Hacking and Franklin on the Functional Complexity of Evidence

After posting my paper here, in the last few days, I've just happened to come across two fabulous statements related to my position. Of course, just when you start to think you're doing something a little bit original, you come across all kinds of people saying basically the same thing.

Ian Hacking, on the first page of the monumental "Experimentation and Scientific Realism":

Experiments, the philosophers say, are of value only when they test theory. . . So we lack even a terminology to describe the many varied roles of experiment.  (Hacking 1982, p. 71)

And Allan Franklin, on the first page of his Selectivity and Discord:

Experiment plays many roles in science.  One of its important roles is to test theories and provide the basis for scientific knowledge.  It can also call for a new theory. . . Experiment can provide hints about the structure or mathematical form of a theory, and it can provide evidence for the existence of the entities involved in our theory. . . it may also have a life of its own, independent of theory: Scientists may investigate a phenomenon just because it looks interesting. Such experiments may provide evidence for future theories to explain. (Franklin 2002, p. 1)

It is a nice surprise to find myself in such good company.  The aim of my paper, of course, is to try to provide a coherent picture of and some terminology for the various roles of evidence.  One of the points that I make in the paper, which I'm not sure Hacking or Franklin would accept, is that there is a useful (functional) distinction to be drawn between observational and experimental evidence.  I suspect they might even say that I leave some roles out of my picture.