Lambda the Ultimate

I'm reminded of the thesis of a recent pop science book on developmental psychology The Scientist in the Crib, which argues that there are close analogies betweeen the ways that scientists form and test theories, and the way infants build their picture of the world. Putting two and two together we conclude child acquistion of social skills must be stopped!

Seriously,you may be right that ethical questions are going to arise where academia meets open source, but I don't think there is much to be learnt from this case; as lawyers put it,bad cases make bad precedents.

The argument in the paper that the entry links sounds fishy to me, but I think a key distinction is the following: studying *programs* vs. studying *programmers*. If you studied a set of programs, each of which, incidentally, had a reasonable number of people involved in them, there wouldn't be as much of an argument that the work you're studying is collaborative and doesn't necessarily reflect on the individual programmers. However, if you did a study where took a set of programmers, and for each of them, analysed their contributions to various programms, then issues of identifiability arise. In short, there's a question here as to what extent, given a specific proposed study, the programmers can be construed as *subjects* in that study.

There's also a public knowledge issue involved; one can argue that the materials being used (open source code) is public knowledge, and moreover, that the author has made it voluntarily so. An analogy that could be drawn is that publishing source code is like publishing an article in a journal, and the extent of the author's protections are the same in each case. If somebody messes up and unjustifiedly says something that does demonstrable harm to a software author, there's laws for these cases.

My feeling however is that there is still going to be some sort of gray line somewhere here, having to do with experimental designs that rely on statistics that correlate measures to identifiable individuals, even if the data involved is public knowledge.

The basic argument is that anyone doing a case study of open source software is doing an experiment on humans, and either needs their permission or needs to make sure that they can't be harmed by the study. Writing a paper that points out flaws or weaknesses in a system can harm the authors of the system, because it casts doubts on their ability and might prevent them from getting a job. Although theatre and film critics attribute flaws to particular people, they have no alternative. However, there is an alternative when we are studying open source projects, because we can describe the projects in such a vague way that the original system is not identifiable.

This is a false analogy. If not disclosing the original system is an "alternative" for such studies, then the equivalent "alternative" for critics is to criticize without disclosing who and what they are criticizing. Clearly this is, in fact, no alternative at all. It is like claiming a proposition is a theorem but refusing to exhibit your proof.

Accountability is an important component of all social activities, and in scientific and professional fields judging the work of peers is not only ethical, it's a moral obligation. The issue really is whether one "casts doubt" unfairly. After all, we don't want incompetent engineers building bridges, or incompetent leaders governing nations. (cough)

A paper that exhibits a flaw in X's system does not necessarily cast doubt on X's competence; if the paper is written impartially and with care to avoid value judgements, as scientific papers should be, then it is the reader not the author who judges X. Consequently, if there is some sort of ethical problem involved in judging others, then it will be one the reader not the author needs to struggle with.

In medical research, the privacy of the participants is paramount. I gather that medical researchers do not divulge the names and vital statistics of their participants - though they will hand over the raw data to panels of experts for review.

Of course, the basic premise of the scientific method is that research results can be duplicated. So even if the raw data is not given, the methodology provided is supposed to be available for others to test.