Thesis, Antithesis, Synthesis: What is Natural Selection?
Duane Meehan
1. An Introduction to Theory of Natural Selection, Original and New.
One pair of any mammal, on the formula of geometric increase, absent natural selection, and unimpeded by scarcity, with all conditions being optimal, “will produce offspring whose numbers would become so inordinately great that no country could support their kind” (Darwin 33). This probable rate of reproduction is a purely statistically determined trend, which trend determination permits census predictions for a future population of mammals. Natural selection theory predicts and explains unique probabilities of fitness of groups of organisms by census (Matthen, 83). And Matthen, who advocates the idea of determinations of fitness by population census, tells us that “in the last analysis, fitness involves sui generis (i.e., “of its own kind”, “alone of its kind” or “unique”) probabilities estimated not by causal, or even mathematical, analysis, but by census” (83). For example, “Even slow-breeding man has doubled in twenty five years, and at this rate, in less than a thousand years, there would literally not be standing-room for his progeny” (Darwin 33). Census prediction of populations of animals is, historically, one of the central descriptive tasks of natural selection theory. Natural selection theory determines how, under actual environmental conditions, populations of organisms successfully achieve extensive reproductive success. Consider this, circa 1859, Darwinian story of census prediction by statistical analysis of one pair of wild elephants breeding under actual environmental conditions, (Darwin will subsequently proffer his thesis regarding the how of successful extensive reproduction, which thesis is given below in a nutshell summary of Darwin):
The elephant is reckoned the slowest breeder of all known animals, and I have taken some pains to estimate its probable minimum rate of natural increase; it will be safest to assume that it begins breeding when thirty years old, and goes on breeding until ninety years old, bringing forth six young in the interval, and surviving till one hundred years old; if this be so, after a period of 740 to 750 years there would be nearly nineteen million elephants alive, descended from the first pair (Darwin 33).
The following nutshell description of the Darwinian notion of natural selection is meant to acquaint the reader with the essence of the original, historical, notion of natural selection, and, to prep the reader for immersion in the troubled waters of current theory of natural selection.
Out in the wild, natural selection is the “means” (Darwin iii) whereby the biosphere sorts which groups of living things will perish, and which will survive and reproduce. In raw nature, the majority of the individuals of every kind of life form are refused survival by a premature death, a death stopping all chances of reproduction. Premature death is the main means by which individual organisms are eliminated from what is usually an initially immense number, or population, of individuals of a new generation. Premature death is the negative side of the natural means of selection, by which vast populations are paired down to a size which nearly matches resources available in the environment.
The positive side of natural selection consists in the cases of those living things which successfully adapt to the hardships of life; survive; reproduce; and, preserve their kind by passing their adaptive characteristics on to the next generation, and, perhaps, onto an increasing number of survivors across said generations. Darwin derived a fundamental positive principle of life-preservation from observing the minority of living things which, out of an initially vast population, survive. Darwin then gives us the how, i.e., the means, of that successful, fruitful, survival, “I have called this principle, by which each slight variation, if useful, is preserved, by the term of Natural Selection” (Darwin 32). Thus, natural selection is that positive/negative principle, operating in nature, by means of which populations of living things differentially survive to reproduce pre-existing traits, which prove efficient for survival, or perish, due to the inheritance of variation(s) in fitness traits which prove to be ineffective for coping with an environment which is, itself, in flux.
Today, one hundred and forty-eight years since the publication of Darwin’s Origin of Species by Means of Natural Selection (1859), the globe teems with theorists and theories, scientists and sciences, seeking precise understanding of what natural selection is. In what follows, the reader is invited to reflect, along with the writer, about the contending structures of two main theoretical positions, i.e., the “dynamical”, and the “statistical” (Walsh 452 ), which are in debate among philosophers of biology, and which claim to be describing what natural selection is. Is the theory of natural selection a statistical theory or a dynamical one? That is, is natural selection theory a theory of actual dynamic environmental causal forces which have effects upon the protoplasm of organisms, or, is natural selection theory properly an acausal state of affairs which comprehends only statistical trends, absolutely random genetic events, and takes probabilistic census of the fitness of extant, and future, ensembles of organisms? That is the question which shall be answered in the course of this essay, for, herein, writer and reader shall, together, reflect upon a recent synthesis of the most theoretically intelligible portions of these opposing positions, which synthesis is an alternative theory of natural selection, and, as my thesis proposes, that synthesis, or synthetic theory, constitutes both an answer to the aforementioned question, which is the crux of the dynamical versus statistical debate in the philosophy of biology, and a resolution of the statistical versus dynamical debate. Said synthesis and synthetic resolution is, without her specifically saying so, proffered by Dr. R. Millstein in her recent essay “Natural Selection as a Population-Level Causal Process” (December 2006), which essay, as a synthetic moment in the history of theory of natural selection, is the central focus of this study.
Before proceeding further I explain a few phrases/notions, either just now employed, or, yet to be employed. “Theoretical intelligibility” is that state of affairs wherein a given theoretical model exhibits no flaws in reasoning which would constitute it as nonsense, e.g., infinite regress; self-inconsistency/contradiction; logical fallacy; incoherence. “Theoretical unintelligibility” is the state of affairs wherein a theory does exhibit any ilk of flawed reasoning which would constitute a given proposition, or theory, as nonsense. I define the phrase “synthetic resolution” as a state of affairs wherein a scholar becomes cognizant of particularly antithetical theoretical structures, represented by differing thinkers within his field of expertise and, he or she subsequently works-out a possible resolution of the antithetical positions, thereafter issuing a written synthesis of the most theoretically intelligible portions of the antithetical arguments. The upsurge of a synthetic resolution of originally antithetical positions is constituted both in the sense that the conflicting theoretical stances are brought into focus by elimination of the theoretically unintelligible portions thereof, (which elimination may consist in making certain concessions that one’s theoretical position does indeed contain a certain measure of flawed and unintelligible reasoning), and, the debate is brought to a synthetic moment (we can speak of a thetic moment wherein a thesis is posited; an antithetic moment wherein that thesis is contradicted; and a synthetic moment, wherein a truce arises as the result the reciprocal acceptance of reasoned argument for and/or against one’s original position), wherein antithetical stances become, for the time being, integrated.
In what is immediately subjoined, I simultaneously arrange the order wherein I shall present evidence and, describe salient characteristics of the three theoretical stances which constitute the thesis, antithesis and, synthesis, originally mentioned in the title. Dynamical theory of natural selection, or, natural selection as a dynamical theory, is the original thesis regarding the means of the evolution of species. The original theory of natural selection is the Darwinian theory, (think of the original thesis of what natural selection is as the “original Coke” flavor of selection theory), and, its central characteristics are causality, i.e., natural selection is on the original theory seen as the “cause” of the origin of all design in living things (Ariew, 1), and, is seen as force, i.e., natural selection per se is a “force”, alike the Newtonian mechanistic (i.e., causal) forces of gravity, or, of the electrostatic force, which mechanistic selective force(s) effect change(s) in the protoplasm of organisms (Matthen, 58). The force that is natural selection is deemed to be both a force in itself, and, to be separable into a series of different and distinctive forces, which can act singly, or combine, to collectively effect structural modification in organisms (Matthen, 58; Walsh, 453). That series of causal forces includes, for instance, selection, mutation, migration, random genetic drift, and, in-breeding (Matthen, 58; Walsh, 453). The several causal forces to which the dynamically-oriented evolutionists subscribe are, according to their theoretical view, the why of evolutionary change in species of organisms. Natural selection as “causal force(s)” is a theoretical construct, subject to being doubted, critiqued, partially corrected, or, completely destroyed, by any antithetical reasoning which might possibly demonstrate the theoretical unintelligibility of natural selection as “causal force”.
I shall further show that the statistical theory of natural selection is the antithesis of dynamical-causal theory of natural selection. Statistical theory of natural selection is antithetical to dynamical theory, because, the statistical view of natural selection holds that natural selection is absolutely not an ensemble of causal forces which constitute means of precipitating effects within the protoplasm of living things (Matthen, 68). The statistical version of natural selection is an acausal (Millstein, 651) and mathematical methodology, concerned with calculation and prediction oftrend within populations of organisms (Matthen, 57). For statistically-oriented theorists natural selection is a trend, and is not a causal process wherein a prior event precipitates a subsequent event (Matthen, 79). Statistical theory of natural selection holds, for example, that natural selection is the operation within populations of organisms of pure random chance, which biologists commonly call “random drift” (i.e., a biological process whereby bodily form arises as the result of pure random and non-teleological events) (Walsh 453), whereby ensembles/populations of organisms, tend to drift, en masse, toward the fittest possible of all design solutions (Ariew, 12) to alleviate the problems of survival within their particular environment; a trend precisely analogous to a mass of gas molecules which tend, en masse, to randomly constitute a particular temperature and pressure within an enclosure (Walsh, 463). For the statistically-oriented natural selectionists, random drift is “probabilistic cause”, i.e., a cause effected by pure chance (Matthen, 81). (Think of the acausal statistical view of natural selection, which is directly antithetical to causal-dynamical theory of natural selection, as the “new Coke” flavor of natural selection theory).
I shall further show, (and this is the thesis of the present essay), that a synthetic theory of natural selection has just now appeared (i.e., 4 Dec 06). Which synthetic theory is the result of the thought of a young philosopher of biology who is engaged in a particular dialectical logic, which logic is the logic of struggle between extant scholars exhibiting significant differences in theoretical stances, which stances ultimately become reconciled into one position, which both preserves and dispenses with elements of the original antithetical positions. This dialectical logic is the mode whereby human history proceeds (Sartre, 15). In the struggle between groups, antagonisms are reconciled into a transformational synthesis of formerly antagonistic elements, which synthesis preserves portions of what were former antagonisms in a new outlook, born out of the original dialectical tensions. (Sartre, 505-563). Scholars in the field of theory of natural selection are assailing what they deem to be the theoretical unintelligibilties of their opposing positions. Theoretical unintelligibility is a state of affairs wherein a given theory exhibits any ilk of fallacious/incoherent thinking. It is out of the conflict of these ascriptions of unintelligibility between scholars engaged in the dynamical/statistical debate, regarding what natural selection is, that concessions are reciprocally made by opposing sides. These sides constitute, when all is said and done, a synthetical resolution, or synthetic theoretical stance regarding what was, theretofore, a zone of pure contention. My thesis posits the claim that Dr. Millstein’s synthetic theory constitutes a synthetic moment of resolution of a debate, current in philosophy of biology, between the dynamical and the statistical versions of natural selection theory. Dr. Millstein’s treatise expresses both what now has to be conceded to the opposition, as having been mistaken, in the dynamical causal theory of natural selection, and, stands firm on that portion of the dynamical theory which she can still support against the onslaught of the purist statisticians. Dr. Millstein predicates her firm causalistic stance upon the results of scientific observations of the montane willow leaf beetle, which observations were seeking, and found, an environmental causal mechanism which can effect beetle protoplasm, which cause-effect could be pointed-to in order to explain and support what, otherwise, was purely statistical beetle census data, which purely statistical data, by itself, according to Millstein, Dahlhoff, and Rank, is insufficient to demonstrate natural selection to be at work (Millstein 634). (Think of this synthetic theory of natural selection as roughly analogous to the state of affairs wherein “original Coke” was created and then abandoned for a “new Coke”; which was then reversed, such that “original Coke” was conserved while the world refused the “new Coke.” Then, the world proceeded to embrace and retain the “new Coke” too, thus creating a synthetic state of affairs wherein both the old and new Coca Cola coexist.)
2. Mainstays of Arguments Con, and Pro, Regarding the Process of Causation in Natural Selection, Including Attendant Concessions to Con, Wherein Synthetic Conclusions Regarding What Natural Selection Is are Drawn.
Within the dialectical/polemical interactions between philosophers of biology who subscribe to antithetical sets of theoretical instrumentations whereby they respectively reason about, and posit solutions to theoretical problems in natural selection, there is struggle and conflict regarding what constitutes an accurate determination of what natural selection is, and, is not. And, within that philosophical conflict, destructions of opposing theoretical stances are effectuated. As one set of theorists have put it, there are “recent attempts to purge the theory (of natural selection) of a causal concept of fitness” (Bouchard, 693) by showing that natural selection is a purely statistical trend (Matthen, 78; Millstein, 627), (precisely like the operation of the second law of thermodynamics, which describes trends toward entropy), toward increased fitness of organisms (Bouchard, 693). And, Bouchard and Rosenberg, for their part, refute, and attempt to reduce to ineffectuality, the statistical trend conception of natural selection, by demonstrating the inapplicability of the second law of thermodynamics to natural selection (Bouchard, 704), upon which thermodynamic law Matthen and Ariew (2002), and, Walsh, et. al. (2004) predicate their statistical and probabilistic (i.e., proceeding by pure chance) theories of natural selection. There is no space in this paper to discuss Bouchard and Rosenberg’s opposition to Matthen and Ariew’s thesis. I will address only the contrary theoretical stance taken by Matthen and Ariew to the fundamental dynamical thesis that natural selection is a causal process, and, how Millstein achieves a synthetic resolution of theoretical difference between causal and purely statistical conceptions of natural selection.
The dynamical theoretical view of natural selection, which maintains selection is a process of causation, wherein external environmental selective forces produce effects within the protoplasm of organisms, which organisms are either somatically structured such that they are fit to survive alongside those environmental forces, or, perish, is under attack by at least two sets of scholars, i.e., Matthen and Ariew (2002), and Walsh, Lewens and Ariew (2002), and, is defended by at least two other sets of scholars, i.e., Millstein (2006) with Dahlhoff and Rank (2000), and, Bouchard and Rosenberg (2004).
As a result of their employment of a particular set of intellectual instruments, in pursuit of a determination of what natural selection is, and is not, Matthen and Ariew (2002), have viably demonstrated the theoretical unintelligibility of the construct of “force”, as that construct is employed by dynamical theorists of natural selection. Matthen and Ariew have painstakingly shown that adaptive forces which produce the an organism’s traits, unlike Newtonian forces, do not add-up, and, thereby, effectually destroy “force” as a theoretical construct employed in dynamical theory of natural selection. In Newtonian mechanics when calculating the fall of a feather one can intelligibly add the force of gravity to wind shear and to the electrostatic force, however, one cannot add up diverse ‘forces of natural selection, like mutation, migration, random genetic drift, selection, and inbreeding’ which, like apples and oranges, do not add up (Matthen 68). The dynamicists, who have been claiming the possibility of adding the above mentioned types natural selection factors, as if they are so many vectors of Newtonian force, which, taken together, determine the design and appearance of adaptive organismic traits, have, via Matthen and Ariew’s incisive reasoning, lost to theoretical unintelligibility, their construct of vectors of force in natural selection.
Matthen and Ariew have demonstrated, via reasoned argument from theory of thermodynamics, that natural selection is not a simple basic process of causal force(s) effecting organic protoplasm, and have done so, with what may be equally destructive force, to the Newtonian mechanistic structure of dynamical/causal theory of natural selection.
For Matthen and Ariew evolution is not caused by natural selection (79). For Matthen and Ariew “the increase of fitness in a population by natural selection” (79) is not a linear “causally connected process ... within which the earlier events caused the later” (79). “The same holds for a spread of adaptive features through a population” (79), i.e., the spread of adaptive features through a population is not a causal process wherein prior events precipitate subsequent events. Rather, natural selection transpires alike “thermodynamic change of state”, as illustrated by a standing pot of water coming to boil on a stove. Let us start by considering the case of a pot of water brought to a boil on a stove. At a phenomenal level, this seems like an orderly process. Heat flows from stove to liquid, the liquid circulates, carrying heat with it; gradually, it comes to a boil. Closer examination shows, however, that the transformation is not an orderly one at all. This is most evident perhaps in the actual phase transition, which is highly jerky. Parts of the liquid heave up and down; bubbles form more or less randomly. When the liquid is actually boiling, its surface is chaotic. A microscopic examination would display similar disorder at earlier stages. The energy transfer from the bottom of the pan to the top is disorderly, with myriad local exceptions (for example, adjacent small regions with unequal temperature with no mutual heat transfer), reversals (energy flow from low- to high-temperature regions), and other fits and starts. Such discontinuities violate the spatiotemporal continuity required of fundamental physical processes, which are strictly law governed. To halt, reverse, delay, or accelerate a fundamental process takes energy and work. But in the pot of water such changes of direction occur spontaneously. This shows that heat flow is not a fundamental process (79). The same is true of evolution. The discontinuities of natural selection are much more evident to the observer than those of thermodynamics, gross enough, in fact, to be recordable by a careful observer. These discontinuities show that natural selection, like thermodynamic change of state, is a time-asymmetric statistical trend instantiated by populations (80).
By taking up a boiling pot analogy, Matthen and Ariew reason that natural selection is not a simple basic linear process of causality, wherein prior events precipitate subsequent events, rather, natural selection is a statistical trend analogous to discontinuous trends transpiring, thermodynamically, in a boiling pot of water. For Matthen and Ariew the second law of thermodynamics, as illustrated by occurrences in a boiling pot of water, as a purely statistically describable phenomenon, is the closest theoretical model to what is actually transpiring in, or as, natural selection. (Given the structure of the state of affairs just described, wherein scholars of natural selection posit a thesis which subsequently encounters the positing of its antithesis, and, in turn, the antithetical position encounters its antithesis, is a multi-directional state of affairs which itself exhibits the entropy described by the second law of thermodynamics! Millstein will impart a synthetic order to this state of affairs.)
As a thinker long engaged within the entropic thetic-antithetic climate of the scholarly polemical dialectic in theory of natural selection, which is an ongoing dialectic among philosophers of biology, Dr. Roberta Millstein, a Ph. D. in biostatistics has arrived at a dialectically achieved realization that natural selection is “a statistical causal” (640) phenomenon. Accordingly, Millstein constructs an alternative theory of natural selection via the synthesis of statistical and causal elements of extant antithetical positions regarding what natural selection is. While other theorists, locked in an either/or conflict over dynamical versus statistical models of natural selection, are busy destroying each others theories, Millstein rationally concedes: 1) natural selection is not a force, i.e., “they (i.e., Matthen and Ariew) have shown that natural selection is not a force” (634), that is, the dynamical notion of “force” is destroyed on account of being theoretically unintelligible at the level of living organisms, and, 2) that natural selection is, indeed, statistical, for, Millstein declares “natural selection is a causal statistical process” (640). However, Millstein does not, will not, cannot, concede that dynamical causality has sustained theoretical destruction, for, according to her, the purely statistically-oriented thinkers “conflate the concept of ‘force’ with the concept of ‘cause’” (634), therefore, Millstein states “but this, (i.e., the showing that natural selection is not a force), fails to demonstrate that it (i.e., natural selection) is not a causal process” (634), and, accordingly, Millstein posits an argument demonstrating causal process to be indubitably operational in natural selection, which argument is supported by evidence garnered by the objective scientific biological field research of evolutionary biologists Elizabeth Dahlhoff and Nathan Rank.
Dahlhoff and Rank have shown a particular “mechanistic” phenomenon, i.e., a causal mechanism (10065) to be working, whereby a particular selective environmental factor (10056), i.e., ambient air temperature, is effecting a causal genetic agent (i.e., a gene a.k.a. the “PGI locus allele”) to precipitate an effect (i.e., heat shock protein production) within the protoplasm of a species of insect known as the montane willow leaf beetle (10056).
Dahlhoff and Rank conducted a series of unique censuses (Matthen, 83), of fractions of different montane willow leaf beetle populations, within the biosphere of certain mountainsides in the California Sierra Nevada range. The evolutionary biologists conducted studies in accord with their custom designed “natural- selection formula” (Matthen, 75), which took into account: ambient environmental air temperatures; resident beetle body temperatures; beetle genetic differences ; differential beetle heat shock protein expression; and, heritability of capacity for heat shock protein expression (Dahlhoff 10056). Rank determined that beetles of certain genetic configuration were, by virtue of their ability to manufacture heat shock protein, (a protein requisite for reconstituting other bodily proteins tend to disintegrate in extreme hot or cold), survived the extremes of mountain temperature. Rank identified a causal mechanistic process wherein certain pre-established genetic types of beetles were fit, heat-shock-protein- wise, to survive exposure to extreme variations in hot/cold. Rank therefore concluded that environmental temperature is a “selective factor” in natural selection of beetle genetic types adapted/able/fit to survive by virtue of a pre-capacity for heat shock protein expression, across a greater spectrum of temperatures than other genetic types of beetle. Here environmental temperature is acting as a “selective factor” upon a causal factor, which causal factor is that part of a beetles’ particular gene mediating manufacture of heat shock protein, whereby the beetle may survive radical extremes of temperature, reproduce, and pass a heat shock efficiency trait onto its progeny. A state of affairs wherein organisms survive by virtue of an advantageous trait, which trait is passed on to the next generation, is precisely natural selection. (Beetles fit to refold unfolding proteins in the most extreme conditions of ambient air temperature possess a genetic variation which is preserved precisely because via that variation survival is achieved, (recall Darwin’s definition, given above, which declares natural selection is the preservation of any slight advantageous variation). The natural selector here is environmental temperature. The selection here is for beetles possessing a pre-established ability to manufacture a given protein, and, against those who cannot manufacture the protein, across a sufficiently wide spectrum of ambient air temperatures which precipitate heat shock stress.
The results of an objective scientific field study, which demonstrated the existence of a mechanistic (i.e., causal) connection transpiring between environment and organism, constitutes a veritable broadsword in the hand of a philosopher of sufficient daring to take command, within a warring fractional theoretical fray, among diverse philosophers of biology. The power of the Dahlhoff-Rank study is that it constitutes solid evidence of a real and objectively demonstrated causal mechanism of natural selection at work in an actual population of organisms, providing a sharp-eyed philosopher of biology with biological evidence whereupon to predicate theory constitutive thought efficient to establish, factually, causal mechanism in natural selection, with the force of objective scientific evidence. What Millstein proceeds, therefore, to ordain, is a population-level construct of causality, and, a statistical causal construct of natural selection which, taken together, constitute both an alternative theory of what natural selection is, and a synthetic resolution of the dynamical versus statistical debate in philosophy of biology.
On Millstein’s synthetic view, natural selection is a statistical causal phenomenon wherein specific causal mechanisms which entail an environmental selective factor(s) (e.g., temperature) effecting a heritable bodily structure (e.g., a gene) to cause the sufficient manufacture, for example, of a given stress protein, or not. Which manufacture, or not, either preserves the life of a given population, or not. And, the differential ability to produce the requisite protein is a population-level phenomenon, for when one is speaking of differences, one is necessarily speaking of differences among ensembles of beings. For Millstein natural selection is a statistical phenomenon because, indeed, trends are transpiring, however, for Millstein those trends are not completely intelligible absent the specific causal mechanism exampled above (634). Millstein’s achievement is, thus, in part, the constitution of an alternate theory of what natural selection is, which alternate theory is a synthetic structure which integrates acausal statistical constructs and the causal constructs of dynamical theory, which integration constitutes a synthetic resolution of differences between differing factions, and, therefore, constitutes resolution of the statistical versus causal process debate in natural selection.
Thus, the answer to the enthymematical question(s) posed in the introduction, i.e., “Is theory of natural selection a statistical theory of a dynamical one?” That is, is natural selection theory a theory of actual dynamic environmental causal forces which have effects upon the protoplasm of organisms, or, is natural selection theory properly an acausal state of affairs which comprehends only statistical trends, absolutely random genetic events, and takes probabilistic census of the fitness of extant, and future, ensembles of organisms? This answer becomes: what natural selection is is a statistical-causal phenomenon, i.e., natural selection is statistical-trend and causal-process, in one. Natural selection is not either causal, or, statistical, it is both causal process and statistical trend.
Works Cited
- Ariew, Andre. “Adaptationism and its Alternatives: Explaining origins, prevalence, and diversity of organic forms.” The Online Writery. 2006. 17 pp. 20 Jan 2007
- Bouchard, F. and A. Rosenberg. “Fitness, Probability and the Principles of Natural Selection.” British Journal for the Philosophy of Science 55 (2004): 693-712.
- Dahlhoff, E. P. and N. E. Rank. “Functional and Physiological Consequences of Genetic Variation at Phosphoglucose Isomerase: Heat Shock Protein Expression Is Related to Enzyme Genotype in a Montane Beetle.” Proceedings of the National Academy of Sciences USA 97 (2000): 10056-61.
- Darwin, Charles. The Origin of Species By Means of Natural Selection. Chicago: U. of Chicago, 1987.
- Matthen, M., and A. Ariew. “Two Ways of Thinking About Fitness and Natural Selection.” The Journal of Philosophy 17 (2004): 55-83.
- Millstein, Roberta L. “Natural Selection as a Population-Level Causal Process.” British Journal for the Philosophy of Science 57 (2006): 627-65.
- Sartre, Jean-Paul. Critique of Dialectical Reason, Volume I Theory of Practical Ensembles. Paris: Editions Gallimard, 1960.
- Walsh, D. M., Tim Lewens, and A. Ariew. “The Trials of Life: Natural Selection and Random Drift.” Philosophy of Science 69 (September 2002): 429-446.