Chickens, Eggs, and Speciation 1

Chickens, Eggs, and Speciation 1




Standard biological and philosophical treatments assume that dramatic genotypic or phenotypic change constitutes instantaneous speciation, and that barring such saltation, speciation is gradual evolutionary change in individual properties. Both propositions appear to be incongruent with standard theoretical perspectives on species themselves, since these perspectives are (a) non-pheneticist, and (b) tend to disregard intermediate cases. After reviewing certain key elements of such perspectives, it is proposed that species-membership is mediated by membership in a population. Species-membership depends, therefore, not on intrinsic characteristics of an organism, but on relationship of an organism to others. A new definition of speciation is proposed in the spirit of this proposal. This definition implies that dramatic change is neither necessary nor sufficient for speciation. It also implies, surprisingly, that an organism can change species during its lifetime.

There may be some deep truth about whether chickens or eggs are more fundamental, but no serious biologist would engage in such a debate, nor (I hope) would any serious philosopher be exercised by the question.

(Philosopher, 1998)

Another smug aperçu to the kindling-basket? So it would seem, for (International Edition, May 26, 2006) reports that David Papineau and two others have solved the puzzle of the chicken and the egg.

Mr Papineau, an expert in the philosophy of science, agreed that the first chicken came from an egg and that proves there were chicken eggs before chickens. He told the UK Press Association that people were mistaken if they argued that the mutant egg was not a chicken-egg because it belonged to the “non-chicken” bird parents. “I would argue it is a chicken egg if it has a chicken in it,” he said. “If a kangaroo laid an egg from which an ostrich hatched, that would surely be an ostrich egg, not a kangaroo egg.” (CNN’s text has been edited for grammar, punctuation, clarity, and sense.)

Papineau’s team also included Professor John Brookfield, a specialist in evolutionary genetics at the University of Nottingham, and Charles Bourns, a poultry farmer who chairs the oxymoronically named trade body, Great British Chicken. CNN says that the team’s research was “organized by Disney to promote the release of the film ‘Chicken Little’ on DVD.”

In this paper, I argue inter alia that chickens came into being at the same time as chicken eggs, at least if you accept Papineau’s criterion. On a closely related criterion, chicken eggs did indeed come before chickens, but this has nothing to do with the first chicken being in the first chicken egg. In fact, on my way of looking at speciation, the first chicken did not come from the first chicken-egg, and the first chicken-egg did not come from the first chicken.

These, however, are merely amusing corrolaries. My primary aim in this paper is to lay the foundations for, and explore the consequences of, a treatment of species-membership and speciation that is mediated by populations. I shall proceed by showing first how the supposition of a first chicken conflicts with certain standard ideas concerning species. Then, in section 3, I attempt to pin these difficulties on a degree of pheneticist thinking in many treatments of species and speciation. The rest of the paper is devoted to reconstructing species and speciation along thoroughly non-pheneticist lines.

The proposed analysis has two significant implications.

First, it turns out on my definition that species membership is relational, and consequently, that an organism can change its species during its lifetime without changing its phenotype in any way that violates its normal developmental pattern. Thus, speciation has nothing to do with “mutant eggs”: the first chickens were previously non-chickens (and came, by Papineau’s stipulation, from non-chicken eggs).

Second, a technical point: my analysis demands a clarification in the cladistic notion of monophyly – I’ll explain this is when the time comes.

More of this later: first let us turn to the general issue of speciation. This discussion occupies the bulk of this paper.

I. The Problem of the First Chicken

1. Papineau’s argument runs like this.

(1) The first chicken came from – hence, was preceded by – an egg E.

(2) An egg is a chicken-egg if and only if “it has a chicken in it”.

Therefore (3) E was a chicken-egg that preceded the first chicken.

On my conception of speciation, this argument is invalid. I shall argue that the first chicken was not a chicken when it was hatched. If this is right, E, the egg from which the first chicken “came” never had a chicken in it.

But I am getting ahead of myself. There is something odd about the very notion of a first chicken, as I shall now attempt to show. In this section, I shall argue that this notion arises from the neglect of a crucial constraint on when two organisms belong to the same species.

Call the “first chicken” Charlie. In order for the chicken-species to have got going, Charlie would have had to reproduce. Since chickens reproduce biparentally, he would have had to find a mate. Call her Charlize. Charlize was not a chicken: it strains credibility that two such similar mutations should occur independently in different organisms. Even if there are special circumstances in which double mutations are likely, the speciation story should not assume such an event, for then it would lose generality. Imposing such a requirement would force us to disregard the many cases of speciation where special circumstances allowing or encouraging multiple simultaneous mutations were absent. A general account of speciation must be able to accommodate single-mutation beginnings.

Let us assume, therefore, that Charlize belonged to Charlie’s parents’ species: she was a pre-chicken. (This assumption is introduced for the sake of simplicity: the argument can be reconstructed with the weaker assumption that Charlize belongs to some species.) On some species-concepts, the game is already over. Organisms that belong to one species are reproductively isolated from organisms that belong to others – the habits, habitat, physiology, and genetics of the organisms that belong to a single species enable them to recombine their genes with those of others of the same species but not with those of any other species. Some accounts of species elevate the possibility of such recombination into a definition. According to a strict construal of the Biological Species Concept, for example, two organisms belong to the same species if and only if their genes (or rather, copies thereof) can recombine on the same genome. By this species-concept, we have already contradicted the assumption that Charlie was a chicken. Since he was able to mate successfully with Charlize, who was a pre-chicken, he was a pre-chicken. It follows on such species-concepts that he was not even a chicken. Obviously, then, he was not the first chicken.[2]

2. Let us waive the demands of the Biological Species Concept, or at least understand them less stringently. Let us allow, at least for the sake of argument, that an occasional chicken can mate successfully with a pre-chicken and produce fertile offspring, but count as a chicken nonetheless. Let Charlie be a chicken then, his successful union with Charlize notwithstanding. Papineau’s solution still does not work. For now the question arises: What about the children? To what species do the offspring of Charlie and Charlize belong?

One possibility is that they are reproductively isolated: that is, they are not normally able to produce fertile offspring, except incestuously with each other or with Charlie. If so, we can define the species chicken as including all Charlie’s descendants up to but not including any who result from a further species-founding mutation in the future. Given this assumption, Charlie was indeed the first chicken – all that Charlize did was to help get the chicken species going, which she did by helping to produce a plurality of birds that can interbreed, but not outbreed.

The possibility just conceded is, however, quite improbable, at least in the case of organisms that reproduce biparentally – and surely we do not wish to exclude them from the discussion. (In fact, my discussion side-steps asexual and self-fertilizing organisms: in this paper, I shall be discussing the complications that biparental reproduction brings.) Charlie himself could mate successfully with a pre-chicken. What prevented his immediate offspring from doing so? If all went as expected, these birds would have had characteristics that ranged between those of Charlie and those of Charlize. Since Charlie could mate successfully with a pre-chicken, those of his offspring that bore his characteristics should have been able to do so as well. And of course Charlize is not isolated from pre-chickens – she is a pre-chicken and need not have chosen one-of-a-kind Charlie. So her characteristics are not going to isolate her offspring from other pre-chickens.[3]

How then can the offspring of Charlie and Charlize have been reproductively isolated? Is this by a second mutation? Again, this seems quite improbable in the absence of special circumstances (though not, of course, impossible). As I urged earlier, we should not construct the speciation story in such a way that it requires two or more founding individuals occurring through independent mutations: either two chicken-making mutations resulting in two originals, Charlie and Charlize, or a Charlie-mutation plus one more that results in Charlie’s immediate offspring or close descendants being reproductively isolated. The problem with the Papineau-team’s account is that it cannot accommodate single-mutation beginnings.

3. Charlie is an innovation, and thanks to Charlize, his genes get passed on. If he is a favourable innovation, his genes will spread across the pre-chicken world. In due course, pre-chickens will become more like him. This is natural selection at work; it is how species become better adapted over time. Here is a parallel: in the last thirty to fifty thousand years, humans have changed in cognitively significant ways. For example, they have somehow acquired an innate and specialized ability to use recursive grammars (cf. Hauser, Chomsky, and Fitch 2002). This is a big change: it leads to all kinds of behavioural specializations in humans. But it does not imply speciation. We would not be right to say that at some point in the process of acquiring a specialized ability to use recursion, a new species was born. This was simply a transformation and “improvement” of the human species itself.

The idea that mere innovation can result in speciation may have some intuitive appeal. However, biologists generally acknowledge that there is a difference between adaptation and speciation. Why they do so will become clearer when I present my own view of the matter. The point that we need to accommodate, though, is that innovation is not by itself tantamount to speciation. The task is to figure out what sorts of changes are required for speciation, as distinct from evolutionary change within species boundaries.

II. The Perils of Pheneticism

1. The egg-before-chicken thesis was first advanced by Roy Sorensen (1992). His argument is somewhat different from that of the Papineau-team. It runs like this.

(4) It is indeterminate where in evolutionary history pre-chickens end and chickens begin.

(5) However, “a particular organism cannot change its species membership during its lifetime.”

(6) Therefore, “the transition to chickenhood can only take place between the egg-layer and the egg,” in other words, this transition traces back to genetic or chromosomal change during reproduction.

(7) The genotype in the egg that hatched the first chicken was already a chicken-genotype, and the egg was thus a chicken egg.

Sorensen is making a point about vagueness: that even if we cannot determine which transitional bird was the first chicken, it is logically necessary, given (5) above, that its genotype existed in the egg from which it hatched.[4] In effect, Sorensen uses universal instantiation – the logic of ‘all’ – to trump the indeterminacy of chickenhood – whichever organism was the first chicken, it is preceded by a chicken egg (by proposition 6), therefore the first chicken was preceded by a chicken egg.[5]

Sorensen takes a view of speciation that is different from Papineau’s in one respect, and the same in another. He says: “Charles Darwin demonstrated that the [first] chicken was preceded by borderline chickens and so it is simply indeterminate as to where the pre-chickens end and the chickens begin.” The allusion is presumably to Darwin’s phenotypic gradualism. Sorensen’s idea is that somewhere in the smudge of gradually evolving organisms, pre-chickens gave way to chickens, but that all through this process, there was a reproductive community of birds, whether this community consisted of pre-chickens, transitional chickens, or chickens. He assumes that different features, or different enough, make for different species. Different enough: that’s where vagueness comes in. Papineau et al take a more saltational line. They seem to think, like Sorensen, that it is Charlie’s difference makes for the distinctness of his species – this is presumably what their talk of “mutant eggs” amounts to – but they think that this difference could have arisen in a single generation.

All of these thinkers seem, then, to be assuming a pheneticist conception of species – the idea that species are defined by similarity. (In note 13, I briefly consider the possibility that they were moved by purely phylogenetic considerations.) By hypothesis, Charlie is dramatically different from his parents – Charlie’s parents failed to “breed true”, to use Sorensen’s description of the case. But this does not acknowledge questions about reproductive barriers between species. This is why these treatments fail to notice the problems of section I.

2. Pheneticism is a mistake.[6] It originates in the harmless but imprecise idea that conspecific organisms resemble each other. However, by defining species in terms of this similarity, pheneticism trips up on the following:

a. The similarity of conspecific organisms is actually not universal. Polymorphisms exist within species, for example, the division of many species into dissimilar sexes. Moreover, similarities between some members of so-called sibling species may actually be closer than those between some members of the same species. For example, the males of one such species may well be more similar to the males of another than to the females of their own species.

b. Species-concept definitions in terms of similarity enshrine as fundamental that which needs to be explained. Both the similarities and the polymorphisms alluded to above – the “population structure” of species, as I shall be calling it – are explained in terms of certain deeper factors. Roughly speaking, the population structure of a species will be attributed to three kinds of factor: first, how that structure helps to carve out a niche that distinguishes a species from its historical predecessors; second, interactions among polymorphic types within a species that contribute to its adaptation to its niche; and third, reproductive integration within a species. The best accounts of species cut the lines between these taxa in way that at least roughly correspond to the fracture-lines of these explanatory factors.

These considerations against pheneticism are more powerful than many contemporary philosophers of biology fully realize. They represent implicit but unnoticed features of taxonomic practice – how males and females, juveniles and adults, larvae, pupae and imagos are all recognized as members of the same species. So-called “disjunctive” accounts are needed for this purpose: accounts of the form “If-female-then-F, if-male-then-G, etc.”, and such accounts do not tell us why males and females etc. are counted as members of the same species.[7]

It has been claimed that the phenetic unification of males and females can be achieved “by refined biometric techniques” (Sokal and Crovello, 1992), but what goes unsaid is that these (essentially disjunctive) techniques have to be rigged in order to serve the “wish to associate males and females that appear to form sexual pairs” (ibid., 37). On what is such a “wish” premised – why should we count interbreeding males and females as belonging to the same species? Pheneticists simply refuse to acknowledge that this can only be based on the role that reproductive integration plays in heredity and in evolution. This lack of groundedness and motivation tells against even weakened pheneticist accounts – for example, the philosophically fashionable Homeostatic Property Cluster Theory, which is a clever and (in certain ways) insightful variation on Wittgensteinian “family resemblance” theory, endorsed by such sophisticated philosophers as Richard Boyd (1991, 1999), Ruth Millikan (1999), Paul Griffiths (1999), and Robert Wilson (1999b).[8]

Pheneticism has no principled distinction between variation within a species and the differences that separate species.[9] This is illustrated in Darwin’s own tendency – for he implicitly employed a pheneticist conception – to treat of species as classifications “arbitrarily given for the sake of convenience to a set of individuals closely resembling each other.” More to our present point, it animates the belief, evinced by the philosophers discussed above, that “macro-mutation”, or saltational change is needed for speciation, while gradual change occurs within a species.