113

I arrived at the printer room this afternoon to collect my printing and I happened to notice that someone was yet to collect a printout of the gravitational waves paper that has been all over the scientific news the last few days/weeks. Of course, I was curious, so I had a peek.

It struck me that there was only one author... until I saw the asterisk indicating a full list of authors could be found at the end of the article.

There I found three full pages of authors. I have no intention of counting the exact number, but a quick estimate by word count suggests there are over 1000 authors from 133 institutions.

You always need to give credit where credit is due, that is a given. But to put it in context, there are at most 7 words of article per author. In fact, I can think of a number of problems coming from so many authors (like who do you contact if you have a question - the list is alphabetical and there doesn't appear to be a designated contact person).

What is the value of listing so many authors and why should (or shouldn't) it happen?

Update: Thanks everyone for the thoughtful answers. As much as I want to choose an answer, I honestly cannot decide between the two most upvoted, so I'm going to abstain from choosing one.

Improve this question
16
  • 28
    Maybe a better question: which author(s) would you remove, and why? My point being, the collaboration decided each of these people made enough of a contribution to be on the author list. The collaboration did the work, so it should be their decision as to who's "in".
    – Ian
    Feb 16, 2016 at 7:58
  • 21
    It may be 7 words/person on average but consider the amount of work that people have put in one way or another. Also what co-authoring means varies across fields. There are some questions here that explored that subject.
    – posdef
    Feb 16, 2016 at 8:04
  • 10
    Maybe adding someone’s name to a paper is a lot cheaper than paying for 3 years of computer programming…..
    – Ian
    Feb 16, 2016 at 16:20
  • 10
    @Ian: Exactly. I've been listed as a co-author on papers that I never saw until they were finished, because I did important parts of the work. And others where my only contribution to the writing was to proofread the result.
    – jamesqf
    Feb 16, 2016 at 18:00
  • 5
    Should change my last name to Aardvarkerson.
    – Nick T
    Feb 19, 2016 at 16:24

6 Answers 6

Reset to default
147

The "point" of 1000 authors is to recognize that 1000 people contributed to the paper at a level sufficient to be considered authors, according to the standards of that particular field.

In some fields (including high-energy physics), large numbers of authors are not unusual. See e.g. the 2015 paper estimating the mass of the Higgs boson with 5,154 authors. This kind of research often involves very large teams spanning multiple institutions. Typically papers published as a result of a collaboration are credited to all members of the team.

For more information see Report by the Working Group on Authorship in Large Scientific Collaborations in Experimental High Energy Physics.

Improve this answer
5
  • 28
    Exactly. LIGO was founded in 1992, and its operation requires compensate for geological distortions, weather... and also careful calibrations, signal processing, theoretical modelling of the events, design of the suspended mirrors, very stable lasers, reliable CCDs... The moment you have a few people working on each topic for such a long period of time, you hit the thousands easily.
    – Davidmh
    Feb 16, 2016 at 8:59
  • 10
    @Davidmh Of course, each of those aspects of experimental design is worth several papers in its own right, and clearly the high-level paper can never hope to describe experimental design in sufficient detail necessary to reproduce. So it would be quite possible for a couple of thousand authors to write a couple of hundred papers together, which are then all cited in the high level paper. That's not the approach in fundamental physics, but it could be.
    – gerrit
    Feb 16, 2016 at 11:53
  • 3
    It is also maybe relevant to mention that not all the authors contribute with actual text in the article. Many of them are doing measurements, building/designing equipment, analyzing data etc. So there is maybe just a handful of people, maybe even just one, that actually writes the paper.
    – Gumeo
    Feb 16, 2016 at 13:56
  • 14
    It makes sense that the kinds of questions that stump the brightest minds for nearly century take teams of 1,000s of people to solve.
    – corsiKa
    Feb 16, 2016 at 18:03
  • 3
    Also, "building/designing equipment" in HEP is a HUGE endeavor.
    – xdhmoore
    Feb 19, 2016 at 4:33
81

It's a high-profile example for something that various people on this Stack Exchange have been preaching for a a long time - authorship norms differ from field to field. This is essentially just a specific case of the Academia.SE mantra of "academia varies more than you think it does".

Your interpretation of "author" seems to be mostly in the strict word sense, as in "a person that co-wrote the paper". This is indeed the norm in many fields, but as you say, in high-energy physics, authorship really has little to do with "authoring the manuscript", and is more related to "involvement with the research that led to the paper". I would speculate that in an research attempt such as the LIGO one, the effort of writing the actual paper seems miniscule in contrast to the work that went into the research, so the question whether somebody contributed words to the manuscript or not will not seem like a particularly relevant one for acknowledging hers or his contribution.

Note that high-energy physics is not the only field with authorship norms that may seem strange to outsiders. In many experimental fields, the lab head / PI is customary the last author on any paper from her/his lab, whether (s)he is involved with the work or not. In some fields, authors are ordered by perceived contribution, in others more or less by seniority, while others don't order at all or do so alphabetically.

why should (or shouldn't) it happen?

The important thing to keep in mind is that it is after all just a norm. Having a long list of co-authors with one or a few designated PIs, project leaders, or contact authors really is not different in practice than having a short list of authors and a long section "Contributions by: A, B, C", as long as everybody has the same understanding. Coming from the outside and telling the physicists that they "are doing it wrong" because their notion of co-authorship isn't the same as in your field seems misguided to me. If it's working for them, let them continue doing it like that.

Improve this answer
9
  • 7
    Looks like they did all sign off on the MS, at least in some capacity: "Then came writing the paper. This involved getting 1,000 researchers to agree on every detail, and took some 5,000 e-mails, says LIGO’s chief detector scientist..." nature.com/news/…
    – Andrew is gone
    Feb 16, 2016 at 13:39
  • 1
    @Andrew My experience has been the same: I sent email to an awful lot of people and teams and got an awful lot of emails back.
    – jakebeal
    Feb 16, 2016 at 13:55
  • 2
    @Andrew Physicist friends told me similar stories, but I guess there is still a difference between 1000 people agreeing with something and/or suggesting small textual improvements and 1000 people actually writing something together, as indicated by the OP's 7 words/person on average.
    – xLeitix
    Feb 16, 2016 at 14:11
  • 7
    "In an research attempt such as the LIGO one, the effort of writing the actual paper seems miniscule in contrast to the work that went into the research" That is exactly right.
    – Chris Mueller
    Feb 16, 2016 at 21:08
  • 3
    It may be important to differentiate between the academics and the technicians as well. The academic folk work on the research with the hope that their name will be on the paper (i.e. - not for profit). The technicians work on the project as their job (they're paid). So, presumably, the ~1000 authors all worked on the project expecting their work would be recompensed with recognition in this way since they weren't monetarily compensated.
    – Ian
    Feb 17, 2016 at 22:22
44

In addition to @ff524's good answer, I can speak to this from personal experience. I have a paper coming out shortly on which I have more than 600 co-authors. The reason that there are so many is because of the nature of the experiment that we conducted---in this case, determining the reproducibility of some critical measurements of cellular behavior.

As such, there were essentially three tiers of authors:

  1. The key organizers who did the bulk of the work running the study and analyzing the data
  2. Secondary contributors to the study organization, execution, and analysis
  3. People from more than 100 teams around the world who actually gathered all of the data

The first two tiers of authors are typical of any paper; for the third tier, in the customs of biology publications, it would generally be quite inappropriate to exclude any person who actually gathered experimental data for a paper. Thus, we have an enormous number of authors on the paper. Many modern scientific experiments have similar scope, in which the contributions of hundreds or thousands of people are required by the simple scale of the project.

Critically, however, if you read the "Author Contributions" section of the paper, which more and more journals are including now, it will tell you these facts about who contributed what. There is some ongoing advocacy for such explanations of contributions eventually simply replacing author lists, since author lists are not very explanatory, essentially giving movie credits for scientific papers. Personally, I think that such an idea is a very good direction to go, though I think that making the transition will not be fast or simple.

Improve this answer
6
  • 9
    "Personally, I think that such an idea is a very good direction to go, though I think that making the transition will not be fast or simple." Nor painless. I suspect many senior people will have a miserable time fairly assessing their actual contributions to some papers carrying their name :D
    – xLeitix
    Feb 16, 2016 at 13:24
  • Many modern scientific experiments have similar scope, in which the contributions of hundreds or thousands of people are required by the simple scale of the project. - Indeed. One of my professors was involved in a research using GPU to sequence DNA, with several teams all over the globe.
    – Mindwin Remember Monica
    Feb 16, 2016 at 13:35
  • 11
    @xLeitix Actually, I think it there is are quite straightforward imports from movies, such as "producer" and "director," that would describe the roles of many senior people very well. "Producer" is an important and valuable role, and the current system often pushes people in this type of role to pretend they have a more technical role, rather than honestly acknowledging the value of finance and management.
    – jakebeal
    Feb 16, 2016 at 13:52
  • 2
    @jakebeal I agree, but quite some cultural change needs to happen before this will become a reality.
    – xLeitix
    Feb 16, 2016 at 14:15
  • 1
    The metadata challenge to citation management in a more-robust, role-aware authorship/attribution system is likely tremendous.
    – hBy2Py
    Feb 18, 2016 at 16:44
38

As a member of the LIGO collaboration I will say that I am proud to be an author on this paper. Everyone on the author list contributed significantly to the incredibly difficult task of building an instrument capable of detecting length fluctuations at the level of 10-19 meters.

You mention 7 words per author, but you are forgetting about the 100's of other papers put out by our collaboration. Among them are some of the biggest advances in the science of precision measurement in the last few decades (understanding and demonstrating squeezed quantum states is probably the biggest). All of our big achievements took large teams of people working for many years, and all of those people deserve to be recognized.

Improve this answer
6
  • 2
    Do you see the 100s of other papers as major parts of the recent one? I'm curious where/how you draw the line between contributing to previously published work and and contributing the recent LIGO paper. (Thanks so much for your work and answering the question!)
    – Ric
    Feb 16, 2016 at 21:15
  • 9
    @Ric We have a few strict rules about what type of papers need to have the full author list which were voted on by voting members of the collaboration and are amended from time to time. The biggest one is that any analysis which uses a significant amount of the dark port signal (where the GWs show up) must have the full author list.
    – Chris Mueller
    Feb 16, 2016 at 21:26
  • 6
    @Ric As for the 100's of other papers being part of the recent one, I'll say this: Any team trying to replicate what we accomplished from scratch would need to understand most of what is in those other papers.
    – Chris Mueller
    Feb 16, 2016 at 21:27
  • 3
    @ChrisMueller Firstly I apologise if my question was a bit discourteous - I didn't mean to diminish anyone's contribution - but I've never seen an author this quite like it (until I saw the links in ff524's answer). Secondly, I'd be very interested to know how on earth you managed coordinate that many people! When talking about groups of scientists I've come across the phrase "herding cats". I think I'll follow this up with another question.
    – Bamboo
    Feb 17, 2016 at 1:09
  • 1
    @Phil I didn't take any offense from your question; I would have wondered the same thing if I hadn't been part of a large collaboration. Management in LIGO can be a bit chaotic at times because, as you suggest, it is like herding cats. The most important part of organization almost happens naturally though; the huge goal is broken down into manageable (and interesting in their own right) goals, and then a group or individual interested in the smaller goal steps forward and runs with it.
    – Chris Mueller
    Feb 17, 2016 at 12:54
24

There's books, and there's movies. A book can be written by one guy in complete isolation. A movie--even if it's just a movie based on a book that one guy wrote--generally requires the collaboration of hundreds.

There are some things in academia that are really just worked out by one person, and some things that take large scale collaboration. In each case, an honest attempt should be made to give credit where credit is due.

Improve this answer
4

This paper is unusually important and comes from an unusual project. Wikipedia reports LIGO is the largest and most ambitious project ever funded by the NSF (see references there).

Seeking the Higgs particle at CERN, for example, allowed many side experiments finding other kinds of results. The LIGO project really had just one goal since 1992. Chris Mueller mentions many papers on progress in designing instruments, and that is true, but they were not exactly partial results. The desired result was to observe gravity waves,and that took over 30 years to do.

So when the thrilling result did come in, it is natural they wanted to credit a lot of people who had invested time and effort in a project which many skeptics had doubted would find anything.

Improve this answer

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .