Philosophy of Biology
Philosophy of Biology
PHILOSOPHY OF BIOLOGY
Biology refers both to the systematic investigation of living things, and to the body of knowledge that is the product of that investigation. Throughout biology's history, however, some important questions debated by biologists have not been so much about the organisms being studied, but about the nature of life, the proper way to investigate it and the form biological knowledge should take. When inquiry shifts from questions about living things to questions about proper and improper ways of asking, or answering, or adjudicating, such questions, it shifts to a philosophical level. One need not, of course, be trained in a department of philosophy to contribute to such an inquiry. Indeed many of the most significant contributions to the subject have been made by people trained in the sciences. Nevertheless such contributions are to the subject designated as philosophy of biology.
One assumption implicit in the very name is that the biological sciences are distinctive enough from other sciences that a general inquiry into the nature of science will not suffice. It was common among logical empiricists to suppose it would—in texts written in that tradition one often finds the biological and social sciences dealt with in chapters late in general books in philosophy of science (Braithwaite 1953, Hempel 1966, Nagel 1961). Two early challenges to this assumption were The Ascent of Life by Thomas Goudge (1961) and The Biological Way of Thought by Morton Beckner (1959).
These two early contributors were followed by a number of introductions to philosophy of biology written in the 1970s and 1980s, most of which were focused narrowly on evolution and genetics, and a standard set of associated philosophical questions (Hull 1974, Ruse 1973, Rosenberg 1985, Sober 2000). But in 1982 an NEH Summer Institute in Philosophy of Biology organized by Richard Burian and Marjorie Grene attracted a group of philosophers ready to focus more or less exclusively on the biological sciences. In 1984 Philosophy of Science devoted a "special issue" to the philosophy of biology; shortly thereafter Michael Ruse played a pivotal role in organizing both a journal (Biology and Philosophy ) and a society (International Society for History, Philosophy, and Social Studies of Biology) devoted exclusively to the biological sciences. Since that time, the scope of research has broadened dramatically, with important contributions focusing on the biomedical sciences, physiology, cell biology, neurobiology, and developmental biology (Amundson 2005; Bechtel and Richardson 1993; Fox Keller 2000; Oyama, Griffiths, and Gray 2001; Robert 2004; Schaffner 1993; Sterelny and Griffiths 1999). And it is now common in general introductions to philosophy of science to have two chapters on philosophy of biology (for example, Salmon et al. 1992; Machamer and Silberstein 2002). In addition, there has been a tendency to integrate advances in the history of biology into these philosophical discussions.
This entry focuses on issues associated with three related biological domains: genetics, evolution, and development. Some of the most interesting recent philosophical work is focused on developmental biology and its relationship to the other two domains just mentioned. But important work is also being done on areas such as ecology, ethology, and neurobiology: each raises its own special philosophical questions.
Darwin, Mendel and a Partial Synthesis
Two publications in the mid-nineteenth century—Charles Darwin's On the Origin of Species and Gregor Mendel's "Experiments in Plant Hybridization"—were to have a lasting impact on the structure of the scientific study of life, an impact still evident in the way philosophers think about biology as a science. Darwin self-consciously characterized the theory of evolution by natural selection presented in his book as one that would provide a theoretical unity to the study of life. As he put it in a letter to the philosopher Sir John Herschel:
… I find so many young and middle-aged truly good workers in different branches, either partially or wholly accepting my views, because they find that they can thus group and understand many scattered facts. This has occurred with those who have chiefly or almost exclusively studied morphology, geographical Distribution, systematic Botany, simple geology & palaeontology.
(burkhardt 1994, pp. 135–136)
Darwin argued that central to explanations in all these domains were a set of "laws," which modern scholars identify as the principles at the core of the theory of evolution by natural selection:
The characteristics of the individual members of a species vary to a greater or lesser degree.
Some of that variation is heritable, transmitted from parents to off-spring.
Populations tend to increase at a geometric rate.
Struggle for existence
Given limited resources, predation, disease, and so on, the tendency to geometric increase is checked, leading to a struggle for survival.
Individuals with advantageous variations tend on average to survive longer and leave more off-spring.
The offspring of parents with advantageous variations tend to have the same advantageous variations.
Darwin used the term "Natural Selection" to refer to the last two principles: "I have called this principle, by which each slight variation, if useful, is preserved, by the term Natural Selection…" (Darwin 1859/1964, pp. 117, 127) And, since the above theory neither provides for the introduction of new variation nor for the divergence within species that actually leads to speciation, he followed the presentation of his theory with lengthy discussions of divergence of character and laws of variation. Crucially, he saw divergence leading to new species and higher taxa as simply long run extrapolations of the same processes that lead to the production of varieties within a species; and he decoupled the causes of new variation from the adaptive needs of the organism.
Modern presentations of the theory often reduce it to a combination of the production of heritable variation and the differential perpetuation of variation. Many philosophical problems emerge from this reduction, and a number of philosophers have been urging a formulation of the theory more in tune with Darwin's.
In 1866, just seven years after Darwin's Origin, a scientifically trained monk published the results of nine years of careful experimentation. Mendel's work was revolutionary both in its methods and its conclusions. Trained at the University of Vienna in experimental physics and statistics as well as botany (where he learned about recent developments in agricultural plant hybridization), Mendel realized that the combination of experimental controls and statistical analysis could be used to solve the puzzles of plant hybridization. In the varieties of pea plants with which he experimented, he established that a number of factors were inherited independently, that if one crossed plants with alternate forms of a factor (for example, green and yellow peas) all their offspring would appear like one or the other (the dominant form), and that the next generation of plants would reveal a ratio of the forms of approximately three dominant to one recessive. Further experimental analysis would reveal that the "dominant" plants were a predictable mixture of pure dominants and plants with a dominant and a recessive factor.
But he went beyond just outlining his experiments and providing plausible inductive inferences from them. In the last section of his paper he suggested an underlying causal mechanism that could account for the observed regularities that he had reduced to a mathematical law of the development of hybrids.
Largely ignored for the remainder of the nineteenth century, the basic idea was developed in quite different ways by many researchers from different disciplinary backgrounds in the first decade of the twentieth century. Scholars have from then until now referred to this theory of inheritance as Mendelian genetics, and the regularities he uncovered experimentally as "Mendel's laws." And because this theory had essentially two distinct components—one related to inheritance, and one related to development—it in theory provided a way of unifying branches of the study of life that Darwin admitted he had not. A number of key steps taken between 1905 and 1920 isolated Mendel's "factors" (genes) to chromosomes and paved the way for generalizing Mendel's principles of inheritance in a mathematical form that would allow their investigation in large populations that do not breed under strictly controlled experimental conditions. This permitted the integration of Mendelian genetics with Darwinism, and it is no surprise that the leading figures in creating this synthesis—J. B. S. Haldane, Ronald Fisher, and Sewall Wright—all had a passion for both mathematics and natural history (Provine 1971, Plutynski 2004, Sarkar 1992). And all had, by quite different routes, fallen under the spell of Darwin's theory of evolutionary descent driven primarily by natural selection.
The basic idea behind their synthesis was remarkably simple: think of populations of organisms in terms of the frequencies of the genes associated with the various traits found in those populations, and think of evolution in terms of gradual changes in those frequencies, under such influences as the migration of organisms in and out of the population, randomly occurring genetic mutations, genetic recombinations of various kinds and, above all, natural selection. Mathematical models were developed which permitted one to predict changes in the genetic make up of future populations given information (or, more often, assumptions) about these variables, the number of alleles of genes for given traits and assumptions about the relative fitnesses of different combinations of these alleles, know as "genotypes." The crucial step in developing these models was, of course, that each of these potential influences on gene frequencies was treated as a quantitative variable—including fitness.
It should be noted that all of these people—and the founder of experimental population genetics, Thomas Hunt Morgan, should be added—treat "genetics" as the study of the transmission of genes in reproduction. Indeed, even in the twenty-first century a synonym for Mendelian genetics is transmission genetics. The Evolutionary Synthesis did not include the study of development or developmental genetics, except for the use of embryological evidence in constructing evolutionary phylogenies. We will come back to this omission later.
Concluding the preface of Evolution: The Modern Synthesis, Julian Huxley declared: "The need today is for concerted attack and synthesis" (Huxley 1942, p. 8). That Synthesis was in the making when he wrote (the key publications by Theodosius Dobzhansky, Ernst Mayr, and George Simpson were published between 1937 and 1944), and by the 1959 centenary of the publication of Darwin's Origin it was declaring itself triumphant. Most of the philosophical issues related to evolutionary biology under discussion in the 2000s are a direct consequence of the form that the "Synthetic" theory takes This entry mentions five that are critical and discusses four of them in detail.
1) The concept of chance in evolutionary theory and the theory's probabilistic nature.
2) Fitness and selection.
3) Units and levels of selection.
4) The nature selection/adaptation explanations.
5) The ontological status of species and the epistemological status of species concepts.
the role of chance
Chance is a contrastive concept; to say that some outcome is chance is typically to deny that it resulted from some cause or other. In evolutionary theory "chance" plays a key role both in discussing the generation of variation and the perpetuation of variation (a distinction owed to John Beatty; see also Sober 1984, ch. 4). Consider the following variation grid, created by considering whether the contribution to fitness of a variation does or does not play a role in either the generation or the perpetuation of that variation:
The uniquely Darwinian position is that the generation of variation is chance in that it is not biased by fitness differences (as it is for Lamarckian theories), but the perpetuation of variation typically is biased by fitness differences. Neutralism, to be discussed shortly, will claim that a significant amount of evolutionary change is due to randomly generated variation that is perpetuated by chance as well.
But now consider the following discussion of chance and selection:
In Darwin's scheme of things, recall, chance events and natural selection were consecutive rather than alternative stages of the evolutionary process. There was no question as to which was more important at a particular stage. But now that we have the concept of random drift taking over where random variation leaves off, we are faced with just such a question. That is, given chance variations, are further changes in the frequencies of those variations more a matter of chance or more a matter of natural selection?
(beatty 1984, p. 196)
In the first two sentences, as often, the generation of variation is characterized as a "chance" process because selection plays no role at that stage—the generation of variations is not biased by the adaptive requirements of the organism. The concept of "random variation" is often used by neo-Darwinians as a synonym for "chance variation" in precisely this sense, as in the following from a product of Morgan's "fruit fly lab" and one of the architects of the evolutionary synthesis:
… mutation is a random process with respect to the adaptive needs of the species. Therefore, mutation alone, uncontrolled by natural selection, would result in the breakdown and eventual extinction of life, not in the adaptive or progressive evolution.
(dobzhansky 1970, p. 65)
The generation of variations is a "chance" process in the sense that the probability assignments are not biased by "adaptive needs" or "fitness."
The remainder of the quotation from Beatty concerns the perpetuation of variations, and in particular how to distinguish variations perpetuated by selection from those perpetuated by another process known as "random drift," in which traits that are selectively neutral may become fixed in a population simply as a result of what statisticians call "errors of sampling." Suppose, for example, that a pair of bats get blown to an island far away from their colony. They mate, and their offspring mate, and a number of genes become fixed in the growing populations simply because they were present in the founding pair, not because they are favored by selection. In the above quoted paper Beatty argues that "it is conceptually difficult to distinguish natural selection from random drift" (Beatty 1984, p. 196). As the entry discusses, this problem arises from a standard way of characterizing "fitness."
Genetic drift plays a critical role in one primary challenge to the neo-Darwinian synthesis, "neutralism," and the concept of chance is often used to draw the contrast. In the following quote, one prominent champion of the neutral theory of molecular evolution characterizes his position:
… the great majority of evolutionary changes at the molecular (DNA) level do not result from Darwinian natural selection acting on advantageous mutants but, rather, from random fixation of selectively neutral or very nearly neutral mutants through random genetic drift, which is caused by random sampling of gametes in finite populations.
(kimura 1992, p. 225)
Here genetic drift refers to a process whereby a selectively neutral allele becomes fixed in a population as a result of a "random (chance) sampling of gametes." This is a rival to neo-Darwinism only because of Kimura's claim that this produces a majority of the evolutionary changes at the molecular level. The contrast between "chance" and "fitness biased" processes is used by Kimura to distinguish means of perpetuating certain variations. We are contrasting two sampling processes.
There is currently a lively debate about whether to characterize this contrast by reference to differences in the sampling processes (Millstein 2002, 2005) or by reference to the expected outcomes of sampling (Brandon and Carson 1996, Brandon 2005). On Millstein's view it is realistically possible for the outcomes to be identical, and thus she seeks to defend a view according to which selection is defined as discriminate sampling (based on selectively relevant differences) and drift as indiscriminate sampling. Both samplings are "probabilistic," of course, but that in no way obviates the above contrast.
fitness and selection
All parties to this dispute are now realizing that wider issues about the nature of probability, explanation, mathematical abstraction, and causation are likely at stake in such disagreements. As one case in point, at least part of the dispute over differentiating drift from selection derives from the tendency to characterize natural selection so that it is indistinguishable from random drift (Brandon 1990, Lennox 1992, Lennox and Wilson 1994).
If we think of selection as a discriminative or biased sampling process, that natural raises the question of the basis of the biasing. Typically, the answer is that it is differences in fitness, the values assigned to different genotypes in the models of population genetics, which some readers will think of as different degrees of adaptation to the relevantly characterized environment.
But as noted above, it is not uncommon to find characterizations of the fitness of a genotype in terms of its relative contribution to the gene pool of future generations—the genotype contributing the larger percentage being the fitter. The expression "survival of the fittest" has essentially been eliminated from any serious presentation of Darwinian selection theory but the concept of "fitness" plays a prominent, and problematic, role. In the mathematical models used in population genetics "fitness" is represented by the variable W. Here is a rather standard textbook presentation of the relevant concepts:
In the neo-Darwinian approach to natural selection that incorporates consideration of genetics, fitness is attributed to particular genotypes. The genotype that leaves the most descendants is ascribed the fitness value W = 1, and all other genotypes have fitnesses, relative to this, that are less than 1. … Fitness measures the relative evolutionary advantage of one genotype over another, but it is often important also to measure the relative penalties incurred by different genotypes subject to natural selection. This relative penalty is the corollary of fitness and is referred to by the term selection coefficient. It is given the symbol s and is simply calculated by subtracting the fitness from 1, so that: s = 1 − W.
(skelton 1993, p. 164)
In this passage evolutionary advantage is equated with reproductive success and fitness is treated indifferently as a quantitative measure of both. But since, as we have seen, natural populations can evolve (via drift) in the absence of natural selection, and since balancing selection may prevent a population from evolving, it is clear that establishing, by measuring different reproductive rates among its members, that the genetic makeup of a population has changed does not establish that natural selection was the source of that change; nor does the fact that no change has been measured establish that natural selection is not operative.
The most widely accepted solution to this problem is to argue that fitness measures a reproductive propensity of organisms (Brandon 1978, Mills and Beatty 1979, Richardson and Burian 1992). Brandon tends to equate fitness in this sense with "adaptedness," and to contrast it with "realized fitness"—differences in realized fitness are explained by differential adaptation to a common selective environment. This suggests that fitness is in some sense relational, enhancing chances of reproducing relative to an environment (Lennox 1992, Lennox and Wilson 1994). In any case, as Millstein has insisted, characterizing fitness as a reproductive propensity raises the question of how to understand this propensity and its organic basis (Millstein 2003).
units and levels of selection
A number of challenges to Darwinian selection theory have emerged since the mid-twentieth century. Those challenges can be placed into two broad categories: (1) proposed limitations on natural selection as the primary cause of evolutionary change; and (2) expansions of the scope of natural selection to include new "targets" and "levels." It will be noted that in neither case is it obvious that the theory itself requires modification in the face of such challenges—in principle these might be nothing more than challenges to the theory's range of application. However, if it turned out that most evolutionary change could be explained without recourse to natural selection, this would be grounds for arguing that evolutionary biology was no longer Darwinian (see Godfrey-Smith in Orzack and Sober 2001.)
Darwin conceived of natural selection as almost exclusively an interaction between individual organisms and their organic and inorganic environments. Taking that as our starting point, we can see two challenges to Darwinism today with respect to the units of selection. One comes from those defending a strong form of genic selectionism, such as G. C. Williams (1966, 1992) and Richard Dawkins (1976, 1982), who argue that selection is always and only targeting genes. Here is a clear statement:
These complications [those introduced by organism/environment interactions] are best handled by regarding individual [organismic] selection, not as a level of selection in addition to that of the gene, but as the primary mechanism of selection at the genic level.
(williams 1992, p. 16)
Dawkins' preferred mode for making the same point is to refer to organisms—or interactors, to use language introduced by David Hull—as the vehicles of their genes (the replicators), in fact vehicles constructed by the genome for its own perpetuation. Neither Williams nor Dawkins deny that there is interaction between phenotype and environment that plays a role in the differential perpetuation of genes. Their argument is that those interactors are part of the "genic selection mechanism," as Williams worded it above.
This view has been extensively challenged by philosophers of biology on both methodological and conceptual grounds (Brandon 1996, ch. 8; Mitchell 2003, ch. 4; Moss 2003, ch. 1; Sober 1984, chs. 3, 7; Sterelny and Griffiths 1999, chs. 4–5), though there are, among philosophers, also enthusiastic supporters (Dennett 1995). In all the give and take, it is seldom noticed how odd it is that defenders of this view claim to be carrying the Darwinian flag (Gayon 1998 and Gould 2003 are exceptions). Yet it is certainly not a position that Darwin would recognize—and not merely because he lacked a coherent theory of the units of inheritance. It is not a Darwinian view because for Darwin it was differences in the abilities of organisms at various stages of development to respond to the challenges of life that had causal primacy in the explanation of evolutionary change. Gene selectionism was explicitly challenged on these grounds by key figures in the Synthesis (for example, Ernst Mayr).
The Darwinian view of the units of selection also has challenges from the opposite direction. In the 1970s a number of biologists working in the fields of paleontology and systematics challenged the Neo-Darwinian dogma that you could account for "macro-evolution" by simple, long-term extrapolation from the processes modeled by population genetics. (The case was enhanced by parallel and contemporaneous developments in embryology and functional morphology that are discussed in the last section of this entry.) Stephen Jay Gould (2003), in a chapter titled "Species as Individuals in the Hierarchical Theory of Selection" combines two conceptually distinct theses: first, the thesis defended by Michael Ghiselin (1997) and championed and refined by David Hull (2001), that species are in a robust sense of the term "individuals"; and second, that there may well be selection among groups of organisms, qua groups. This approach brings us to the brink of problem (5) on the list, how to understand the species category and species as taxa, questions discussed only briefly.
Gould exemplifies one approach to group selection—the unit of selection is always the individual, but there are individuals other than individual organisms that are subject to selection. A very different result emerges if one assumes that groups of organisms such as demes, kin-groups, or species, though not individuals, are nevertheless possible units of selection. Adding to the conceptual complexity, some researchers propose that "group selection" be restricted to the process whereby group-level traits provide advantages to one group over another, in which case there are strict conditions delimiting cases of group selection, while others focus solely on group level effects. Thus a debate analogous to that earlier discussed regarding the definitions of "fitness" emerges here—by group selection do we mean a distinct level of causal interaction, or merely a tendency within certain populations for some well defined groups to displace others over time? (For further discussion, see Sterelny and Griffiths 1999, 151–179; Hull 2001, 49–90.) It is now common to characterize "selection," "interactor," and "replicator" abstractly and to specify the conditions under which an entity is properly identified as a unit of selection. This allows one to leave it an open and essentially empirical question whether, under the right conditions a particular "unit" could be subject to selection. With the modular picture of development that is emerging, the "developmental module" will likely be added to the list.
selection, adaptation, and teleology
Perhaps the central promise of Darwinism, and the reason it was rightly seen as a challenge to the Argument from Design for a benevolent creator, was that it provides a scientific explanation for both phylogenic continuity and adaptive differentiation by means of the same principles. The nature of "selection explanations" is a topic to which much philosophical attention has been devoted in recent years (Allen, Bekoff, and Lauder 1998; Sober 1984). How does one account for the apparently teleological character of explanation by natural selection?
The appearance of teleology is certainly present in Darwinian explanations, and has been since Darwin spoke of natural selection working solely for the good of each being. The appearance of teleology stems from the ease with which both evolutionary biology and common sense take it for granted that animals and plants have the adaptations they do because of some benefit or advantage to the organism provided by those adaptations. But in what sense can the adaptive advantage be the cause of the presence of the adaptation?
Some insist it cannot (Ghiselin 1997). Others argue that such explanations are actually masked appeals to the past effects of selection (see the papers in Allen, Bekoff, and Lauder, section 3). This entry sketches a case that shows selective explanations of adaptations are robustly teleological (see Lennox 1992, 1993; and the papers in Allen, Bekoff, and Lauder 1998, section 1).
Are the functions performed by confirmed adaptations a central and irreducible feature of explanations of the presence of those adaptations? If the answer is yes, the explanations are teleological. Take the following example. In research combining painstaking field work and laboratory experimentation, John Endler demonstrated that the color patterns of males in certain Caribbean guppy populations resulted from a balance of mate selection and predator selection. For example, he demonstrated that a group of males with a color pattern that matched that of the bottoms of the streams and ponds they populated except for bright red spots have that pattern because a common predator in those populations, a prawn, is color blind for red. Thus red spots did not put their possessors at a selective disadvantage, and were attractors for mates\(Endler 1983).
Their pattern of coloration was a complex adaptation serving the functions of predator avoidance and mate attraction—and it is an adaptation, as that term is used in Darwinism, only if it is a product of natural selection (Williams 1966, Brandon 1985, Burian 1983). In order for it to be a product of natural selection, there must be an array of color variation available in the genetic/developmental resources of the species wider that this particular pattern but including this pattern. Which factors are critical, then, in producing differential survival and reproduction of guppies with this particular pattern in a shared homogeneous environment? The answer would seem to be the value -consequences this pattern has compared to others available in promoting viability and reproduction. In popular parlance (and the parlance favored by Darwin), this color pattern is good for the male guppies that have it, and for their male offspring (Binswanger 1990, Brandon 1985, Lennox 1992). This is a robust version of "consequence etiology" accounts of selection explanations (Bekoff, Allen, and Lauder 1998, section 1), which stresses that selection ranges over value differences which are causally relevant to one among a number of color patterns having a higher fitness value. Selection explanations are, then, a particular kind of teleological explanation, an explanation in which that for the sake of which a trait is possessed, its valuable consequences (avoiding predation, attracting mates), account for the trait's differential perpetuation and maintenance in the population.
species and taxonomy
Darwin at one point in the Origin says that he considers the term "species" one that is given arbitrarily, for convenience. He based that comment on a review of the taxonomic work of his day, and a similar review today would have the same result. Equally competent taxonomists will disagree about whether to rank a group of similar organisms as members of the same species or as members of two distinct species. This issue takes on philosophical import because speciation—the "origin of species," to use Darwin's language—is taken to be the key step in the evolution of life. One would hope to have a clear way of deciding, at least in principle, when that step has been completed! But every attempt to give a clear account of what makes a taxonomic unit a member of the species category runs up against rather compelling problems. Surrounding this topic, which has generated an enormous literature, are both epistemological issues regarding the basis for our species concepts and ontological issues about the nature of species. Interested readers should consult the work of Marc Ereshefsky (1992, a collection of essays defending various views of the species category) or Kim Sterelny and Paul Griffiths (1999, chapter 9; a readable and current overview of the issues).
In standard texts in the philosophy of biology in the 1970s and 1980s (as well as in most of the more technical journal articles) genetics played a key role in the discussion of two philosophical topics: reductionism, and the structure of evolutionary theory (Hull 1974, ch. 1; Schaffner 1969; Ruse (1973), ch. 10; Rosenberg 1985, ch. 4). The discussion began by importing a theory of reduction that had been developed with physical theories in mind, and asking whether, on such models, there had been a reduction of Mendelian or transmission genetics to the molecular level. This model, developed most clearly by Ernst Nagel (1961), imagined two theories formalized with axioms and laws. Reduction would require that the laws of one theory be, in some clear sense, deducible from the fundamental laws of the other, as, with appropriate corrections, Kepler's planetary laws could be from those of Newtonian celestial mechanics. Typically, this would also require that the key concepts in the two theories be interdefinable. This model was developed into a "general reduction/replacement model" by Kenneth Schaffner (1969) in a paper in which he argued for the potentially application of such a model to the case of genetics.
David Hull (1974) pointed out that a critical problem in the way of achieving this goal was that the two theories had essentially different goals and domains—hereditary transmission of differences versus genetic input into the biochemistry of development. This suggested to him not only the impossibility of a reduction, but its irrelevance to biology. All parties to this discussion concluded that one needed a much more elaborate account of both biological theory structure and explanation to even try to answer the question.
A number of recent discussions have stressed that understanding both biological investigation and explanation in terms of mechanisms and their operations provides a more realistic picture of fields such as neurobiology and molecular biology (see Machamer, Darden, and Craver 2000; Waters 1994, 2000). It may also provide a more tractable notion of "reduction" in terms of "underlying mechanisms." Detailed histories of the development of genetics played a very important role in this discussion, and thus it was one important area driving the integration of history and philosophy of biology. A fine review of that topic, as well as a carefully hedged defense of genic reductionism which takes into account the complex, interlevel nature of typical biological theories, can be found in Schaffner 1993, chapter 9 (and see Waters 1994 for a somewhat different defense).
One of the puzzles that emerges from reviewing the literature on genetics and reduction is that Hull's point, mentioned above, about the fundamentally different aims of molecular genetics and transmission genetics only really gets serious consideration once developmental biology comes to the fore in the 1990s (see Waters 1994 and the papers in Beurton, Falk, and Rheinberger 2000). This is still very much a discussion in process, so the entry touches on some of the philosophical and historical questions being raised about different uses of the gene concept (or, alternatively, different gene concepts) in evolutionary and developmental contexts.
The traditional gene concept associated with Mendelian genetics that formed the basis of evolutionary biology was important because it was the basis of heritable differences in populations. Genes, or more precisely alleles, were the sources of heritable variation in populations, and thus provide "the material basis for evolution." In the context of developmental biology, however, the focus of research has always been on the genes as sources of deep relationships among species within and even across phyla. Here is a succinct expression of the difference:
In the Modern Synthesis of population genetics and evolution, genes become manifest by differences in alleles that are active in conferring differential reproductive success in adult individuals. The gene is though to act as a particulate, atomic unit. In current syntheses of evolution and developmental genetics, important genes are manifest by their similarities across distantly related phyla, and they are active in the construction of embryos. These developmental genes are thought to act in a context-dependent network.
(scott gilbert, in beurton, falk, and rheinberger 2000, p. 178)
In a defense of genic selectionism George Williams argues that genes should be understood as units of information.
Only DNA provides the durable archive for most of the earth's organisms. This constraint should not blind us to the fact that it is information we are concerned with, and that DNA is the medium, not the message. A gene is not a DNA molecule; it is the transcribable information coded by the molecule.
(williams 1992, p. 11)
Williams praises philosophers for adopting the distinction between replicators and interactors discussed earlier, but he is critical of them for regarding "replicators as material objects and miss[ing] the codex concept" (Williams 1992, p. 12). Notice that what this approach does is allow nominal acceptance of advances in our understanding of cellular mechanisms at the molecular level while continuing to treat the gene as an "atomic" unit differentiated by reference to phenotypic differences. That is, development can continue to be "black boxed" by taking the gene to be any selectively relevant bit of the "codex" for the organism. In principle it should allow Williams to take on board a suggestion made by Sterelny and Griffiths; on grounds that lots of things get replicated in reproductive cycles "gene selectionism should be generalized to 'replicator selectionism'"(Sterelny and Griffiths 1999, p. 69).
However, taking this approach also raises a new set of concerns, namely those involved with the application of concepts from information theory in the characterization of genes and gene action. This way of talking became extremely popular after the "breaking of the genetic code" in the 1960s. Complementary strands of this "double helix" consist of only four bases, two purines (Adenine [A] and Thymine [T]) and two pyramidines (Guanine [G] and Cytosine [C]); and since proteins consist of polypeptide chains made from only twenty different amino acids, if DNA is to contain the "instructions" for synthesizing all the possible proteins, the simplest possible "code" would be one in which three bases combined to specify each amino acid. The bases came to be represented as the "letters" of the genetic "alphabet"; they combine into syllables, words and "reading frames"—the book of life! The coded script is "transcribed" and "translated"; there is an "encoding" and "decoding" process; and with the discovery of the complexities of DNA transcription, it is not surprising that terms like "editing" and "proofreading" got added. It is not inevitable that these metaphors should lead researchers to present the genome as both the architect and the blueprint for building an organism—but it was natural. (For a compelling story of the history, see Keller 2000.)
Are there problems with it? A number of philosophers of biology think so, and they are discussed in this entry in two parts. The entry first discusses those problems that are not specific to developmental biology, and then discusses the philosophical debates around Developmental Systems Theory and "evo-devo."
The aforementioned quotation from George Williams (1992) establishes that some evolutionary biologists want to take the information metaphor one step further, and allow it to float free from its source in the discovery of the relationship between DNA sequence and amino acid differences. Genes are units of information, pure and simple. The value of doing this is that it allows one to avoid the troubling fact that the causal complexity of the processes involved in biological development make it quite meaningless to talk about some relatively short and self-contained DNA sequence as a gene for anything other than an amino acid, perhaps.
What could "information" be, in this case? It seems not to be information in the sense of mathematical information theory. One suggestion is to see it from a "teleosemantic" point of view; that is, genes are something akin to units of "meaning," their meaning being what they are present for, the phenotypic trait whose selection insures the replication of that gene (Sterelny and Griffiths 1999, pp. 82–92). Critics have argued that this simply severs completely the causal connection between DNA sequences and phenotypic traits. And it looks as if the original impulse for gene selectionism gets lost. It looks like the interactor—whether it is a colony, an animal, or a gamete—is the only serious causal determinant of differential reproduction. "A purely functional notion of a gene, untied to anything constant at the molecular level, is not a definition suitable for gene selection theory, whatever its other uses might be" (Sterelny and Griffiths 1999, p. 90).
A very different defense of the language of information in biology would in fact tie it very tightly to the detailed machinery of molecular biology, and therefore takes seriously the role of the analogies based on this language in the development of molecular genetics, such as treating DNA as a "reading frame" made up of triplet units, which then allows one to see certain mutations as analogous to "frame shifts" that create nonsense (Maynard Smith 2000, p. 184). But even here, the sense of "information" that seems relevant to the analogy is again a semantic notion tied to meaning and intentionality, not that of the 'signals', 'channels', and 'sources' of "information theory." And thus all the problems associated with that notion are still present. (See the replies to Maynard Smith in Godfrey-Smith 2000; Sarkar 2000.) There is consensus here, however, that the language associated with codes and information storage and retrieval was extremely important in the development of molecular biology. Insofar as there is disagreement, it is over whether these metaphors have outlived their usefulness and are now in fact the source of significant misunderstanding.
The Challenge of Development
Much recent philosophy of biology has focused on the process of development. There are at least two reasons for this. First, the model of the gene described in the previous section is deeply problematic, and one response was the philosophical defense of a developmentalist alternative based on the work of Susan Oyama, known as Developmental Systems Theory or DST. In an important and productive exchange on "the developmentalist challenge" to this sort of "genetic primacy," Ken Schaffner (1998) identified eleven theses of DST, and focused on four with which he thinks serious researchers in biomedical molecular genetics would, in one form or another, agree. The basic idea is that development is a product of a complex time series of interactions among many cellular and extra-cellular factors, among which "genes" (and the quotation marks are important) are just one. As a consequence, DST denies what I have called "information theoretic determinism." In so far as the information metaphor has value (this is currently much disputed), it is applicable only to the developing system—genes carry information, if at all, only as aspects of developmental systems.
Another obvious consequence of DST is the rejection of various common themes of behavioral genetics (Lewontin 1995). Schaffner's approach to these DST theses is to compare them with work done on a "simple model system," one of the organisms at the center of human genome research, the nematode worm C. elegans. This worm became a model organism due to Sidney Brenner adopting it to investigate the development, from zygote to mature adult, of an entire nervous system (and behavior). It was ideal for many reasons, not the least of which was its simplicity—a nervous system containing only 302 neurons forming roughly 5,000 synapses. Schaffner compares the DST theses about genes, development and behavior with the results of the massive, worldwide research assault on C. elegans to see how they hold up. This entry cannot follow the details, but one can see from his eloquent conclusion that at least some of the DST argument is acceptable to him:
Characterizing simple "genes for" behaviors is, accordingly, a drastic oversimplification of the connection between genes and behavior, even when we have the (virtually) complete molecular story. The melody of behavior represents no solo performance—it is the outcome of an extraordinarily complex orchestra—and one with no conductor.
(schaffner 1998, p. 247)
This paper was the target for responses from philosophers of biology and biologists more or less sympathetic to DST (Gilbert and Jorgensen 1998, Wimsatt 1998) and Schaffner was given the last word in reply. This selection of papers constitutes the best introduction to the DST reply to gene-centered research (see Waters 2005).
Independently of philosophical discussion of DST, there are compelling reasons for philosophers to be interested in evolutionary developmental biology, or "Evo-devo." Given the long history of both developmental biology and evolutionary biology, and the long history of their interactions, one might wonder why the goal of integration has appeared on the horizon only in the twenty-first century. The answer is a complex of historical, philosophical, and biological components.
According to one historical narrative (Beurton, Falk, and Rheinberger 2000; Burian 2005), the rapid development of new investigative techniques in molecular biology, driven in part by the medical and agricultural potential of the methods of genetic modification, and the field of "genomics" that evolved along with the Human Genome Project, provided the means for investigating development at the molecular level. This gave rise to a number of quite revolutionary discoveries; this entry notes only two: (1)The "Hox" regulatory genes encode a special sort of protein with a stretch of amino acids known as a "homeodomain." These proteins attach to quite specific segments of DNA, regulating the expression of a series of genes. These proteins act in concert and have "modular" effects on such things as organ formation, body segmentation and bilateral duplication of body parts (a clear introduction can be found in Burian 2005, chs. 11 and 12). (2) Molecular genetics is providing a highly complex, "interactive" picture of gene regulation. It will be noted that in the description of the Hox genes it became clear that certain proteins were responsible for their regulation. In fact all sorts of signals, some coming from within the cell and some from the extracellular environment, play a role in gene expression. This is now so widely accepted that philosophers and historians refer to it as "the interactionist consensus." According to this picture, genes are one of many interacting factors all of which must play their roles in order to give rise to an organism—the study of this interactive process is termed "epigenetics," though it is unclear to what extent its practitioners understand development as a truly epigenetic process (Robert 2004, ch. 1).
On this view, the integration of evolutionary and developmental biology will be—is being—effected by the long overdue integration of molecular genetics, and the molecular understanding of development, into evolutionary studies. At least one advocate of this view (Burian 2005) has stressed the modularity of this view of development, and the implication of the semi-autonomy of these developmental modules (body segmentation and bilateralism, organ systems, limb structure, and so on) for the way evolution can possibly work.
There is another way of viewing the history, being developed in different ways by Alan Love and Jason Robert (Love 2003, Robert 2004). Love focuses on what the proponents of Evo-devo claim their investigations can do that the current evolutionary "synthesis" cannot. Many proponents of this field put the explanation of evolutionary novelties such as feathers, tetrapod limbs, or jaws as the central contribution of development to evolution. While they are happy to concede to population genetics and ecological genetics the explanation of gradual evolutionary changes in traits associated with one or a few Mendelian genes, they argue that the explanation of the appearance of novelties at particular phylogenetic junctures requires an understanding of the network of changes in the organization of developmental resources needed to produce the novel structure, and an understanding of its functional morphology. The history of work on evolutionary novelties focuses attention on a number of research programs in developmental biology, functional morphology and paleontology, all focused on understanding the first appearance of novel structures and behaviors—and all more or less ignored by the evolutionary synthesis.
Jason Robert argues for the primacy of the organism as it develops from zygote to maturity, and thus for a seriously "top down" or "whole to part" view of developmental causation. This allows us to see the analytic tools that allow us to understand the details of the developmental mechanisms as a first step, with true understanding of development coming when we have an integrated understanding of how those mechanisms interact. Pretty much everyone looking at this rapidly developing area of biology agrees with the following sentiment from Burian:
During the next few decades, I believe, biologists will highlight the roles played in constructing organisms by dynamic regulatory systems above the level of the genome. The result will be a nonvitalist but much more holistic, vision of the organism, one that places the integration of the organism at the focus of attention. In short, our new understanding of the apparatus regulating gene expression has undermined classical genetic determinism.
(burian 2005, p. 243)
As Robert as pointed out, this prediction for the future sounds remarkably like a return to the "organismic" biologists, such as E. S. Russell, writing in the 1920s and 1930s, against the then rising tide of a population genetic centered evolutionary synthesis (Robert 2004). There are, of course, critics of this viewpoint. While the aforementioned text indicated that Schaffner's review of C. elegans research encouraged him to accept, at least in a modified form, some of the theses of "the developmentalist challenge" to genetic determinism, the modifications were significant. And some would likely say he has gone too far, arguing that what we have in this new molecular understanding of development is a vindication of reductionism (Waters 1994, 2004, 2005; Rosenberg 1985, discussed in Robert 2004, pp. 12–15).
Evo-devo once again brings into focus the question of the unity of biology as a science. As stressed earlier, one thing that the evolutionary synthesis provided for philosophers of biology was an image of how the biological sciences could be unified that was decidedly unlike the standard models based on the physical sciences. The attempt to unify evolutionary biology and developmental biology may complicate that image considerably. The fields omitted from the synthesis share key concepts (for example, gene, homology) with evolutionary biology, but appear to deploy them in very different ways. Moreover, the methods of investigation in functional morphology, developmental biology and population genetics or ecology are extremely different. The central problems and questions to be answered are very different, because the basic research agendas of the fields are very different. A field that focuses on "the production of the tetrapod limb" and a field that thinks of populations as gene pools of heritable variation being sampled by selection do not appear to look at organisms in the same way (Amundson, in Orzack and Sober 2001; Love 2003). As this proposed "synthesis" or "integration" takes place, philosophers of biology can both test their models of theoretical unification against the accomplishment of evo-devo, and can provide its advocates with ideas about adequacy conditions for a successful integration. One thing appears certain at this point: evo-devo specialists who have explicitly written on this topic see a special set of problems that will require an integration of concepts and techniques from evolutionary biology and developmental biology; they do not imagine one field being gradually "reduced" to another.
What, then, are the logical and conceptual prerequisites for such an integrated investigation? If we look back to where we started in this entry, it will be recalled that the "integration" of Darwinism with the Mendelian genetics of populations, required the concepts of "fitness" and "selection" to be reshaped into a mathematical form; and what began as a cytologically and developmentally based genetics eventually "black-boxed" development in the interests of focusing on the transmission of genetic "information" from one generation to the next. These changes, in the interests of integration or synthesis, gave rise to a host of philosophical problems. Perhaps, with philosophers and historians inextricably involved with this new synthesis, at least some problems can be avoided.
Allen, Collin, Mark Bekoff, and George Lauder, eds. Nature's Purposes: Analyses of Function and Design in Biology. Cambridge, MA: MIT Press, 1998.
Amundson, Ronald. The Changing Role of the Embryo in Evolutionary Thought. Cambridge, U.K.: Cambridge University Press, 2005.
Ayala, Francisco. "Teleological Explanations in Evolutionary Biology." In Nature's Purposes: Analyses of Function and Design in Biology, edited by Collin Allen, Mark Bekoff, and George Lauder. Cambridge, MA: MIT Press, 1998.
Beatty, John. "Chance and Natural Selection." Philosophy of Science (Special Issue: Philosophy of Biology) 51 (2) (1984): 183–211.
Bechtel, William, and Robert Richardson. Discovering Complexity. Princeton, NJ: Princeton University Press, 1993.
Beurton, Peter, Raphael Falk, and Hans-Jörg Rheinberger, eds. The Concept of the Gene in Development and Evolution. Cambridge, U.K.: Cambridge University Press, 2000.
Binswanger, Harry. The Biological Basis of Teleological Concepts. Los Angeles: ARI Press, 1990.
Braithwaite, Richard B. Scientific Explanation. Cambridge, U.K.: Cambridge University Press, 1953.
Brandon, Robert. Adaptation and Environment. Princeton, NJ: Princeton University Press, 1990.
Brandon, Robert. "Adaptation Explanations: Are Adaptations for the Good of Replicators or Interactors?" In Evolution at a Crossroads, edited by Bruce Weber and David Depew. Cambridge, MA: MIT Press, 1985.
Brandon, Robert. Concepts and Methods in Evolutionary Biology. Cambridge, U.K.: Cambridge University Press, 1996.
Brandon, Robert. "The Difference between Drift and Selection: A Reply to Millstein." Biology and Philosophy 20 (1) (2005): 153–170.
Brandon, Robert, and Richard Burian, eds. Genes, Organism, Populations: Controversies over the Units of Selection. Cambridge, MA: MIT Press, 1984.
Brandon, Robert, and Scott Carson. "The Indeterministic Character of Evolutionary Theory: No 'No Hidden Variables' Proof but No Room for Determinism Either." Philosophy of Science 63 (3) (1996): 315–337.
Burkhardt, Frederick, et al. The Correspondence of Charles Darwin. Vol. 9. Cambridge, U.K.: Cambridge University Press, 1994.
Burian, Richard. "Adaptation." In Dimensions of Darwinism, edited by Marjorie Grene. Cambridge, U.K.: Cambridge University Press, 1983.
Burian, Richard. The Epistemology of Development, Evolution, and Genetics: Selected Essays. Cambridge, U.K.: Cambridge University Press, 2005.
Darden, Lindley, and J. Cain. "Selection Type Theories." Philosophy of Science 56 (1989) 106–129.
Darwin, Charles. On the Origin of Species: A Facsimile of the First Edition (1859). Cambridge, MA: Harvard University Press, 1964.
Dawkins, Richard. The Extended Genotype. Oxford: Oxford University Press, 1982.
Dawkins, Richard. The Selfish Gene. Oxford: Oxford University Press, 1976.
Dennett, Daniel. Darwin's Dangerous Idea. New York: Simon & Shuster, 1995.
Depew, David, and Bruce Weber. Darwinism Evolving. Cambridge, MA: MIT Press, 1995.
Dobzhansky, Theodosius. Genetics of the Evolutionary Process. New York: Columbia University Press, 1970.
Endler, John. "Natural and Sexual Selection on Color Patterns in Poeciliid Fishes." Environmental Biology of Fishes 9 (2) (1983): 173–190.
Ereshefsky, Marc. The Units of Evolution: Essays on the Nature of Species. Cambridge, MA: MIT Press, 1992.
Gayon, Jean. Darwinism's Struggle for Survival. Cambridge, U.K.: Cambridge University Press, 1998.
Ghiselin, Michael. Metaphysics and the Origin of Species. Albany, NY: Student University of New York Press, 1997.
Gilbert, Scott, and Erik Jorgensen. "Wormwholes: A Commentary on K. F. Schaffner's 'Genes, Behavior, and Developmental Emergentism.'" Philosophy of Science 65 (2) (1998): 259–266.
Godfrey-Smith, Peter. "Information, Arbitrariness, and Selection: Comments on Maynard Smith." Philosophy of Science 67 (2) (2000): 202–207.
Godfrey-Smith, Peter. "Three Kinds of Adaptationism." In Adaptationism and Optimality, edited by Steven Orzack and Elliott Sober. Cambridge, U.K.: Cambridge University Press, 2001.
Goudge, Thomas. The Ascent of Life. Toronto: University of Toronto Press, 1961.
Gould, Stephen Jay. The Structure of Evolutionary Theory. Cambridge, MA: Harvard University Press, 2003.
Griffiths, Paul, and Robin Knight. "What Is the Developmentalist Challenge?" Philosophy of Science 65 (2) (1998): 253–258.
Hempel, Carl. The Philosophy of Natural Science. Englewood Cliffs, NJ: Prentice-Hall, 1966.
Hull, David. The Philosophy of Biological Science. Englewood Cliffs, NJ: Prentice-Hall, 1974.
Hull, David. Science and Selection: Essays on Biological Evolution and the Philosophy of Science. Cambridge, U.K.: Cambridge University Press, 2001.
Hull, David, and Michael Ruse, eds. The Philosophy of Biology. Oxford: Oxford University Press, 1998.
Huxley, Julian. Evolution: The Modern Synthesis. London: George, Allen and Unwin, 1942.
Keller, Evelyn Fox. The Century of the Gene. Cambridge, MA: Harvard University Press, 2000.
Keller, Evelyn Fox, and Elisabeth Lloyd, eds. Keywords in Evolutionary Biology. Cambridge, MA: Harvard University Press, 1992.
Kellert, Stephen, Helen Longino, and Kenneth Waters, eds. Scientific Pluralism. Minneapolis: University of Minnesota Press, 2005.
Kimura, Motoo. "Neutralism." In Keywords in Evolutionary Biology, edited by Evelyn Fox Keller and Elisabeth Lloyd. Cambridge, MA: Harvard University Press, 1992.
Lennox, James G. "Darwin Was a Teleologist." Biology and Philosophy 8 (4) (1993): 408–421.
Lennox, James G. "Philosophy of Biology." In Introduction to the Philosophy of Science, edited by Merrilee Salmon et al. Englewood Cliffs, NJ: Prentice-Hall, 1992.
Lennox, James G. "Teleology." In Keywords in Evolutionary Biology, edited by Evelyn Fox Keller and Elisabeth Lloyd. Cambridge, MA: Harvard University Press, 1992.
Lennox, James G. "Teleology by Another Name: A Reply to Ghiselin." Biology and Philosophy 9 (4) (1994): 493–495.
Lennox, James G., and Bradley Wilson. "Natural Selection and the Struggle for Existence." Studies in History and Philosophy of Science 25 (1) (1994): 65–80.
Lewontin, Richard. Human Diversity. New York: Scientific American Library, 1995.
Love, Alan. "Evolutionary Morphology, Innovation, and the Synthesis of Evolutionary and Developmental Biology." Biology and Philosophy 18 (3) (2003): 309–345.
Machamer, Peter, Lindley Darden, and Carl Craver. "Thinking about Mechanisms." Philosophy of Science 67 (1) 2000: 1–25.
Machamer, Peter, and Michael Silberstein. The Blackwell Guide to the Philosophy of Science. Oxford: Blackwell Publishers, 2002.
Maynard Smith, John. "The Concept of Information in Biology." Philosophy of Science 67 (2) (2000): 177–194.
Millikan, Ruth Garrett. "In Defense of Proper Functions." In Nature's Purposes: Analyses of Function and Design in Biology, edited by Collin Allen, Mark Bekoff, and George Lauder. Cambridge, MA: MIT Press, 1998.
Millstein, Roberta. "Are Random Drift and Natural Selection Conceptually Distinct?" Biology and Philosophy 17 (2002): 33–53.
Millstein, Roberta. "Selection vs. Drift: A Response to Brandon's Reply." Biology and Philosophy 20 (1) (2005): 171–175.
Mitchell, Sandra. Biological Complexity and Integrative Pluralism. Cambridge, U.K.: Cambridge University Press, 2003.
Mitchell, Sandra. "Functions, Fitness, and Disposition." In Nature's Purposes: Analyses of Function and Design in Biology, edited by Collin Allen, Mark Bekoff, and George Lauder. Cambridge, MA: MIT Press, 1998.
Moss, Lenny. What Genes Can't Do. Cambridge, MA: MIT Press, 2003.
Nagel, Ernst. The Structure of Science. New York: Harcourt, Brace and World, 1961.
Orzack, Steven, and Elliott Sober, eds. Adaptationism and Optimality. Cambridge, U.K.: Cambridge University Press, 2001.
Oyama, Susan, Paul Griffiths, and Russell Gray. Cycles of Contingency: Developmental Systems and Evolution. Cambridge, MA: MIT Press, 2001.
Plutynski, Anya. "Explanation in Classical Population Genetics." Philosophy of Science (Proceedings of the 2002 Biennial Meeting of the Philosophy of Science Association) 71 (5) (2004): 1,201–1,214.
Provine, William. The Origins of Theoretical Population Genetics. Chicago: University of Chicago Press, 1971.
Robert, Jason. Embryology, Epigenesis, and Evolution: Taking Development Seriously. Cambridge, U.K.: Cambridge University Press, 2004.
Rosenberg, Alexander. The Structure of Biological Science. Cambridge, U.K.: Cambridge University Press. 1985.
Ruse, Michael. The Philosophy of Biology. London: Hutchinson and Company, 1973.
Salmon, Merrilee, et al. Introduction to the Philosophy of Science. Englewood Cliffs, NJ: Prentice-Hall, 1992.
Sarkar, Sahotra, ed. The Founders of Evolutionary Genetics: A Centenary Reappraisal. Dordrecht: Kluwer Academic Publishers, 1992.
Sarkar, Sahotra. Genetics and Reductionism. New York: Cambridge University Press, 1998.
Sarkar, Sahotra. "Information in Genetics and Developmental Biology: Comments on Maynard Smith." Philosophy of Science 67 (2) (2000): 208–213.
Schaffner, Kenneth. Discovery and Explanation in Biology and Medicine. Chicago: University of Chicago Press. 1993.
Schaffner, Kenneth. "Genes, Behavior, and Developmental Emergentism: One Process, Indivisible?" Philosophy of Science 65 (2) (1998): 209–252.
Skelton, Peter. Evolution: A Biological and Palaeontological Approach. Harlow, U.K.: Pearson Education Ltd., 1993.
Sober, Elliott. The Nature of Selection. Cambridge, MA: MIT Press, 1984.
Sober, Elliott. The Philosophy of Biology. 2nd ed. Boulder, CO: Westview Press, 2000.
Sterelny, Kim, and Paul Griffiths. Sex and Death: An Introduction to the Philosophy of Biology. Chicago: University of Chicago Press, 1999.
Waters, Kenneth. "Genes Made Molecular." Philosophy of Science 61 (2) (1994): 163–185.
Waters, Kenneth. "Molecules Made Biological." Revue Internationale de Philosophie 4 (214) (2000): 539–564.
Waters, Kenneth. "A Pluralist Interpretation of Gene-centered Biology." In Scientific Pluralism, edited by Stephen Kellert, Helen Longino, and Kenneth Waters. Minneapolis: University of Minnesota Press, 2005.
Williams, George C. Adaptation and Natural Selection: A Critique of Some Current Evolutionary Thought. Princeton, NJ: Princeton University Press, 1966.
Williams, George C. Natural Selection: Domains, Levels, and Challenges. Oxford: Oxford University Press, 1992.
Wilson, David Sloan. "Individual Selection and the Concept of Structured Demes." In Genes, Organism, Populations: Controversies over the Units of Selection, edited by Robert Brandon and Richard Burian. Cambridge, MA: MIT Press, 1984.
Wimsatt, William. "Simple Systems and Phylogenetic Diversity." Philosophy of Science 65 (2) (1998): 267–275.
Wright, Larry "Functions." In Nature's Purposes: Analyses of Function and Design in Biology, edited by Collin Allen, Mark Bekoff, and George Lauder. Cambridge, MA: MIT Press, 1998.
Wright, Larry. Teleological Explanation. Berkeley: University of California Press, 1976.
James Lennox (2005)