Authorship inclusion

Who should be on the author list of a paper? There have been some claims that some people put students on papers not because of their significant contributions but because of the need to pad their resumes.

For patents, the patent office requires a substantial contribution for someone to be named an inventor, and adding someone who does not deserve to be in the inventor list can be grounds for dismissal of a patent. There is a general sense in the CS community that authors should make substantial contributions, but there is little sense for how that contribution should be defined.

Here are some questions that have arisen:

• At what point does having "too many" authors dilute a paper to the point where those who are included do not benefit professionally from the inclusion?
• Should there be explicit guidelines about what constitutes a contribution?
• Should authors have to check a box affirming that they made an individual contribution?

What is the threshold for significance? One could imagine something developed over the period of years, in which various parties have spent anywhere from a few hours to many months or years of effort in contributing to the overall effort. Is there a threshold (5%? 10?%) of the overall effort that someone should contribute in order to qualify as a "coauthor"?

This subject was the topic of active email discussion prior to WOWCS and will no doubt be fleshed out.

-- FredDouglis - 01 Apr 2008

While I was reading the discussion of authorship I was reminded of a lunch conversation I had today with a bunch of physicists. It seems that somebody's son at MIT made an agreement with another physics grad student that they would take turns being first author on papers, and flipped a coin to start the alternation. The other guy won, so he was first author on "here's how to build the instruments for our experiment." The son got to be first author on the results paper, which has caused a huge stir. So the son has earned the notoriety for something that both contributed to about equally.

Geoff Kuenning geoff@cs.hmc.edu http://www.cs.hmc.edu/~geoff/

Similarly, I just interviewed someone who had a paper where I noticed that he was 1st author, pulled out of alphabetical order, yet the paper ended with a tiny section claiming that the first two authors contributed equally. Struck me as very odd. I think at the least the community should have standards that say something like "authorship is alphabetic and when it is not alphabetic order it implies a precedence relation" (which I think most people currently use) or a novel but equally valid standard like "authorship shall be in random order" or, as in the case you cite, rotated. But when some people rotate and others don't, it gets awfully confusing.

-- FredDouglis - 02 May 2008