Homology, an Unresolved Problem

Let us recall that the taxonomic reality studied by systematics is divided at its basic level into two general aspects, viz. taxonomic and partonomic. In the former, diversity of organisms is investigated by the totality of their characters; this diversity is ordered by grouping organisms into taxa and building taxonomic systems. In the latter, disparity of the properties of organisms is investigated, and this is ordered by grouping the organismal properties into partons (= merons) and building partonomic systems. The “puzzle” issues concerning the taxonomic aspect were considered in previous sections of this chapter; an issue of this kind is considered here concerning the partonomic aspect. This consideration is facilitated by the fact that the problems associated with the structure of taxonomic and partonomic aspects of the diversity of organisms is similar in many ways.

Each parton represents a class of equivalence of certain organismal properties considered indistinguishable with reference to the respective class-forming parameter. As units of partonomic classifications, the partons can be of different kinds. In systematics and related disciplines, of paramount importance is recognition of their status as either natural or artificial; in one of the most commonly used traditional terms they are designated as homologies and analogies. The procedure of separating partons is partonomization; accordingly, two components are distinguished in it, viz. homologization and analogization.

Without going into detail, the basis for a general classic understanding of the homologies and analogies can be presented as follows. The homologies are formed by certain “internal” causes of an “essential” kind: they can be interpreted as manifestations of the same essences of organisms, as elements of their general body plans or as derivatives of a certain archetypal (ancestral) structure. In contrast to this, the analogies are formed by the “external” causes which are accidental with respect to the organisms: the influence of environmental conditions is most usually meant, and consideration aspects fixed by a researcher also belong to this class of causes. Thus, homologies express a certain “deep” affinity of partons, while analogies reflect their “superficial” similarity. This explains the great attention that is traditionally paid to the task of revealing homologies and distinguishing them from analogies in solving particularly both taxonomic and partonomic tasks.

For systematics, the significance of partonomization is determined by the fact that partons serve as the common basis for identifying characters by which organisms are described, compared, and classified. In the simplest version, there is a one-to-one correspondence between them: one parton corresponds to one character, and elements of this parton correspond to the modalities of the respective character. In a more complex variant, characters may correspond to several partons—for example, describing the ratio of different parts of an organism. From a formal point of view, all partons are equivalent as the bases for distinguishing characters; with this, homologies yield homologous characters, while analogies yield analogous characters.

Such a differentiated evaluation of the homologies and analogies and characters associated with them is rooted in the natural-philosophical understanding of the Natural System as a network of the natural affinity of organisms manifested in their “essential similarity” by “essential characters.” This affinity was understood either natural-philosophically (scholastics, taxonomic “esotericism”) or genealogically (Darwin, Haeckel), while the “essential characters” became understood as homologies, in contrast to the “accidental characters” equated with analogies. In contemporary taxonomic schools of thought, such a distinction is inherited by typology, phylogenetics, and partly phenetics. However, in biomorphics and partly biosystematics, their assessment is different: analogies may be of greater importance than homologies for identifying life forms.

Contemporary understandings of the ways of defining the homology concept in its general meaning are so diverse that there seems to be no sign of a common agreement: this gave the zoologist Gavin de Beer reason to declare that “homology is an unresolved problem in biology” [de Beer 1971]. This homology problem, which is comparable in its importance and unresolvedness with the species problem (discussed in Section 6.7), shapes the homology puzzle in systematics. Its main content, as in the case of the species problem, lies in the impossibility of defining homology in a trivial unified way [Bock 1989; Brigandt 2002; Hossfeld and Olsson 2005; Jamniczky 2005; Kleisner 2007; Pavlinov 2012, 2018; Pavlinov and Lyubarsky 2011; Minelli 2016]. Several special books are devoted to this problem [de Beer 1971; Voigt 1973; Hall 1994; Sanderson and Hufford 1996; Bock and Cardew 1999; Timonin 2001; Wagner 2014].

The general metaphysical context of an understanding of the partonomic structure of organisms, and thus the homology problem, is provided by a hierarchical whole-part relationship [Woodger 1952; Jardine 1967; Bertalanffy 1968; Ghiselin 1981,2005; Rieppel 1988b, 1992; Lyubarsky 1996; Pavlinov 2012, 2018]. It assumes (a) the subdivision of a whole (organism, archetype) into parts and the existence of certain relations between both (b) different parts of the same whole and (c) the respective parts belonging to different wholes, with the latter (d) being, in their turn, parts of a higher-order whole.

The beginning of the contemporary concepts of homology and analogy was laid by the zoologist Richard Owen [Hall 1994] (see Section 2.4.4). His concept is based on an idea of the archetype (partly in the sense of Goethe) as a general principle of the structural organization of an ideal super-organism, which is partitioned into basic structural elements, homotypes [Owen 1848]. Owen defined homology as the correspondence of elements of the same homotype, which made them ‘‘the same,” and analogy as the absence of such correspondence. This general scheme, as applied to particular organisms, is detailed in three ways: (a) general homology is a correspondence of the main structural elements (parts, organs, etc.) in a given organism to archetypal homotypes; (b) serial homology is a correspondence of repeating elements along a segmented organism that realize sequentially the same homotype; and (c) special homology is a correspondence of structural elements in different organisms that realize the same archetypal homotype. Taking the vertebrate body plan as an instance, we have: its different parts (fins, limbs, glands, blood vessels, etc.) as different general homologs, paired limbs as serial homologs, and variants of the forelimb in fishes, birds, and mammals as special homologs.

According to Owen, whose natural philosophy was close to “Biblical Platonism” (see Section 2.4.4), an ideal archetype is single and all-encompassing for a certain group of organisms, therefore distinguishing its structural elements (partons) as homologs and analogs is also the only possible one. The Natural System of organisms is a result of different modes of realizing this ideal archetype and the embodiment of its homotypes into the corresponding special homologs of particular organisms. This implies the above-mentioned priority status of the homology: Owen himself paid most attention to homology, while he lost interest in analogy. Such an attitude was inherited by most of the explorers of the homology problem: it is believed that all comparative morphology is the “science of homology” [Remane 1956].

Owen’s typological concept subsequently underwent significant changes caused by its different interpretation. The phylogenetic one gave an idea of phylogenetic homology and led to dividing it into inherited from ancestors (homophyly of Haeckel, homogeny of Lancaster, complete homology of Gegenbaur, genetic homology of E. Wilson) and acquired as a result of parallel evolution (homoplasy of Lancaster, incomplete homology of Gegenbaur, latent homology of Osborn, homoiology of Plate). A special phylogenetic interpretation of anatomical correspondences was proposed by the zoologist Edward Cope, who borrowed his concepts of homologous and heterologous series from chemistry and “inscribed” them into his theory of polyphyletic evolution [Cope 1887]. Simultaneously with the phylogenetic view, an ontogenetic understanding of homology was developed based on the ideas of K. Baer’s epigenetic typology. On this basis, a detailed interpretation of serial homologies developed during ontogenesis was elaborated [Bronn 1858; Haeckel 1866]; this general understanding was reinforced by the term ontogenetic homology [Mivart 1870]. As a result, by the end of the 19th century, three main concepts of homology existed, viz. typological, phylogenetic (in several versions), and ontogenetic (embryological). This fragmentation continued further: in the 1920s-1930s, up to five main “homologisms” appeared, two decades later there were about ten of them, and in two more decades their number increased to several dozen [Blacher 1976]. In phylogenetics, special attention was and is paid to the delineation of homogenies and homoplasies,[1] which turns out to be more than problematic [Lankester 1870; Osborn 1902; Spemann 1915; Söderström 1925; Hubbs 1944; Sanderson and Hufford 1996; Pavlinov 2012; Minelli and Fusco 2013].

In contemporary ideas of homologies, two main generalized interpretations can be distinguished: structural (stationary) and transformational (dynamic, generative). They differ mainly by the ontic interpretation of the nature of interrelations between homologous elements: they are united either through their structural “sameness” or through their sequential transformation, respectively. Some experts contrast these two interpretations; in an extreme version, it is considered impossible to include the structural consideration of homology in studies of transformations of biological forms [Naef 1931; Kalin 1945; Borhvardt 1988; Shatalkin 1990b]. Others attempt to consider them not mutually exclusive but complementary and to combine them in one way or another [Rieppel 1985. 1988b; Brigandt and Griffiths 2007; Pavlinov 2012, 2018]. This intention is fixed by biological or synthetic concepts of homology: the latter is interpreted as a canalized development of structural correspondences [Wagner 1989; Szucsich and Wirkner 2007].

An emphasis on the transformational aspect suggests an important transition from the traditional homology of definitive structures to the homology of the processes [Bertalanffy 1968; Laubichler 2000; Gilbert and Bolker 2001; Scholtz 2005, 2010; Minelli and Fusco 2013]. In this case, the homology of structures can be correctly defined in terms of certain processes that generate them: this approach leads to the recognition of partons as process homologs [Hall 1992, 1995, 1996; Gilbert and Bolker 2001; Minelli 2003; Kleisner 2007]. Extending homology to include both developmental processes and structures as their results allows the introduction of the general concept of organizational homology as a correspondence between morphoprocesses and morphostructures [Müller 2003; Kleisner 2007]. Two types of process that generate such homologies have been considered since the end of the 19th century, phylogeny and ontogeny. Homologies defined by them are sometimes contrasted and considered incompatible [Wagner 1989, 1994; Shatalkin 1990b; Rieppel 1992]. More promising seems to be their joint consideration within the framework of the general concept of “evo-devo,” which gives an understanding of a homolog as a structural unit capable of separate evolutionary development due to its ontogenetic (quasi)autonomy [Wagner 1994, 2014; Laubichler 2000; Amundson 2005; Brigandt 2007; Suzuki and Tanaka 2017]. The correspondence of definitive structures resulting from either the same or different generative (both ontogenetic and evolutionary) pathways was proposed to designate syngeny and allogeny, respectively [Butler and Saidel 2000].

In the most recent studies on process homology, special emphasis is given to the regulator genes (Hox, MADS, etc.) that affect the formation of basic morphostructures in the early stages of ontogenesis [Holland et al. 1996; Schierwater and Kuhn 1998; Galis 1999; Shatalkin 2003, 2012; Scholtz 2005; Davis 2013]. Accordingly, it was proposed that genetic homology should be recognized as a special category [Hossfeld and Olsson 2005]. It is assumed that homology of these genes in animals makes it possible to homologize structures traditionally considered paradigmatic analogs, such as wings in insects and birds [Shatalkin 2003]. Taking a less optimistic view, especially taking into account that MADS genes are also found in plants [Niklas 1997; Ng and Yanofsky 2001), things are not as simple and unambiguous as they seem. It has been shown that homologous genes are responsible for non-homologous (in a traditional sense) morphostructures, while homologous (in the same sense) morphostructures can be regulated by non-homologous genes [Striedter and Northcutt 1991; Wray and Abouheif 1998; Wray 1999].

At a molecular level of organization, there are specific subtleties in the understanding of homology: in this case, a distinction between orthology and paralogy is of importance [Williams 1993; Hillis 1994; Wheeler 2001, Sonnhammer and Koonin 2002; Freudenstein 2005; Wagner 2007, 2014]. Orthology is a kind of special homology that corresponds to the homogeny of morphological structures: orthologous genes are regions of the same macromolecule in two organisms (groups of organisms) inherited from their closest ancestor. Paralogous genes in molecular biology (as opposed to morphology; see above) are duplicated regions; if they occur within the same organism, they correspond to the serial homology; those in different phyletic lineages superficially correspond to homoplasy. In addition, xenology is distinguished to designate alien DNA or RNA fragments included in the genome as a result of horizontal transfer [Gogarten 1994].

Modern research is characterized by an expansion of the application of the general concept of homology to include consideration of functions of organisms [Roth 1982, 1988, 1991; Wake 1992; Gilbert and Bolker 2001; Love 2007; Matthen 2007; Tetenyi 2013]. The following examples are worth mentioning here: biochemical reactions constituting the Krebs cycle; viviparity that repeatedly appeared in the evolution of animals; and behavioral stereotypes. Such interpretations bring specific problems in the homology puzzle: for example, in the case of behavioral stereotypes, it is not clear whether they should be considered conjointly with the morphological structures that perform them or independently of these structures.

An important part of the modern content of the homology puzzle, in contrast to its classical versions, is an understanding that distinctions between homologies and analogies are context-dependent and thus relative. These contexts may be set conceptually by different definitions of homology (which is self-evident); by higher-order hypotheses, within which the respective structures and functions are homologized; by specific aspects considering the structural and functional organization of biological objects; etc. [Roth 1988, 1991; de Pinna 1991; Brigandt 2002; Pavlinov 2012, 2018]. For instance, in phylogenetics, a distinction between homogenies and homoplasies depends on the contexts of particular phylogenetic hypotheses [Hennig 1966; Wiley 1981; Pavlinov 2005]. In the “new typology,” homologies and analogies are distinguished cognitively depending on a particular research task [Meyen and Shreyder 1976; Meyen 1978; Lyubarsky 1996].

Another important aspect of the relative character of the boundaries between homologies and analogies is determined by the recognition of their quantitative and fuzzy character. This means that, instead of their strict delineation, it may be more correct to consider varying “degrees of homology/analogy,” according to which there may be complete and partial homologies [Gegenbaur 1865]. This is most characteristic, for example, of ontogenetic interpretations: the more similar developmental trajectories are, the more homologous are both corresponding morphoprocesses themselves and morphostructures generated by them [Sattler 1994; Minelli 2003]. Specific methods for determining homologies in numerical taxonomy make them strictly quantitative: the mutual similarity of the structures is considered to be a measure of the degree of their homology [Smirnov 1959; Sneath and Sokal 1973]. The same is true for the homologization of macromolecules through alignment and assessment of the similarity of their sequences [Hillis 1994; Doyle and Davis 1998; Wheeler 2001, 2016; Morrison et al. 2015]. A quantitative character distinguishing between homologies and analogies also occurs when considering judgments of them as hypotheses that may be more or less plausible.

An important part of the homology puzzle is an elaboration of homology criteria for solving the particular tasks of homologization. For anatomical structures, a sufficiently consistent system of criteria is developed on a typological basis [Remane 1956], which goes back to the theory of analogs by E. Geoffroy de Saint-Hilaire [Geoffroy Saint-Hilaire 1830]. This includes three main criteria: of special quality, of position, and of intermediate forms. The first two are primary and allow distinction between two descriptive forms of structural homology, compositional and positional, respectively [Jardine 1967; Sluys 1996; Minelli and Fusco 2013]; the third is secondary to them. According to the criterion of special quality, structures are homologous if they coincide in some essential “internal” properties (for instance, tissue composition). The criterion of position (connectivity) implies that structures in different organisms are homologous if they occupy the same position among the same structures that have already been proved to be homologous. The criterion of transitional forms (continuity) is addressed if two previous criteria do not provide quite clear recognition of homology; it is especially relevant for the transformational homology concept. An auxiliary criterion of congruence gives preference to the hypothesis of homology of a structure that is more consistent with other hypotheses of homologies of other structures within the context of a particular higher-order (phylogenetic, etc.) hypothesis [Patterson 1982; de Pinna 1991]; this is a kind of logical judgment by analogy (see Section 3.6 on the latter).

Currently, several ideas of how to solve the homology problem and puzzle are considered, and these ideas can be divided between two opposites. One of them presumes rejection of the very general concept of homology, inventing different particular concepts and corresponding terms [Borhvardt 1988; Shatalkin 1990b]; this position goes back to early attempts to interpret homology in a phylogenetic manner (see above). It is sometimes justified by reference to scientific pluralism [Heather and Jamniczky 2005; Kleisner 2007], by which the homology problem is quite comparable with the species problem [Ereshefsky 2001b; Pavlinov 2012] (see next section). On the other hand, it is proposed that we should think of a totality of structural, functional, and developmental correspondences of partons, which might be represented in the form of a complex hierarchical system integrated by the concept of hierarchical (deep), or combinatorial (factorial) homology [Minelli 1998, 2016; Scotland 2010; Minelli and Fusco 2013]. According to this, at different levels of hierarchically stratified supra- and intra-organismal diversity, different but interconnected particular concepts of homology can operate, shaping a single conceptual system of metahomology [Kleisner 2007]; this viewpoint agrees in general with the idea of a conceptual pyramid (see Section 3.2.3).

  • [1] In contemporary cladistics, homogeny is incorrectly identified with the whole homology, thus ignoring other particular interpretations of the latter.
< Prev   CONTENTS   Source   Next >