The New Systems Biology in the light of EcoEvoDevo is good science:

How to view the problems and prospects of biological organization

William Wimsatt

The new systems biology is a worthwhile, perhaps marvelous, opportunity arising jointly from new systematic knowledge of the genomes of biological organisms, and their genetic phenomes—the gene activity patterns implicit in "gene chips" and other massively parallel assay technologies, to determine the detailed workings of the cell. But I will argue that the value of this new cross-disciplinary matrix or new discipline in the making is amplified by drawing more connections, not fewer, with evolutionary and developmental biology, and also with ecology.

I first review six basic properties of living systems, and how these are centrally connected with evolution and development, and why they must remain central topics of investigation for the New Systems Biology. Many were first conceptualized before Darwin, so one might claim these as independent intuitions about living systems utilized by evolutionary studies and by physiology and its descendant, NSB. True! But this does not vitiate claims about the pivotal roles of evolutionary and developmental perspectives throughout biology, including NSB. They need not always be the subject of constant attention by investigators pursuing NSB, but this does not show their irrelevance in understanding the origins and nature of living systems.

I next consider a problem with the analysis of functional organization. Delineating functional organization is and will remain a central tool in the analysis of living systems. The systems we study demand it, and our cognitive processes favor it as a good heuristic. However, looking at the processes through which it evolves (through a series of piecemeal accretions which a programmer would describe as "kluges" or "patches" by which earlier adaptive organization is modified or co-opted for new emerging or amplifying functions) raises questions about how we describe and analyze it. We do some investigation about how a system and its parts, experimentally or conceptually isolated, work and break down under some range of conditions [much smaller than the range under which they could be investigated], and from this infer its "functional wiring diagram" [as if at that point "god" gives us its "design drawings."] I argue that part of why we can do this as successfully as we do (and that question needs further examination!) is due to another property, which is both physiological and developmental, and of central evolutionary significance: the differential generative entrenchment of different parts of the causal network comprising the organism's life cycle. This differential entrenchment has consequences for selection, and thereby predicts relative evolutionary stability of structure and function, making such inferences tractable. (Thus, we can look for other closely related organisms that may be more accessible or more easily analyzed). Analysis of adaptive structures for evolutionarily significant properties like plasticity, robustness, redundancy, canalization, modularity, and entrenchment are things that NSB can contribute—in fact centrally—to evolutionary and developmental studies.

Finally I want to argue that both good science and the search for generality and robust predictability requires that we not always be in a hurry to "merge the molecular and cellular levels", except in circumstances which benefit from that. But there will always be circumstances which will not, so "Nothing-but-ism" is inappropriate. And we must also recognize that more detail is not always better, and that prediction, as a goal, should be secondary to understanding, except in technology, where sometimes prediction matters more than understanding—even be it local and conditional—if industry can secure the conditions for locally viable trend extrapolation. But then recent practices of unregulated short-term maximizing industry should lead us to consider whether this is sufficient after all!