Abstract
On the basis of distinctions between those properties of entities that can be defined without reference to other entities and those that (in different ways) cannot, this note argues that non-trivial forms of frequency-dependent selection of entities should be interpreted as selection occurring at a level higher than that of those entities. It points out that, except in degenerately simple cases, evolutionary game-theoretic models of selection are not models of individual selection. Similarly, models of genotypic selection such as heterosis cannot be legitimately interpreted as models of genic selection. The analysis presented here supports the views that: (i) selection should be viewed as a multi-level process; (ii) upper-level selection is ubiquitous; (iii) kin selection should be viewed as a type of group selection rather than individual selection; and (iv) inclusive fitness is not an individual property.
This is a preview of subscription content, log in via an institution to check access.
Similar content being viewed by others
Notes
This formulation is not intended to suggest that all genic selectionists believe that it is not also possible to hold that some other unit (or level), for instance, the genotype may also consistently be regarded as the only unit (or level) of selection. For instance, Kitcher and Sterelny (1988) accept pluralism of frameworks (genotype-based or gene-based accounts of selection) but, as is the case of almost all genic selectionists, not pluralism within frameworks.
Moreover, in the era of classical genetics, it was typical to identify alleles entirely on the basis of phenotypic differences (see Haldane [1942]) and, from that perspective, a coarse-grained classification of alleles, as used here, is entirely legitimate.
For more detail, see Wimsatt (1980).
I am intentionally using the locution “genotypic” rather than “individual” to keep distinct the genetic (allele → locus → gene complex → genotype, etc.) and structural (molecule → organelle → cell → tissue → organ → individual → group, etc.) hierarchies. For a discussion of the relevance of this distinction to the units of selection problem see Sarkar (1994, 1998); the explicit distinction goes back at least to Williams (1985).
Note that Sterelny and Kitcher also develop other arguments against Sober’s (1984) defense of genotypic selection in the case of heterosis. These other arguments are not addressed in this note, partly because it builds a different case against genic selectionism than the one developed by Sober.
A very weak sense of definability is being used here: being definable is taken to mean being computable, that is, if X can be defined using Y, then X can be computed using Y. To ward off metaphysical quibbles that are not relevant to the philosophical issue under debate here, definability is not being construed linguistically.
There is a different sense in which fitnesses are almost never unqualified individual properties and are usually viewed as relational properties—they typically depend on the environmental context. But that sense of non-individuality in not the relevant one here.
Waters (1991) has noted other possible types of allelic fitnesses but it is not clear how these could be used to model heterosis.
To avoid unnecessary confusion, note that the individuals referred to in this paragraph and the next are individual entities in any ontology, not necessarily biological individuals.
Sterelny and Kitcher (1988), p. 358; emphasis added. Dawkins’ thesis in The Extended Phenotype is similarly described: “there are often alternative, equally adequate representations of selection processes and that, for any selection process, there is a maximally adequate representation which attributes causal efficacy to genic properties (1988, p. 358; emphasis added).”
The power set, \(\wp (\Upsigma ),\) is the set of all subsets of \(\Upsigma.\) Note that \(\Upsigma \in \wp (\Upsigma )\) because \(\Upsigma \subseteq \Upsigma\) and the null set, \(\phi \in \wp (\Upsigma ).\) We will need the former, but not the latter, fact about \(\wp (\Upsigma ).\)
Alternatively, the set of multiple-locus genotypes can be modeled as the Cartesian product of the sets of individual genotypes.
This is one possible explication of what is often called an “intrinsic” property (see Swoyer [2000, § 7.15]).
A physical example may also help clarify what contextual individual properties are: consider an electric charge, e +, in an electromagnetic field caused by other charges. The force felt by the charge, e +, may be viewed as beingcaused by these other charges (those committed to a purely particle ontology of classical physics would insist that it must be so viewed). However, the force may also be viewed as being caused by the local electromagnetic field at the point in space-time where the charge, e +, happens to be. Thus, it depends on the context but is still not a relational property. What happens to e + may depend on events elsewhere in space-time but only as mediated by the local electromagnetic field: locality is thus central to the individuation in this case, and the interactions are not permitted to move faster than the speed of light.
For expository simplicity A i A j is being treated as different from A j A i —this choice does not affect the results.
A similar point was made by Wimsatt (1980).
For more on this point, see Uyenoyama and Feldman (1980).
References
Arnold AJ, Fistrup K (1982) The theory of evolution by natural selection: a hierarchical expansion. Paleobiology 8:113–129
Damuth J, Heisler IL (1988) Alternative formulations of multi-level selection. Biol Phil 3:407–430
Godfrey-Smith P, Kerr B (2002) On Price’s equation and average fitness. Biol Phil 17:551–565
Haldane JBS (1942) New paths in genetics. Random House, New York
Keller L (ed) (1999) Levels of selection in evolution. Princeton University Press, Princeton
Lewontin RC (1970) The units of selection. Annu Rev Ecol Syst 1:1–18
Maynard Smith J (1982) Evolution and the theory of games. Cambridge University Press, Cambridge, UK
Michod RE (1999) Darwinian dynamics: evolutionary transitions in fitness and individuality. Princeton University Press, Princeton
Okasha S (2004a) The averaging fallacy. Biol Phil 19:167–184
Okasha S (2004b) Multilevel selection and the partitioning of covariance: a comparison of three approaches. Evolution 58: 486–494
Sarkar S (1994) The selection of alleles and the additivity of variance. In Hull DL, Forbes M, Burian RM (eds) PSA—1994: Proceedings of the 1994 Meeting of the Philosophy of Science Association, Vol. 1. Philosophy of Science Association, East Lansing. pp 3–12
Sarkar S (1998) Genetics and reductionism. Cambridge University Press, New York
Sober E (1984) The nature of selection. MIT Press, Cambridge, MA
Sober E, Wilson DS (1998) Unto others: the evolution and psychology of unselfish behavior. Harvard University Press, Cambridge, MA
Spencer R-J, Janzen FJ, Thompson MB (2006) Counterintuitive Density-Dependent Growth in a Long-Lived Vertebrate after Removal of Nest Predators. Ecology 87:3109–3118
Sterelny K, Kitcher P (1988) The return of the gene. J Phil 85:339–361
Swoyer C (2000) Properties. In Zalta EN (ed) Stanford Encyclopedia of Philosophy.http://www.plato.stanford.edu/archives/win2000/entries/properties/
Uyenoyama M, Feldman MW (1980) Theories of kin and group selection: a populations genetics perspective. Theor Popul Biol 17:380–414
Waters CK (1991) Tempered realism about the force of selection. Phil Sci 58:533–573
Williams GC (1985) A defense of reductionism in evolutionary biology. Oxford Surv Evol Biol 2:1–27
Wimsatt WC (1980) Reductionistic research strategies and their biases in the units of selection controversy. In: Nickles T (eds) Scientific discovery: case studies. Reidel, Dordrecht, pp 213–259
Author information
Authors and Affiliations
Corresponding author
Additional information
For comments on an earlier version of this analysis, thanks are due to James Justus, Alexander Moffett, Samir Okasha, and an anonymous referee (who provided very detailed criticisms).
Rights and permissions
About this article
Cite this article
Sarkar, S. A note on frequency dependence and the levels/units of selection. Biol Philos 23, 217–228 (2008). https://doi.org/10.1007/s10539-007-9092-8
Received:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s10539-007-9092-8