Join 3,562 readers in helping fund MetaFilter (Hide)


Why do papers in academic journals so often have more than one author?
March 27, 2014 4:20 PM   Subscribe

Why do papers in academic journals so often have more than one author?


These preprints I'm looking at often have four, five, seven, eight authors.
Geez! What gives?

What's wrong with one person writing one paper?
posted by shipbreaker to Education (27 answers total) 6 users marked this as a favorite
 
People have different skills, training, abilities, etc. For example, I'm on a paper with my undergrad supervisor, and the grad students I was helping out. I did the analysis and literature review. My supervisor took my undergrad paper and turned into something good enough to publish. The grad student taught me the observational skills I needed and got the funding to pay me while I was working for them. The paper really couldn't have happened without each of us.
posted by hydrobatidae at 4:24 PM on March 27 [1 favorite]


Most science is collaborative and everyone who substantially contributes to the study is typically listed as an author, although the writing is done by one or two people at most. I've been an author on papers because I did the field work, or had the idea and got the grant but didn't do the fieldwork. Or sometimes because I just help people out with their studies. Once I got a co-authorship for literally going to a site and ID'ing a single organism. Also if you work with a statistician they are often listed as an author. Same for programmers etc.
posted by fshgrl at 4:25 PM on March 27 [17 favorites]


A stupid idea can take a long time to reveal its true nature if you just mull it over inside your head.

All the best ideas in my research experience have been improved through discussion with other smart people. These people are called "coauthors".
posted by L0 at 4:25 PM on March 27 [5 favorites]


Because academic research is often done in teams. In the most typical cases, there are a team of graduate students or postdocs who are doing the actual work, and then there are one or more professors supervising the work and potentially contributing more theoretical aspects; these are both worthy activities, and doing both is worthy of recognition.

Additionally, being an author on a paper is rewarded - authoring more papers improves your chance at the next step, whether the next step is a degree or a tenured position; since it's free to put someone on as an author, you generally want to err on the inclusive side and keep everybody happy.

There is a cultural difference from field to field, though; in the engineering papers I usually write, the coauthors are generally limited to a couple of people who did the actual work and detailed supervision, two or three authors, with others thanked in the acknowledgements. I've just started publishing in medical journals as well with some folks from the faculty of medicine, and I couldn't pick half the names on our most recent paper out of a lineup of two, and since I did the research, I don't know what their role actually was - but it could have been coming up with grant money or something like that.
posted by Homeboy Trouble at 4:31 PM on March 27 [7 favorites]


At least in the sciences, one paper often represents over a year's work but am entire research lab. There is simply that much work out in to the experiments covered in the paper. As many of the people who contributed intellectually, physically, and financially as can be credited should be.
posted by maryr at 4:32 PM on March 27 [1 favorite]


This comic is fairly cynical but gets across the different kinds of scientific work that can warrant 'authorship'. (Which as mentioned above is a big goal when you're in grad school and afterwards.)

In my lab, the general rule of thumb is that whoever writes the first draft of the paper becomes the first author (probably means it's 'your project'/you did the bulk of the hands-on work), and the head of the lab (if involved) is the last author. But the ordering of authors (and who gets to be an author) is a weird and field-specific thing.
posted by heyforfour at 4:34 PM on March 27 [2 favorites]


Because the rewards in academia (and, increasingly fields related to academia) are decided almost solely on authorship, and giving it out is a mildly unethical, but cheap, way to reward friends and collaborators. Authorship, therefore has seen inflation to rival Zimbabwe's currency, with people getting credit for all sorts of things, from visiting a field site once to having a conversation with somebody. Here are some guidelines on how authorship is decided:

How do you decide authorship order? (Ecology)
A Graduate Student's Guide to Deciding Authorship (American Psychological Association)
Bringing Order to Authorship (The Scientist)
posted by one_bean at 4:34 PM on March 27 [1 favorite]


In the sciences, papers generally come out of research groups, which will have a group head, post-doctoral researchers, doctoral candidates and often a few undergraduates. The ordering of the credits is significant but field-specific -- in the scientific fields where I've asked about authorship norms, a publication history generally starts with back-of-the-list credit and slowly moves further forward, but if something really interesting comes out of individual research work within the group, it can mean an early front-end credit, although the process of turning it into a paper will still be collaborative.
posted by holgate at 4:36 PM on March 27 [2 favorites]


One other thing: having a more prestigious co-author can help you get stuff published in better journals. Or if you are publishing in a speciality area you probably want at least one author with some standing in that area.It shows your work has been pre-vetted, I guess.
posted by fshgrl at 4:38 PM on March 27 [3 favorites]


I'm a theorist, not an experimentalist, and in my field we have strict alphabetic ordering (yay for not having squabbles about who is "first author").

But only two of my publications are solo. One of those is my thesis. Why? Well, working alone is not only not easy, I just plain don't like it much. I prefer to bounce ideas off other people, have people check my thoughts, provide their own expertise and knowledge of the literature, etc.

I don't think any of my publications have authors that weren't strongly involved in the work on them. But the work I do is almost always collaborative, so there are multiple authors (2-5 or so).
posted by nat at 4:39 PM on March 27 [4 favorites]


What's wrong with one person writing one paper?

Yes, it would be relatively easy for one person to write a 10-20 page paper, but the text on the page is just a small part of the paper. The authors are the people who came up with the ideas, did the experiments, had expertise in the different fields and experimental techniques that produced the results, etc. In many cases they wrote the sections that they had particular expertise in, but more commonly, authorship is a reflection of everyone who contributed to the work.
posted by deanc at 4:41 PM on March 27 [2 favorites]


Although this article is a serious work of linguistics, the fourteen authors cited are a parody of the multiple-authors phenomenon in the sciences. (The "authors" are anyone with whom the real author had a relevant conversation at some point.)
posted by thomas j wise at 4:48 PM on March 27


My husband is in a hard science, and always does research in a team. Usually there are a couple of PhD experimentalists in the lab who run the experiment, a senior academic who supervises them, usually doing at least some of the design of the experiment, plus troubleshooting when it doesn't work, someone who gets the funding for the research, and then a theoretician or two like my husband who runs simulations and computational models to try to make sense of the experimental data they collect. Sometimes there will be a further person or two who contributes to the actual writing of the paper that describes the research, especially if the main team doesn't have a native English speaker on it. It makes perfect sense that all of those people will be credited as authors - otherwise anyone left out would feel like their time and effort spent on the project was "wasted" (since future grant funding, success in job applications, and distribution of departmental funding will depend on number of papers published).

I am in a social science/humanities field, and we don't collaborate on paper authorship so often, since it is mostly "ideas-based" rather than experimental, so we don't tend to work in teams. Sometimes if we have a paper that crosses disciplinary boundaries, we might work together with another academic in that discipline to make sure that our ideas make sense to readers from both areas, and then co-authorship is common. The other time it is common for us is if someone gets invited to contribute a chapter to a book or to a themed journal article, but doesn't really have time to do it, and then they'll ask a more junior colleague if they want to "come in on" the project, which basically means doing most of the work in exchange for the co-authorship credit. But that junior person would probably never have been invited to contribute otherwise, so it's a reasonably good deal. Pretty much all my co-authored papers are like that.
posted by lollusc at 5:11 PM on March 27 [1 favorite]


Just to mention one more thing, remember that authorship is the primary way that credit is recognized amongst academics, so when conducting very large experiments it's necessary to recognize lots of people who might not have written down words on the page but nonetheless contributed to the existence of the work.

For example, check out the author list on the CMS Higgs discovery paper, or pages 26-38 of the ATLAS Higgs discovery (pdf). These are rather extreme examples, but this is relatively common on smaller levels too.
posted by kiltedtaco at 5:31 PM on March 27 [1 favorite]


In my field (and I think most hard sciences, I don't know) the research advisor is always on the paper in addition to the actual author who did the work. So even if you are super abnormally independent in lab it's still two people on your paper.
posted by gerstle at 5:51 PM on March 27 [2 favorites]


If you saw the sheer number of emails a productive science collaboration can generate, you wouldn't be so surprised by the number of authors a paper can have. A paper typically isn't a single person sitting down and going, "Okay, I've thought about this really hard. Now it's time to do SCIENCE!"

Instead, imagine many hours of involved conversation, many more hours of staring at forlorn scribbles on chalkboards and notebooks, more involved conversation and gut checks, failed attempts to write code, mediocre attempts to troubleshoot code, frustrated attempts to get code to work on a computing cluster, panic and/or confusion at the initial results, a lovely fleeting moment of AH-HA! leading to frantic attempts to draft something halfway coherent, scrapping draft halfway through to go in a completely new direction after someone realizes your original idea was already published fifteen years ago, and then, after many revisions, arguments, headaches, and cups of coffee, finally submitting a finished paper. And this is for a theorist! Spreading the pain around collaborators is one of the only ways any of this would get done.

A paper isn't just a good idea. It's the end result of a ton of hard work.
posted by Diagonalize at 5:59 PM on March 27 [3 favorites]


Even outside the physical sciences, coauthorship is common because writing is a long, complex process. What I tell newish graduate students in political science is that when they read a published article, they shouldn't mentally compare their own work to it. Published articles represent hundreds to thousands of man-hours of work, and quite commonly five or more cycles of major revision in response to conference appearances, rejections at previous submissions to journals, and at least one revision on its final submission.

People coauthor because that environment is difficult to navigate and the work is easier to manage with one or a few coauthors. Also coauthorship is, well, kind of fun, and often a welcome break from the intensely solitary way (non-lab) social-science research often takes place.
posted by ROU_Xenophobe at 6:17 PM on March 27 [1 favorite]


It's the end result of a ton of hard work.

And tea breaks. In the hard sciences, tea breaks are decompression and debriefing and batting about ideas and that's worthy of author credit.

It's also worth thinking about papers in research-group science as snapshots of "what the group is doing", which may focus on particular individuals' work because that's where the most paper-worthy research has come from, but research groups have a pretty well-defined focus within the field. If you look at a group's collected publication history (an example in organic chemistry), you'll see lots of variations on a theme in terms of the paper titles and subject matter, as well as rotations of authorship.
posted by holgate at 6:26 PM on March 27


It's hard to imagine many papers in my field, ecology / environmental science, being written by a single author, unless they were some kind of review paper or short note. Not these days, anyway. Typically authors on my papers include; the guy who obtained the grant and supervised the project, the guy who went out into the field for months collecting the data, the woman who sat down at a computer analyzing the data and doing statistics, a further expert in the field who had input into what it all means, and the woman who actually sat down and started writing it (assuming that guy isn't also one of the above). At least.
posted by Jimbob at 6:48 PM on March 27


Unlike the rest of the sciences, in mathematics you do sometimes see single-authored papers.

For myself, I have several single-authored papers (including the two from my thesis) and a number of multiple-authored papers. Some of those are just with one other author. Some are with a group. One group (5-6 folks, depends on the semester) meets weekly at the university and works on stuff, and eventually we write it up. One group paper (5 authors) was the result of a conference workshop.

I prefer to work collaboratively---it's more fun, you have someone else to be accountable to, there's more than one brain to solve the hard problems, there's someone else to share the writing.

In mathematics, authorship order is strictly alphabetically. (Although there's an apocryphal story about using the opposite ordering of the Zucker-Cox theorem...but apparently they just went with alphabetical, despite the snickering from the back row.)
posted by leahwrenn at 7:13 PM on March 27


So here's the process on a paper I was working on today:

We're doing a study that needs some chemistry and some microbiology.

My team does the chemistry, I'm the PI, I have a junior scientist, Z, who does the instrumental analysis, she has a tech, K, who does the sample prep. The tech also went into the field with another of our senior scientists, S (was supposed to be me, but I got dragged into something else at the last minute and couldn't go on the collection trip). Z has just delivered to me her part of the paper, a dozen figures and tables, as well as part of the method section.

On the microbiology side, there's the other lab head (Co-PI), G, and his lab tech B. B did most of the work on their side, and has written up their methods and results and sent them to me as well.

My job is now to pull the results together and write the paper. G, B, and Z will all be heavily involved in revising the drafts, but we'll likely also give it to K and S for review. It will also go to an emeritus who will read it over and give his comments too. Finally, the ms will get approved by my boss and sent up my line management for approval. By the time we're done, at least four people will have done major revisions on the document and will each have contributed text. Three to four more will have done substantive reviews.

I will likely be first author: my grant is paying for all of this, and I'll likely have written the most text. B, Z and G will all be co-authors as well. K will likely get a co-authorship, as he's contributed a lot. If one of the other reviewers contributes a strong viewpoint or an interesting analysis, we would add them as well. The other reviewers would be thanked in the acknowledgements.

In all, the resulting publication will have five to seven authors, each with substantive contributions to the work. Upwards six or so other colleagues may see the paper prior to submission, and (possibly) provide comment. Study design took a couple of months, mostly by me and G. The experimental work has taken about three months. I expect the writing to take a month to six weeks. In all, we're talking about 500 hours of labour total, for a relatively small study. The study cost is on the order of $30k out of pocket (how much we charge the research account), with labour costs not included.
posted by bonehead at 8:31 PM on March 27 [2 favorites]


One person did all the grunt work, another person answered all of the grunt worker's questions, another person did the stats, so and so procured the money, etc.
posted by oceanjesse at 11:05 PM on March 27


Why do papers in academic journals so often have more than one author?

Because so often more than one person has done the work. If you're able to design, plan, execute, analyse, and write up enough experiments for a paper without any intellectual or other input from anyone at all then sure, you can be the only author. But this is rare, in the experimental sciences at least.

As examples, I have one paper which is just me and my three PhD supervisors. They all gave intellectual input during the project but I did all the work and all the writing (plus the project was my idea). But as a PhD student, i.e. a trainee even if an advanced one, I needed input and supervision during this process. It's very rare for a PhD student to publish without their supervisors, many places flat out won't allow it, and personally I felt that mine earned their co-author credit even if I was the only one running ELISAs at 4 am. This is as close as a single author paper as I could get during that project and there were four authors.

Whereas my other paper has nine authors (me, the 3 supervisors, five others). Again I did most of the work and all of the writing but there were specialised types of analysis I couldn't do. So there were four extra authors for that and the paper really couldn't have been done without their expertise and input. And then this was a collaborative project building on something else so I had a bunch of input and training at the start, so an extra author gets added for that too. Suddenly I'm up to quite a list of authors without even trying, and this wasn't even a particularly wide-ranging project overall.

Some places give co-authorship to anyone that contributed experiments, even students or technicians who are just doing what they are told. Other places really expect a significant intellectual input to be listed and the students and techs go in the acknowledgements. Some places allow authors for no other reason than getting the funding, other places find that unethical. People get co-authorship for providing specialised cell lines, mouse strains, or experimental drugs. Sometimes it's just politics. Collaborators get added to everything for example. Or that post-doc who the boss wants to move on gets a co-authorship to make it easier for them to find a new job. All of this is why first and last authorship can end up being really important in certain fields or situations, because everything else is fuzzy.

Many journals these days require a paragraph in the acknowledgements listing who did what work. If you want to know more about authorship find some of those journals and take a look through this section. Then it becomes clear why so many authors are needed.
posted by shelleycat at 11:57 PM on March 27


In my field (and I think most hard sciences, I don't know) the research advisor is always on the paper in addition to the actual author who did the work. So even if you are super abnormally independent in lab it's still two people on your paper.

Yup. Speaking as a grad student, I'm about to submit a manuscript wherein I designed all the methods myself, completed all the fieldwork myself (funded entirely out of my own pocket and not from my lab's funds, since my dissertation project is not connected at all to my lab's current grants), did all of the stats myself, and wrote the text myself, and my advisor is still an author on the paper because she is my advisor (her authorship represents the fact that she placed her seal-of-approval on each step of the process as I carried out the project).

But generally, you earn a spot as a co-author when you've contributed some combination of the following to the project: (1) Came up with the idea; (2) Got the money to do the project; (3) Carried out the gruntwork; (4) Did the analysis; (5) Wrote the actual text of the paper. Depending on which of these you contributed and who you're working with, you may need to have done more than one of these to qualify for co-authorship. (For instance, my advisor tends not to include people as authors solely because they carried out the gruntwork, and requires they also contribute one of the other factors; other people are more lenient in this regard.) Various journals have guidelines on what contributions they consider sufficient to justify co-authorship.
posted by pemberkins at 1:58 AM on March 28


In the corporate world any papers we produce are the result of a project with multiple people working on it. Then When the paper needs to be written, sections get handed out and the whole thing gets assembled. And thats how i'm a 8th author and will be permanently spammed by shady conferences for life.
posted by TheAdamist at 4:40 AM on March 28 [1 favorite]


In my field it's usually one person who writes the first draft of the paper (almost always the person who led the project and did the majority/plurality of the work, unless there's a significant language barrier), and then improvements and edits happen with the help of folks who already contributed to the project. Scientific writing is hard — trying to boil a complex project down to a few pages in the clearest, simplest, most honest way takes dozens of cycles through the editing process. But the credit still usually gets distributed on the basis of the work it took to create the content, not the writing.

The ordering of names in my field works roughly like this: person who did the majority/plurality of the work (intellectual and physical); person who did the second-most work; third-most; other folks who contributed code for data analysis, made samples, or built custom equipment for the experiment; advisor of non-group authors; primary investigator (the professor who leads the team and provides guidance, edits, and support). Sometimes people who you casually discussed the data/theory with get acknowledged, but they're not usually on the author list unless they significantly contributed to the project. But again, this ordering depends on the field.

So when you read a paper in my field, you'd probably start by looking at the LAST name on the paper — that tells you what group it comes out of. Depending on where you sit in your career, you'll want to have your name at the beginning or end of a lot of papers. Where those papers get published also makes a huge, huge difference when it comes to determining someone's scientific cred.
posted by you're a kitty! at 7:44 AM on March 28


Everyone who worked on the research project in whatever capacity (lab techs, field techs, data analysis, etc.) gets listed as an "author" even if they didn't write a word of the actual final paper. It's how people get credited for their work in academia -- academic CVs are so long because people list every paper they're a co-author of to show a history of the type of research they've worked on.
posted by Jacqueline at 12:00 PM on March 28


« Older In reading about codependency ...   |  Do you remember bunches of ban... Newer »

You are not logged in, either login or create an account to post comments