One of the scientist key policy is always to refer to people who did the first work (as it is pointed out by the "hard blogging scientists" manifest). It is due to the fact that researchers want to share free science with everybody (at least at little cost), and that recognition is a form of remuneration (in a similar way, Eric S. Raymond explain such mechanism for hackers, in his essay "The Cathedral and the Bazaar").

Recently is increasing a (rather small, but interesting) controversy about the authorship of the Ant Colony Optimization (ACO) idea.

The "orthodox" seminal paper about ACO is a technical report, written by Dorigo, Maniezzo and Colorni, submitted in 1991. As technical report are not a rigourous publication, perhaps a more pertinent citation would be the corresponding paper published in the proceedings of the First European Conference on Artificial Life, in 1992, or the Dorigo's PhD thesis, completed in 1991. Dorigo et al. refers to a work by Denebourg, Pasteels and Verhaeghe ("Probabilistic Behaviour in Ants: a Strategy of Errors?", published in 1983), as the key paper that give them the ACO idea.

Nowadays, M. Dorigo has become the reference researcher on ACO, for instance, the ANTS conference is helded every two years in his lab, in Brussels.

Where is the controversy ?

It seems that there is an oldest paper talking about using ant colonies behavior to design algorithms. This paper, written by Manderick and Moyson (The collective behavior of ants: An example of self-organization in massive parallelism) and published in 1988 in the proceedings of the AAAI Spring Symposium on Parallel Models of Intelligence, also refers to the work by Deneubourg et al., but is not cited by Dorigo et al..

Well, this does not seems to be a big problem, everyone can miss a paper in the huge amount of ideas published every days.

I first read the question of the authorship of ACO in the "habilitation" thesis of Pierre Collet (published in 2004), and it recently rised up with the modification of the ACO article on Wikipedia, the 11th of august, by an anonymous editor (with the nick "MoyMan"), who add the Manderick paper to the references. The 15th of august, Daniel Angus posted an email on the ACO mailing-list (and a post on his laboratory blog), asking "why this paper has not received more attention from the field?". The same day, a contributor called "MDorigo" delete the reference on Wikipedia, in order to give M. Dorigo the authorship. But a few hours later, the first addition was reintroduced.

In my opinion, M. Dorigo just saw the addition as a kind of vandalism, trying to introduce a fake reference, as it sometimes occurs on Wikipedia, and did a revert without verifying the validity of the source. Note that M. Dorigo did not enter an edit war after the revert, which prove his bona fide.

The funniest (well, I should say "sadiest") about this story is that people are already talking about it around me... Why such gossiping? I think this is because M. Dorigo has a strong reputation. I personaly think that such reputation is incontrovertible when you begin to be known for your work. And M. Dorigo is certainly known for his work, as it is a very good one, acknowledge by his peers.

Finally, let say that the controversy does'nt have any basis, as M. Dorigo point out in a message on the ACO mailing-list, replying to the question of D. Angus. Firstly, if the Moyson and Manderick paper is about using ant colonies behavior in the AI field (about using the theory of complex dynamical systems to design massively parallel systems, to be sharply exact [1]), it is not about using it to tackle optimization problem. Secondly, M. Dorigo cite the papers that gave him the key idea, and Moyson & Manderick one is not concerned, even if it is interesting. And finally, Bernard Manderick read the paper before its submission, and didn't felt necessary to tell to M. Dorigo that his paper should be cited.

This stupid story (when people are gossiping without having all the informations, it is always stupid) is about an insignifiant event. Only ten persons in the world will feel concerned. But I find it interesting for the following reasons:

  • this show that scientist can have an ego, and are finally as subjective as "humans": when a researcher gain popularity, he also gain criticisms, despite objectivity;
  • it illustrate how much a reputation is decisive when judging other's work;
  • and how much transparency is necessary;
  • finally, this show that, despite criticisms, Wikipedia can be a place for researchers to exchange informations, even if it can be in a rather rough manner.

Notes

[1] Here is the abstract of the Moyson & Manderick paper:
Massive parallelism needs a new methodology. The theory of complex dynamical systems provides such a methodology in those cases where massively parallel systems can be modelled as complex dynamical systems.
In particular for AI, complex dynamical systems are intersting in another respect: they have the ability to self-organize. Moreover, this self-organizational behavior can be adaptive. Important is that the emergence of self-organization is not programmed or a consequence of external instructions but results from local interactions at the microscopic level and the interplay between the system and its environment.
The new methodology and the emergence of self-organization are illustrated by the adaptive response of ant colonies to their environment. This behavior is simulated by an intrinsically parallel algorithm. Also, a mathematical model of the behavior is proposed and used to fix the parameters in the simulation. For these parameter values, self-organization is observed in the ant colony.