F. Eugene Yates
The Logic of Life
Self-Organising Systems

Oxford University Press. 1993
 

Physics is about simple  beings and becomings, characterized by uniformity and generality: all electrons in the universe are alike; there are few kinds of quarks and only four basic forces—perhaps only one. 

Biology, in contrast, presents diversity and specialness of form and function, and sometimes a striking localness of distribution of its objects. Biological systems are complex  by any definition of the term. 

Physics is a strongly reductionistic science, and has prospered in that style; but conceptually biological sciences now suffer from permeation by a mechanistic reductionism in the guise of two limiting and inappropriate metaphors: 
1  the dynamic metaphor of organisms as machines and 
2  the ‘information’ metaphor, of life as a text written on DNA. 

It is here argued that both metaphors are false and destructive of conceptual advances in the fundamental understanding of complex living systems that self-organize, grow, develop, adapt, reproduce, repair and maintain form and function, age, and die. 

The rise of the sciences of complexity offers a fresh, non-reductionistic avenue toward the nature, origin, and fabrication of life. One of the new avenues is homeodynamics. 

BIOLOGICAL ORDER

Driven by discoveries in biology over the past twenty years, mathematics and the natural sciences have accepted ‘complexity’ as a technical subject. 

Old philosophical issues such as the basis for ‘emergent properties’ and the conflict between reductionistic and holistic frames of reference are being reexamined in the light of new theories about complex systems. Figures of thought brought to bear on complexity and the names of some of the leading contributors include 

general systems theory (Bertalanffy); 
computation (Turing, von Neumann); 
cybernetics (Wiener); 
chaotic dynamics (Lorenz, Ruelle, Takens, Smale, Feigenbaum); dissipative structure theory (Prigogine); 
catastrophe theory (Thom); 
synergetics (Haken); 
homeokinetics (Iberall and Soodak); 
network thermodynamics (Katchalsky, Oster, Perelson); 
relational biology (Rashevsky, Rosen); 
and information/dynamic complementarity (Pattee). 

Interest in non-linear mechanics, qualitative dynamics, general bifurcation theories, and differential topology runs strongly in most of these approaches (Yates 1993). 
Until recently the philosophy of science focused on physics and the nature of physical theories. Now biology sparks studies in the philosophy of science, because it is generally recognized that physicists attempt to explain simple systems by means of universals, whereas biologists study complex systems that are special. However, the philosophy of science from a biological perspective is limited by the scarcity of predictive theories in biology; the most compelling biological theory is contingent and historical — some version of Darwinian evolution. Furthermore, biological order is unlike order in physics or mathematics. Biological order is remarkable not for its degree, but for its specialness. It is a functional order that serves to correlate relevant biochemical and physiological events (Careri 1984); but it is difficult to formulate mathematically the condition of invariance that must be fulfilled, which can be stated broadly as the need to keep the characteristics of one species constant during all the transformations that give rise to biochemical events during development. In contrast, in the case of a crystal lattice the spatial order is best expressed by the presence of correlations among the positions of equal atoms, and this order is further characterized by a condition of invariance toward the space transformations allowed by the symmetry class of the lattice in question. In functional order the correlations must be formed among the times at which different events occur. Event correlators are required, and these may be biomembranes. 
Physicists make the Assumption of Simplicity—that in spite of the mathematical and other complications that may veil our vision, Nature is simple, both in composition of material objects and in rules for change. 
Biologists, on the other hand, take complexity as a given for the systems of interest to them. Although there is no universal agreement as to what constitutes a complex system, at the heart of the concept is some kind of non- reducibility—the behaviour we are interested in evaporates when we try to reduce the system to a simpler, better understood one (Stein 1989; Yates 1993). Furthermore, biological systems are inherently, fundamentally, and profoundly non-linear. 

There are at least three approaches to describing living systems: 
(1) structural (focus on form); 
(2) informational (focus on codes and messages); and 
(3) functional (focus on dynamics). 
Genetic chemistry unites the structural and the informational, leaving a dynamic residue unexamined. The dynamics are not Newtonian. Newtonian mechanics classically rests on models with holonomic constraints and exact differentials, and is therefore non-generic. Paradoxically, biology is more general than physics (Rosen 1991). Life depends on non-holonomic constraints, and when models are made, inexact differentials appear. 
I shall address some epistemological and technical issues that must be resolved if we are to bring the power of the natural sciences to bear on questions concerning the origins and nature of life. Rosen (1987a,b, 1991) and Howard Pattee (1982, 1988, 1989) have illuminated the way. Their contributions to our understanding of the nature of life from physical/philosophical/mathematical perspectives have recently been clearly presented by Cariani (1991). Purely computational, logic-driven strategies for finding new functions (for example, artificial intelligence, evolutionary simulation strategies, computationally based artificial life) are incapable of generating true novelty, which is the outstanding characteristic of biological evolution. They are trapped by their syntax. 

The logic of life, I suggest is not syntactic (formal), but semantic (relational); the accompanying stability of the living organism, sufficient overall to secure the advent of the next generation, is unlike that of any machine that does work (Newtonian machine) or processes information algorithmically (Turing machine). 
LIVING SYSTEMS ARE HETERARCHICALLY
AS WELL AS HIERARCHICALLY ORGANIZED.

FIG. 9.3. A hierarchical view of life. The levels are not all disjoint sets; some elements participate in more than one level (for example, tubulin is both molecule and cytoskeleton). Gaia is the fanciful vision of Earth as mega-organism (Lovelock 1988). 
 

Figure 9.3 represents a view of terrestrial life as hierarchically organized. Except for the extremes (quarks at the ‘bottom’ and Gaia at the ‘top’, with extraterrestrial life possibly beyond), the diagram is conventional. But it does not adequately express the organization of an organism. I learned from four years of servicc in the US Navy that although the organizational diagram for the command units of which I was a member showed the admirals at the top and people like me at the bottom, in fact the chief petty officers (somewhere in the middle of the vertical hierarchy diagram) and the admirals had about equal effective rank when it came to getting anything to happen. Functionally, they were side-by-side in a heterarchy. 

FIG . 9.4. A heterarchy with interacting elements (at the vertices). This projection is arbitrarily chosen to be a Schlegel polyhedron in three-space for a 24-cell heterarchy. (Reprintet with permission from Banchoff 1990, p. 118.) 

Figure 9.4 expresses this view, supposing that a living organism has (for the purposes of illustration) 24 ‘cells’ or elements interacting. They are all of equal rank. The diagrams have no top or bottom, and represent an object that can be viewed equivalently from any perspective. Such a representation seems to me to capture living systems as networks of co-operativity better than does the more conventional and limited hierarchical view of their organization (Fig. 9.3). 
 
 

WHAT IS LIFE ? 

Whatever science of complexity finally emerges, it must ultimately displace cybernetics, general systems theory, artificial intelligence, information theory, and control theory—all of which have failed to account for the stunning diversity of biological structures, functions, and trajectories. I close with an excerpt from Rosen’s very thoughtful work, Life itself (1991, pp. 279-80). 

What is life? . . . Contemporary biology gives two kinds of answers to this question. In somatic terms the answer is: Life is a machine, a purely syntactic device, a gadget, to which a reducionistic strategy may be universally applied. In evolutionary terms, on the other hand, life is what evolves; the evolutionary process itself, which takes us from gadget to gadget, is devoid of entailment, the province of history and not of science at all. 

To me, neither of these answers, either separately or together, serves to answer the question. If somatically an organism is a machine to be understood in purely syntactic, reductionistic terms, then life is only a matter of putting its fractions back together. But as we all know, it is literally not that simple . . . Evolution, entailed or not, has from the beginning concerned itself only with origin of species of life; it does not bear on life itself. There are many good reasons for wanting to be a reductionist, but unfortunately these have nothing to do with answering the question. One reason is that reduction, syntax, has since the time of Newton been identified with science itself, and that any deviation from its prescribed algorithmic progression from earlier to later, any shred of function or finality, any manifestation of semantics, is mysticism . . 

And, of course, reducionism carries with it the lure of unification; of having to know ultimately only one thing, one principle, from which everything else syntactically follows . . . No (finite) concatenation of syntactic models of an organism yields something which must be an organism. Or, to put it otherwise, every concatenation of such models can be realized by something which is not alive. We do not seem to be able to stop at any finite point in this modeling process and say that we are done. From this it is only a step to realizing that something is special about material systems whose properties can thus be syntactically exhausted at some finite point, a point at which we have reached a ‘largest model’ . . . Organisms are not in this class of systems; as we have seen, this is one way into the world of complex systems, systems that have no such largest syntactic model....A corollary is that complex systems cannot be exhauste by reductionistic fractionation either. Just as we cannot concatenate syntactic models to obtain an organism, we cannot, for the same reason , concatenate reductionistic fractions to get an organism. 

Something else is needed to characterize what is alive from what is complex. Rachevsky provided this in his idea that biology was relational, and that relational meant (as we stated it) throwing away the physics and keeping the organization. A rough analogue would be: throwing away the polypeptide and keeping the active sites. Organization in its turn inherently involves functions and their interrelations; the abandonment of fractionability, however, means that there is no kind of 1 to 1 relationship between such relational, functional organizations and the structures that realize them. These are the basic differences between organisms and mechanisms or machines. 

As I was pondering the possible relevance of chaotic dynamics to the problem of self-organization in biology (a very faddish pursuit), I came across a poem by Robert Frost, first published in 1949, entitled Pertinax. I give it in its entirety below. 

Let chaos storm! 
Let cloud shapes swarm! 
I wait for form. 



 

HOME      BOE      SAL     TEXTE