Whatever definition is used, there is no shortage of complexity in biology, medicine, or of greatest relevance for this forum, evolutionary medicine. As is made abundantly clear in the first-ever textbook of evolutionary medicine by Gluckman et al. (2010), in seeking to understand the evolutionary origins of human disease susceptibility there are profound challenges in charting genetic, developmental, and environmental influences on phenotype, and the interactions among these sources of variation can be exceptionally difficult to disentangle. Therein resides one reasonable justification for reading Melanie Mitchell’s book, Complexity: A Guided Tour (2009).
Mitchell is a lucid guide to many of the concepts that form the substance of “the sciences of complexity,” which remains a field of highly uncertain boundaries (see below). She provides accessible and clear explanations for such topics as chaotic dynamical systems, information, information processing in living systems, fractals, computation, computer modeling, networks, scaling relationsips and power laws, cellular automata, genetic algorithms, evolution, and molecular genetics. She is sensitive to the multiple senses in which most of these key terms can be used productively, as is made especially clear in chapter 7 on complexity, in which she assesses no less than nine or ten definitions of the concept while alluding to still more. The varying emphases of these different notions of complexity and their incompatibilities are effectively delineated. As Mitchell notes at the end of the chapter, there may be no one measure or definition of complexity that captures all of its “interacting dimensions.”
Mitchell does not give any indication of familiarity with the concept of polythetic categories (http://evomed.org/?p=135). Although merely noting that the concept of complexity (like the concepts of evolution, gene, information, and life) is probably best thought of as a polythetic category does not immediately eliminate the challenges associated with characterizing it, this notion can facilitate acceptance of the impossibility of devising a single definition, formula, or characterization that is superior for all applications and in all circumstances. Complexity, like life, is complex.
Another positive feature of the book is the author’s willingness to offer alternative views after presenting some of the standard findings and interpretations of complexity-related studies, the importance of which have often been exaggertated. After discussing, in chapter 10, Stephen Wolfram’s 2002 book, A New Kind of Science, and Wolfram’s assertions for the revolutionary nature of its content, Mitchell at least acknowledges the skepticism expressed by many competent commentators towards Wolfram’s claims. Similarly, after describing research on the applications of game theory to understanding cooperation in biological and social sciences, Mitchell describes studies that challenged some of the settled conclusions from earlier work. This openness to deflating the hype associated with complexity science is exceptionally refreshing in the context of earlier books by both investigators and science writers whose views on the centrality of the “sciences of complexity” sometimes seem incompletely tethered to reality. Nevertheless, Mitchell herself seems insufficiently critical with respect to some of the claims and some leading lights of the field.
Mitchell’s initial discussion of molecular biology (chapter 6) is very brief, presented at a very basic level, and contains a couple of gratuitous quotes from her graduate mentor, Douglas Hofstatder, a brilliant and creative computer scientist but not a significant contributor to any area of molecular biology. In chapter 18, her references to transcription are worded so as to leave the impression that the process of generating mRNA is carried out by RNA (as opposed to the protein, RNA polymerase, which she appropriately cites when first discussing this phenomenon in chapter 6). Also in the later chapter, Mitchell strains to explain inter-species differences solely in terms of differences in “genetic switches.” While promoters and other non-coding DNA elements that serve to regulate gene expression undoubtedly contribute importantly to the phenotypic differences among species, at this stage of relatively modest levels of insight, it is too early to completely discount the relevance of other sorts of differences, including those in amino acid sequences of proteins, in nucleotide sequences of genes encoding RNA molecules of several categories, in post-translational modifications of proteins, and in epigenetic modifications of DNA.
Chapter 5, on evolution, begins with a thoughtful history of the idea of evolution and effectively places Darwin’s contributions in perspective. This chapter also covers Mendel’s central contributions, the chief developments and thinkers responsible for the Modern Synthesis, and challenges to thereto. Although easy to follow and generally accurate, Mitchell’s treatment of evolution devotes excessive space to Gould’s critique of the Modern Synthesis.
The treatment of Stuart Kauffman’s work on gene regulation and gemone evolution (in chapter 18) reflects his relative prominence among scientists of complexity, as opposed to his considerably more modest reputation among the simple scientists whose primary interests are real organisms and real molecules. Mitchell does however acknowledge that there are many individuals who are highly critical of Kauffman’s results and his interpretations of those results. I will give one example, part of which Mitchell discusses.
Kauffman claimed (1995) that his theoretical treatment of genetic regulatory systems as Boolean networks combined with the estimate for the number of genes in the human genome predicted the number of different cell types in humans, which Mitchell gives (without a reference) as 256. The author notes that Kauffman’s calculation assumed that there were about 100,000 genes in the human genome and that the current estimate is closer to 25,000 genes. What Mitchell may not be aware of is that as the estimate for the number of genes in the human genome has dropped the number of cell types identified, at least if the immune system is at all reprsentative, has skyrocketed. Where once there were two types of T lymphocytes, there are now probably as many as ten to fifteen (and perhaps more) depending on how fine-grained the analysis of gene expression patterns.
Although claims for the revolutionary nature of the insights derived from the theory-driven study of complex systems, especially as embodied by work pursued in connection with the Santa Fe Institute, have frequently been hyperbolic, many of the ideas are both stimulating and worth knowing. Mitchell’s overview of the field, my criticisms notwithstanding, is the clearest, most accessible, and most clear-eyed single source I have encountered for becoming familiar with the disparate styles of investigation and challenging concepts that have come from looking for common principles underlying the behaviors of systems ranging from cells and computer programs to economies and whole ecosystems.
Gluckman, Peter, Beedle, Alan, and Hanson, Mark. Principles of Evolutionary Medicine. Oxford University Press, Oxford, 2009.
Mitchell, Melanie. Complexity: A Guided Tour. Oxford University Press, Oxford, 2009.
Geenspan, N. Boundaries of categories, categories of boundaries, and evolution. http://evomed.org/?p=135
Wolfram, S. A New Kind of Science. Wolfram Media, Champaign, IL, 2002.
Kauffman, S. At Home in the Universe: The Search for the Laws of Self-Organization and Complexity. Oxford University Press, New York, 1995.