sign in sign up
<<
Home , Halteres, Homology (biology), Homoplasy

Homology and Introductory article : A Article Contents . The Concept of up to Darwin . Adaptationist, Taxic and Developmental Concepts of Philosophical Perspective Homology . Developmental Homology

Ron Amundson, University of Hawaii at Hilo, Hawaii, USA . Character Identity Networks

Online posting date: 15th October 2012

Homology refers to the underlying sameness of distinct Homology is sameness, judged by various kinds of body parts or other organic features. The concept became similarity. The recognition of body parts that are shared by crucial to the understanding of relationships among different kinds of organisms dates at least back to Aristotle, organisms during the early nineteenth century. Its but acceptance of this fact as scientifically important to developed during the first half of the nineteenth importance has vacillated during the years between the century. In an influential definition published in 1843, publication of Darwin’s Origin of Species and recent times. defined homologues as ‘the same in For much of the twentieth century, homology was different animals under every variety of form and function’ regarded as nothing more than evidence of common (Owen, 1843). Owen had a very particular kind of struc- descent. In recent times, however, it has regained its for- tural sameness in mind. It was not to be confused with mer importance. It is regarded by advocates of evo- similarities that are due to shared function, which were lutionary as a source of evidence termed ‘.’ are not homologous to for developmental constraint, and evidence for biases in wings, even though they look similar and perform the same phenotypic variation and the commonalities in organic function. Because the similarities are ascribed to functional form. The recent growth in understanding of develop- need, they are identified as analogues. However, the wings mental has made possible a new theory about the of , the front fins of porpoises, the forelegs of and the arms of are all homologous. Even though mechanical causes of homology, which was impossible they look different and function in different ways, they are under the naı¨ve views regarding genetic causation that the same in the sense of homology (see Figure 1). were available during most of the twentieth century. This Some body parts bear very little resemblance to their new theory, proposed by Gu¨nter Wagner (2007), states homologues in other species. The and , tiny that homologies of character are due to the development bones in mammals, are homologous with the bones that of a certain kind of genetic regulatory network called a form the jaw joint in reptiles. A distinct but related concept character identity network. that Owen termed serial homology identifies repeated elements within an individual organism. All of the ver- tebrae in an individual’s body are serial homologues, as are The Concept of Homology up to their right and left limbs. An interesting pre-Darwinian discovery of serial homology was the identification of the Darwin parts of flowers (, , etc.) with the of plants. The concept of serial homology has varied in its The organic world is amazingly diverse, but patterns of importance within evolutionary thought, in ways that will unity can be seen throughout. has the be discussed in ‘Character Identity Networks’. See also: double task of accounting for both the diversity and the Aristotle of Stagira; Homology in Character ; unity of . These two aspects are tied to the two most Owen, Richard central concepts in evolutionary theory: and A heated biological debate during the early nineteenth homology. Adaptation to diverse situations is believed to century concerned the relative importance of the principles account for the diversity of life forms. Relations of hom- of unity of type versus conditions of existence. Unity of ology express their unity. type expressed the consistency of generalised categories (types) that reflected the organisation of nested sets of eLS subject area: Bioethics & Philosophy homologies within comparative . Owen proposed the archetype as the generalised form of ver- How to cite: tebral skeletons, and sometimes spoke of the archetype as Amundson, Ron (October 2012) Homology and Homoplasy: A an abstract platonic idea. Conditions of existence, in con- Philosophical Perspective. In: eLS. John Wiley & Sons, Ltd: Chichester. trast, expressed the diverse of organisms to DOI: 10.1002/9780470015902.a0003445.pub2 their environment and to internal physiological needs.

eLS & 2012, John Wiley & Sons, Ltd. www.els.net 1 Homology and Homoplasy: A Philosophical Perspective

Whale Lion Bird

Figure 1 Homologous characters can have different shapes and functions. of seven species show that corresponding body parts have similar designs but can serve different functions, from swimming to flying. Hence, functional necessity cannot explain the similarity of homologous characters. Reproduced from Wagner (2007) p. 475, with permission of Nature Publishing Group.

Now termed adaptationists, the advocates of conditions of homology from causal explanation to summary of existence included British natural theologians such as the evidence was reinforced by the Darwinian E. R. Lankester. followers of William Paley. Adaptationists claimed that Lankester disliked Owen’s terminology, which sounded types were scientifically disreputable. A true understanding mystical to many English-speaking theorists. In 1870, of biology required attention only to function. Lankester argued that Owen’s term homology should be Many twentieth century scholars have attributed the replaced by the term homogeny, which would be defined as antievolutionism of the early nineteenth century to the characters that ‘have a single representative in a common allegedly mystical ‘typological thinking’ of unity of type ancestor’ (Lankester, 1870). Most evolutionists have advocates. But nineteenth-century typologists were very retained the term homology, but used Lankester’s histor- diverse in their views on metaphysics and on evolution. ical definition in place of Owen’s, which had simply said They ranged from mystical to materialistic and even that homologues were ‘the same organ in different animals included T. H. Huxley, Darwin’s strongest supporter. _’ without offering any definition of sameness. In place of Typologists were united not by metaphysical or anti- analogy, Lankester offered homoplasy, a term that desig- evolutionary commitments, but by a belief in the import- nates any biological similarity that does not fit the histor- ance of homology over adaptation. Evolutionary debates ical definition of homology. Because analogy requires within the twentieth century reiterated the pre-evolution- evidence for an adaptive cause of the similarity, homoplasy ary debates regarding the relative importance of homology has come to seem the simpler concept. The term homology versus adaptation as the most basic unifying principles of is still widely held to involve Lankester’s historical concept: biological understanding. See also: History of Compara- ‘A feature in two or more taxa is homologous when it is tive Anatomy; Huxley, Thomas Henry derived from the same (or a corresponding) feature of their One of Darwin’s great successes was his use of the well- common ancestor’ according to Ernst Mayr (Mayr, 1982). established facts about homology as evidence for common See also: Lankester, Edwin Ray ancestry. Darwin rhetorically transformed Owen’s arche- The actual criteria used for recognising homologies have type into an ancestor, and showed how nested sets of changed little between pre-Darwinian days and modern homologies are just what we would expect from genea- times. One is the principle of connections originally logical relationships. Darwin considered ‘the homological articulated by Geoffroy Saint-Hilaire in the 1820s. Hom- argument’ as one of his best arguments for common ologous body parts might differ greatly in shape and even ancestry. Still, Darwin’s replacement of Owen’s archetype composition, but can be recognised by their topological with an ancestor did reduce the status of homology within connections to the body parts around them. A second cri- biological theory. For Darwin, homology was merely evi- terion is the embryological origin of a body part. Body dence for ‘the fact of’ evolution (meaning descent with parts that differ in adults are sometimes observed to modification). So homology was no longer involved in develop out of the same parts in early embryos. Embry- causal explanation. Instead, it counted only as post ology allowed the discovery of the ear bone/jaw bone hoc evidence for . This downgrading of homology above. Here, the homological sameness is

2 eLS & 2012, John Wiley & Sons, Ltd. www.els.net Homology and Homoplasy: A Philosophical Perspective inferred from the similarity of topological connections, treated with great scepticism. One reason for this scepti- even though the bones themselves are very dissimilar. A cism was the difficulty in identifying such , but third traditional criterion is the existence of transitional another was the belief that interspecies sterility guaranteed forms of the character in adults of related species. that no possible mechanisms could maintain the kinds of Lankester’s historical definition of homology has been developmental stability necessary for such things as favored, especially in the English speaking world, because archetypes and homologies to exist. On this view, homo- it seems more concrete. A real entity (an ancestor) rather logues indicated no current biological causes, no causally than an ideal archetype is named as the source of the active constraints. This view is discussed in detail in homologues. But in fact Lankester’s definition is vague in (Amundson, 2005), and will receive no more attention here. precisely the way that Owen’s was. What exactly consti- As will be seen, today’s evidence does not favour this view. tutes sameness? Lankester’s historical definition defines the Research in the 1990s revealed huge numbers of genes that sameness of two contemporary homologues (say a bird’s were identifiable as homologues across immense taxo- and a ’s fin) in terms of the sameness each of nomic ranges, and revealed the complexity of the processes them bears to the body part of an ancestor (a reptile’s by which the expression of genes is controlled by other ). But this only pushes the problem back. What factors (genetic and nongenetic). For most of the twentieth constitutes the sameness of today’s wing with the ancestral century, genes themselves were regarded as responsible for forelimb? We sometimes speak loosely of birds’ wings the traits of organisms. It is now known that differences having descended from forelegs. But this is truly among organisms cannot be understood as differences nonsense, if taken literally – body parts do not descend among genes, but as differences in the expression of genes, from other body parts. Each organism generates its body most of which are shared among organisms. See also: parts anew during embryogenesis. The historical definition Homologous, Orthologous and Paralogous Genes does successfully explain why the pattern of homologies is a A second, contemporary competitor to the develop- nested hierarchy rather than some other pattern. But it mental theory of homology comes from modern system- does not, any more than Owen’s definition, explain what atics. The method called ‘’ has become the ‘same’ means. Lankester speaks of homologies being dominant approach to this field, and it uses the term ‘represented in’, and Mayr speaks of them ‘corresponding homology to refer to any trait that is shared within a to’ body parts of a common ancestor. Representation and monophyletic group of species (all species descended from correspondence are merely synonyms for sameness. If we a single common ancestor). A synonym for homology in agree that patterns of homology reflect patterns of ances- this meaning is the technical term ‘synapomorphy’. If the try, the historical definition of homology adds nothing term homology is used in this sense, there is no point in more. It does not explain, for example, why homologies asking for a developmental explanation because the term is should be expected to occur at all in the . defined by mere sharing of traits. This criticism is simply See also: Morphological Evolution: Epigenetic Mechanisms a semantic matter. If a developmental theory of homology is successful, it is only necessary to invent a term (as Lankester tried to do) to separate the two meanings. So the Adaptationist, Taxic and author will proceed to investigate the developmental the- ory and its recent history. See also: : Relevance Developmental Concepts of to the Twenty-first Century Homology

The present article is directed primarily at the develop- Developmental Homology mental view of homology, because that view is at the pre- sent time the most novel and creative of the alternatives The developmental interpretation of homology directs (especially from a philosophical perspective). However, the attention back toward anatomy, and in recent competitors should be acknowledged. One competitor is years toward developmental genetics. However, there are the tradition of adaptationist scepticism regarding hom- important barriers to any explanation of homology that ology. According to this view, homologies are (as Darwin founds the homological relation on shared embryological saw them) mere evidence of descent, not indications of or genetic origins. The problem is the frustratingly loose developmental commonalities. The view was prominent relation between homologous characters and the embryo- among proponents of the ‘Modern Synthesis’ in evo- logical and genetic processes that create them. One would lutionary theory, which dominated evolutionary thought think that the mechanistic basis of homology is the sharing during most of the twentieth century (Mayr and Provine, of developmental processes. Not so – at least not so in any 1980). The view considered developmental biology as in direct sense. A large number of characters that are principle irrelevant to evolution. was unquestionably homologous are known to arise from believed to operate on phenotypes, via the genotypes that distinct embryological and genetic causes. These will be were believed to directly produce them. The possibility that termed ‘paradoxical’ homologues. homological traits were produced by homological genes In his classic 1971 pamphlet on homology, De Beer began (identical genes operating in different species) was also the discussion of paradoxical homology with the alimentary

eLS & 2012, John Wiley & Sons, Ltd. www.els.net 3 Homology and Homoplasy: A Philosophical Perspective canal of . This was the most obvious case of indicating a pre-scientific worldview, with nonmaterial homology one could imagine (De Beer, 1971). But the tis- ‘ideas’ acting as causal agents in the production of bio- sues from which this system is built can come from the roof logical form. The author has argued that this is a serious of the embryological gut cavity (as in sharks), the floor misconstrual of nineteenth century biology (Amundson, (lampreys), the roof and floor () or the blastoderm, the 2005). The term ‘ideal’ often meant no more than hypo- lower layer of the embryonic disc (reptiles and birds). It can thetical. Such things as archetypes and bauplans (body be assumed that this range of alternative origins must have plans) certainly were hypothetical, and were known to be come about by means of evolutionary innovations in early open to further scientific investigation. Still, it has been at ontogeny (assuming that the common ancestor of all ver- least 150 years since the puzzle of homology arose, and a tebrates already possessed an alimentary canal). More cases material explanation of the relation has been lacking. are readily available, but they received little attention from Recently, however, a new proposal has changed the early evolutionists. This may have been because they did not landscape. It arises from the modern recognition of the fit within the von Baerian tradition in embryology, which importance, not of expression itself (as a gene pro- claimed that the earliest stages of ontogeny were the most ducing a ) but of the control of gene conservative. See also: Baer, Karl Ernst von expression, which may take place by the influence of other The von Baerian view is broadly correct, but it is strik- genes (transcription factors for example) or environmental ingly mistaken in detail. Evolutionary innovations in early influences. Among the factors that control ontogeny often leave the existing commonalities of later in development are gene regulatory networks (GRNs), sets ontogeny (including adult homologies) undisturbed, even of genes and their together with environmental after radical reconstructions of their embryological origins. influences that control the expression of a developmental Dramatic examples come from the introduction or the loss gene. The distinction between GRNs and the genes they of indirect development, such as the introduction or loss of control can be used to construct a promising new account larval stages. Homologies of adult organisms typically of homology. See also: Genetic Networks remain unchanged, even when a larval stage is interposed or eliminated in a subgroup. For example, the body seg- ments of all orders of are clearly homologous. Character Identity Networks However, grasshopper adult segments arise from direct development, whereas the adult body parts of dipteran flies Gu¨nter Wagner has proposed such a developmental- (indirect developers) arise from imaginal discs within the genetic account of homology (Wagner, 2007). Wagner larva. The same body parts have different embryological considers characters to be the principle kind of homology, origins. Genes share in this indirection; those that specify and character states as a less-central topic. The importance the development of segments in are only of characters is that they can retain their identity even when expressed in the nervous systems of grasshoppers (Wagner, their state (details regarding their form) is changed. Win- 2007). Adult phenotypes seem to possess their own manner ged insects share the characters forewing and hindwing, of organisation, such that phenotypic characters can per- associated with the second and third segments, respect- sist (keeping their homologous identities) even during the ively. Either character can take on different character evolutionary modification of their own ontogeny. What states. Basal flies have wings in both positions. Dipteran could be responsible for this eerie manner of persistence? flies have wings in the forewing position, but If these cases were not bad enough, how about regener- (balancing organs) in the hindwing position. have ation, where an adult gecko’s tail is lost to a predator and wings in the hindwing position, and wing covers called then regrown from adult tissue? The tissue of origin for the elytra in the forewing position. Nevertheless, the characters regenerated tail is quite different from the tissue that gave forewing and hindwing persist even though the character rise to the original tail. How about experimental manipu- state changes. Wagner points out that the genetic under- lations that produce embryological inductions of body pinnings of characters differ in kind from those of the parts in completely ‘unnatural’ ways? Hans Spemann character states. demonstrated that certain species of frog can accept a sur- Characters are specified by a particular kind of GRN gically transplanted eye cup, which will then induce a to that Wagner calls character identity networks (ChINs), form in the ectoderm that was replaced above the trans- which involve transcription factors such as abdA and UBX planted eye cup. Again, the lens that covers the trans- in the forewing and hindwing, respectively. Character planted eye cup is produced out of surgically manipulated states can change in evolution, but their homological embryonic ectoderm, not the ectoderm that produced the identity as forewing (for example) is under the control of a lens that the frog was born with. Ordinary and regenerated ChIN that persists through the change. tails were surely homological, as are the lenses that cover Something is missing, however, from this description of ordinary and transplanted eye cups. These last problems the ChIN. The notion of GRNs is well established, and could be dismissed as special cases by the sceptics, but ChINs are a special case of these. In these networks, gene surely not the earlier cases. See also: Spemann, Hans products (transcription factor proteins) interact to control In Owen’s day, homology was often claimed to be an the expression of developmental genes. However, these ‘ideal’ relationship. This has often been criticised as transcription factors are used in other developmental

4 eLS & 2012, John Wiley & Sons, Ltd. www.els.net Homology and Homoplasy: A Philosophical Perspective processes than the particular ChIN at issue. What could concepts like ‘the urodele ’ that the author has argued keep a ChIN stable over the immense times (hundreds of are typological in the same sense as archetypes, hom- millions of years) that homologies persist? Wagner has an ologies, and bauplans (Amundson, 2005), and also such answer. He observes, first of all, that transcription factors venerable examples as Owen’s vertebrate archetype. At do not remain functionally identical through evolutionary last, we have a mechanistic model for the ideal entities of time. (By the ‘function’ of a transcription factor we mean the past two hundred years. Moreover, if Wagner is correct, the causal interactions that the particular molecule can ChINs can also explain such cases as regeneration, have with binding sites in DNA and with other develop- experimental embryological inductions and serial hom- mental molecules.) The parts of the molecules that bind to ology. The next step must be the gathering of additional DNA must remain consistent, or else the dramatic proofs evidence. See also: Evolutionary Developmental Biology: of genetic homology (e.g. Gehring’s shocking demon- and Constraint stration that the gene that triggers eye development in a Even if Wagner is mistaken about the genetic basis of gene can also trigger eye development in Drosophila) homology, he has at least shown how the resources of would have been impossible (Gehring and Ikeo, 1999). But modern developmental genetics can solve problems that the part of the that binds to DNA is only a fraction were seen as irreducibly and egregiously metaphysical of the entire molecule. Wagner proposes, secondly, that the during most of the twentieth century. For the better part of partial variation in function of these molecules comes from two centuries, advocates of the importance of homology the coadaptation of the molecules within a ChIN to each have been stigmatised as ‘typological’ and ‘idealist.’ other. Because ChINs are internally coadapted, they are Without embarrassment, all can be typologists today. able to persist even while the genes that produce the char- acter states, and the genes that produce earlier develop- mental events, are allowed to vary. To put it in a different way, the ChIN is able to persist even when the genetic References events that are upstream (earlier in ontogeny) or down- Amundson R (ed) (2005) The Changing Role of the Embryo in stream of it (later in ontogeny) are modified. It is the co- Evolutionary Thought: of Evo-Devo. New York: Cam- adaptation of the molecules within ChIN’s that produces bridge University Press. the paradoxical persistence of homology even during evo- De Beer GR (ed) (1971) Homology, An Unsolved Problem. Oxford: lutionary changes in the origins, and the consequences, of Oxford University Press. the homologous characters. Gehring WJ and Ikeo K (1999) Pax 6: mastering eye morpho- Presumably the alimentary canals of the ancestors of genesis and eye evolution. Trends in Genetics 15(9): 371–377. vertebrates were controlled by ChINs that persisted while Gould SJ and Lewontin RC (1979) The spandrels of San Marco the sources of the tissue that produced the alimentary and the Panglossian paradigm: a critique of the adaptationist canals of different phyla were changing. The ChINs of programme. Proceedings of the Royal Society of London B 205: dipteran body parts persisted even when they became 581–598. associated with imaginal discs in larvae rather than the Mayr E (ed) (1982) The Growth of Biological Thought. Cambridge, directly developing bodies of ancestral flies. ChINs explain MA: Harvard University Press. the paradoxical nature of homology. Mayr E and Provine W (eds) (1980) The Evolutionary Synthesis. Wagner began his examination of the developmental Cambridge, MA: Harvard University Press. Owen R (ed) (1843) Lectures on the and basis of heredity in the late 1980s, at a time when advocates Physiology of the Animals. London: Longman of the importance of development for evolution were Brown Green and Longmans. beginning to challenge the Modern Synthesis view that Roth VL (1984) On homology. Biological Journal of the Linnean development was irrelevant (Gould and Lewontin, 1979; Society 22: 13–29. Wagner, 1989). Other authors were attempting similar Van Valen L (1982) Homology and causes. Journal of theories (Van Valen, 1982; Roth, 1984). These early 17: 305–312. attempts to deal with the paradoxical homologues showed Wagner GP (1989) The biological homology concept. Annual signs of desperation. Roth picturesquely proposed ‘genetic Review of Ecology and Systematics 20: 51–69. piracy’ to explain the paradoxical cases, and others used Wagner GP (2007) The developmental genetics of homology. concepts such as ‘continuity of information’ and ‘devel- Nature Reviews Genetics 8: 473–479. opmental constraint’ to dodge the paradoxes. Of course, one can only work with what they have, and it would be unfair to criticise these efforts. But it can now be seen that such concepts are not much more than summaries of the Further Reading paradox. Wagner’s new proposal seems to be the first that Hall BK (ed) (1999) Homology. Novartis Foundation Symposium offers a direct, evidentially based, and unequivocal 222. Chichester, UK: John Wiley and Sons. explanation of how and why homologies can persist in such Russell ES (1916) Form and Function. Chicago: University of a surprising pattern. If homologies are based on internally Chicago Press. coadapted GRNs, there exists some hope for a mechanistic Wagner GP (ed) (2001) The Character Concept in Evolutionary cashing in of formerly ‘idealistic’ concepts. These include Biology. New Haven, CT: Yale University Press.

eLS & 2012, John Wiley & Sons, Ltd. www.els.net 5