had followed the lead of Hutchinson’s “Homage to Santa Rosalia”, he would have called his book “Homage to Manhattan, Kansas”, since his inspiration to write the book “originated form a rather bad case of diarrhea that I acquired while I was conducting reserach near a little garbage dump on the outskirts of Manhattan, Kansas”. So, maybe this post should be named “Homage to Central New York”, where I acquired my stomach virus from extended family.
**My plan was to look at the first issue of the journal in each year, but, in 1996 and 2006, the first issue had a Special Feature. In those two years, I looked at the second issue instead, just to avoid any possible confounding effects of Special Feature papers.
***If there was only one author, I said that the corresponding author was the first author. In the early days when there wasn’t always an email address, if there was only one author, I still said the first author was the corresponding author. If there was more than one author and no one had an email address listed, I said the corresponding author was not designated (“ND”).
****I heard that Daphne Fairbairn made a similar plot for Evolution and presented it at the Evolution meetings. My search for it revealed this tweet but not a writeup of those results. If someone knows of one, I’d love to see it!
*****I did not always heed this advice.
Dynamic Ecology 
So, I think it’s obvious that the expectations regarding what information is encoded by author order, differ wildly by field and venue, and perhaps even by institution. Obviously, much hilarity ensues.
Corresponding author though, I have always assumed should be set to the person who has a combination of the most authority regarding the work, and the most durable contact information. I would almost never assign corresponding author to a trainee, for the simple reason that their contact information is fairly ephemeral, and I can’t be sure that two years down the road, they’ll be paying the slightest bit of attention to correspondence at the contact address they have today (heh – if you have the same luck with trainees that I do, it’s hard enough to get them to pay attention to “old people” communication modalities like email even when that email is their job).
As I see it, that’s not necessarily a position of prestige, but it’s absolutely must be a position of responsibility, with as long-as-possible a guaranteed lifetime, and as long-as-possible durable access to the original data and understanding of the complete project, should, in the future, correspondence be necessary.
Via twitter: https://twitter.com/FrancoisKeck/status/838784641964392449
(Update: unfortunately it’s not loading the preview! Click through for a plot of the number of authors per article over time)
Here’s a blog post with the figure and more on how Francois made it: http://www.pieceofk.fr/?p=479
I was drafting an email to my coauthors on just this when your post popped up. Seems to me the role of corresponding author has morphed into that of the Administrative Author, who might be the same as the lead or Senior. Who’s willing to slog through the publisher’s interminable online document manager, who’s willing to corral the criticisms, assign responses, and all the other nonscience parts of writing a science paper? Should this be the lead author or junior person? If they are trainees such as PhD students who are moving into a professional or academic science role need to know the drill, certainly. But when their career paths are uncertain, no. I just worked with a brilliant driven young woman in ?Colorado USA who did a great job with the experiment, organizing and analyzing the data, writing the paper, … and then took a teaching job in Ghana which led to another in Tanzania. I don’t think she would be too responsive to author inquires.
I used to chafe at why all authors weren’t allowed to give emails, since that would include the odds of at least one of us having a functional address after more than a few years. But now, it’s usually easy enough to find someone by search engine, as long as they want to be findable.
I have seen the corresponding author role abused by a senior author as a means to enforce complete control over the lead author, who in that case had graduated and wasn’t under Senior’s thumb quite so much. No correspondence, no article. It’s been described as the “White Bull” problem. Primack et al’s post and article is worth a read, although I don’t want to have go as far as they recommend on author agreements. “Some of the most problematic cases occur when professors and graduate students are co-authors on a paper. Because professors have much greater power and experience in these situations, unethical and selfish professors can manipulate authorships to their advantage, dictating who will be co-author without consulting everyone involved. In extreme cases, professors can take primary or even exclusive credit for work done primarily by their students, or even block their students from submitting papers.” https://www.elsevier.com/connect/co-authors-gone-bad-how-to-avoid-publishing-conflicts. Also, today’s Retraction Watch had a story of retractions because the authors didn’t include the Senior author. One wonders’ whether its intellectual contributions or control/credit for the boss at stake in some of these authorship disputes.
Have to admit to being on a bit of a protest over massive mega-author lists, such as Gleick. These aren’t authors, they are signatories. I won’t spell them out, and if the journal style police won’t let it through, I won’t cite them. For example: Gleick, P.H. and 252 co-authors. 2010. Climate Change and the Integrity of Science. Science. 328(5979): 689-690. http://dx.doi.org/10.1126/science.328.5979.689.
I’m trying to lead an article with a similar issue, although I’ve got a more manageable list 15 or so author/signatories. Be interested what people think of author/signatories. Is it a dubious practice, akin to gift authorship? How else might I approach it.
Thanks for taking time for another thoughtful post. I hope these collected posts are more durable than email addresses of corresponding authors.
In terms of durability: I’ve been working (slowly) to compile the main points into a manuscript to give it more durability, and to make it easier for people to share with colleagues.
Thanks for the link to the Primack article! And your points about huge author lists and lack of email addresses for all authors are related. Trish Morse from AmNat said that part of why they stopped automatically adding email addresses for all the authors is that, on a paper with very large author strings, sometimes the actual article didn’t start until the second page.
One of my senior colleagues in ecology who publishes frequently with a group of geochemists often gets last author on the interdisciplinary papers, as he rarely contributes much more other than guidance and obtaining grant funding. To the geochemists, this makes sense that he’s last in the authorship, as he has done the least, yet to the ecologists, it appears that he is the senior author on all of the papers coming out of the group. He’s well aware of this win-win situation!
That’s convenient!
In my opinion authorship order debate is a symptom of a bigger problem that either no one understands or no one wants to talk about it for whatever reason. Why one coauthor needs to be distinguished from the rest by establishing a “prestige position” in authorship order?
There’s almost never a debate about who should be the first author. That’s because everyone knows that the first author works the most and leads the paper. But debates surface about the remaining authors. Why?
In theory all coauthors should contribute to a paper equally. But in reality many coauthors make questionable contribution (don’t ask for reference please!). So those coauthors who work harder try to find a way to distance themselves? Or may be some just find themselves in an “identity crisis” situation? What’s the root cause of this temptation to be distinct in a large author group?
I think ethical standards need to be higher when it comes to contribution of coauthors. In an imaginary world where every single author makes exactly equal contribution, authorship order should not have mattered.
“In theory all coauthors should contribute to a paper equally.”
Why do you say that?
“There’s almost never debate about who should be the first author”
In my experience, this is not always the case, for at least two different reasons. First, there are collaborations that involve truly equal partnerships, with sometimes also a pair of leaders and other coauthors making lesser but still meaningful contributions. Second, take the case that someone completes some research, including drafting a manuscript. But subsequently, maybe new data are added, analyses are revised, and some amount of rewriting occurs, with no/minimal involvement of the person who pulled the project through its early stages (in my case this was a student who didn’t continue on in academia). Is that person still first author? Or more accurately, at what point in the revision does the paper stop being primarily that original person’s work and primarily someone else’s? These are cases where the first author may not be clear-cut, and fortunately they were resolved pretty quickly and without obvious hard feelings, but they did require discussion.
I am also curious why you assert that “all coauthors should contribute to a paper equally” and look forward to your response to Jeremy. I hope you’ll also clarify whether you really mean all authors (including the first author), or all authors after the first author.
@ Jeremy and JW:
All coauthors will contribute equally – isn’t it the default expectation?
We generally give away the first position to someone who leads the paper and/or does the most work. But there is no accepted norm about where to put someone who contributes the least. If A, B and C contribute 55, 40 and 5 percent, shouldn’t they be listed in that order?
Or how about any of the following somewhat radical but not-completely-unheard-of ways:
1) Author contribution statement goes right after affiliations, clearly designating who contributed what and by how much? Currently some journals require a kind of meaningless author contribution statement – “A” did this “B” did that. And this comes towards the end, not sure how many people care reading this. This should come upfront right after affiliations and with measurable indicators.
I think Nature’s guideline (http://www.nature.com/nature/authors/gta/) is a right step in that direction:
“If authors regard it as essential to indicate that two or more co-authors are equal in status, they may be identified by an asterisk symbol with the caption ‘These authors contributed equally to this work’ immediately under the address list. If more than three co-authors are equal in status, this should be indicated in the author contributions statement”
Also, a quick “equal contributions and credit” search in Google Scholar returns a bunch of interesting papers. Looks like in medical research it’s a kind of a big deal!
2) Lab director’s authorship statement can be made clear – e.g. supervised the research; unless they contribute to manuscript development. Not sure finding money is a great authorship criteria.
3) Make different font size for different coauthor names depending on their relative contribution? This sounds bizarre, but this kind of practice does exist. Look at one example of the font size of a James Patterson coauthor: https://www.amazon.com/Middle-School-Australia-JamesPatterson/dp/0316272620
Patterson as a first author isn’t enough – they want to make it clear, by printing names in smaller fonts, that the coauthors make smaller contributions (is there any alternative interpretation?). It’s bizarre to think this in scientific papers, nevertheless makes perfect sense.
I know I sound like out of line – but just wanted to share honest thoughts.
Thank you for your further comments, you don’t sound out of line at all!
“All coauthors will contribute equally – isn’t it the default expectation? ”
I’ve never heard of that being the default expectation.
“But there is no accepted norm about where to put someone who contributes the least. If A, B and C contribute 55, 40 and 5 percent, shouldn’t they be listed in that order?”
Yes, the lack of a universally-accepted norm is unfortunate. Until fairly recently, listing authors in order of their overall contribution was standard practice in ecology (https://dynamicecology.wordpress.com/2013/03/13/how-do-you-decide-authorship-order/). I favor that norm, but it seems not to be the norm any longer. According to a poll Meg did a little while back, ecology seems to be shifting to the biomedical model in which lab PIs are last (“senior”) authors, a status to which some importance is attached. Some ecologists also are starting to attach some importance to “corresponding author” status. See https://dynamicecology.wordpress.com/2016/06/27/last-and-corresponding-authorship-practices-in-ecology-part-1/ and https://dynamicecology.wordpress.com/2016/06/28/what-factors-influence-views-on-last-authorship-in-ecology/.
Authorship standards also are dropping; people who would merely have been acknowledged a decade or two ago tend to get coauthorship today. See https://dynamicecology.wordpress.com/2015/02/12/is-the-peg-model-paper-an-indicator-of-changing-authorship-criteria/
I used to think that author contribution statements were a solution to these issues. But almost no one reads them, except for trivial, unimportant purposes: https://dynamicecology.wordpress.com/2016/07/28/views-on-authorship-and-author-contribution-statements-poll-results-part-1/ and https://dynamicecology.wordpress.com/2016/08/08/views-on-authorship-and-author-contribution-statements-poll-results-part-2/. Perhaps that will change in future, but so far I see no sign of it.
The previous two links also have some (puzzling) poll results on what contributions ecologists think are “authorial”. For instance, should you be a co-author just by virtue of being the head of the lab in which the work was conducted, or by virtue of having obtained the funding for the work? Opinions differ. My own view is that you shouldn’t be a co-author just because you’re the head of the lab or just because you obtained the funding.
I agree that indicating author contributions with font sizes is a bizarre idea that would work in a scientific context. It’s done on the book you link to because James Patterson is famous and his co-author isn’t. James Patterson fans are numerous and will buy any book he writes, so the book is marketed by featuring his name prominently, whether or not he did the bulk of the writing. I definitely would not want to see scientific papers printed so as to accord extra prominence to the most famous author!
Pingback: Should figures go at the end of a manuscript or appear in line? And where should figure legends go? | Dynamic Ecology
Pingback: Last and corresponding authorship in ecology: a series of blog posts turns into a paper | Dynamic Ecology
Pingback: Who should be a co-author? Rules and etiquette of academic authorship – Ramblings of an ecóloga