Did Epstein ever respond to your & Derr's response paper? Well, there are 2 ways I know of that they can help you understand a referent. There may be more.
1) Parallax 2) Expressibility If I have 1 totally opaque model of a referent, I'm limited to (2 - Expressibility), establishing what the model can and can't do and how that may be different/similar to what the referent can/can't do. The model might have more expressibility than the referent (i.e. it's phenomena space "covers" that of the referent). It might have less. Or they might overlap. Because we usually only use models because we can't fully search the referent's behavior space, we'll be more or less certain about such overlap/coverage. But regardless, that expressibility does tell us something about the referent. Even if all it does is suggest experiments we might do on the referent. If, however, I have 2 opaque models, then through parallax, we can do (2) even better. And if I have N opaque models, where some of them overlap the referent better than others, then I can bin them into equivalence classes where some equivalence classes of models do a better job than others. And that allows me to *argue* (not prove or establish or know, or whatever other strong word you might want to use) that the referent's *structure*, mechanism, generators ... are probably similar to the structure of the models in those higher fidelity equivalence classes. As N→∞, this argument by analogy gets stronger. And because these opaque models are *models*, you can develop non-opaque models that have the same (or similar) expressibility to their opaque counter parts. So, even if you can't say that the referent definitely *has* the structure of a non-opaque model with the same expressibility as a class of opaque models, that triplet relation allows you to develop fine-grained manipulation studies of the referent and ... thereby better understand the referent without ever really ever understanding the structure inside the opaque models. On 1/15/20 11:49 AM, thompnicks...@gmail.com wrote: > Could you say more about the heuristic value of obtuse models? -- ☣ uǝlƃ ============================================================ FRIAM Applied Complexity Group listserv Meets Fridays 9a-11:30 at cafe at St. John's College to unsubscribe http://redfish.com/mailman/listinfo/friam_redfish.com archives back to 2003: http://friam.471366.n2.nabble.com/ FRIAM-COMIC http://friam-comic.blogspot.com/ by Dr. Strangelove
