I was amused by the synchronicity of my going straight from a discussion of YAGNI on the XP mailing list to reading the following in Politics of Nature (pp. 195–196; emphasis in original):

As soon as we agree to differentiate the past from the future no longer through detachment but through reattachment, political ecology begins to profit differently from the passage of time. Unlike the other forms of historicity that preceded it, it can confide the questions it has been unable to answer today to the restarting, tomorrow, of the process of composition. It need not claim that the things it does not know at time t are nonexistent, irrational, and definitively outdated, but only that they are provisionally excluded beings on the path toward appeal, and that it will find these beings in any event on its way to t + 1, since it will never be rid of them. In other words, it no longer uses any of the three labels that the moderns have always used up to now to characterize their development: the struggle against archaism, the front of modernization, the utopia of a radiant future. It is required to devote itself to a meticulous triage of the possible worlds, of the cosmograms, always to be begun anew. Irreversibility has changed direction: it no longer finds itself in the abolished past, but in the future to be recommenced.

Let us retain from the sciences the word “experiment,” to characterize the movement through which every collective passes in this way from a past state to a future state, from good sense to common sense. Public life has striven up to now to imitate Science and to await the salvation of reason: Why would it not try to imitate the sciences a bit by borrowing the experimentation that is incontestably their greatest invention? An experiment, as etymology attests rather well, consists in “passing through” a trial and “coming out of it” in order to draw its lessons. It thus offers an intermediary between knowledge and ignorance. It defines itself not by the knowledge that is available at the start, but by the quality of the learning curve that has made it possible to pass through a trial and to know a little more about it. Experiments, as any researcher worthy of the name knows quite well, are difficult, uncertain, risky, and never allow recourse to reliable witnesses who would be available from a catalog, as it were. They can fail; they are difficult to reproduce; they depend on instruments. A bad experiment is not one that fails, but one from which the researcher has drawn no lesson that will help prepare the next experiment. A good experiment is not one that offers some definitive knowledge, but one that has allowed the researcher to trace the critical path along which it will be necessary to pass so that the following iteration will not be carried out in vain.

Post Revisions: