Too many authors

Last year, I was moved to write this post by my amazement at the length of the author list of a paper in Science. Soon after that, I came across a different article with over 100 authors, and suddenly that list of 56 didn’t seem quite so impressive. But as I recently learned, the distinction between 50 and 100 authors turns out to be completely meaningless—the folks at CERN have destroyed the competition.

Maybe destroyed isn’t a strong enough word. Let’s say they murdered the competition, and then blew up the playing field. This paper (currently in press and open-source from Nuclear Physics B) has over 3000 authors. I didn’t get an exact count; I opened the list as a comma-delimited spreadsheet and started to calculate the number of names, but then I remembered that I have a job.

A cartoon of the ATLAS detector, 150 feet long and 300 feet underground

To understand this situation, you have to know a little bit about CERN and the LHC. CERN  is a massive laboratory in Geneva, with thousands of employees devoted to various high-energy physics experiments that take place inside particle accelerators. The most noteworthy of these particle accelerators is the Large Hadron Collider, which began successfully smashing two 3.5 TeV proton beams together in March 2010. Each proton travels at nearly light speed, and when two of them encounter each other in a head-on collision (which happens a few hundred million times per second) they release a debris trail of elementary particles that can be detected and analyzed.

Studying elementary particles in this manner has been commonplace since the first synchrotrons were built in the 1950’s, including UC Berkeley’s own beloved and recently-decommissioned Bevatron, but never before have the particles been moving so fast. The extra energy involved in the collisions gives us hope of observing the Higgs boson, the most elusive of the elementary particles and one that would explain a great deal about the fundamental workings of matter and energy, if only we could catch one.

Simulated creation of a Higgs boson

There are several detectors along the length of the LHC, including one called ATLAS (A Toroidal LHC ApparatuS). ATLAS was built by an international team of collaborators, whose goal was to measure a broad range of signals from the proton-proton collisions in the LHC.  Together, this group of over 3000 scientists claims authorship of all published results from the measurements made with this detector. ATLAS is not alone in their communal approach to publishing. The CMS collaboration (Compact Muon Solenoid), which works on another general purpose detector in the LHC, has similarly mind-boggling author lists.

On one hand, I can understand the intent to give credit where credit is due. It’s only fair that a paper based on ATLAS or CMS data should acknowledge the contributions of all the scientists who played a role in building the detector. On the other hand, such a long author list makes it impossible to identify the individuals who should be held accountable for the actual text of the publication (even the correspondence email address is to a generic ATLAS account). I am certain that far fewer than 3000 people were responsible for organizing the results and developing the conclusions of this particular paper; I would only request that those people be noted. In my experience, the purpose of an author list is to clearly identify who should receive credit or take responsibility for mistakes in the specific work being presented.

One final thought: perhaps this paper is simply a sign that we are moving away from the era where publication records are the currency of a scientist’s career. As the low-hanging fruit—scientific questions that can be answered by an individual stroke of genius—disappears, it seems that the profession of science is becoming more difficult all the time. These days, collaboration is not only beneficial, but necessary. The experiments at ATLAS could never have been done by a group of 10 or 20 people, each one angling to get a first-author paper from the project. It’s hard to tell what science of the future will look like (wiki journals?), but if the papers pouring out of the LHC right now are any indication, it will certainly be a different landscape.

P.S. You’ll thank me for abbreviating the citation below. In the PDF version of the article, the full list of authors and affiliations took up 24 pages.

ResearchBlogging.orgATLAS Collaboration (2011). Measurement of the differential cross-sections of inclusive, prompt and non-prompt J/ψ production in proton–proton collisions at sqrt(s)=7 TeV Nuclear Physics B : 10.1016/j.nuclphysb.2011.05.015

Leave a Reply


  1. Therese

    It definitely is a different system, but people are judged based on talks they give at meetings instead of papers (I have numerous friends who have worked/work at CERN)– it’s still about individual contribution. Same thing goes for instrumentation in astronomy; you go by conference proceedings (which are published), as people don’t publish papers on pure machinery. Purely collaborative science is an interesting concept, but I don’t think it will happen with the current academic hierarchy in place.

  2. Anna Goldstein

    Thanks for your insight, Therese. I have wondered before about the role of conference proceedings… it’s interesting that they are ignored in many fields while being crucially important in others.

  3. Anna Goldstein

    By the way, speaking of credit where credit is due: I should have mentioned that Anthony Fu was kind enough to point out to me both this paper and the Neanderthal paper last year. He is a constant source of links and amusement.

  4. Julia

    I stumbled across this blog and think its fantastic. I am a graduate students in Physics in the UK and wish we had something like this at my university.
    I was part of the CMS (compact muon solenoid) project as a summer student and found these massive publication lists rather irritating! Though it still depends, if you do Monte Carlo modelling and phenomenology you can still write a paper with just a few authors. It tends to be the experimental results where everyone gets acknowledged, which I can understand because the run up till you finally get to collect your data is so slow and involves a large number of people!
    But in the end I didn’t do particle physics and I am very happy now in my average of 4 author paper world .

  5. Pingback: On “New Forms of Scholarly Communication” | Shreds and Patches

  6. nick

    This is going to become more and more commonplace and author list is probably going to be replaced with institutions list.

  7. Pingback: “Encouraging” solo authorship would be a really bad idea | Scientist Sees Squirrel

  8. Tip

    There’s an easier way to figure out the number of authors. Copy and paste the list of names into a Word document, add a comma after the final name, replace all commas with some other character, and there in the number of replacements displayed in the dialog box you have the exact number of authors. It’s 3,308.