In the previous post I described the debate among evolutionary biologists over the so-called adaptive hypothesis. Some biologists believe that natural selection has the power to drive evolution in adaptive directions, and that most changes that we observe in organisms are there because they confer some adaptive benefit. Other biologists believe that most of the changes we see in organisms over time are due to neutral, non-adaptive processes.
You don’t need to take my word for the existence of this debate. Michael Lynch, an eminent evolutionary biologist, lays out the case against the power of natural selection in a paper called “The Frailty of the Adaptive Hypothesis,” 1 published a few years ago for an evolutionary symposium. In it he argues that natural selection is neither a necessary or sufficient explanation for organismal complexity. Rather, he contends that many of the genomic complexities of multicellular organisms are the result of the passive non-adaptive forces that I outlined in my previous post.
For example, mobile genetic elements tend to proliferate, inserting themselves more or less at random into new chromosomal positions, leading to what Lynch calls genomic bloat. In bacterial species with large populations and rapid generation time, this excess baggage is rapidly purified by natural selection. In contrast, among multicellular organisms, with their small effective population sizes and longer generation times, natural selection is usually not strong enough to prevent the accumulation of neutral or weakly harmful insertional mutations.
However, when it comes to answering how organisms deal with the continual genetic disruption such insertional mutagenesis would produce, Lynch has no answers. He offers no explanation of how non-adaptive forces can produce the functional genomic and organismal complexity we observe in modern species.
Lynch believes that evolution is the result of the four forces he lists, mutation, recombination, drift, and natural selection, with the first three non-adaptive forces having a major shaping role. He must accept that natural selection is enough to generate coordinated functional complexity, even in the face of these non-adaptive processes, because after all, what else is there?
Yet it’s clear that other evolutionary biologists are beginning to recognize that the problem is larger than Lynch realizes.2 They understand they are up against some hard limits — everything from how many mutations can be required before a new selectable function is achieved 34, to how much time is available to produce the highly divergent body plans of the Cambrian explosion 56. They continue to come forward with new proposals for how to generate innovation by purely naturalistic means 7. But all these new ideas suffer from the same flaw — they are undirected, stochastic processes themselves, and likely to suffer from the same failings as the four forces of Michael Lynch.
The problem is clear when we consider the limit mentioned above — how many independent mutations are within reach of an evolutionary search. Recent papers have suggested that anything beyond two mutations may be impossible unless special scenarios are introduced 34. Yet a single enzymatic innovation can require many changes and still result in a very poorly functioning enzyme 8. So how do we account for the huge diversity of proteins existing in nature — over a thousand distinct protein folds that can be subdivided into more than 3400 distinct families of similar proteins, according to the most recent count 9?
The problem just keeps growing exponentially the higher up the scale of biological complexity you go. Consider the information in a single cell. Even the simplest minimal cell needs to be able to reproduce itself, use and store energy, and make new cellular components not available in its environment, all of which require the complex cellular systems of replication, transcription, translation and metabolism. To organize and control such systems requires an enormous amount of information.
The only source that is able to generate this level of specified, complex information 10 is intelligence. We know from our own experience that intelligent agents can develop systems to store information and recall it when needed. Intelligent agents can reconfigure non-living things (and to a limited extent, living things) in purposeful ways, to modify or produce new functions as desired and to coordinate those functions into a working whole.
Obviously, biologists recognize the information storing and processing capacity of cells. What they are now realizing is how much information is needed to produce living systems. This is knowledge Darwin didn’t have. We need to begin to take account of it in any theory of life’s origin and change over time.
1 Lynch M (2007). The frailty of adaptive hypotheses for the origins of organismal complexity. Proc Natl Acad Sci USA 104:8597-8604.
3 Durrett R and D Schmidt (2008). Waiting for two mutations: with applications to regulatory sequence evolution and the limits of Darwinian evolution. Genetics 180:1501-1509.
5 Morris SC (2006). Darwin’s dilemma. Phil. Trans. R. Soc. B 361:1069-1083.
6 Meyer S (2004). The origin of biological information and the higher taxonomic categories. Proc Biol Soc Washington 117:213-239.
7 Koonin EV (2007). The cosmological model of eternal inflation and the transition from chance to biological evolution in the history of life. Biology Direct 2:15. doi:10.1186/1745-6150-2-15.
8 Graber R et al. (1999) Conversion of aspartate aminotransferase into an L-aspartate β-decarboxylase by a triple active-site mutation. J Biol Chem 274: 31203-31208.
10 Dembski W (1998) The Design Inference: Eliminating Chance through Small Probabilities. Cambridge University Press, ed. B Skyrms.