Selective Relativity

(This is a rough draft of a paper I am writing for my phil. of biology class. Comments and criticisms are certainly appreciated.)

In this paper I will describe and argue for a position which I will call selective relativity. I follow C. Kenneth Waters in arguing that the debate surrounding which is the uniquely correct level at which selection operates derives in part from a deep-seated conceptual confusion. This confusion is in not realizing that all notions of fitness, selective force, etc. follow from rather than constrain the biologist’s identification of a unit of selection. The only things which constrain the biologist’s choice in unit are pragmatic concerns and I will defend this claim against gene-centrist arguments and counterexamples. Finally, I will demonstrate that all such objections are based in the same conceptual confusion mentioned above.

· Units of Selection as a Pragmatic Choice

With his theory of special relativity, Albert Einstein demonstrated that the very idea of a uniquely correct inertial frame of reference was completely meaningless. The only reason one could have to think otherwise, it was shown, was due to a conceptual confusion in thinking that some concepts could be applied independent of an inertial frame of reference from which they are necessarily derived. If we call the inertial frame of reference “X” and the concepts of velocity, length, simultaneity, etc. “Y”, his ideas can be formulated as follows:

a) Y only has meaning with respect to some X.

b) After X is in place, Y follows logically from it.

c) Truth and falsity can only be applied to Y with respect to some X, which is held constant.

d) Holding Y constant while varying X leads to confusion and error.

e) There is no uniquely correct X.

f) Different X’s with their corresponding Y’s are simply different descriptions of or perspectives on the same physical phenomenon.

g) The scientist’s choice in X is constrained solely by pragmatic concerns.

These exact same points hold within evolutionary biology if we simply redefine Y as forces of selection, evolutionary fitness, etc. and X as the frame of reference which is created by drawing a conceptual line between some selected unit and it surrounding, selecting environment. I suggest, then, that the “units/levels of selection” debate is based in the same conceptual confusion upon which the debate concerning the absolute inertial frame in reference was, namely in not recognizing (a) and (b). Accordingly, searching for the correct unit of selection in evolutionary biology will be no more profitable than the physicist’s search for the cosmic ether was.

The argument presented above for selective relativity is explicitly an argument from analogy: in physics, points (c) – (g) follow from the truth of (a) and (b), therefore, since (a) and (b) hold true in evolutionary biology points (c) – (g) must follow as well. In his paper “Tempered Realism about the Force of Selection” C. Kenneth Waters argues to the same end, but by different means: “Only after this division [between unit and environment] is made can we speak meaningfully about the forces of selection and the levels at which they impinge.” (Waters, 571) Due, however, to limitations space and scope, a detailed analysis of his arguments will not be given here. What concerns us for the time being is primarily a firm grasp of selective relativity, (c) – (g), and the premises upon which it is founded, (a) and (b), in order to defend the view from gene-centric criticisms. Such criticisms, it will be shown, are primarily aimed at producing counterexamples, especially to (g).

· Additional Constraints

The gene-centrist’s argument that the individual gene is the proper unit of selection as opposed to the individual organism depends upon (g) being false for one reason or another. One way of viewing selective relativity is in terms of what Wilson calls “cheap individualism” only generalized to all level and units (Wilson, 67); selection as it applies to the gene, gene-complex, cell, trait, organism or group is valid in every case by very definition. What the gene-centrist objects to is not that some unit is the correct one simply by definition, for Williams’ argument for gene-centrism makes this very move. (Williams, 44; Dawkins, 33) Rather, he objects to the claim that every unit wins by definition. Thus, in order to make their argument that the gene and only the gene is the correct unit of selection stick, they must assume that X is constrained by something beyond mere pragmatic concerns.

This something else is stability. Y correctly applies to some X only if X is sufficiently stable across numerous generations.

“One necessary condition [for natural selection] is that the selected entity must have a high degree of permanence and a low rate of endogenous change, relative to the degree of bias (differences in selection coefficients)… [there must be] a favorable or unfavorable selection bias equal to several or many times its rate of endogenous change.” (Williams, 43, 44)

The gene-centrist claims that the only biological entity which is able to meet this criterion of stability is the individual gene, or something close to it. It is only genes which are sufficiently stable across numerous generations in virtue of their being replicators, entities which achieve multi-generational longevity by way of replicating themselves with high fidelity and fecundity. (Dawkins, 191) With the sole exception of the individual gene, all other biological entities are too ephemeral to meet this criterion and therefore do not qualify to be X.

This account contradicts selective relativity on a number of points; at least (e), (f) and (g) to be specific. According to the gene-centrist, since X is constrained by the stability criterion, X is not constrained solely by pragmatic concerns. While he does not go so far as to argue that there cannot possibly be any other biological entity in addition to the gene which passes the stability criterion, it is argued that in actuality the gene is the uniquely correct X. Furthermore, since the individual gene is the uniquely correct X, all other potential X’s must be describing the physical phenomenon incorrectly, meaning that they are describing a distinct physical phenomenon altogether as being true.

It should be noted that the gene-centrist’s criticism does not directly address the premises or the structure upon which the argument for selective relativity rests. Nowhere does the gene-centrist argue that (a) or (b) is not the case. Rather, a counterexample is provided to (e), (f) and (g) which are supposed to be the consequences of (a) and (b). If the gene-centrist’s claims are true, then this would be strong reason to believe that something must be wrong with either the structure of the argument or its premises. The gene-centrist will likely reject the application of the of the argument to evolutionary phenomenon since to reject the structure of the argument itself would be to undermine special relativity as well. I will respond to the gene-centrist’s counterexample by demonstrating that it is a paradigm case of the conceptual confusion which selective relativity is aimed at avoiding.

· Conceptual Confusion

The gene-centrist’s argument violates the premises upon which our argument from analogy is built, (a) and (b), and therefore falls prey to the predicted confusion and error, (d). Let us first consider how it violates (a) and (b). Here is Williams’ full quote:

“The essence of the genetical theory of natural selection is a statistical bias in the relative rates of survival of alternatives (genes, individuals, etc.). The effectiveness of such bias in producing adaptation is contingent on the maintenance of certain quantitative relationships among operative forces. One necessary condition is that the selected entity must have a high degree of permanence and a low rate of endogenous change, relative to the degree of bias (differences in selection coefficients).” (Williams, 43)

The stability of the selected unit is necessary, according to Williams, in order to ensure the statistical bias produced by various relationships. Thus, what is required for evolution by natural selection is not merely the stability of the selected unit. If such were the case then one could indeed simply look for the most stable biological entity as the gene-centrist does. Rather, what is required is stability in the relationship between the selected unit and its selecting environment. Accordingly, the necessary stability cannot simply be read off of the selected unit alone. Instead, stability, operative forces and bias are all aspects of Y which must follow from rather than constrain by preceding X.

This tendency to see Y as being, to some degree, independent of any X leads the gene-centrist to make a fundamental error in their argument against those who see selection as operating at the level of the organism. One unit, the organism, is seen as being in an unstable relationship with the environment which corresponds to that particular unit. This relationship is then compared to that which another unit, the gene, has with an environment that does not correspond to it, namely that of the organism. (The common appeal to interactors/vehicles is all too helpful in facilitating this confusion.) Once the individual gene is seen within its own corresponding environment (which includes not only the environment external to the organism, but everything which is internal to the organism while remaining external to the individual gene as well) the stability of the unit/environment relationship no longer appears near as stable as it once did. As the gene-centrist himself is wont to point out, each generation the individual gene finds itself in a completely different genetic environment due to the genetic shuffling native to sexual reproduction. In other words, once one sees stability as an aspect of Y which follows from X rather than the other way around not only does the counterexample to selective relativism vanish, but the gene-centrist’s argument against selection at other levels loses its force as well.

· Replication, Reproduction and Instantiation

There still remains significant work to be done, however, in order to fully dissolve the gene-centrist’s argument against selective relativity. While the gene-centrist’s appeal to stability may rest upon a conceptual confusion, the above response seems a little too quick and easy to conclude that, pragmatic concerns aside, there are no constraints at all on the unit of selection. It certainly seems that replication, or at least reproduction is essential to being a unit of selection and this is what the gene-centrist’s criticism is really getting at. If such is the case, then replication or reproduction would indeed be constraints to X in addition to those of pragmatic concerns. Thus, (g) still seems to be in trouble.

This objection gets its strength from two sources of conceptual confusion: (1) the failure to appreciate the type/token distinction, and (2) the mistake of viewing some aspect of Y as existing independent of some X. Let us first address the lack of appreciation for the type/token distinction. Individual tokens of some gene last no longer than the organisms which carry them. The supposed longevity of the type GENE lies in its ability to be tokened, by way of replication, with sufficient frequency in the population. There is no reason why this same reasoning cannot also be applied at the level of the organism. Thus, the type DOG achieves its multigenerational longevity by being tokened, this time by way of reproduction, with sufficient frequency in the population. The supposed stability of both cases follows from simply holding the type constant across generations and nothing more.

With such a view in mind, we are now equipped to address the confusion of viewing some X as following from some aspect of Y rather than the other way around or as viewing Y as having meaning independent of some X. Fitness and selective forces, according to the type/token account of replication and reproduction, are simply the relationships which hold between the frequency with which a type is tokened and the frequency with which the type ceases to be tokened within some population; fitness is simply the growth rate of tokening-frequencies. By this account, it is irrelevant how a type comes to be tokened, just so long as it is tokened. (Harms) Accordingly, fitness and selective force, as aspects of Y, generalize to non-reproducing types as well. In fact, to be a type which is tokened in some population simply is to be a potential X along with some Y which necessarily follows from it. In fact, it is in virtue of every X being a type rather than a token that every X meets the stability criterion.

An example which demonstrates how non-reproducing X’s can have corresponding Y’s is in order. Consider the case of the queen-bee and the worker-bee. There are a number of types at work in this phenomenon which can be correctly described from any number of perspectives. The swarm of bees interacts with its environment in order to created more swarms of bees. The individual bees are interacting with their individual environments in order to create more individual bees. The queen is interacting with her environment of workers in order to create more queens. The individual worker is interacting with his environment composed of other workers and a queen in order to create more workers by way of the queen. The bee-genes are interacting with their genetic environment, which includes all the other individual bees along with their corresponding bee-genes, in order to create more bee-genes. Strangest of all, the individual bee hive is interacting with its environment, which consists of all the bees, to create more bee hives! The genes replicate, the queens reproduce, the workers do neither, the swarm splits and the hive gets built. Each X has its own corresponding Y which logically follows from it and describes the frequency with which X is tokened. No particular X is uniquely correct in this account, for each X simply describes the same phenomenon from a different perspective.

· Conclusion

Concepts such as fitness, selective force, stability, etc. follow from rather than constrain the biologist’s choice in unit of selection and all arguments for some unit(s) of selection to the exclusion of any others rest on a conceptual confusion about this point. The scientist’s identification of a unit of selection is constrained solely by pragmatic matters, such as whether he has an interest in the unit or how difficult it is for him to define, identify and/or quantify the unit, not on whether the potential X is a replicator or a reproducer. Viewing natural selection from different units of selection is simply viewing the same phenomenon from different perspectives and searches for the uniquely correct unit of selection are tantamount to quests for cosmic ether.

Bibliography

Dawkins, Richard (1989) The Selfish Gene, 2^nd ed.

Harms, William (2004) Information and Meaning in Evolutionary Processes

Waters, C. Kenneth (1991) “Tempered Realism about the Force of Selection”. Philosophy of Science, 58 (1991) pp. 553-573

Williams, George C. (1966) Excerpts from Adaptation and Natural Selection. Reprinted in Sober, Elliot ed. (2006) Conceptual Issues in Evolutionary Biology

Wilson, David Sloan (1989) “Levels of Selection: An Alternative to Individualism in Biology and the Human Sciences”. Reprinted in Sober, Elliot ed. (2006) Conceptual Issues in Evolutionary Biology