[The] conclusion [that evolution is a fact], even if well founded, would be unsatisfactory, until it could be shown how the innumerable species inhabiting this world have been modified, so as to acquire that perfection of structure and coadaptation which most justly excites our admiration.
—Charles Darwin
Abstract
The common opinion has been that evolution results in the continuing development of more complex forms of life, generally understood as more complex organisms. The arguments supporting that opinion have recently come under scrutiny and been found wanting. Nevertheless, the appearance of increasing complexity remains. So, is there some sense in which evolution does grow complexity? Artificial life simulations have consistently failed to reproduce even the appearance of increasing complexity, which poses a challenge. Simulations, as much as scientific theories, are obligated at least to save the appearances! We suggest a relation between these two problems, understanding biological complexity growth and the failure to model even its appearances. We present a different understanding of that complexity which evolution grows, one that genuinely runs counter to entropy and has thus far eluded proper analysis in information-theoretic terms. This complexity is reflected best in the increase in niches within the biosystem as a whole. Past and current artificial life simulations lack the resources with which to grow niches and so to reproduce evolution’s complexity. We propose a more suitable simulation design integrating environments and organisms, allowing old niches to change and new ones to emerge.
This is a preview of subscription content, log in via an institution to check access.
Similar content being viewed by others
Notes
For a skeptical review of this consensus opinion, identifying culprits, see McShea (1991).
Throughout, we mean by “evolutionary system” any system which has the three basic ingredients required for evolution—inheritance, with variation, under selection. This includes systems of biological organisms and, arguably, their aggregates in species and genera. It cannot include a total biosystem, however, if only because there is no non-trivial population of them subject to selection.
This explanation builds upon Stanley (1973), who applied the concept of “passive diffusion” to Cope’s rule of size increase.
When we say that the diffusion argument trivializes complexity growth, we are not saying that diffusion is trivial or unimportant—it is, in fact, essential (cf. Lynch 2007). Diffusion in evolution is based on mutation, chromosomal cross-over, etc., and is the very stuff of that variation upon which selection acts. So, without diffusion evolution simply stops. What is trivial is its use as a full and complete explanation for complexity growth: there is more to complexity than what diffusion alone can explain.
For example, most of the contributions to the special issue on the evolution of complexity of Artificial Life, introduced by Gershenson and Lenaerts 2008, are devoted to organismic complexity.
No one has previously produced such a model (other than that of McShea (1994), which suffers from its genome being epiphenomenal). Most A-Life simulations introduce implicit biases against complexity and so cannot be models of passive diffusion. But another impediment is just that no researcher is likely to confuse passive diffusion models with a serious attack on the open-ended evolution challenge in the first place. In any case, for this research we have produced such a model, with a genome that genuinely grows in complexity over time and with that complex genome playing a causal role in the inheritance of complexity over the generations. In consequence, the simulation runs progressively more slowly across those generations. Unsurprisingly, as a proper implementation of an unbiased random walk, our model’s results are qualitatively the same as McShea’s, showing that its epiphenomalism is immaterial.
In consequence of this trivialization, we will not refer to Bedau’s classes of unbounded evolution and instead use “unbounded”, “creative” or “open-ended” evolution simply in the sense of satisfying the Strong AOC.
After all, the central concept of adaptation—“the design or suitability of an object for a particular function” (Gould 2002, p. 117)—directly refers to functionality. A few examples of those taking biological functionality as integral to biological complexity include Bronowski (1970), Wicken (1979), Dawkins (1986), Kampis and Csányi (1987), Carroll (2001), Adami (2002). McShea himself, in McShea (2000), takes functional complexity as primary and proposes diversification measures as proxies for functional complexity. Indeed, in many of McShea’s past papers on biological complexity his preference for diversity over functional measures has been motivated simply by the difficulty in identifying biological functions and without any claim of the latter’s irrelevance. Heylighen (1999) sensibly points out that the root Latin word complexus means a twisting together of separate elements, suggesting a dual concept of diverse parts that are functionally interrelated.
Here we are not denying that evolution may operate at various cladistic levels or at the level of ecosystems (e.g., Swenson et al. 2000). We are simply pointing out that it does not operate at the highest conceivable level, that of the biosphere as a whole.
There have been parallel discussions of biology and entropy, exemplified by Schrödinger (2003), finding that individual organisms fight entropy, that the developmental process from egg to adult to senesence and death runs against the tide of entropy for the time of one lifespan. Schrödinger focused on the complexity and organization of organisms, but that is quite different from the complexity of organisms involved in the Weak AOC. Schrödinger’s organisms were individuals, whereas the “organisms” of the Weak AOC are really species. The issues raised are distinct: for Weak AOC, the issue is how diffusion through speciations moves species through the genotypic design space; for Schrödinger, the issue is how ontogenesis produces ordered bodies rather than chaos.
But see Marshall and Jacobs (2009) for a Permian example closer to pure exponential growth.
One reason for this may well be that mass extinctions select for generalists amongst species; cf. Jablonski (2001).
An alternative view held by some is that the apparent successive increases in biodiversity are an artifact of biased availability of paleontological data.
However, there certainly are also rapid adaptive radiations in such cases as well (Kocher 2004).
There are many two-part message length approaches to statistical inference, which collectively have delivered a variety of effective computational tools for pattern discovery in data mining. MML is one of the earliest to have been developed and is the one most clearly aligned with the rapid growth of Bayesian methods in applied computer science.
It’s worth noting that if we use the more appropriate notion of niche complexity, namely graph complexity, the exponentiality of niche complexity growth will only intensify, since the number of possible graphs is superexponential in the number of nodes.
References
Adami C (2002) What is complexity? Bioessays 24:1085–1094
Badii R, A Politi (1997) Complexity: hierarchical structures and scaling in physics. Cambridge University, Cambridge
Bambach RK (2006) Phanerozoic biodiversity mass extinctions. Annu Rev Earth Planet Sci 34:127–155
Bar-Yam Y (1997) Dynamics of complex systems. Westview, Boulder
Bedau M (2006) The evolution of complexity. In: Symposium on the making up of organisms. École normale supérieure
Bedau M, McCaskill J, Packard N, Rasmussen S, Adami C, Green D, Ikegami T, Kaneko K, Ray T (2000) Open problems in artificial life. Artif Life 6:363–376
Bedau MA, Packard NH (1991) Measurement of evolutionary activity, teleology, and life. In: Langton CG, Taylor C, Farmer JD, Rasmussen S (eds) Artificial life II. Addison-Wesley, Reading, pp 431–461
Bedau MA, Snyder E, Packard NH (1998) A classification of long-term evolutionary dynamics. In: Adami C, Belew R, Kitano H, Taylor C (eds) Artificial life VI. MIT Press, Cambridge, pp 228–237
Benton MJ (2001) Biodiversity on land and in the sea. Geolog J 36:211–230
Bronowski J (1970) New concepts in the evolution of complexity. Synthese 21:228–246
Bullock S, Bedau MA (2006) Exploring the dynamics of adaptation with evolutionary activity plots. Artif Life 12:193–197
Carroll S (2001) Chance and necessity: the evolution of morphological complexity and diversity. Nature 409:1102–1109
Chaitin GJ (1979) Toward a mathematical definition of “life”. In: Levine RD, Tribus M (eds) The maximum entropy formalism. MIT Press, Cambridge, pp 477–498
Channon A (2006) Unbounded evolutionary dynamics in a system of agents that actively process and transform their environment. Genet Program Evolvable Mach 7:253–281
Chase JM, Leibold MA (2003) Ecological niches. University of Chicago, Chicago
Dawkins R (1986) The blind watchmaker. Longman, New York
Dawkins R (1989) The evolution of evolvability. In: Langton C (eds) Artificial Life. Addison-Wesley, Sante Fe, pp 201–220
Dawkins R (1996) Climbing mount improbable. W.W. Norton, New York
Dawkins R (1997) Human chauvinism. Evolution 51:1015–1020
Dorin A, Korb K (2010) Network measures of ecosystem complexity. In: Artificial life XII: Proceedings of the ninth international conference on the simulation and synthesis of artificial life. MIT Press, New York, pp 323–328
Dorin A, Korb KB (2007) Building virtual ecosystems from artificial chemistry. In: European conference on artificial life. Springer, Berlin, pp 103–112
Dorin A, Korb KB, Grimm V (2008) Arificial-life ecosystems. In: Eleventh international conference on the simulation and synthesis of living systems. MIT Press, New York, pp 173–180
Gershenson C, Lenaerts R (2008) Evolution of complexity. Artif Life 14:241–243
Gotelli NJ, Colwell RK (2001) Quantifying biodiversity: procedures and pitfalls in the measurement and comparison of species richness. Ecol Lett 4:379–391
Gould S (1989) Wonderful life. Norton, New York
Gould S (1996) Full house. Random House, New York
Gould S (1997) Self-help for a hedgehog stuck on a molehill. Evolution 51:1020–1023
Gould SJ (2002) The structure of evolutionary theory. Harvard University Press, Cambridge
Heylighen F (1999) The growth of structural and functional complexity during evolution. In: The evolution of complexity. Kluwer Academic, Dordrecht, pp 17–44
Hinegardner R, Engelberg J (1983) Biological complexity. J Theor Biol 104:7–20
Hutchinson GE (1944) Limnological studies in Connecticut, VII. A critical examination of the supposed relationship between phytoplankton periodicity and chemical changes in later waters. Ecology 25:3–26
Hutchinson GE (1957) Concluding remarks. Cold Spring Harbor Symp Quant Biol 22:415–427
Hutchinson GE (1978) An introduction to population ecology. Yale University, New Haven
Jablonski D (2001) Lessons from the past: evolutionary impacts of mass extinctions. Proc Nat Acad Sci 98:5393–5398
Jones CG, Lawton J, Shachak M (1997) Positive and negative effects of organisms as physical ecosystem engineers. Ecology 78:1946–1957
Kampis G, Csányi V (1987) Notes on order and complexity. J Theor Biol 124:111–121
Kocher TD (2004) Adaptive evolution and explosive speciation: the cichlid fish model. Nature Rev Genet 5:288–298
Korb KB, Dorin A (2009) A simulation of niche construction. Technical Report 2009/2, Bayesian Intelligence. http://www.bayesian-intelligence.com/
Leibold MA (1995) The niche concept revisted. Ecology 76:1371–1382
Lynch M (2007) The frailty of adaptive hypotheses for the origins of organismal complexity. Proc Nat Acad Sci 104:8597–8604
Marshall C, Jacobs D (2009) Flourishing after the end-permian mass extinction. Science 325:1079–1080
Maynard Smith J, Szathmáry E (1995) The major transitions in evolution. W.H. Freeman, New York
McMullin B (2002) John von Neumann and the evolutionary growth of complexity: looking backwards, looking forwards. Artif Life 6:347–361
McShea D (1991) Complexity and evolution: what everybody knows. Biol Philos 6:303–324
McShea D (1994) Mechanisms of large-scale evolutionary trends. Evolution 48:1747–1763
McShea D (1996) Metazoan complexity and evolution: is there a trend? Evolution 50:477–492
McShea D (1998) Possible largest-scale trends in organismal evolution: eight “live hypotheses”. Annu Rev Ecol Syst 29:293–318
McShea DW (2000) Functional complexity in organisms: parts as proxies. Biol Philos 15:641–668
McShea DW, Brandon RN (2010) Biology’s first law. University of Chicago, Chicago
Odling-Smee F, Laland N, Feldman M (2003) Niche construction. Princeton University, Princeton
Papentin F (1980) On order and complexity I. J Theor Biol 87:421–456
Papentin F (1982) On order and complexity II. J Theor Biol 95:225–245
Ray TS (1990) An approach to the synthesis of life. In: Artificial life II. Addison Wesley, pp 371–408
Roman S (1992) Coding and information theory. Springer, New York
Ross S (1996) Stochastic processes. Wiley, New York
Ruiz-Mirazo K, Umerez J, Moreno A (2008) Enabling conditions for ‘open-ended evolution’. Biol Phil 23:67–85
Schrödinger E (1943/2003) What is life? Cambridge University, Cambridge
Sepkoski JJ, Jablonski D, Foote MJ (2002) A compendium of fossil marine animal genera. Paleontological Research Institution, Ithaca
Shannon C (1948) A mathematical theory of communication. Bell Syst Tech J 27:379–423–623–656
Signor P (1994) Biodiversity in geological time. Am Zool 34:23–32
Simberloff DS, Wilson EO (1969) Experimental zoogeography of islands: the colonization of empty islands. Ecology 50:278–96
Standish R (2010) Network complexity of foodwebs. In: Artificial life XII: proceedings of the ninth international conference on the simulation and synthesis of artificial life. MIT Press, New York, pp 337–343
Stanley SM (1973) An explanation for Cope’s rule. Evolution 27:1–26
Stout A, Spector L (2005) Validation of evolutionary activity metrics for long-term evolutionary dynamics. In: Beyer HG, O’Reilly UM (eds) Proceedings of the genetic and evolutionary computation conference. ACM Press, New York, pp 137–142
Swenson W, Wilson D, Elias R (2000) Artificial ecosystem selection. Proc Nat Acad Sci 97:9110–9114
Wallace CS (2005) Statistical and inductive inference by minimum message length. Springer, New York
Ward P (1995) The end of evolution: dinosaurs, mass extinction and biodiversity. Weidenfeld and Nicholson, London
Wicken J (1979) The generation of complexity in evolution: a thermodynamic and information-theoretical discussion. J Theor Biol 77:349–365
Yang X (2002) Characterization of multispecies living ecosystems with cellular automata. In: Standish RK, Bedau MA, Abbass HA (eds) Artificial life VIII: proceedings eighth international conference on artificial life. MIT Press, New York, pp 138–141
Acknowledgments
We would especially like to acknowledge the contribution of Mark Bedau, both for originally stimulating this research and for extensive discussion of its content. We would also like to acknowledge helpful discussions with Martin Burd, Suzanne Sadedin and Tim Taylor and material assistance from the LSE Centre for Philosophy of Natural and Social Science and the Tilburg Centre for Logic and Philosophy of Science. Finally, we would like to thank our anonymous referees for helping us to improve our discussion.
Author information
Authors and Affiliations
Corresponding author
Rights and permissions
About this article
Cite this article
Korb, K.B., Dorin, A. Evolution unbound: releasing the arrow of complexity. Biol Philos 26, 317–338 (2011). https://doi.org/10.1007/s10539-011-9254-6
Received:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s10539-011-9254-6