5

THE TREE OF LIFE RECONSIDERED

JEAN GAYON*

INTRODUCTION

Darwin referred to his theory in The Origin of Species as “descent with modification through natural selection” (Darwin 1859: 459). This theory was comprised of two parts: “descent with modification,” or in modern terms the general “pattern” of evolution; and “natural se-lection,” or the main explanatory hypothesis that accounts for the modification of species.

Darwin organized the whole book in such a way that the two parts are always tightly connected. Nevertheless, there are good reasons for taking the distinction between them seriously. First, Darwin was perfectly aware of its importance, as shown by a letter he wrote to Asa Gray, dated May 11, 1863: “Personally, of course, I care much about natural selection, but that seems to be utterly unimportant compared to the Creation or Modification” (Burkhardt 1985-2007, vol. XI, 403; italicized and bold characters in text). Another reason is that the scientific community did not react in the same way to the two parts of Darwin’s theory. “Descent with modification” was quickly re-named “evolution” by Darwin’s readers and by Darwin himself, who introduced it in the last edition of The Origin (1872).1 This part of the theory was so successful and so rapidly adopted in the last third of the nineteenth century2 that it became common to call it “the general fact of evolution.” In contrast, the explanation of this “general fact”

* Université Paris 1-Panthéon Sorbonne, Institute of History and Philosophy of Science and Technology (IHPST), 13 rue du Four, 75006 Paris, France; <gayon@noos.fr>

I thank Rachel Bryant (University of Toronto) for her linguistic revision of this pa-per, and for her fruitful comments. I also thank Camilo Cela Conde for his comments and Jorge Martínez Contreras for his invitation to publish in this stimulating volume.

1 There, “evolution” appears eight times, as an abbreviation for “descent with modi-fication” of the previous editions (p. 201, 202, 215, 282, 424; source: Darwin 2002-9).

2 Peter Bowler says that this may well have been one of the most spectacular and rapid examples of “paradigm shift” in the history of science (Bowler 2003).

6 JEAN GAYON

through natural selection has been subject to unending controversies since 1859.

This paper is devoted exclusively to the first half of Darwin’s the-ory, commonly referred to as “Darwin’s tree” because of the famous branching diagram given in The Origin of Species. In the first part of the paper I will take for granted that this diagram expressed what we would call a “pattern” today. I will raise the question whether the gen-eral pattern of evolution proposed by Darwin has been challenged or not since 1859. In the second part of the paper, I will question the common claim that Darwin’s tree depicted only an evolutionary pat-tern, not a process.

DARWIN’S TREE OR “THE GENERAL FACT OF EVOLUTION”

Darwin’s tree-like diagram

Darwin, in the Origin of species, did not provide an explicit definition of what he meant by “descent with modification,” which he probably found self-evident. Nevertheless, he had a rather restrictive interpre-tation of this idea — we might say he had a model of it. This model is expressed in the famous branching diagram given in chapter 4 of The Origin (fig. 1). This diagram relies upon a series of conjectures:

1] Species are subject to modification.2] This modification is slow and gradual.3] Many species go extinct.4] The species that do not go extinct usually split into other species.5] Once two or more species have split, they diverge indefinitely.6] The diagram applies to all taxonomical levels from the lowest local

varieties to the most encompassing groups. As a consequence: 7] Classification is entirely determined by genealogy. Therefore, there

are as many “ranks” as there are branching events of ramification, and all taxonomic categories used in systematics (species, genera, families, etc.) are arbitrary.

Let us now come to the image. Did Darwin explicitly accept and justify the tree image? In fact the 1837 “transmutation notebooks”

THE TREE OF LIFE RECONSIDERED 7

show that he hesitated between two images, the “tree of life” and the “coral of life.” “The tree of life should perhaps be called the coral of life, base of branches dead; so that passages cannot be seen” (Note-book “B”: p. 25). The image of a coral was appealing because it fit-ted better with a genealogical classification in which most species are extinct. In a tree, the branches are most often alive. In a coral reef, only the upper part is alive, while most of the structure below is dead. Some authors have taken the coral image quite seriously, and even proposed that the origin of the 1859 diagram was in fact a draw-ing of a real coral that Darwin had made when he was in Patagonia (Bredekamp, 2008). This is questionable, but there is no doubt that a

Figure 1. Darwin 1859 : contra p. 117. In chap. IV, the letters “A, B, C... L” represent “the species of a genus large in its own country.” The roman num-bers represent intervals between an arbitrary high number of generations.

They produce local varieties which modify and diverge during many genera-tions. For example, at the end of the process, species I leads to eight species. In chapter XIII, the letters “A, B, C... L” represent “allied genera” which will lead at to a greater number of genera. The same reasoning can be applied

to any desired level of taxonomic categories.

!

8 JEAN GAYON

coral could be an attractive image. Maderspacher, who has reviewed the most recent literature on this subject, writes: “The branching pat-tern of the coral is less hierarchical than that of a tree and it lacks a trunk. It thus would avoid the problem of directedness or teleology” (Maderpascher, 2006).

Nevertheless, Darwin finally chose the image of a tree, as shown by the following passage in The Origin of Species:

The affinities of all the beings of the same class have sometimes been repre-sented by a great tree. I believe this simile largely speaks the truth. The green and budding twigs may represent existing species; and those produced dur-ing each former year may represent the long succession of extinct species. At each period of growth all the growing twigs have tried to branch out on all sides, and to overtop and kill the surrounding twigs and branches, in the same manner as species and groups of species have tried to overmaster other species in the great battle for life (Darwin, 1859: 129).

In these sentences, Darwin applies his diagram to specific classes. But at the end of the chapter he goes further applying it to the en-tirety of the history of living beings:

The limbs divided into great branches, and these into lesser and lesser branches, were themselves once, when the tree was small, budding twigs; and this connexion of the former and present buds by ramifying branches may well represent the classification of all extinct and living species in groups subordinate to groups (Darwin 1859: 129-130).

At the end of the chapter, Darwin explicitly uses the term “tree of life”:

As buds give rise by growth to fresh buds, and these, if vigorous, branch out and overtop on all sides many a feebler branch, so by generation I believe it has been with the great Tree of Life, which fills with its dead and broken branches the crust of the Earth, and covers the surface with its ever branch-ing and beautiful ramifications (Darwin 1859: 130).

There is no doubt, therefore, that Darwin endorsed the image of the tree. Darwin’s diagram was not only “tree-like.” It was presented as a tree.

Within a short time this diagram became the paradigmatic repre-

THE TREE OF LIFE RECONSIDERED 9

sentation of organic evolution as an established fact, the “general fact of evolution.” Its sudden and dramatic effect constitutes one of the most spectacular examples of a paradigm-shift in the history of sci-ence, to use Thomas Kuhn’s famous terminology. However, the char-acterization of Darwin’s tree-of-life as a “general fact” of nature is mis-leading. This diagram, with the non-trivial postulates it illustrates, is a theoretical construction. The kind of theory involved is not a causal theory. It is a heuristic device, a plausible bet about the general form and pace of the phenomena of which evolution is a result, and so calls for explanation. Darwin’s theory of “descent with modification,” as summarized in his diagram, is a low level theory, but nevertheless a theory. In modern terms, it is a conjecture about the general pattern of evolutionary phenomena.

Contesting Darwin’s tree

Now the question is: has the scientific consensus about the so-called “fact of evolution,” as represented by Darwin’s diagram, been signifi-cantly threatened since Darwin’s time? There have been at least three serious challenges to Darwin’s tree-of-life picture. Each of these chal-lenges, or classes of criticisms, relates to the shape of the tree. The first two have been recurrent since the time of Darwin. The third is recent.

The first kind of criticism can be found among those biologists or paleontologists who have advocated a non-gradual or saltationist representation of the origin of species. Since Thomas Henry Huxley, saltationist theories have been numerous and varied. They have been held by morphologists, paleontologists, geneticists, and developmen-tal biologists. Today, the most famous example is represented by El-dredge’s and Gould’s theory of “punctuated equilibria” (Eldredge and Gould, 1972). This theory postulates that evolutionary change occurs primarily during events of speciation or cladogenesis.

Saltationist theories of evolution entail a significant alteration of Darwin’s tree. Instead of branches that progressively diverge, the sal-tationists’ trees have the shape of successive candelabras. The most fa-mous example is Eldredge’s and Eldredge’s “tree” in their 1972 paper (fig. 2). This kind of criticism, which focuses on the level of species (or at least on low taxonomic levels) challenges postulates 2, 5, listed

10 JEAN GAYON

above: gradualness of change, divergence of species as such, and in-definite divergence after splitting.

A second class of criticism consists in saying that evolution does not happen only at the level of varieties and species but also at that of

!

Figure. 2. Eldredge and Gould, 1972 : 113. Original legend: “Three-dimensional sketch contrasting a pattern of relative stability (A) with a trend (B), where speciation (dashed lines) is occurring in both major lineages. Morphological change is depicted here along the horizontal

axes, while the vertical axis is time. Though a retrospective pattern of directional selection might be fitted as a straight line in (B), the actual pattern is stasis within the species, and differential success of species exhibiting morphological change in a particular direction.”

THE TREE OF LIFE RECONSIDERED 11

higher taxonomic levels. This kind of criticism has also been popular among morphologists and paleontologists ever since Darwin. Accord-ing to it, modification and multiplication of species are superficial phenomena, not the core of the evolutionary process. An example of this criticism was given by the French morphologist Louis Vialleton in 1929. He thought that major groups arise rather suddenly. Therefore he argued that at higher taxonomic levels, Darwin’s tree should be replaced by a “bush” with parallel twigs arising from a given level:

Genealogical trees, by illustrating in a simple way a number of infinitely com-plex relations in the real world, were very useful to the transformist theory... The first diagrams were quite satisfying. They strikingly represented the evo-lutionary unfolding of beings and seemed to be able to provide their natu-ral relationships that had been sought for such a long time. But disillusion came fast... Taking into account the time of appearance of various forms... it soon appeared that genealogical trees looked quite different from how they looked before. Instead of showing a more or less elongated trunk during the period when only the supposed initial form existed, they had the shape of bushes whose twigs stemmed from approximately the same height above the ground, and moved away in a parallel way from each other, each of them be-ing more or less ramified (Vialleton, 1929).

To illustrate these doubts about Darwinian trees, Vialleton gave two different representations of the genealogical tree of mammals. One was borrowed from the American paleontologist Henry Fair-field Osborn (fig. 3); the other was Vialleton’s representation of ex-actly the same data and interpretation (fig. 4). Vialleton’s represen-tation emphasizes the idea that something decisive occurs at high taxonomic ranks. Vialleton thought that “transformism” (his name for Darwinian evolution —i.e. gradual modification and splitting of species) was only a superficial phenomenon, whereas genuine “evo-lution” was the generation or “creation” of major morphological novelties in groups that suddenly appeared and remained approxi-mately stable thereafter.

More recently, Eldredge and Gould’s punctuated equilibria have provided a totally new version of the old idea that evolution is not confined to the level of species. Their famous diagram also illus-trates the hypothesis that a given evolutionary trend in a taxonomic group (either stasis or directional evolution) might well result from

12 JEAN GAYON

changes above the level of species rather than from changes within the species (see fig. 2). Although this way of thinking is far from Vialleton’s typological approach (and, in that sense, is much closer to Darwin’s thought), it exemplifies another manner of advocating evolution above the species level, and consequently of challenging postulate 6.

The third class of criticism of Darwin’s diagram is more recent and more radical. The suspicion is that this diagram represents only a fraction of the real pattern of the history of life. Darwin’s image rep-resents the conviction that a tree of irreversible divergence of species adequately and sufficiently represents the history of biological diver-sity. An increasing number of phenomena challenge this view of spe-cies as (material) monads “without any doors or windows,” to freely pastiche Leibniz’s famous formula. Among these new phenomena, lateral gene transfer and symbiosis are the major ones.

Since the end of the 1990s, lateral gene transfer has been known to be a major phenomenon among prokaryotes (Eubacteria and Ar-

!Figure 3. Vialleton 1929 : 183. Original legend “Arbre généalogique des mammifères (d’après Osborn)” [Genealogical tree of mammals (from

Osborn)]. Vialleton refers to Osborn, 1917.

THE TREE OF LIFE RECONSIDERED 13

chaea). Prokaryotes are single-celled organisms without a membrane-bound nucleus. Since Carl Woese’s pioneering work in the 1970s and 1980s, they have been divided into two “domains”: Bacteria and Ar-chaea. Eukaryotes constitute the third “domain,” including all single-cell and multicellular organisms with a membrane-bound nucleus. At the beginning, the application of molecular techniques to prokary-otes generated the hope of reconstructing the unique and universal tree of life (Woese, 1987). But an increasing number of anomalies led to the discovery that gene transmission among prokaryotes is not only

Figure 4. Vialleton 1929 : 181. Original legend: “Arbre généalogique des mammifères” [Genealogical tree of mammals].

14 JEAN GAYON

vertical, it is also horizontal. The magnitude of this phenomenon has been fiercely debated for almost twenty years. It is now clear that it is important enough to cast serious doubts on the very existence of a unique tree representing the natural relationships among all cellular organisms or, at least, on the possibility of reconstituting this tree if it exists (for a review, see J. O. McInerney, J. A. Cotton, and D. Pisani, 2008). This is due to the fact that almost all (if not all) prokaryotic genes have been subject to lateral gene transfer. Whether this turn-over of genes in Eukaryotes is significant remains a debated issue. Nevertheless, we now observe an intense controversy between two ways of representing the genealogical pattern of the history of life: tree and network (the best and most comprehensive information on this can be found in Ragan, McInerney, and Lake, 2009).

It is presently uncertain whether the “network of life hypothesis” will be able to successfully combat the “tree of life hypothesis.” In reality, it is likely that both the tree and the network will have to coexist in a plu-ralistic view of evolutionary patterns, where several histories are told using several methodologies (Doolittle & Bapteste, 2007). In actual fact, evolutionary biologists now seem to admit that Darwin’s diagram, which focuses exclusively at the level of species, encapsulates only one fraction of the history of life. In the case of Prokaryotes at least, lateral gene transfer has been important enough to obliterate, and perhaps overcome vertical transmission at a large historical scale.

Another process that jeopardizes Darwin’s representation of gene-alogy is symbiosis. Symbiosis has probably been a major evolutionary process, especially at the level of cell evolution (Margulis and Sagan, 2002). Symbiosis is much rarer than lateral gene transfer, but when it occurs, it may have dramatic effects because it implies the coexistence of two full genomes. Symbiosis seems to have been a key process in the emergence of a number of major groups of unicellular eukaryotic organisms, and of Eukaryotes themselves. Symbiosis does not chal-lenge the notion of irreversible divergence of species, because sepa-rate organisms that fuse initiate new species that diverge from others. But fusion events between organisms introduce complications that cannot be assimilated by an exclusively tree-like pattern of phylogeny. Figure 5 illustrates the kind of anastomosis that results from such con-jectures in a special group (algae). Figure 6, a hybrid between a tree and a network, summarizes the way evolutionary biologists represent the general pattern of evolution today.

THE TREE OF LIFE RECONSIDERED 15

To sum up, the first aspect of Darwin’s theory (“descent with modi-fication”) was successful in imposing a view of the history of life in terms of genealogy and modification, but the particular model that Darwin offered has been repeatedly disputed. To be sure, the tree diagram has proven immensely fertile and remains the basic intuition behind most of the work on phylogenies. But it is not true that the diagram has remained unchanged. Several of the postulates embod-

!

Figure 5. Palmer, 2003. Role of symbiosis in the origin of the diversity of algae and their plastids.

16 JEAN GAYON

ied in Darwin’s representation appear not to be false but to express a rather simplified view. At least three of the seven postulates that I have listed have been seriously challenged: 2 (gradualness of modifi-cation), 5 (irreversible divergence of species), and 6 (uniform validity of the tree-like diagram at all levels. Modern evolutionary biologists are not in agreement regarding the assertion that modification of species is always slow and gradual (against 2). Indefinite divergence after splitting is not strictly false, but it is only one part of the story of life because important horizontal relations must also be taken into account (against 5 and 6).

DARWIN’S TREE: PATTERN OR PROCESS?

I now raise the question whether Darwin’s diagram represents a “pat-tern” or a “process.” Although this is a modern distinction, it can shed light on what Darwin really intended to express through his diagram and on whether there is continuity between Darwin’s approach to phylogenetic trees and contemporary approaches.

The pattern/process dichotomy emerged in the theoretical con-text of cladistics, a particular theory of classification also called “phy-

!

Figure 6. McInerney and al., 2008. Diagrammatic representation of the entire “network of life.”

THE TREE OF LIFE RECONSIDERED 17

logenetic systematics.” It became standard after the publication of Joel Cracraft and Niles Eldredge’s book Phylogenetic Patterns and the Evolutionary Process (Eldredge and Cracraft, 1980). After the publica-tion and the translation into English of Willi Hennig’s Phylogenetic Systematics (Hennig [1950] 1966), a growing number of systemati-cists became unsatisfied with the modern synthesis’ approach to clas-sification. They reproached the “evolutionary systematicists” (e.g. Ernst Mayr or George Gaylord Simpson) for basing their method for elaborating classification upon hypotheses about the evolution-ary processes.

In this context, a pattern is a schema that gathers data using a meth-od that is independent from any explanatory theory. Conversely, a pro-cess is invoked for the purpose of explaining a certain kind of change. For instance, if we calculate the genetic distance between chimpan-zees, gorillas and humans, and if we conclude from this that chimps and humans are more closely related to each other than they are to gorillas, we’re making a claim based on a pattern, ordinarily expressed through a dendrogram (Sober, 1991). But if we propose a hypothesis about the environmental conditions and the selective pressures that may have a led to bipedalism in the human lineage (the hominin), then we say something about a process. Mutation, recombination, natural selection, and extinction are typical evolutionary processes. Therefore, when the cladists say that phylogenetic systematics should limit itself to elaborating “patterns; they mean that classification and phylogenetic inference should be totally independent of any a priori hypothesis about evolutionary processes.

This distinction seems intuitively easy to understand, but it has generated extremely difficult problems among the cladists them-selves. For instance, for the classical cladists, a phylogenetic tree is a classification; in fact, it is the only admissible kind of classification in systematics. But for the “pattern cladists,” classification and phylog-eny should be carefully distinguished. Cladograms should be free of any reference to evolution, and phylogenetic trees are interpretations that come only after the cladograms. I will not go further into that discussion (see Panchen, 1992). It will be enough to keep in mind the notion of evolutionary pattern as a schematic representation that is as free as possible from particular hypotheses about evolutionary processes. Let us then consider Darwin’s tree-like pattern from that perspective.

18 JEAN GAYON

There was of course nothing like a pattern/process distinction in Darwin’s writings. But there are good reasons to retrospectively apply this distinction to him. Let us first consider the question of words. Darwin used both the words “pattern” and “process” but used only one of the two in the modern evolutionary sense. In The Origin of Species, “pattern” applies only to concrete morphological structures, not to the tree-like diagram. But he did use the word “process” in the common sense of a temporal sequence of events obeying certain “laws.” For instance, Darwin characterizes variation, natural selection, divergence of character and extinction as processes. Therefore, it is interesting to examine whether or not Darwin makes use of processes when commenting on his tree-like diagram.

In The Origin of Species Darwin devotes two long and quite different comments to his tree-like diagram. These appear in chapter 4, which is devoted to natural selection (Darwin, 1859: 116-126), and in chap-ter 13, which is devoted to classification (Darwin, 1859: 413-434).

Let us first consider the chapter on natural selection. Many authors, including me, have characterized the diagram given in this chapter as representing a pattern (e.g. Sober 1991, Gayon, 2008, 2009). How-ever, Darwin himself had a very different view. Throughout the eleven pages of comments about his diagram, Darwin refers repeatedly to processes. Furthermore, he never speaks of his diagram as an ideal-ized way of representing data but as an idealized way of representing what should happen in evolution if his explanatory hypotheses are right. The following are the key points, taken in the order in which Darwin presents them.

Darwin first explains that the “the modified descendants of any one species will succeed by so much the better as they become more diversified in structure, and are thus enabled to encroach on places occupied by other beings” (Darwin, 1859: 116). The increasing num-ber of these varieties will itself be a result of natural selection. Natural selection should indeed preserve the beneficial variations, or those adapted to new circumstances. In the diagram, this is illustrated by the small clusters of dotted lines at each level. These clusters repre-sent the joint effect of variation and natural selection.

Then Darwin conjectures that the most divergent variations should be preserved, because they allow the species to maximize the number of places occupied in nature (Darwin, 1859: 119). This is symbolized in the diagram by a general tendency of the most successful lineages

THE TREE OF LIFE RECONSIDERED 19

to repeatedly generate “marked varieties” whose characters are more and more differentiated. This, again, is a process.

Darwin also explains that the more improved branches will destroy the less improved ones and replace them: “this is represented in the diagram by some of the lower branches not reaching to the upper horizontal lines” (Darwin 1859: 119). In other words, Darwin here contends that competition and differential elimination occur not only within local varieties, but also between “well-marked varieties” or “races” (this is not, however, a process of natural selection).

This principle of intergroup competition is also extended to the level of larger taxonomic groups (species, genera, families, etc.), which are subject to differential extinction (Darwin, 1859: 121). In time, this entails that the number of highly diversified groups is lim-ited. It can be easily seen on the diagram that only a few among the initial groups reach the superior level. Only three out of eleven initial groups have descendants at the end of the process.

Finally, Darwin considers the embarrassing case of species that do not change and still persist (Darwin, 1859: 124). Such cases raise a delicate question for taxonomists. For example, look at “F”: this species does not change, but it finds itself in an intermediate posi-tion between the mean types of descendants from “A” and “I.” This means that the adaptive histories of species do not coincide with their genealogical histories. Adaptive differentiation in the most success-ful groups tends to cloud over the underlying purely genealogical sequence.

We see therefore that Darwin’s comments on his diagram are satu-rated by hypotheses about evolutionary processes such as variation, natural selection, inter-group competition, and extinction. There-fore, the tree-diagram presented in the chapter on “natural selection” (not a coincidence, because natural selection is the key process that triggers all other processes evoked by Darwin in his comments) can-not definitely be interpreted as a pattern in the sense used by modern evolutionary biologists.

Chapter 13 of The Origin of Species offers a new series of comments about the tree-like diagram (pp. 420-434). The tone of these com-ments is quite different from those of chapter 4. Darwin begins by saying that “all true classification is genealogical” (Darwin, 1859: 420). Then he argues that the “grades of difference between the de-scendants from a common parent” do not provide a secure basis for

20 JEAN GAYON

classification, because they result from the adaptive effect of natural selection (Darwin, 1859: 433). The reason for this is that adaptive dif-ferentiation tends to hide the genealogical nexus, which is the only thing that counts when it comes to classification:

We have no written pedigrees; we have to make out community of descent by resemblances of any kind. Therefore we choose those characters which, as far as we can judge, are the least likely to have been modified in relation to the conditions of life to which each species has been recently exposed (Darwin, 1859: 425)

Consequently, Darwin recommends that the systematicists neutralize as much as possible adaptive history in order to reconstruct a true genealogical classification. He also recommends that they rely upon rudimentary organs and embryological characters for the purpose of classification, as such organs and characters reveal genealogical re-lationships often hidden by the adaptive features of the organisms (Darwin, 1859: 433). Except for these practical recommendations, Darwin had no definite method of phylogenetic inference to pro-pose. This is indeed why it would be anachronistic to attribute him a notion of pattern. It was only in the second half of the twentieth cen-tury that methods of phylogenetic inference reasonably independent from particular evolutionary hypotheses were proposed. But Darwin was deeply aware of the different theoretical levels at which the theo-ry of natural selection —the theory of a process— and the theory of classification operated.

CONCLUSION

Let us retrospectively contemplate the fate of Darwin’s tree-like dia-gram. Darwin carefully distinguished two possible senses of “tree.” In the first sense, Darwin’s tree was a methodological device for rep-resenting the relationships between species of a given group of or-ganisms in an evolutionary perspective. In the second sense, Darwin also spoke of “The Tree of Life” (capital letters), that is to say, the idea of the entire genealogical structure of all living beings, past and present.

THE TREE OF LIFE RECONSIDERED 21

As for the methodological device, we have seen that the shape of the tree has been discussed again and again since its first proposal. We have also seen that in the past twenty years, this device has been seriously challenged. Irreversible branching trees of species conflict with two major classes of phenomena: horizontal transfer of infor-mation (lateral transfer of genes or other cellular components) and symbiosis.

Prokaryotic evolution also challenges the very idea of the Tree of Life. A number of specialists in molecular evolution today say that this tree is either impossible to describe, or even fictional, because of the particular processes of variation that are encountered in bacteria, archaea, viruses, and possibly in other groups, especially unicellular eukaryotes and fungi (see the remarkable review by Bapteste et al., 2009).

Finally, in all cases, we may observe that the changing representa-tions of the overall pattern of the history of life have been affected by considerations about processes, such as natural selection in Darwin, species selection in Eldredge and Gould, and horizontal transfer and symbiosis in contemporary work on unicellular organisms.

REFERENCES

Bapteste, E. et al., 2009, “Prokaryotic Evolution and the Tree of Life are Two Different Things,” Biology Direct, 4:34 (electronic reference: doi:10.1186/1745-6150-4-34).

Bredekamp, H. 2008, Les coraux de Darwin — Premiers modèles évolutionnistes et tradition de l’histoire naturelle, Dijon, Les Presses du Réel.

Burkhardt, F. et al., 1985-2007, The Correspondence of Charles Darwin, 15 vols., Cambridge, Cambridge University Press.

Eldredge, N. and J. Cracraft, 1980, Phylogenetic Patterns and the Evolutionary Process: Method and Theory in Comparative Biology, New York, Columbia Uni-versity Press.

Darwin, C., 1859, On the Origin of Species by Means of Natural Selection, or the Pres-ervation of Favoured Races in the Struggle for Life, London, John Murray.

——, 2002-2009, The Complete Work of Charles Darwin Online, http://darwin-online.org.uk/.

Doolittle, W. Ford and E. Bapteste, 2007, “Pattern Pluralism and the Tree of

22 JEAN GAYON

Life Hypothesis,” Proceedings of the National Academy of Sciences USA 104 (7: 2043-2049.

Eldredge, N. and S. J. Gould 1972, “Punctuated Equilibria. An Alternative To Phyletic Gradualism,” in T. J. M. Schopf (ed.), Models in Paleobiology, Free-man, San Francisco, pp. 82-115.

Gayon, J., 2008, “‘Is a New and General Theory of Evolution Emerging?’ A Philosophical Appraisal of Stephen Jay Gould’s Evaluation of Contempo-rary Evolutionary Theory,” in W. J. González, Evolutionism: Present Approach-es, La Coruña, Netbiblo: 77-105.

——, 2009, “Mort ou persistance du darwinisme? Regard d’un épistémolo-gue,” Comptes Rendus Palevol 8: 321-340.

Gould, S. J., 2002, The Structure of Evolutionary Theory, Cambridge, The Belknap Press of Harvard University Press.

Hennig, W. [1950], 1966, Phylogenetic Systematics, translated from German by D. Dwight Davis and Rainer Zangerl, Urbana, University of Illinois Press.

Maderspacher, F., 2006, “The Origins of Early Evolutionary Imagery,” Current Biology, 16 (13): R476-R478.

McInerney, J. O., J. A. Cotton, and D. Pisani, “The Prokaryotic Tree of Life: Past, Present... and Future,” Trends in Ecology and Evolution, vol. 3, nº 5: 276-281.

Margulis, L. and D. Sagan, 2002, Acquiring Genomes. A Theory of the Origins of Species, New York, Basic Books.

Panchen, A. L., 1992. Classification, Evolution, and the Nature of Biology, Cam-bridge, Cambridge University Press.

Osborn, H. F., 1917, The Origin and Evolution of Life, on the Theory of Action, Reaction and Interaction of Energy, New York, C. Scribner’s Sons.

Palmer, J. D., 2003, “The Symbiotic Birth and spread of Plastids: How Many Times and Whodunit?,” Journal of Phycology, 39: 4-11.

Ragan, M. A., J. O McInerney, and J. A. Lake, 2009, The Network of Life: Ge-nome Beginnings and Evolution, special issue of Philosophical Transactions of the Royal Society of London: 2169-2289.

Sober, E., 1991. Reconstructing the Past. Parsimony, Evolution, and Inference, Cam-bridge, MIT Press.

Vialleton, L., 1929, L’origine des êtres vivants. L’illusion transformiste, Paris, Li-brairie Plon.

Woese, C. R., 1987, “Bacterial Evolution,” Microbiology Reviews 51: 221-271.