I agree revisiting Rosen with fresh eyes could be helpful. What I take from his work may be quite different from what others have, I think. I recently reread some of his "Essays on Life Itself", finding mostly his challenging questions for mathematical theory, and not the implications for a logic of ecosystem behavior I now see following from them. I also now can't find where he discusses his own interpretation of the "Rosen diagram" for the relation between science and nature, so I don't have in his own words what it meant to him.

To me one of his main challenges to mathematics was his observation that mathematical science has only been able to master convergent sequences, but nature seems to mostly rely on divergent ones (open ended and undetermined patterns rather than closed ones). It's certainly true, and as profound and deep a scientific question as one might find. What it led me to was an important general case in point, observing that growing systems in nature initially display incomplete divergent developmental patterns, that can only resolve with the addition of new patterns, coming from interaction with other levels of internal or external pattern environments.

My interest in his operational diagram for "how science works" started from how it looks to be inconsistent with the ruling scientific theory of how nature works. It shows the relation between science and nature as associating "causation in natural systems" with the "formal implications of science", mediated by the scientific practices of "encoding and decoding" what we find. One implication I took from that is that for science to have anything to study it really must acknowledge that the causations of nature need to be considered as having their own meanings and existence. Another implication I found in it is that nature is then implicitly the common subject of study for all the sciences, so multiple sciences could compare notes to learn from each other, as it were, to find how differing paradigms of interpretation deal with same subjects.

Both of those seem quite antithetical to the modern philosophy of science offered by Popper and Bohr, etc, holding that science can only study data and represent nature as the equations that statistical analysis converges on, and so imply statistically that the universe works by too. As I see it, the Pattern Language transformation of the sciences (my sense of PLAST), builds on that interpretation of Rosen, for addressing the problem that different paradigms of inquiry find different design patterns. One would like that to open up new ways for observers of natural design patterns to compare notes, and start untangling the great mess created by some hundreds of years of quite exclusionary reductions of nature to convergent mathematical approximations.

Could that even catch on as an idea, though? At this point I'm not at all sure it's inevitable anyway, as it seems so immediately disdained and discarded by the social groups that would need to support it. Then I also wonder if any of this would be of interest to Rosen? I'm not quite sure about that either. ;-)