Systems Biology, Holism and Reductionism

A new paper, “So what do we really mean when we say that systems biology is holistic?” [1] discusses the relationship between holism, reductionism and systems biology.

Although the author does provide a historical overview of various interpretations of ‘holism’ and ‘reductionism’, the end result remains: a lack of clarity, specificity and agreement over the meaning of these terms. In addition, the term ‘systems biology’ is arguably equally nebulous.

For my part, there is perhaps nothing more useless in science than arguing between two putatively contrary ill-defined concepts (holism and reductionism), especially if the argument is in relation to how those concepts apply to a third ill-defined concept (systems biology).

Even the author states:

We may not know exactly what modern holism in systems biology is – although we can perhaps generalise that it is usually explanatory ontological antireductionism with some tendencies to epistemological antireductionism – but we do know that it is against reductionism. That might be the end of the discussion, were it not for the fact that it is not entirely clear if we know what reductionism is either.

Rosen’s work arises during the discussion, but the author appears to misstate Rosen’s view:

Although neo-reductionism has had a low profile among systems biologists and biologists in general, a new kind of holism, Relational Biology, has attracted attention, mostly among those who are dissatisfied with traditional molecular biology but also sceptical about the explanatory capabilities of modern versions of holism. Developed over some years by Robert Rosen and a small band of disciples, relational biology does not dispense with the hierarchy of the Vienna Circle, but rather inverts it. Rosen, based on some earlier similar ideas by Elsasser, claimed that physics, by virtue of its application to homogeneous molecular structure is in fact not the fundamental science, but actually a special case. Biology, as the science of the complex, is the lowest level in the layer model.

The notion of a hierarchy (and its inversion) is unhelpful with respect to understanding Rosen’s view, nor did Rosen argue that biology was more fundamental than physics. Also, “Biology, as the science of the complex” can be a somewhat confusing phrasing.

Rosen realized and proved [2] (and Louie proved in even more detail [3]) that if all the models of a given system are simulable (i.e. Turing-computable) then the analytic models of the system coincide with the synthetic models of the system. That is, all the information one can have about such systems can be found in its synthetic models alone. Moreover, the entire collection of these synthetic models can be combined into one model, which is called the largest model of the system. As an immediate result of this coinciding, the properties of such systems are entirely fractionable and can thus possess no “emergent” properties. Rosen defined this class of systems to be the class of mechanisms or simple systems.

By contrast, the class of complex systems are those systems which possess at least one nonsimulable model and, as a consequence, possess at least one analytic model which does not coincide with a synthetic model. That is, there is at least one model (i.e., analytic model) of the system which describes properties of the system for which there is no corresponding model of an assemblage of parts of the system (i.e., synthetic model). In simpler terms, in such a case the system possesses properties which cannot be described by an assemblage of the properties of its parts.

This means of classification allows for a precise definition of ’emergent property’, if one so desires.

It is important to point out, as Rosen did repeatedly, that complex systems can still possess many analytic models which coincide with synthetic models: “The failure of a system to be a mechanism does not at all mean that has no mechanical models; indeed, in some sense, these form a subcategory of the category of all its models.” If this were not the case, then all the obvious successes of reductionistic-oriented science to date would not have been possible. Thus one must be careful when stating that biology is “the science of the complex”.

As a corollary: the existence of an abundance of simple models in a system cannot, by itself, logically entail that simple models alone are exhaustive of all models of the system

The class of organisms, Rosen argued persuasively, form a subset of the class of complex systems; however, organisms are not the entire class of complex systems.

Rosen also proved that the very structure of the Newtonian paradigm (and its quantum mechanical variant) is such that it inherently treats physical systems as simple systems: that is all that this paradigm can encode (or “see”) about any system, simple or complex.


To put all this together, in the Rosennean picture:

  • The class of all natural systems can be bifurcated into two non-intersecting subsets: simple systems and complex systems.
  • The class of organisms are a subset of the class of complex systems.
  • Modern physics is founded on a paradigm which can treat systems only as simple systems.
  • Therefore, physics needs to alter/enlarge its paradigm if it is to more fully model  complex systems, such as organisms.
  • Necessarily, some of those models will be nonsimulable.
  • Likewise, systems biology will need alter/enlarge its paradigm to embrace nonsimulable models.


So, there is no “hierarchy” in the Rosennean view, nor is biology in any sense more fundamental than physics. Instead, the complex system nature of biological organisms is symptomatic of the need for more encompassing approaches in both biology and physics. As Rosen said, “Perhaps the first lesson to be learned from biology is that there are lessons to be learned from biology.” [4]



[1] Gatherer, D. 2010. “So what do we really mean when we say that systems biology is holistic?”. BMC Systems Biology. 4(22).DOI:10.1186/1752-0509-4-22

[2] Rosen, R. 1991. Life Itself: A Comprehensive Inquiry Into The Nature, Origin, and Fabrication of Life. Columbia Univ. Press

[3] Louie, A. 2009. More Than Life itself: A Synthetic Continuation in Relational Biology. Ontos-Verlag.

[4] Rosen, R. 2000. Essays on Life Itself. Columbia Univ. Press.

Bookmark the permalink.

One Response to Systems Biology, Holism and Reductionism

  1. I seem to remember (although I don’t have a copy of Life Itself to hand and am only going by my notes), that in the Prolegomena (about p.15 or shortly before) to Life Itself, Rosen does restate Elsasser’s view about physics being the special rather than the general science, so I don’t think that I do misstate Rosen in this regard. Perhaps, since you certainly do have the book, you can find this passage and confirm if I have over-egged the pudding on this one.

    As for the fatuousness of arguing between reductionism and holism, my article wasn’t intended to stoke up the debate but rather to stimulate systems biologists to think before making glib declarations of holism (which quite a few are).