The politics of the author list

In lecture on November 12, 2012 at 12:00

Most published scientific articles display two or more authors. This does not necessarily mean that more than one person was involved in writing the article. The “author list” of scientific articles, like marriage in traditional cultures, is usually decided by unwritten political rules rather than love and the personal preferences of the people involved. Here’s an overview.

The first candidates to the author list are the following:

  • The instigator is the person who first proposes the main idea presented in the article, or who provides the “spark” that triggers the start of the authoring process. There is typically only one person in this case.
  • The principal investigator is the person who develops the initial idea into a fully-fledged scientific contribution, usually in the form of implementation, formal proof, prototype, etc., and subsequently confirms via evaluation whether the contribution is ready to be published. There may be (and usually are) multiple principal investigators.
  • The writer is the person who writes the text of the article. The role of the writer is to crystallize thoughts into words that make sense to outsiders to the work, and to shape the overall structure of the argument. There may be multiple writers.
  • The editor is the person who drives the authoring process. The role of editor is often played by the writer(s) directly for small articles, but sometimes another person takes the responsibility. This is often necessary because the authoring process is like a mini-project that needs to be managed: the ideas need to be collected, analysed, filtered; the document needs to be structured, formatted, typeset; the mistakes must be corrected. When more than one person is involved, they need to be coordinated and the editor takes this role.

The “owners” of the article who have the final word on the author list and eventual submission are the principal investigators and the writers. They should appear first in the author list. If there is more than one person playing these roles, then the order of first authors is decided according to another rules described below. The editor(s) then follow in the list, and finally the instigator(s). The internal order for editors and instigators is decided by the article owners.

Next to these main roles, additional people may be mentioned in the author list, up to the discretion of the article owners:

  • The reviewers of working drafts of the article. Reviewers intervene during the authoring process by reading early versions of the article and making comments about its readability, stylistic or linguistic issues, the validity of the argument, or by suggesting potential additional arguments. Reviewers should be at least thoroughly and gratefully thanked in a so-called “acknowledgements” section at the end of the article. In addition to acknowledgements, the article owner(s) may choose to list some reviewers just before or after the editor(s) in the author list if they consider their contributions were essential.
  • The scientific peers of the authors. Often, the work reported in an article is realized in the context of a larger research project supported by colleagues, work partners or peers in the scientific field. Even though these related people may not be involved in the authoring process, it is sometimes desirable to associate them to the article by means of the author list, as a way to signal to outsiders the structure of the peer network. This artificial grouping of people who are not directly contributors of a scientific work is akin to the role of traditional marriage to form family alliances. The position of the additional author in the author list should then be decided by the article owner(s) depending on the value they expect to receive back from the alliance: first, second or last position for a strong alliance, middle position for a weak alliance.

Finally, whenever there are multiple people who play the role of principal investigator and writer, their order as first authors is decided as follows:

  • if the individuals involved mutually agree by consensus to an order, that order prevails;
  • otherwise, they should refer to a third party whom they consider senior and impartial, to arbitrate.

These are the rules that decide the author list of most scientific articles.


Of course, as with all rules there are exceptions. The two main exceptions are as follows.

A special category is formed by scientific articles that serve as summary or overview of large projects, when many people are involved. “Many” here is meant to mean more than 5, and numbers up to 20-50 are possible. These projects typically produce a lot of scientific output, described in multiple articles that each detail one specific aspect of the project. At the very end of the project, it is then customary to provide an overview of the entire work, with citations to the individual articles that provide details. This overview is itself an article. The author list in such an overview article is usually composed of the names of all the people involved in the project, in alphabetical order, with the “project leader” either as first or last author depending on whether the leader actually contributed research (yes: first author; no: last author).

The other exception is formed by articles where the author list is decided by managerial fiat or plain bullying. This occurs either when the organization where the article owner(s) work(s) is hierarchical and a superior decides arbitrarily to change the rules, or when a colleague uses blackmail or emotional pressure to force their ways. Although this may be allowed by the particular organization, and actually occurs in practice for some articles, it is largely frowned upon in the scientific community. Moreover, the person who gives themselves a more prestigious position in the author list in this way is often eventually recognized as a fraud and quickly loses scientific credibility. This form of peer pressure ensures that this exception does not occur too often.


After I phrased the explanation above, a question came up: how does a group decide who plays which role, and how to prevent undue attributions? Indeed, writing an article often works like a Sunday barbecue party: one or two friends start organizing, and then by word of mouth people come by with their meat and expect to get a turn at the grill. If the people involved are friends or friendly acquaintances, and there is enough coal, then all is well and a merry partying ensues. But there is such a thing as a crowded barbecue and party crashers.

In an ideal world, a scientific article starts with the instigator and principal investigators inviting people to join them in the writing enterprise. They co-opt a writer, and from this point the article owners can pick and choose contributors to help them. Again, in an ideal world all the contributors who were explicitly picked actually contribute work and make the article better.

In practice, three situations are likely to occur.

The first situation occurs when one of the key authors turns out to be incompetent while the work is being done. Either their contributions contain too many errors, or they waste the time of their colleagues with fruitless arguments, or they play “busy” while not actually contributing anything significant. In this situation, the other authors should first consider whether they still want to give a place to this person in the author list as per the “alliance” rule described above. This is possible, but it has a cost. It is possible or sometimes likely that the slacker will eventually get a bad reputation in the scientific field. In this case, all the scientific articles where this person was listed as a co-author will transitively be tainted by his bad reputation, too. Is is thus often advisable to not use the “alliance” rule with slackers and simply demote them from the author list.

The second situation occurs when the key authors disagree on the structure or argument of the article. They may have valid but different views as to the direction of the research and the phrasing of the argument. When this occurs, it is safer to either abort the authoring process or split the article into two different articles. Attempting to reach consensus is difficult in this case, and more often than not results in an article which is unclear and unfocused. Instead of wasting time in producing an article likely to be eventually rejected, the time of the authors would be better invested in starting in a new direction without the disagreement.

The third situation occurs when a person wishes to “jump on the bandwagon,” ie. wishes to promote themselves as a key author after the authoring process is already on its way. This often occurs when the onlooker, often a peer scientist, recognizes the potential future success of a scientific publication and desires to obtain a share of the resulting fame, although they were neither instigator, writer or principal investigator. This form of opportunistic extension of the article authorship can be further separated into two sub-cases.

In the first, “good” case, the candidate co-author first signals their desire to participate to the article owner(s), together with a concrete proposal as to how they wish to contribute to the work. Based on their credibility and the quality of this proposal, the article owner(s) decide whether they accept the extra help and explicitly says so. Collaboration can then start (or not, depending on the decision).

In the second, “bad” case, the candidate co-author simply starts acting as if they were co-opted already and starts to contribute directly. This means they start to push either text, if they want to play the role of writer, scientific ideas/results, if they want to play the role of investigator, or comments and corrections, if they want to play the role of reviewer. This puts the existing collaborators in a difficult position: if they simply refuse the contribution of the jumper, it will look as if they are foolishly wasting a valuable opportunity; if they accept the contribution, the jumper will have effectively forced his way into the enterprise. This behaviour can thus be considered as a form of bullying, and should be denounced so that the bully understands their behaviour is not acceptable.


In my (short) career as an academic writer, I have seen all these situation occur. I also know that most newcomers have trouble understanding and playing with these rules, who are often left unwritten. Quite often, “beginners” will unknowingly let themselves be bullied by a more senior colleague who will unfairly reap the fruit of their work by tweaking the author list against the rules. I hope that my summary in this post will help prevent this situation.

  1. Are the authors of CS papers always listed alphabetically (and so there’s no particular significance for the “first” author, as opposed to most of the other sciences)?

  2. No, am alphabetical order is even actually unusual.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: