Shishkin, 2006. Individual development and lessons of evolutionism

ONTogenesis, 2006, vol. 37, No. 3, pp. 179‑198 UDC 576.12(575.53591.3) INDIVIDUAL DEVELOPMENT AND LESSONS OF EVOLUTIONISM ©2006 M. A. Shishkin Paleontological Institute of the Russian Academy of Sciences 117997 Moscow, GSP‑7, Profsoyuznaya St., 123 E‑mail: shishkin@paleo.ru Received by the editorial office 30.08.05. Final version received 14.12.05. “For any problem there is a solution that is simple, obvious, and wrong.” Albert Einstein “Strive for simplicity and do not trust it.” Alfred North Whitehead The current crisis of evolutionism was predictable from the start, because the pre‑formational model of development, embodied in the idea of discrete inheritance, contradicts the systemic properties of ontogeny. Accordingly, the principle of selection of hereditary factors cannot explain evolution. The synthetic theory built on it contains irresolvable contradictions in its key concepts. According to an alternative epigenetic theory, derived from the integrity of the living organization, inheritance is a product of selection and expresses a teleonomic direction of development toward a stable final state. The union of the genetic concept of evolution with the recognition of developmental integrity is fundamentally impossible. The dominance of genetic views on evolution is rooted not in their logical justification but in conformity with the mechanistic tradition of the 18th–19th centuries. For the same reason, evolutionary biology as a whole typically equates particular linear dependencies with evolutionary regularities. Following this path in contemporary searches for a “new evolutionary synthesis” predestines them to failure. Evolutionary interpretation of genetic generalizations is possible only on the basis of their description in terms of individual development. Keywords: system of individual development, inheritance, evolution, stability, regulation, reductionism. INTRODUCTION The current state of evolutionary concepts is far from the elegance and persuasiveness it possessed in the eyes of most researchers a few decades ago, during the period of unchallenged dominance of the Synthetic Theory of Evolution (STE). Increasingly, calls are made for a “new evolutionary synthesis.” Such an outcome was initially predictable, and its causes lie not in present‑day progress in biology. One can say they are embedded in historically formed features of collective scientific thinking. In the one and a half centuries since the emergence of Darwinian doctrine, a main direction of evolutionary thought has been clearly outlined. This is the neo‑Weismannian view, according to which evolution is based on the selection of discrete hereditary factors (genes). On the other hand, during the same period a modern paradigm of theoretical biology has accumulated. This is a sum of comparative and experimental generalizations that, even without being tightly linked, are widely accepted and form the basis of biology textbooks. Comparing these two things raises the inevitable question: how did the first manage to become established in the presence of the second? Their incompatibility seemed obvious 80 years ago, just as it does now. Analyzing this problem constitutes our main task. Its solution lies in the fact that scientific explanations are ultimately determined not by logical requirements but by the researcher’s initial conceptual choice. This is merely deduction from a general rule that the character of the read information always depends on the specifics of the recipient’s perception. As Lyubyshev (1925) put it, theories are not built from facts; rather, facts are fitted into a system based on theory. The nature of a theory’s postulates determines not only its set of working concepts (often devoid of meaning outside the theory) but also the criteria by which it distinguishes essential phenomena from “informational noise.” Only a change of postulates (“scientific revolution”: Kuhn, 1975) makes a fundamentally different explanation of the same circle of phenomena possible. Before examining the impact of these factors on the evolutionary thought of the past century, we must identify several key generalizations that allow us to understand the choice at stake. The early period of post‑Darwinian evolutionism was based on embryological generalizations, primarily on E. Haeckel’s biogenetic law and K. Bär’s law of embryonic similarity, which Darwin himself saw as confirming his theory. From the 1920s onward, this role passed to genetics, leading to the emergence of STE, in which evolution is reduced to mutational processes and changes in allele frequencies in populations under selection. By the definition of Schmalhausen (1968, p. 20), the genetic theory of evolution is the absorption of Darwinism by genetics. Two defining features of this theory are important for us. First, it lacks a unified concept in which fundamental notions would be causally interrelated. Such a task is not even posed. In STE, inheritance and natural selection are understood as two independent evolutionary factors. The former is given the primary role, while the participation of the latter (mainly as a filter for mutations) is not essential, for example, in the random fixation of a fluctuation in gene composition in populations or in the emergence of monofactorial differences between races or morphs. Thus, the early genetic rejection of the creative role of selection (Morgan, 1937, 1938) left an indelible imprint on their subsequent mechanical “synthesis,” often manifesting as new recurrences (Golubovsky, 1981; Gilbert et al., 1997, p. 338). The second feature of STE is a violation of the methodological setting rightly declared by its authorities (Dobzhansky, 1951; Mayr, 1981) and requiring that all manifestations of biological order have an evolutionary explanation. In fact, the concept of inheritance in STE has only a tautological explanation as an immanent property of hereditary factors (genes). Even if it is replaced by a description of DNA replication and matrix synthesis, nothing is clarified, because we are dealing with complex regulated systems that are clearly not reducible to physico‑chemical processes. The obvious inability of STE to explain the causes of self‑organization in living beings generated skepticism among many researchers regarding the very possibility of explaining the latter on the basis of selection. As the main source of biological order they see internal laws of formation and development of living systems, while selection at best is recognized as a factor acting on the products of systemic self‑organization (Salthe, 1993; Kauffman, 1993, 1995). It is not surprising that since the early 1980s calls have been heard for a more refined selectionist theory in which embryological generalizations would again take their proper place. It is now widely recognized that the ontogeny contains those complex genotype‑phenotype relationships without which description of macro‑evolutionary changes is impossible. Understanding grows that the limited, specific, and discrete pathways of phenotypic realization cannot be directly expressed in the language of mutations. With this in mind, hopes are expressed that integrating the genetic theory with the regularities of individual development will lead to a more effective synthesis than that achieved in the past century (Alberch, 1982; Gould, 1982a,b; Maddison et al., 1982; Vrba & Eldredge, 1984; Reff & Koffman, 1986; Thomson, 1988; Gilbert et al., 1997). In fact, this usually means not a revision of the genetic evolutionary theory but merely its supplementation. It is assumed that STE remains adequate for describing microevolutionary events, but explaining macroevolutionary processes, due to their autonomy, requires data from embryology (Gould, 1982b; Vrba & Eldredge, 1984). Nevertheless, a significant shift in dominant selectionist thinking is evident. Even more radically it is expressed in the view of the prevailing role of systemic self‑organization mechanisms relative to mutations and selection (Kauffman, 1993, 1995). TWO VERSIONS OF SELECTIONISM AND INDIVIDUAL DEVELOPMENT It is easy to see that the aforementioned hopes for improving the evolutionary paradigm are linked to incorporating into it those aspects of biological knowledge that form the basis of the theory of stabilizing (canalizing) selection of Schmalhausen‑Waddington (Schmalhausen, 1940, 1941, 1968a,b, 1982; Waddington, 1957, 1966); the latter was formalized by the author under the name epigenetic theory of evolution (ETE) (Shishkin, 1984, 1987, 1988, 2003; Shishkin, 1989, 1992). According to this concept, the immediate object of evolution is not genes but whole developmental systems, whose fluctuations are stabilized as irreversible changes. At the individual level, the material of selection consists of carriers of divergent morphogenetic reactions (morphoses) realized by the system when conditions deviate from the norm. Selection for a preferred aberration, realized by non‑identical carriers, turns it into an inheritable change, gradually replacing the former norm. The stability (heritability) of the norm here rests on regulatory interactions within the system created by selection and channeling a particular developmental trajectory. Thus, inheritance in this theory is not a partner of natural selection but its product, acting as an integral property of normal development. Evolutionary changes begin with the phenotype and spread as they become stabilized toward the genome, not the reverse. Evolution is viewed here as a payment for repairing ontogenetic stability of the living system, and natural selection as a means by which the system seeks a new equilibrium to replace the lost one. Selection is merely the biological expression of a mechanism that transforms an open system through successive cycles of correcting its state. If one appreciates the significance of the biological generalizations that constitute the foundation of this theory (see below), the current shift in evolutionary thinking becomes quite predictable. Yet, just as predictably, the anticipated “new synthesis,” as outlined above, is unrealizable because the synthetic and epigenetic theories rest on mutually exclusive foundations. To clarify the alternative nature of these two concepts, let us return to some elementary notions. The task of evolutionary theory, from Lamarck onward and despite certain objections (see, e.g., Vorontsov, 2004, p. 334), is to explain organic teleology. Teleological, or teleonomic, behavior of systemic objects is called such when it manifests as a tendency toward equilibrium maintained by self‑regulation of arising deviations. For living organisms this equilibrium state is their standard organization (adaptive norm), realized during ontogeny. Its resistance to disturbance, i.e., its capacity to persist in time and space (across generations and populations), is the expression of biological teleology, or adaptability (Wake et al., 1983; Shishkin, 1987). Another synonym for these notions, perhaps unexpected for the modern biologist, is inheritance. Historically this concept predates genetics. Borrowed from breeding practice, it denotes the transmission of phenotypic traits from parents to offspring; but “transmission” here means nothing other than the stability of the realization of such traits in the offspring’s ontogeny. In the words of Gurvich (1923), the problem of inheritance is the problem of specific development. Thus, the decisive question for evolutionary theory is: what is meant by the mechanism of living organization’s realization? In the history of biology two alternative ways of interpreting it are known: pre‑formational and epigenetic. In the first case it is assumed that individual organismal properties are determined by independent factors in the germ cell, i.e., development can be described with “terms borrowed from the final result” (Gurvich, 1944, p. 148). It is accepted that even when multiple interactions among such factors (genes or other determinants) are allowed during development, each still plays its specific role in phenotype realization, which in principle can be isolated. This reductionist approach objectively underlies genetic conceptions of development and evolution, despite frequent declarative attempts to abandon it (cf. Svetlov 1972, 1978; Shishkin, 1987). It is labeled the “principle of integrated gene activity,” implying that the mosaic of such actions ultimately produces an adaptive adult organization (Goldschmidt, 1940; Goodwin, 1982). In this approach two features are self‑evident: a) systemic properties of development are ignored, because the whole is summed from the parts; and b) the developmental process loses independent interest for the theory, becoming essentially a carrier of predictable properties from hereditary factors to the phenotype. Accordingly, STE manuals often omit the concept of ontogeny altogether. In contrast, the epigenetic (systemic) approach underlying ETE implies an increase in the diversity of the factors themselves during development and an asymmetry between them and the properties of the adult organization. The developmental outcome reflects the constitution of the entire germ cell and cannot be reduced to the properties of its elements (Gertwig, 1895). Whether determination of embryonic primordia occurs very early in ontogeny (in mosaic development typical of most basal invertebrates) or, conversely, late (in regulatory ontogeneses), it is in all cases predetermined by the determination of the whole properties of the oocyte or germ cell (Svetlov, 1964, 1978; Goodwin, 1982). This factor of wholeness is usually regarded in embryology as an anisotropic biological field that controls developmental course and is transformed in the process (Gurvich, 1944; Svetlov, 1978; Belousov, 1979; Goodwin, 1982). This model logically excludes the notion of a initiating role of genes in development and is consistent with what is actually known about the mechanisms governing their differential activity. As Olegov (1967, p. 198) said, any example of ontogeny convinces that “genes are not dictators on which the course of events depends, but rather clerks working according to established traditions.” Experiments transplanting inactive nuclei from adult cells into enucleated oocytes at various stages of maturation show that gene expression is stimulated by the cytoplasm and determined by its state at the corresponding stage (Gordon, 1977). This allows us to speak of the whole properties of the germ cell as a controlling factor of expression. Equally important evidence of systemic control of ontogeny comes from experiments inactivating the zygote nucleus in various animal groups, showing that development without genome participation can reach the gastrula stage or even later phases (Tokin, 1977; Dondua, 1979). (Further development halts due to depletion of transcripts supplied by the maternal cytoplasm and the impossibility of synthesizing needed products in the absence of a genome.) From this it follows that the program of normal morphogenesis is embedded in the overall organization of the embryo, not in chromosomal units. For these systemic conceptions of ontogeny, the relationship between the developing organism and its genome is comparable to a person rereading the same book at different stages of life. The informational value of the book depends on the “literacy” of the reading system, which in this case is the organism transformed during development (cf. Bertalanffy, 1969). Such a view of ontogeny, as noted, underlies ETE. It implies that neither development nor its evolution can be substantively described as the summed result of elementary initial factor activity. The distinction between the two described approaches to understanding the mechanism of ontogeny can be presented as alternative ways of hitting a target. For a firearm, hit accuracy depends entirely on the sight’s precision at the moment of firing; the projectile’s flight adds nothing but deviating “noise” (analogous to environmental influences on gene expression). The start and end of the process are linked by a linear relationship. In contrast, for a homing missile the initial flight direction is set approximately, and its aiming accuracy improves gradually through successive cycles of auto‑correction of its trajectory. This is a model of a dynamic system transforming its state toward a final equilibrium. This is exactly how a canalized ontogeny behaves, showing a progressive reduction of variability toward the adult stage (Baer, 1828; Svetlov, 1964; cf. Waddington, 1957, p. 32). Although the second model matches the real properties of individual development, the neo‑Darwinian STE was built on an alternative (linear‑causal) model. Why this happened is, as already stated, the main aim of our analysis. EPIGENETIC THEORY OF EVOLUTION: HISTORICAL FOUNDATIONS AND CURRENT ROLE First, let us recall old empirical generalizations that objectively point to the stability of the final developmental outcome and thus to the systemic nature of the ontogenetic process. From comparative embryology and paleontology it has long been known that the adult organizational plan or its individual features can persist in evolution despite increasing changes in the way they are realized in individual development. The observed dynamics of historical ontogenetic transformations have been described in the literature under many names—such as acceleration (Sore, 1887), condensation (Haeckel, 1866), pre‑emptive segregation (Lankester, 1877), adulation (Jägersten, 1972), rationalization (Schmalhausen, 1940), straightening (Müller, 1940), embryonization, etc.These ratios seemingly indicate that, during evolutionary development, “nature strives to reach the end by the easiest means” (Balfour, 1885). Although, within the framework of 19th‑century linear determinism, such a conclusion was seen as a teleological attempt to “put the cart before the horse” (Conklin, 1905, p. 110), it has found numerous empirical confirmations. It follows with obviousness that the adult organization is historically more stable than the mode of its ontogenetic realization. Experimental biology soon confirmed this rule also for a particular type of ontogeny, showing that artificial disturbances of early development can, within wide limits, be regulated back to normal in later stages. This regularity is known as the “Roux rule,” or Driesch’s principle of equipotentiality (Roux, 1896; Driesch, 1908). Evaluating these generalizations retrospectively in the light of subsequent developments in evolutionary thought, it is easy to notice that they contain two key prohibitions concerning the interpretation of evolution adopted by genetic theory. 1. The mechanism of ontogeny, by ensuring the stability of the adult stage, thereby protects it from “forced” change under the direct influence of alterations in the initial developmental factors; it is directed toward neutralizing such influences through self‑regulation. 2. The final stage of development is equipotential (invariant) with respect to a wide range of variations in its elementary initial factors and, consequently, cannot be described in terms of those factors. This corresponds to the general rule that system parameters cannot be described in terms of dynamic variables characterizing its lower hierarchical levels. Regardless of embryological experience, the second of these rules repeatedly entered the field of chromosome genetics in its attempts to understand genotype–phenotype relationships. It suffices to mention the discovery by Chetyverikov (1926) of genetic heterogeneity in natural populations beneath the veneer of the wild phenotype. From this, it seemed only one step away to recognize that the mechanism of evolution cannot be substantively described in reductionist terms. Yet even today we do not speak of a full awareness of this conclusion. If, for the Modern Synthesis, selection for the preservation of phenotypic norm implies a continuous change in genotype–phenotype ratios (Olenov, 1976), then genetic thinking is directed toward seeking symmetry between the first and the second. Embryological conclusions about the regulatory properties of individual development, together with the doctrine of natural selection, form the basis of the Schmalhausen‑Waddington concept. Another indispensable source is the revolutionary work of Goldschmidt in physiological genetics. It led to the view of development as a reactive system possessing holistic properties that are not described in terms of genes (Goldschmidt, 1938; 1940, pp. 218‑219). Goldschmidt’s experiments on phenocopying mutational effects showed that any developmental aberration, regardless of its cause, ultimately characterizes only the properties of the given system and does not exceed its capacity for response. Interestingly, this latter position is sometimes recognized at an empirical level also within traditional genetics and the Modern Synthesis as the conclusion that no mutation takes an individual beyond its species affiliation (Dubinin, 1966, p. 240; Mayr, 1968, p. 432). Evidently, this is not realized as a rejection of the very idea of evolution by elementary mutations. On the contrary, Goldschmidt clearly understood this, asserting that the essence of evolution lies in systemic transformations of ontogeny. His well‑known hypothesis of “systemic mutations” (discussed for many decades without grasp of its true motives) was merely an unsuccessful attempt to find a mechanism for such transformation. Goldschmidt’s generalizations also vividly demonstrated the illusory nature of a principled separation between hereditary and non‑hereditary changes (by the criterion of genetic determination)—a notion that has strangely persisted in biologists’ minds since A. Weismann. The oddity lies in the fact that, with the emergence in the 20th century of the concept of reaction norm, it seemed self‑evident that any realized trait is an expression of this inherited norm under specific developmental conditions. Yet only Goldschmidt’s experiments first provided evidential support for this, allowing the conclusion that all developmental deviations share a common quantitative nature, namely mismatches in the parameters of morphogenetic reactions. Thus opened the way to understand that phenotypic heritability is merely an indicator of how regulated (canalized) its ontogenetic trajectory is within a given developmental system, as formulated in the Modern Synthesis (Waddington, 1957; Schmalhausen, 1982). That the experimental generalizations we have considered constitute the foundation of the Modern Synthesis follows obviously from its concepts of inheritance, development, and evolutionary mechanism. Recall some of its formulations. “…We consider the stability of a trait not as a property of the gene but as an expression of the interdependence of parts in the correlational systems of the developing organism… Stabilizing selection is the primary integrating factor of evolution” (Schmalhausen, 1982, p. 174; italics ours). “The creative role of natural selection is the creation of hereditary mechanisms in the form of a system of interdependent reactions that ensure reliable development… Apparently, all organismal reactions initially arise in connection with environmental factors that trigger a morphogenetic response” (ibid., p. 214). “Hereditary stability of organization rests on the complexity of a historically formed system of regulatory correlations, not on the robustness of hereditary material” (ibid., p. 218). The soundness of any theory is tested by how successfully it explains facts and how well those facts are predicted by deductions from its basic propositions. In this sense, the Modern Synthesis appears to be the only adequate interpretation of Darwinian doctrine. Major empirical generalizations of population genetics that it either cannot explain without hard‑to‑test assumptions (such as frequency‑dependent selection or genotype‑environment interactions) or treats as informational noise, on the contrary, are logically predictable for the epigenetic concept. For example, this concerns the stability of the standard (wild) phenotype compared with anomalies (“genetic homeostasis” in the language of the Modern Synthesis), heterogeneity of natural populations, the possibility of altering the genetic basis of the phenotype by selection (especially evident in sex‑determination shifts), or the non‑Mendelian inheritance of variation in natural populations. Mendelian inheritance itself receives a consistent explanation in the Modern Synthesis, unlinked to the inevitable genetic need to ascribe pure‑line properties to natural aberrations (Shishkin, 1987; see also below, pp. 186‑187). Conversely, the epigenetic concept predicts those categories of phenomena that traditionally served Lamarckism as proof of inheritance of traits primarily induced by the environment. First among these is the parallelism of geographic races and modification‑induced variability—a phenomenon stripped of evolutionary content in the eyes of neo‑Darwinists (Dobzhansky, 1951). The same applies to facts employed by the Morgan‑Boldwin hypothesis—a direct predecessor of the idea of stabilizing selection. It must be especially emphasized that, by setting the explanation of the stability of living organization as its task, the Modern Synthesis for the first time offered an explanation of organic purposiveness, which should constitute the aim of evolutionary doctrine. Both main alternatives to the Modern Synthesis—Lamarckism and neo‑Darwinism (the Modern Synthesis)—despite their apparent polarity, occupy essentially identical positions on this issue, namely they effectively evade answering it. Indeed, if in the eyes of Lamarckians purposiveness, i.e., stability, is an a priori property of all realized changes, then in neo‑Darwinism, where this phenomenon is expressed through the synonymous concept of inheritance (see above), its essence is understood analogously—as an immanent property of hereditary factors that does not require evolutionary explanation. The contrast between the explanatory potential of the epigenetic theory and the uncertain place it occupies in evolutionism even today may at first glance seem strange. A major role was played by the historical context of its emergence. This theory did not arise directly from the evolutionary ideas of the late 19th century, in which embryology was at the forefront. It emerged from genetic research that secured the victory of neo‑Darwinism and initially relied on the system of operative concepts it had developed. Yet the language of any theory reflects its original premises, and attempts to use it in a different worldview lead to an eclectic mixing of concepts and principles. Such practice inevitably obscures for the reader the alternative character of explanations offered by the new theory. This happened with the Modern Synthesis as well. Although its founders introduced a number of key concepts reflecting the systemic properties of development as the object of evolution (epigenetic landscape, ontogenetic trajectories and their canalization, developmental autonomy, stability as an expression of regulatory correlations, etc.), they nevertheless continued to employ reductionist genetic notions without interpreting them in the language of their own theory. A vivid example is Schmalhausen’s “Factors of Evolution” (1968a), in whose content it is difficult to see an alternative to neo‑Darwinism unless one reads the author’s earlier works on stabilizing selection (1940, 1941, 1982). Among the elements of dualism in Schmalhausen’s views we note only a few. First, stabilizing selection (despite several statements about its universality) was regarded by him merely as one form of selection alongside the driving, or direct, form. Yet the cause of evolution is always a disturbance of stability (Spencer, 1899). The transition of a physical system from one stable state to another is impossible without an intermediate nonequilibrium phase, during which relaxation periods of system oscillations lengthen and its susceptibility to parameter changes increases (Volkenstein, 1984). To assume that new stable states can arise ready‑made—because there can be no other substrate for “direct selection”—means believing in the spontaneous emergence of purposiveness. In other words, in some cases stability (purposiveness) is interpreted by Schmalhausen as the result of selection creating regulatory developmental correlations, while in other cases its random mutational emergence is allowed—in full accordance with neo‑Darwinian understanding of evolution. From this it is also clear that the “selection of mutations” allowed by Schmalhausen (1968a, 1982) is a notion devoid of content in the language of his theory. According to it, mutations lack a constant expression without which such selection would be impossible. Any phenotypic variant realizable within the given developmental system is merely an expression of one of the system’s holistic reactions. Accordingly, the material of selection is not mutations but heterogeneous individuals—iso‑aberrants that realized the developmental variant that proved most viable under new conditions. “A new trait arises not as a mutation but as the response of the genotypes of a heterogeneously hereditary population to a new environmental factor” (Kamsilov, 1967, p. 113). Thus, the idea of “selection of mutations” in Schmalhausen is another relic of neo‑Darwinian thinking. The consequences of such linguistic fuzziness and eclecticism in some of its generalizations were entirely predictable. It is not surprising that the doctrine of stabilizing selection was often characterized as one of the achievements of the contemporary (i.e., neo‑Darwinian) evolutionary synthesis (see, e.g., Schmalhausen, 1968a, editorial note p. 10). Similarly, the alternative nature of Waddington’s (1957) views relative to neo‑Darwinism remained poorly understood by contemporaries. Proponents of the Modern Synthesis, such as Mayr (1968, p. 432), may agree that species organization is determined by its epigenetic system, yet they understand the latter reductionistically, i.e., as a specific set of genes (ibid., p. 480). The same applies to the interpretation of Waddington’s experiments on genetic assimilation, which showed that an unstable anomaly initially caused by developmental disturbances can be transformed by selection into a stable (heritable) trait. Such facts, demonstrating the selective nature of the inheritance mechanism, are interpreted by neo‑Darwinists (Mayr, 1968, 1974; Ruse, 1977; Grant, 1980) in their customary way—as the result of selection lowering the threshold for expression of a “hidden” trait controlled by a particular gene, through the creation of an appropriate genotypic environment. TRADITIONS OF SCIENTIFIC THINKING AS THE BASIS OF NEODARWINISM However, the main reasons for the rather ineffective assimilation of the epigenetic theory by contemporary evolutionists, in our view, lie much deeper. As noted at the outset, they are embedded in the very features of collective scientific thinking as a historical process. Their essence is clearly revealed in the turning points of scientific thought that mark crises of the prevailing theoretical paradigm. Several key factors can be distinguished. First of all, the evolution of theoretical representations bears little resemblance to a linear ascent from lesser to greater knowledge. It is more comparable to the course of ontogeny, where the process of stepwise determination proceeds from a seed consisting of successive choices that each conclude a phase of uncertainty (“sensitive period”). Each such step, once made, narrows the potential spectrum of further differentiations. Likewise in science, the initial conceptual choice channels the direction of inquiry for whole generations of researchers, who already accept it without discussion as an “obligatory postulate of human thought, not merely one approach to nature” (Lyubyshev, 1925, p. 16). As in morphogenetic processes, it does not matter whether this choice was right or wrong; the crucial point is that it became part of the collective consciousness and thereby set the vector for subsequent events. Consequently, the entire language of the emerging theory is predetermined by this choice and usually lacks content beyond it. If a theory ultimately reaches a dead end, the only way out is a revision of the original choice—that is, a return backward (Shishkin, 1989, 1992). “Scientific progress often requires the restoration of old truths and their reinterpretation” (Gould, 1982a, p. 344). In development, an analogous shift is the replacement of its previous primary determination with an alternative, possible only as a new choice in a later ontogenetic cycle. In the realm of science, as Max Planck famously said, truth triumphs only with a change of generations. What, then, determines the choice of a theory’s initial principles? For the post‑Darwinian stage of evolutionary thought, the answer is obvious. The decisive factors were traditional conceptions of methodology and the tasks of scientific knowledge. For the mechanistic mindset dominant in natural science of the 18th–19th centuries, a synonym for truly scientific explanation was the identification of linear relationships between phenomena and their immediate causes. The sole aim of cognition was analysis, i.e., the dissection of reality (Bertalanffy, 1969). (This stance is reflected, among other things, in the fact that early experimental embryology called itself “the mechanics of development.”) Accordingly, the impossibility of a linear‑causal explanation was taken as equivalent to acknowledging the action of unknowable forces—as clearly illustrated by Driesch’s (1908) idea of entelechy governing ontogenetic development. Thus the dominance of the reductionist approach to explaining biological problems was pre‑determined. It is therefore unsurprising that all speculative precursors of the chromosomal theory of inheritance—the idea of pan‑genesis by C. Darwin, the theory of idioplasma by K. Negeli, the theory of germ plasm by A. Weismann—were built on the same reductionist foundation, i.e., they assumed the existence of elementary particles that individually determine the properties of the adult organism. This was true regardless of whether such ideas were advanced by Lamarckians or by selectionists. If the interpretation of theoretical biology’s own experience proceeded according to the laws of logic—i.e., by seeking a non‑contradictory reconciliation of accumulated generalizations—the fate of Weismann’s theory would have become a turning point, demonstrating the impossibility of constructing an evolutionary theory on the principle of particulate inheritance. The fact that reality turned out otherwise vividly demonstrates how self‑sufficient pre‑existing mental settings can be when choosing a theory’s foundations, irrespective of known prohibitions on that choice. The most instructive point follows. Weismann’s theory (Weismann, 1892) appears to be the only example in the pre‑Schmalhausen history of evolutionism where mechanisms of inheritance, development, and natural selection were logically linked. Unlike most later neo‑Darwinists, Weismann recognized that from the principle of mosaic inheritance necessarily follows the recognition of mosaic (pre‑formed) realization, which obliges the presentation of a corresponding developmental model. In Weismann’s concept, the latter was represented by the theory of non‑Mendelian division, explaining somatic differentiation and assuming a complete dissociation of developmental outcomes from the pathways of germ‑plasm inheritance.But these concepts turned out to be in sharp conflict with the discoveries of experimental embryology at the turn of the 19th–20th centuries, which revealed the key role of regulation in development and thus the irreducibility of its mechanisms to mosaic determination of organismal properties. Consequently, the Weismannian theory of inheritance failed precisely as a theory of pre‑formed realization. This lesson, it seemed, should have been decisive for understanding the nature of G. Mendel’s laws after their rediscovery at the beginning of the 20th century. Logic demanded the recognition that Mendelian factors controlling the modality of traits in hybrids of pure lines cannot be identified with material particles, because that would entail a return to the preformationist model of development. Only one correct way of interpreting these factors is possible—as epiphenomena, i.e., expressions of relations between real entities (cf. Holubovskyi, 1982). According to Kamshilov (1935, p. 141), the Mendelian gene is not a corpuscle but an expression of the difference between two genotypes, that is, a feature of the genotype as a whole manifested in a distinctive type of development. Within the ETÉ framework, the Mendelian gene is defined in the same way as a relation between two alternative parental creodes combined in the hybrid’s developmental system. This is a symbol of the morphogenetic shift required for the choice between creodes at their point of dichotomy, i.e., a comparative characteristic of two stabilized developmental types, not an element of hereditary substance. The choice of locus whose activity determines this switch is set by the stabilization of parental types and can be altered by selection (Shishkin, 1987). The theoretical impossibility of equating Mendelian genes with chromosomal loci followed not only from the conclusions of experimental embryology. The attempt to resolve this problem within Mendelism itself led to the same conclusion. In the views of such Mendelian authorities as Johannsen and Baur (Baur, 1919, p. 121; Johannsen, 1926), the gene was merely a convenient abstraction characterizing the difference between two idioplasms and referring to a specific comparison (cf. Kamshilov, 1935; Holubovskyi, 1982). Accordingly, phenotypic properties, according to Johannsen, are not independent effects of separate genes in the Morganist sense; he regarded the phenotype as an expression of the complex interactions between genotype and environment, in which each trait reflects the reaction of the entire zygotic constitution (Johannsen, 1926; Johannsen, 1933). Nevertheless, in 20th‑century biology the interpretation of hybrid analysis genes as units of hereditary substance, independently determining phenotypic traits, prevailed. The parallelism between the pattern of homologous chromosome segregation and combination and Mendelian laws, demonstrated by V. Setton and T. Boveri and adopted by T. Morgan’s school, was enthusiastically taken as a self‑evident foundation for such an interpretation—as finally found objective confirmation of the particulate nature of inheritance. Yet no one asked under what conditions such correspondence arises. In fact, proper Mendelian inheritance observed in crosses of pure lines is far from typical for “raw” natural variability, as numerous experiments have shown (see, e.g., Balkashyna, Romashov, 1935). In other words, chromosomal behavior laws do not, in general, have an ordered expression in the pattern of trait inheritance. Even after Fisher’s (Fisher, 1930) conclusion that simple Mendelian segregation is a product of modifier selection (and not an original inheritance regularity), the entrenched belief in the chromosomal nature of Mendelian genes remained unshaken. Logical arguments go unheard where they conflict with traditional expectations, and those expectations, in this case, demanded a reductionist explanation of inheritance. The concretization of Mendelian genes, equating a systemic property of development with a chromosomal segment and signifying the triumph of “factorial preformationism” (Svetlov, 1978), could not leave Morganian genetics untouched. It placed it before an insoluble dilemma in assessing the nature of gene–trait relationships. On the one hand, Mendelian analysis, on which its judgments about inheritance are based, assumes a direct projection of genes onto traits. “All modern concepts of inheritance are based on the recognition of special carriers of all organismal features, contained in the gametes” (Filippenko, 1924, p. 10). On the other hand, attempts by genetics to evaluate the role of chromosomal genes in individual development (starting with early sex‑determination studies that led to the idea of genetic balance) clearly demonstrated the impossibility of such linear correspondences. This forces genetics to declaratively distance itself from their recognition, attributing it to early Mendelism or even to amateurish notions (see, e.g., Dubinin, 1966a; Timofeyev‑Resovsky, Ivanov, 1966). Most often it is claimed that ultimately each trait is determined by the entire genotype; yet this view is still sometimes considered extremist (Koročkin, 2002a, p. 15). Morgan also experienced difficulty with this contradiction, which is especially understandable given his extensive experience in experimental embryology. He noted that the particulate character of inheritance implies only the independence of its units with respect to mutation, combination, and crossing‑over, but does not exclude the possibility that each trait is determined by the whole embryonic plasma (Morgan, 1924, p. 232). In fact, from this contradiction only one way out is possible: the recognition that, from the outset, two different concepts of the gene are at stake—that the Mendelian factor and the chromosomal segment are distinct entities. If this contradiction was at least implicitly recognized in Morganian genetics, the synthetic theory of evolution built upon it did not notice it at all, because the organism as a whole and the problems of its ontogenetic implementation are not part of its substantive concepts. Here a linear correspondence between chromosomal genes and their phenotypic markers, evaluated by their contribution to fitness, is assumed. Without this assumption, analysis of changes in allelic composition of populations—what constitutes the essence of evolutionary process for the STE—would be impossible. In recent decades, as the theory’s difficulties have grown, its proponents more often speak of this principle as an extreme simplification, yet such an assessment does not capture the essence of the matter. In reality, the issue is the untenability of the very idea of factorial preformationism that underlies the theory. Traditional searches for linear determination of hereditary properties are linked not only to historical reasons. In any natural science, the initial research period is characterized by the predominance of analytical methods and, consequently, a largely reductionist approach to describing causal relations; the systemic regularities behind them are recognized only later. Thus, for example, the emergence of molecular genetics began with a revival of the same notion of a direct gene–trait correlation in the form of the “one gene–one enzyme” principle. As expected, further research revealed the ambiguity of the relationships between the coding matrix and the functional protein molecule with its spatial structure. Poly‑variability is observed at all stages of synthesis, from transcription to post‑translational modifications. In particular, the “phenotype” of a protein molecule is modified by the environment and, conversely, can be realized through different pathways. The same DNA matrix can be read into different transcripts and, accordingly, into different enzymes depending on the state of the translation apparatus (Inge‑Vechtomov 1976; Mikhailova et al., 1981; Holubovskyi, 1985). These relationships allow us to consider that protein evolution proceeds in principle like that of macromolecules—that selection here begins with the stabilization of their final functional changes (determined by protein conformation). The first step in this process may be a change in the sensitivity of the molecule’s active sites to their substrate, which is then fixed by rearrangements in other parts of the amino‑acid chain and later by changes in preceding synthesis steps (Woodington, 1970; Olenov, 1977; Shishkin, 1987, 1988). Among other reasons influencing methodological choices in evolutionary constructions, one may also name an irrational factor: the specificity of national mentality. Here the role of Anglo‑American ways of thinking in the formation and spread of reductionist principles of the STE is discussed. The rationalism characteristic of this type of thought dictates the search for simple algorithms that link observed facts in the most economical way, primarily using mathematical methods. This stance has its merits but carries the risk that simplification for analytical convenience becomes a priority, ignoring prohibitions known in the field. The STE is a instructive example of this. Another biological example with the same roots is the current fetishization of the cladistic method in phylogenetics, where the most “economical” model of relationships among taxa (with a minimum of parallelisms and reversals in character space) is a priori considered the most accurate reflection of their kinship. All of the above illustrates the previously mentioned general pattern, according to which the character of read information is determined by the perceiver’s mode of perception. Thus, black‑and‑white vision excludes an adequate representation of the color spectrum. In science, the same determining role is played by initial theoretical postulates, regardless of how consciously they are held. Their fetishization—rather than critical comparison with other generalizations—sooner or later leads to logical contradictions even within the system of views they once underpinned. The example already discussed above concerned the understanding of Mendelian factors. A similarly strange phantom of collective consciousness in modern biology is the notion of differences between hereditary and non‑hereditary traits, inherited from Weismann. Historically both concepts refer to phenotypes and denote differences between traits that are stably reproduced in offspring and those that appear only under external influences. Weismann (Weismann, 1892) separated these concepts with an impenetrable wall, calling hereditary changes units of embryonic plasma and non‑hereditary those pertaining to somatic cells. Another distinction, proposed by Johannsen (Johannsen, 1926), assumes that hereditary properties characterize the genotype, or the norm of reaction of the zygote, as a whole. All features realized in development are non‑hereditary (since they are not directly transmitted) but are genotype‑dependent, i.e., they are expressions of the norm of reaction under specific conditions. This construction is logically flawless, yet it violates the original understanding of inheritance as a characteristic of organisms, not genotypes. It allows, on equal footing, to describe all phenotypic properties as hereditary (Timofeyev‑Resovsky, Ivanov, 1966), or as non‑hereditary (Kamshilov, 1935, 1967), or as both simultaneously (de Beer, 1930). Thus, two different understandings of hereditary properties emerged: as phenotypic traits or as properties of embryonic plasma (or its units). However, the view that is objectively embedded in the STE and is widely shared by biologists is based on a strange combination of these incompatible approaches, namely: the distinguishing features of hereditary traits are simultaneously recognized as both stability of expression (though subject to fluctuations for internal reasons) and “genetic determination”—in contrast to labile non‑hereditary traits induced solely by the environment. Such a division is permissible only as a provisional simplification in experiments; as a theoretical stance it is meaningless, being equivalent to believing in purely “paratypic” traits not conditioned by the norm of reaction. SOURCES OF THE MODERN CRISIS OF EVOLUTIONISM A feature of theoretical biology, as we have seen, is a relatively low level of demand for logical consistency of its constructions. This is primarily due to the diversity of the organic world and the difficulty of identifying universal regularities within it. Therefore, biology usually does not set the task of creating a strictly unified theory. The very notion of “synthesis,” so popular in late evolutionism, merely emphasizes that the discussion concerns the amalgamation of factors whose causal links are obscure and even deemed unnecessary. A typical example in this sense is the STE with its mechanical combination of hereditary variability and natural selection. This isolation of biological generalizations, whereby their interrelations remain outside critical analysis, has dramatic consequences. In any other natural science it is hard to imagine that a simple reformulation of a problem would lead to a fundamental change in explanatory approach and that alternative explanations could be simultaneously recognized. Such a situation cannot persist for long. Yet in biology it essentially predetermined the main trajectory of evolutionary thought in the 20th century. We again refer to the circumstance that normal (stable) development and inheritance are the same phenomenon, so embryology and genetics essentially study the same object. Until the early twentieth century they remained essentially a single science (Gilbert et al., 1997, p. 325). However, their conclusions turned out diametrically opposite (cf. Vogt, 1934). In this context we can only briefly repeat what has already been said. Experimental embryology (as well as Goldschmidt’s physiological genetics) demonstrated the systemic character of individual development, showing that the general is determined earlier than the particular (Gertwig, 1895; Driesch, 1908; Svetlov, 1978); thereby excluding the possibility of describing developmental results in terms of a mosaic of independent hereditary units. In contrast, chromosomal genetics arrived precisely at such a description—that is, a preformationist model of development linking phenotypic properties with independent genes (while formal attempts by genetics to free itself from this interpretation are set aside). The untenability of the particulate inheritance idea had long been evident to embryologists. “Regarding the properties we must ascribe to the inheritance apparatus, we can judge… by analyzing the process of their realization. Such analysis lies beyond the possibilities of Mendelian genetics…” (Gurvich, 1944, p. 94). “…It is hard to cite another… case… where criticism of the theory is based not on its insufficiency but on logical untenability and even absurdity of the conclusions derived from its basic premise” (ibid., p. 98). “The chromosomal theory of inheritance as a basis for developmental theory proves completely unsuitable” (Svetlov, 1978, p. 212). “The very concept of a genetic program as a guiding factor of embryonic development is highly problematic and poorly reconciled with genetic and embryological evidence… Specific gene products do not determine either universal or specific constraints that find expression in organic form” (Goodwin, 1982, pp. 45‑46, 49). One can state with confidence that if 20th‑century biology had been better prepared for a non‑contradictory interpretation of its findings, chromosomal genetics would never have had a chance to become the foundation of evolutionary thought. Clearly, the simplicity and visual appeal of the reductionist explanation of inheritance, as well as the simplicity of the proposed method of its analysis (seemingly convincingly demonstrated on pure‑line material), played a decisive role in the monopolization of evolutionary theory by genetics, which acquired the status of a “nearly exact science” (Filippenko, 1924). The development theory standing on alternative, i.e., systemic, positions, still insufficiently formalized, could not compete with genetics, partly because the very notion of systemic processes was only in its infancy. Consequently, the study of individual development, which dominated theoretical biology at the end of the 19th century, fell out of the view of evolutionary doctrine for a long time—until the emergence of the Schmaltz‑Woodington doctrine. From the foregoing it becomes evident that genetic generalizations can be integrated into evolutionary theory only on the basis of their interpretation as systemic properties of individual development. This idea has been expressed, directly or indirectly, by many authors (Kamshilov, 1935; Goldschmidt, 1940; Gurvich, 1944; Svetlov, 1978; Shishkin, 1987, 1988), and such a task is essentially implied by epigenetic theory. The first steps in this direction were often made unconsciously within chromosomal genetics. Thus, since the 1930s it has repeatedly been emphasized that dominance, recessivity, pleiotropy, and the very specificity of a gene’s phenotypic manifestation express not the gene’s properties but features of ontogenetic processes (Goldschmidt, 1933; Kamshilov, 1935, 1967; Gaisinovich, 1967; Mitrofanov, 1977; Holubovskyi, 1982). Alongside the central collision in evolutionary thought (the antagonism between genetics and embryology) other evidence can be cited that it is poorly inclined to make sense of its own historical experience. One example is the neutralist theory of molecular evolution (Kimura, 1985).Based on the ambiguity of the correspondence between a protein’s function and its primary structure, this theory concludes that the process of amino‑acid substitutions in proteins (in their functionally less significant parts) is not controlled by natural selection and is governed by its own laws. However, the discovery of such ambiguity does not constitute anything fundamentally new compared to what embryology and genetics have long established regarding the organism as a whole. Their research has shown the equifinality of developmental outcomes in relation to variations in the characteristics of living organization at many different levels. This is expressed, for example, in the “neutrality” of allelic variations uncovered beneath the wild phenotype (the Chertëvko rule), the lack of order in mitoses within a symmetric embryo (Gurvich’s “normalization” principle), the multiplicity of embryonic states that allow regulation toward the norm when development is disturbed, and so on. From the standpoint of the Theory of Evolutionary Ecology (TEE), all these phenomena are inevitable consequences of selection for the stability of the normal phenotype. This stability means that the elementary components of the developmental process vary within limits that permit the norm to be achieved. The neutralist theory deals with the same systems, but only limited to the process of synthesizing individual functional molecules. Its interpretation of the nonlinear relationships within these systems as evidence of non‑Darwinian evolution is a reliable indicator that Darwinian doctrine is here identified with the Theory of Structural Evolution (STE), whose ideology is built on the principle of unambiguous causality. FROM MULLER AND HAECKEL TO THE “NEW SYNTHESIS”: A UNIFIED METHODOLOGY OF EVOLUTIONARY CONSTRUCTIONS All that has been said about historically formed features of evolutionary thinking defines its one main fatal trait. It lies in the search for linear dependencies where none actually exist, namely: between organismal properties and the laws of historical transformation of living organization. In fact, everything we observe in the development, structure, and functioning of organisms is, by definition, a product of evolution—that is, direct or indirect manifestations of systemic order created and reshaped by natural selection. It is in these concrete manifestations that researchers usually try to see a direct expression of the evolutionary mechanisms themselves. Thus, the products of evolution are presented as its raw material, and its consequences as the original driving causes (Shyshkin, 1987). Strangely, it is rarely recognized that this approach underlies almost all of the most demanded post‑Darwinian evolutionary syntheses that rely on experimental and comparative studies. Here lies the key to evaluating these ideas, including those now linked with hopes for achieving a “new synthesis” (see below). The mixing of causes and effects is equally traditional for both embryology and genetics when their conclusions are applied to understanding evolution. Both disciplines share as their object the properties of normal individual development and the specific space of its aberrations. Accordingly, all their generalizations reflect the structure and behavior of this dynamic system under various internal states and external influences. This holds true whether we speak of matrix synthesis or embryonic induction, DNA organization in chromosomes or types of blastomere cleavage, etc. In other words, everything we observe in development at different levels of analysis is the outcome of the historical formation of a given organizational variant, not a direct reflection of the causes of its appearance. The conflation of these concepts in embryology, beginning with E. Joffroy‑Saint‑Hilaire, F. Müller, and E. Haeckel, has long become almost a rule. Following it, the repeatedly transformed ontogenetic record of historical events is constantly identified with the events themselves, despite the formal acknowledgment of evolution in ontogeny (cf. Shyshkin, 1984, 1987). It suffices to mention various schemes of “embryonic modes of evolution,” according to which, for example, divergence of the ontogenies of two forms at a certain developmental point should indicate a punctuated emergence of one from the other by switching the developmental trajectory at that point. Following the same logic, heterotopies in the way the same structure is laid down in modern organismal groups are often taken as direct evidence of its independent origin in those groups. A classic example is E. Gaup’s notion of a “capsular” nature of the amphibian auditory ossicle in contrast to the gyroid one in amniotes—a view widely shared at the beginning of the 20th century. On the same basis, differences in the number of homologous bone anlagen among modern vertebrates are constantly used by embryologists as self‑evident proof of independent origins of those bones through the fusion of different combinations of original definitive precursors. The same conceptual substitution underlies the genetic approach to explaining evolution. The mechanisms of phenotypic determination uncovered by genetics, expressed in terms of gene activity and their products, actually embody the historical rationalization of ontogeny as a whole. Yet genetics treats them as genuine evolutionary factors. This rests on the belief that organismal traits are determined by independent genes and, consequently, that evolution is the result of gene mutations. These views, in turn, are founded on the recognition of the material (chromosomal) nature of Mendelian factors and the universality of Mendelian inheritance laws. The inadequacy of both of these latter positions has already been discussed (cf. Shyshkin, 1987). Mendelian inheritance is an epiphenomenon of alternative variants of norm stabilization, not a property of the hereditary substance. A Mendelian factor is merely a symbol of the magnitude of quantitative shift in the system’s state required for choosing among possible stabilized developmental pathways. How this switching occurs depends on the pathways that stabilize parental phenotypes, but it is not a specific function of any single locus. The same systemic nature characterizes all epigenetic effects caused by chromosomal mutations, differing only in that the choice occurs within the space of aberrations permitted by the system (Goldschmidt, 1938, 1940). We are again dealing with products of evolution, not its “raw” material. Note that the notion of evolution by mutation implies two assumptions: (1) each mutation has a specific effect on the phenotype, even if mediated by gene interactions; (2) mutations generate new stable changes. Both are wholly based on equating particular ontogenetic events with general developmental and evolutionary regularities. The first of these propositions rests on demonstrating a direct link between mutations and anomalies in laboratory lines—a effect sustained by selection on those cultures. By default, genetics extrapolates this to the entire space of relationships between chromosomal genes and traits (although formally this is often denied). The fallacy of such a conclusion is evident from the fact that the effect of any mutation can be phenocopied, i.e., induced externally without genetic change, yet it displays the same morphogenetic nature. As Goldschmidt rightly concluded (Goldschmidt, 1938, 1940), this shows that all realized developmental variants characterize the properties of the reactive system, not the specificity of elementary disturbances. In other words, the diversity of mutation effects is limited by system properties. This view is further supported by the fact that the same anomalies arise from very different genetic disturbances—both in terms of chromosomal location and qualitative nature, including point mutations, deletions, duplications, etc., for example the Bar, vestigial, or yellow phenotypes in Drosophila (Goldschmidt, 1938, 1955; Tymofeyev‑Resovsky & Ivanov, 1966; Olenov, 1976). Thus, no mutation allows one to go beyond the range of deviations that is achievable even without it within a given developmental type. Moreover, none of them possesses a constant phenotype that could serve as a target of selection. Accordingly, selection on any deviating phenotype does not preserve “identical mutants” but yields a sample of individuals with the most diverse individual genetic constitutions. This is demonstrated by numerous experiments analyzing natural anomalies in populations (Balkashina & Romashov, 1935; Dubinin et al., 1937; Hlubovsky et al., 1974; Olenov, 1976). Summarizing, we can say that “evolutionary novelties” potentially linked to mutation action are in fact neither novelties nor properties of mutations; they merely describe a historically formed type of ontogeny. The belief that evolutionary laws can be directly extrapolated from observed ontogenetic properties (understood mainly as mechanisms of gene expression) also underlies modern efforts to create a “new synthesis” that would unite genetics and embryology. The impossibility of describing organismal morphology at the level of genes responsible for protein synthesis led to the hypothesis that special regulatory genes play the primary role in implementing structural plans (King & Wilson, 1975; Lewis, 1985; Reff & Kofmen, 1986; Gilbert et al., 1997). These genes are linked to the ability to switch development onto alternative pathways and to create morphogenetic fields that determine the expression of “subordinate” genes. Development is described as a hierarchy of regulatory‑gene action, and morphological evolution as the result of changes in these genes. Such a modernization of earlier views remains entirely within the traditional pre‑formist framework of genetics. Developmental properties are reduced to gene properties; only now, instead of direct trait determination, it is assumed that higher‑order regulatory genes control the choice of developmental pathways, ensuring stepwise determination via other genes. Yet the understanding of evolutionary mechanisms still relies on a direct historical interpretation of ontogenetic events. Specifically, the distinction between normal and mutant ontogenetic variants is taken as a reliable model of an evolutionary event. A typical example of this approach is the discussion of the evolutionary role of homeotic mutations in insects, i.e., changes in regulatory genes that cause the development of a structure’s serial homolog in place of the normal one. A textbook case is the role of the Bithorax (BX‑C) and Antennapedia (ANT‑C) homeotic gene complexes in the transition from ancestors resembling myriapods to winged insects (Lewis 1963, 1985; Reff & Kofmen, 1986). In particular, it is thought that the successive acquisition of BX‑C genes led to the suppression of limb development in abdominal segments of early winged insects and later (in Diptera) also to the suppression of wing development on the posterior thorax. This reconstruction is based on analyzing differences between mutant developmental types and the normal variant revealed by successive disruptions within the BX‑C complex. These differences manifest as repression of alternative (atavistic) developmental pathways of segments, for example, the transformation of abdominal segments into a median‑thoracic‑like form and the restoration of wings on the posterior thorax. In sum, the correlations obtained between the type of genetic disturbance and phenotypic change are interpreted as a mirror “counter‑imprint” of an evolutionary event. In other words, the event is conceived as the stepwise acquisition of normal alleles of this homeotic complex, leading to successive alteration of the original segment morphology. Here one is forced to recall the still unheard warning of Goldschmidt (Goldschmidt, 1938, p. 310; 1955, p. 98) that the very concept of a “normal” locus gene governing a normal process is nothing more than a dubious extrapolation from the action of its mutant allele that disrupts that process. Surprisingly, these evolutionary reconstructions do not take into account that the effects of homeotic mutations are phenocopied just like any other (e.g., in dipterans—aristopedia, bicaudal, bithorax phenotypes, etc.; see Goldschmidt, 1938, 1955; Waddington, 1957; Reff & Kofmen, 1986), and therefore they have a systemic (not specifically mutant) nature. The nature of the developmental parameter disturbance underlying such anomalies may be well known; for instance, the aristopedia phenotype reduces to altered segmentation rates of the embryo regardless of the damaging agent (Goldschmidt, 1938). The fundamental significance of these facts, which once led Goldschmidt to the concept of a holistic reactive system with an invariant space of possibilities, remains unappreciated by the architects of the “new synthesis” and is reduced by them to the action of similar morphogenetic products in cases of mutation and phenocopy (Reff & Kofmen, 1986, p. 229). Meanwhile, the key question is how this system property, attributed to a specific gene, arose and whether the mechanism identified by the authors as evolutionary was always as we observe it today. Conscious or unconscious belief that causal links in evolution and ontogeny are identical is common in genetic thinking and many other similar manifestations. One typical example is the identification of chromosomal rearrangements with evolutionary mechanisms. Central to this is T. H. Ono’s idea of evolutionary novelties arising through duplications (repeats) and subsequent divergence of identical genes, which, according to the hypothesis’s author, should lead to the creation of families (clusters) of homologous nucleotide sequences with differentiated functions. This scenario, now widely accepted, is postulated for both regulatory homeobox‑containing genes (mutations of which cause homeotic transformations) and for groups of structural genes responsible for synthesizing homologous proteins or subunits, such as globin, cytochrome families, etc. Yet this hypothesis rests on no other grounds than (1) the reality of duplication as one type of chromosomal disturbance in germ cells and (2) the presence in eukaryotic genomes of clusters of similar functional units, from which it is inferred that the former is a symmetric cause of the latter. This is an approximate analogue of concluding that white coloration in polar animals is a consequence of albinism mutations. The arbitrariness of such a conclusion about the role of duplications has already been noted (Soidla, 1983; Shyshkin, 1987; Rogal, 1997). From the epigenetic theory’s perspective, the appearance of a stable homolog of a former locus is not the result of its duplication nor the beginning of a new matrix function, but rather the final stage of its establishment, associated with the stabilization of the process that produces a new final product (Shyshkin, 1987). A number of geneticists share this view. As Olenov (1977) expressed it, nature does not wait for a gene to duplicate; the necessary product first appears by alternative routes, for example by incorporating an existing analogue into another regulatory group. In other words, a new adaptive function precedes the emergence of the corresponding matrix (cf. Soidla, 1983, p. 335). Consequently, the start of the process should be considered selection in favor of a new useful modification of an already existing molecular type. Stabilization of this post‑translational change inevitably affects earlier synthesis stages, i.e., the choice between the old and new adaptive variants first shifts to the transcriptional level and then leads to the emergence of separate coding DNA matrices for each variant. From this standpoint, the possibility of various ways of editing the same RNA transcript (Hlubovsky, 1985; Korochkin, 2002a) appears as an intermediate stage of such evolutionary transformation. The whole process of forming a locus‑homolog seems analogous to the course of stabilizing changes in macro‑ontogeny. Thus, neither chromosome structure features, nor matrix‑synthesis mechanisms, nor the connection of individual loci with potential developmental switches contain a direct method for reading the history of their own formation. The same applies to all other manifestations of developmental order. Here we almost verbatim repeat Kryzhansky’s (1939) longstanding conclusion about the impossibility of a direct historical interpretation of ontogenetic events. As already noted, genetic thinking in its evolutionary conclusions follows a logic directly opposite to that inherited from 19th‑century evolutionary methodology. This logic remains unchanged even in attempts to create a “new synthesis,” in which changes in regulatory genes are viewed as the main driver of evolution (Reff & Kofmen, 1986; Gilbert et al., 1997). The essence of this logic is clearly expressed in the following reasoning: “Regulatory elements of development are genes whose mutant alleles cause … switching … from one … pathway to another… Thus, evolutionary change occurs by modification of the genetically determined developmental program” (Reff & Kofmen, 1986, pp. 352‑353). In other words, the emergence of holistic regulatory properties of development is attributed here to those specific genes whose mutation reveals them experimentally. ON THE PATHS OF SEARCHING FOR A “NEW SYNTHESIS” Some ideas employed in attempts to construct a “new synthesis” require further consideration. Its initiators recognize that the embryological concept of the morphogenetic field as an integral developmental factor has historically been an antagonist to the genetic program concept, according to which development results from a mosaic of gene interactions (Gilbert et al., 1997). Yet they see this conflict as only a temporary and surmountable stage in understanding the driving forces of development and evolution.{ "translated_text": "Thus the discussion concerns the attempt to reconcile the incompatible, namely: the epigenetic (systemic) and preformationist (reductionist) models of development as the basis of a new evolutionary theory. In practice these attempts result in the preservation of the latter model, only in a more complex form. The fact that development is a cascade of events of increasing complexity is reduced here to a hierarchy of gene expression acts. Morphogenetic fields, taken in this concept for different stages of development, are thought of only as products of regulatory gene activity and serve as linking elements between stages of gene action, i.e., those higher‑order genes that create a given field and those that are subsequently activated by it (Gilbert et al., 1997). This is merely another version of the traditional belief that ontogeny is the process of realizing properties predetermined by genes.\nThe refusal to recognize a factor of wholeness that governs development inevitably turns any variant of such a \"synthesis\" into a set of isolated evolutionary mechanisms, only declaratively linked. As already noted, the underlying notion is that morphological evolution is associated with a different part of the genome than evolution at the protein level (King, Wilson, 1975). Accordingly, genes are divided into structural (providing material for development) and regulatory (controlling the construction plan). Among the latter, Raff and Coffman (1986, pp. 358, 363) distinguish pathway‑switch genes and those that control tissue differentiation. Finally, among the former these authors identify genes that generate positional information during oogenesis and those that interpret it in later development. This factor‑splitting route drives theoretical thought in a direction directly opposite to the intended goal. (In this light, the examples of moving away from such views toward systemic explanations of evolution not reducible to mutational changes do not seem surprising; see Kauffman, 1993, 1995.)\nOn the other hand, if the principle of a cascading increase in the number of genes expressed during development is taken to its logical conclusion, we inevitably arrive at the conclusion that the apex (origin) of this hierarchical pyramid must correspond to a single regulatory gene that initiates development. Then the question arises: are we not thereby returning to the idea of a holistic embryonic field?\nCurrent searches for a new synthesis, which began in the early 1980s, have been from the start influenced by two mutually exclusive ideologies. One, as mentioned above, is linked to the recognition of systemic regularities of individual development. The other, stemming from molecular biology, is linked to the discovery that regulatory genes responsible for mutational homeotic transformations are homologous to each other and contain regions with stable base sequences (homeoboxes). Establishing correspondences of these gene groups across different organisms and, in some cases, functional similarities—e.g., roles in establishing the anterior‑posterior body axis in insects and vertebrates (McGinnis, Kramlauf, 1992)—led to a new explosion of preformationist ideas about development and evolution, without parallel since the rediscovery of Mendel's laws.\nCentral to these convictions is the hypothesis that all morphological diversity, at least in the animal kingdom, arose on the basis of expression of shared homeotic genes. Accordingly, the emergence of functionally similar organs in different groups is explained by homology of the corresponding processes, i.e., the activity of similar genes and their products that provide homologous signal‑transduction pathways in different morphogenetic fields. As an example, the development of limbs in insects and vertebrates is cited. According to Gilbert et al. (1997, p. 336), \"it seems that nature decided once and for all how to create appendages.\"\nThus this hypothesis of universal pathways for morphological novelties implies a sequence: homologous genes → homologous molecular processes operating in different fields → heterogeneous morphological products. As a result, the proposed statements about evolutionary relationships between structural types acquire a completely mystical character, because they require no (and usually impossible) verification based on morphological criteria. Therefore, examples intended to demonstrate structural commonality on the basis of \"homology\" of the involved processes actually prove nothing beyond the probable participation of those processes themselves—except perhaps in cases where some similarity of the final result is also observed. The latter case can theoretically include the formation of the anterior‑posterior body axis in different animal types. It is considerably harder to say the same about the suggestion that the involvement of the gene *Rax‑6* in eye development of mammals, cephalopods, and insects indicates a common photoreceptor cell origin (Quiring et al., 1994). Even more tenuous are parallels based on the assumption that the same protein system controls limb‑axis formation in insects and vertebrates (Gilbert et al., 1997), especially in light of the fact that it also functions in the eye‑antenna disc of insects. Finally, it is impossible to discern what conclusions about structural kinship can be drawn from the fact that epidermal determination in mammals, terminal segments in *Drosophila*, and genitalia in nematodes are linked to the activity of the same tyrosine‑kinase receptor.\nAll these examples rather demonstrate the opposite of what is desired, namely that pathways of morphological evolution (as well as evolution in general) are not determined by a specific set of genes. It is striking that the discussed notions are presented by their authors as evidence of a new way of thinking that acknowledges a leading evolutionary role of morphogenetic fields and is advanced \"as an alternative to the monopoly of the genetic model of evolution and development\" (Gilbert et al., 1997, p. 338).\nCONCLUSION\nTo summarize, evolution of organisms is a systemic transformation of their development that cannot be described on a reductionist basis, i.e., expressed as a sum of effects of independent initial factors. Such an explanation of biological phenomena, traditional for the natural sciences of the past, was conditioned by the acceptance of universal linear determinism in the physical world. The fact that it remains dominant in genetics and in the genetic evolutionary theory (which sees organisms as the outcome of a mosaic of hereditary units) has its reasons. One is the researchers' drive to find simple and easily formalizable regularities in the properties of studied objects. However, the main issue lies in the methodology of inquiry itself. General system properties do not have a direct correlation with particular dependencies within it; the latter are determined only statistically with respect to system parameters. Meanwhile, any direct observation of biological objects reveals only those concrete dependencies, which can be described in terms of linear causality (e.g., the relationship between a locus mutation and its phenotypic effect under given developmental conditions and genotype specifics). Mechanically equating these particular relationships with general laws of living organization is the principal source of errors characteristic of theoretical biology in general and genetics in particular.\nThe relative and unstable nature of such dependencies within a whole can sometimes be uncovered (by comparison) quite easily. This happened at the end of the 19th century with experimental embryology, which, setting out to find mechanical causes of development, soon confronted the fact of holistic control of the process (the entelechy of G. Driesch, which gave rise to the concept of the biological field). For genetics this path was longer, culminating in R. Goldschmidt's fundamental generalization of development as a reactive system governed by a single embryonic plasma that determines the entire space of potentially realizable final outcomes. These ideas became the cornerstone of the epigenetic theory of evolution of Schmalhausen‑Woodington. Yet, because of the traditional eclecticism of biological thought and the rarity of critical comparisons of the concepts created, systemic syntheses of embryology and genetics remain little used. They exist in the shadow of prevailing reductionist views that equate particular manifestations of biological order with mechanisms that create that order, as discussed above. At the same time, the rapid growth of experimental knowledge through molecular disciplines objectively stimulates the continual emergence of such mechanistic constructions, essentially repeating in a new language what has long been proposed. In this situation, when the original postulates of genetics clash with its own systemic generalizations, the arbitrariness of fetishizing the former becomes increasingly evident.\nRecall some of the internal contradictions discussed earlier:\n- Inheritance in genetic concepts does not require evolutionary explanation, contrary to the declared need for such an explanation for every biological phenomenon (cf. Dobzhansky, 1951).\n- Dividing traits into hereditary and non‑hereditary on the basis of genetic determination is incompatible with the concept of a reaction norm (cf. Johannsen, 1926).\n- The principle of a one‑to‑one correspondence between gene and trait (a prerequisite for Mendelian analysis) contradicts the principle that a trait is defined by all genes, derived from K. Bridges' balance hypothesis and R. Goldschmidt's views on the developmental system.\n- Acceptance of the universality of Mendelian inheritance contradicts the conclusion that it depends on selection (Fischer, 1930).\n- The idea of evolution by mutation contradicts the conclusion that no mutation takes an individual beyond its species affiliation (Dubinin, 1966; Mayr, 1968).\n- The notion of a genetic program does not agree with the conclusion of cytoplasmic control of gene expression (Gordon, 1977).\nToday's searches for a \"new evolutionary synthesis\" undertaken by genetics without revising its own foundations merely expand the domain of reductionist interpretations applied to systemic phenomena. The new object of such interpretation has become the concept of regulation. In fact, regulation denotes a system's property of maintaining its equilibrium through interactions of its parts via feedback loops. Such system behavior (called purposive or teleological) by definition cannot be ascribed to material particles within it. Yet this reduction is the main principle now linked to the renewal of evolutionary theory. As we have seen, it is embodied in the notion of regulatory genes directing morphological evolution, with the expectation that this will allow description at the genetic level (King, Wilson 1975; Raff, Coffman, 1986; Gilbert et al., 1997). In this sense, the concept of regulation in genetics merely repeats the fate of the concept of inheritance, which actually denotes systemic stability of organization rather than a property of discrete material carriers. The extreme expression of this reductionism is the interpretation of biological teleology as a property determined by DNA (Koro​chkin, 2002, p. 13).\nIn our final assessment of evolutionism's prospects, we start from the view that despite its mosaicism and disorder, collective biological thinking nonetheless constitutes a single system with internal feedback loops. Consequently, it is capable—albeit slowly—of historically inevitable self‑correction based on critical comparison of its own theoretical generalizations. Ultimately this must inevitably lead to the recognition of the incompatibility of systemic properties of development (transformed by natural selection) with the idea of particulate inheritance. This process has only just begun, and its completion lies far ahead. Such a reassessment will not entail a literal revision of genetics' factual achievements. It is also likely that the established language of the discipline will remain the most adequate for describing the phenomenology of its specific studies. However, the future evolutionary theory, built on epigenetic principles and viewing the change of organic forms as a systemic transformation of the mechanism that implements them, will necessarily require interpreting all major genetic generalizations in terms of individual development.\nREFERENCES\nBalashkina E.I., Romashov D.D. Genetic structure of *Drosophila* populations // *Biol. J.* 1935. Vol. 4. No. 1. pp. 81‑106.\nBeloousov L.V. Holistic and structural‑dynamic approaches to ontogeny // *J. Gen. Biol.* 1979. Vol. 40. No.4. pp. 514‑529.\nVolkenstein M.V. The essence of biological evolution // *Successes of Physical Sciences* 1984. Vol. 143. No. 3. pp. 429‑466.\nVorontsov H.H. Development of evolutionary ideas in biology. Moscow: KMK, 2004. 431 p.\nGaisinovich A.E. The birth of genetics. Moscow: Nauka, 1967. 196 p.\nGordon J. Regulation of gene function in animal development. Moscow: Mir, 1977. 196 p.\nGertwig O. Contemporary controversial issues in biology. Vol. 1. Evolution or epigenesis. Moscow, 1895. 125 p.\nGilbert S.F., Opitz D.M., Raff P.A. The new synthesis of evolutionary biology and developmental biology // *Ontogeny* 1997. Vol. 28. No. 5. pp. 325‑343.\nGolubovsky M.D. Some aspects of the interaction of genetics and evolutionary theory // *Methodological and Philosophical Problems of Biology*. Novosibirsk: Nauka, 1981. pp. 69‑92.\nGolubovsky M.D. Critical studies in genetics // *Alexander Alexandrovich Lyubyshev*. L.: Nauka, 1982. pp. 52‑65.\nGolubovsky M.D. Organization of the genotype and forms of hereditary variability in eukaryotes // *Methodological Problems of Medicine and Biology*. Novosibirsk: Nauka, 1985. pp. 135‑152.\nGolubovsky M.D., Ivanov Yu.P., Zakharov I.K., Berg R.L. Study of synchronous and parallel changes in gene pools of natural populations of fruit flies *Drosophila melanogaster* II *Genetics*. 1974. Vol. 10. No.4. pp. 73‑81.\nGoldschmidt R. Genetics and the physiology of development // *Nature* 1933. No.5‑6. pp. 124‑133.\nGrant V. Evolution of organisms. Moscow: Mir, 1980. 407 p.\nGurvich A.G. Lectures on general histology. Moscow; Petrograd: Gosizdat, 1923. 175 p.\nGurvich A.G. Theory of the biological field. Moscow: Soviet Science, 1944. 156 p.\nDondua A.K. Reprogramming of developmental control in early ontogeny of multicellular animals // *J. Gen. Biol.* 1979. Vol. 40. No. 4. pp. 530‑543.\nDubinin N.P. Evolution of populations and radiation. Moscow: Atomizdat, 1966a. 743 p.\nDubinin N.P. Foundations of population genetics // *Current Issues of Modern Genetics*. Moscow: MSU Press, 1966. pp. 221‑265.\nDubinin N.P., Romashov D.D., Heptner M.A., Demidova Z.A. Aberrative polymorphism in *Drosophila fasciata* Meig. (syn. — *melanogaster* Meig.) // *Biol. J.* 1937. Vol. 6. No.2. pp. 311‑354.\nInge‑Vechtemov S.G. Genotype system // *Physiological Genetics*. L.: Medicine, 1976. pp. 57‑113.\nJohannsen J. Elements of precise study of inheritance and variability. Moscow; L.: Selkhozgiz, 1933. 410 p.\nKamshilov M.M. Is pleiotropy a property of a gene? // *Biol. J.* 1935. Vol. 4. No. 1. pp. 113‑144.\nKamshilov M.M. Role of phenotype in evolution // *Genetics*. 1967. No. 12. pp. 108‑116.\nKimura M. Molecular evolution: the neutral theory. Moscow: Mir, 1985. 398 p.\nKoro​chkin L.I. Biology of individual development. Moscow: MSU Press, 2002a. 263 p.\nKoro​chkin L.I. Ontogeny, evolution and genes // *Nature* 2002. No. 7. pp. 10‑19.\nKryzhansky S.G. The principle of recapitulation and conditions for historical understanding of development // *In memory of Acad. A.N. Severtsov*. Moscow; L.: USSR Academy Press, 1939. pp. 281‑366.\nKuhn T. The structure of scientific revolutions. Moscow: Progress, 1975. 288 p.\nLyubyshev A.A. On the nature of hereditary factors // *Proceedings of the Biological Institute of Perm State University*. 1925. Vol. 4. Append. 1. pp. 1‑142.\nMayr E. The zoological species and evolution. Moscow: Mir, 1968. 597 p.\nMayr E. Populations, species and evolution. Moscow: Mir, 1974. 460 p.\nMayr E. Evolution // *Evolution*. Moscow: Mir, 1981. pp. 11‑32.\nMitrofanov V.G. Physiological bases and evolution of dominance // *Problems of Experimental Biology* / Ed. Belyaev D.K. Moscow: Nauka, 1977. pp. 21‑31.\nMikhailova H.H., Simarov B.V., Tikhomirova V.L., Inge‑Vechtemov S.G. Effect of increased magnesium and sodium ions on translational and post‑translational levels in yeast *Saccharomyces cerevisiae* // *Genetics Research*. L.: LGU Press, 1981. pp. 5‑76.\nMorgan T. Structural foundations of inheritance. Moscow; L.: Gosizdat, 1924. 310 p.\nMorgan T. Gene stability // *Selected Works on Genetics*. Moscow; L.: Selkhozgiz, 1937. pp. 243‑255.\nMorgan T. Experimental bases of evolution. Moscow: Biomedgiz, 1938. 250 p.\nMüller F. For Darwin // F. Müller, E. Hekkel. The fundamental biogenetic law / Ed. Schmalhausen I.I. Moscow: USSR Academy Press, 1940. pp. 61‑166.\nOlenov Yu.M. Cellular inheritance, cell differentiation and carcinogenesis. L.: Nauka, 1967. 309 p.\nOlenov Yu.M. Molecular biology and the theory of natural selection // *J. Gen. Biol.* 1976. Vol. 37. No. 1. pp. 7‑17.\nOlenov Yu.M. Problems of molecular genetics. L.: Nauka, 1977. 207 p.\nRogale I.G. Concept of macromolecule evolution as the basis of organismal evolution // *Historical‑Biological Research*. Moscow: Nauka, 1997. Vol. 11. pp. 67‑84.\nRues M. Philosophy of biology. Moscow: Progress, 1977. 320 p.\nRaff R., Coffman T. Embryos, genes and evolution. Moscow: Mir, 1986. 402 p.\nSvetlov P.G. On holistic and elementalist methods in embryology // *Archive of Anatomy, Histology and Embryology*. 1964. Vol. 46. No. 4. pp. 3‑26.\nSvetlov P.G. Ontogeny as a teleological process // *Same journal*. 1972. Vol. 63. No. 8. pp. 5‑16.\nSvetlov P.G. Physiology (mechanics) of development. L.: Nauka, 1978. Vol. 1. 279 p; Vol. 2. 262 p.\nSoidla T.R. Features of eukaryotic genes // Inge‑Vechtemov S.T. Introduction to molecular genetics. Moscow: Higher School, 1983. pp. 328‑335.\nSpencer G. Foundations of biology. St. Petersburg: Ioganson Publishing, 1899. Vol. 1. 456 p; Vol. 2. 380 p.\nTimofeev‑Resovsky N.V., Ivanov V.I. Some questions of phenogenetics // *Current Issues of Modern Genetics*. Moscow: MSU Press, 1966. pp. 114‑130.\nTokin B.P. General embryology. Moscow: Higher School, 1977. 509 p.\nWoodington K.H." }Fundamental biological concepts. On the way to theoretical biology. Moscow: Mir, 1970. Vol. 1. Prolegomena. pp. 11–38. Filippenko Yu.A. Heredity. Moscow; Petrograd: Gosizdat, 1924. 252 p. Chetyreikov S.S. On certain aspects of the evolutionary process from the perspective of modern genetics // J. Exp. Biol. 1926. Series A. Vol. 2. No. 1. pp. 3‑54. Shishkin M.A. Individual development and natural selection // Ontogeny. 1984. Vol. 15. No. 2. pp. 115‑136. Shishkin M.A. Ontogeny and evolutionary theory // Evolution and biocenotic crises / Ed. by Tatarinov L.P., Rasnitsyn A.P. Moscow: Nauka, 1987. pp. 76‑123. Shishkin M.A. Evolution as an epigenetic process // Modern Paleontology / Ed. by Menner V.V., Makridin V.P. Moscow: Nedra, 1988. Vol. 2. pp. 142‑169. Shishkin M.A. Two alternative approaches to understanding the evolutionary process // Abstracts of the XI International Conference on Plant Phylogeny. Moscow: Center for Wildlife Protection Publishing, 2003. pp. 112‑114. Shmal’gauzen I.I. Paths and regularities of the evolutionary process. Moscow; Leningrad: USSR Academy of Sciences Publishing, 1940. 231 p. Shmal’gauzen I.I. Stabilizing selection and its place among evolutionary factors // J. Gen. Biol. 1941. Vol. 2. No. 3. pp. 307‑354. Shmal’gauzen I.I. Evolutionary factors: Theory of stabilizing selection. Moscow: Nauka, 1968a. 451 p. Shmal’gauzen I.I. Cybernetic questions of biology. Novosibirsk: Nauka, 19686. 223 p. Shmal’gauzen I.I. The organism as a whole in individual and historical development. Moscow: Nauka, 1982. 228 p. Alberch P. Developmental constraints in evolutionary processes // Evolution and Development / Ed. Bonner J.T. Berlin: Springer, 1982. pp. 313‑332. Baer K. Über Entwicklungsgeschichte der Thiere. Königsberg: Kupfertaf, 1828. Vol. I. 271 p. Balfour F.M. Address to the Department of Anatomy and Physiology of the British Association for the Advancement of Science // The Works of F.M. Balfour / Eds. Foster M., Sedgwick A.L.: Macmillan, 1885. Vol. 1. pp. 698‑713. Baur E. Einführung in die experimentelle Vererbungslehre. Berlin: Bornträger, 1919. 410 p. Beer G. de. Embryology and evolution. Oxford: Clarendon Press, 1930. 116 p. Bertalanffy L. General system theory. New York: Braziller, 1969. 289 p. Conklin E.C. The organization of cell lineage of the ascidian egg // J. Acad. Nat. Sci. 1905. Vol. 13. pp. 1‑119. Cope E.D. The origin of the fittest. New York: Macmillan, 1887. 467 p. Dobzhansky T. Genetics and the origin of species. New York: Columbia University Press, 1951. 364 p. Driesch H. The Science and philosophy of the organism. L.: Black, 1908. Vol. 1. 329 p.; Vol. 2. 381 p. Fisher R. Genetical theory of natural selection. Oxford: Clarendon Press, 1930. 272 p. Goldschmidt R. Physiological genetics. New York: L. McGraw Hill Book Co., 1938. 375 p. Goldschmidt R. The material basis of evolution. New Haven: Yale University Press, 1940. 436 p. Goldschmidt R. Theoretical genetics. Berkeley: University of California Press, 1955. 563 p. Goodwin B.C. Development and evolution // J. Theor. Biol. 1982. Vol. 97. pp. 43‑55. Gould S.J. Change in developmental timing as a mechanism of macroevolution // Evolution and Development / Ed. Bonner J.T. Berlin: Springer, 1982a. pp. 337‑346. Gould S.J. The meaning of punctuated equilibrium and its role in validating a hierarchical approach to macroevolution // Perspectives in Evolution / Ed. Milkman R. Sunderland: Sinauer, 1982b. pp. 83‑104. Haeckel E. Generelle Morphologie der Organismen. Berlin: Georg Reimer, 1866. Vol. 1. 574 p.; Vol. 2. 462 p. Jägersten G. Evolution of the metazoan life cycle. L.: Academic Press, 1972. 282 p. Johannsen W. Elemente der exakten Erblichkeitslehre. Jena: Fischer, 1926. 736 p. Kauffman S.A. The origin of order: self‑organization in evolution. New York: Oxford University Press, 1993. 709 p. Kauffman S.A. At home in the universe. The search for laws of self‑organization and complexity. New York: Oxford University Press, 1995. 321 p. King M.-C., Wilson A.C. Evolution at two levels in humans and chimpanzees // Science. 1975. Vol. 188. pp. 107‑116. Lankester E.R. Notes on the embryology and classification of the animal kingdom: comprising a revision of speculations relative to the origin and significance of germ layers // Quart. J. Microsc. Sci. 1877. Vol. 17. pp. 399‑454. Lewis E.B. Genes and developmental pathways // Amer. Zool. 1963. Vol. 3. pp. 33‑56. Lewis E.B. Regulation of the genes of the bithorax complex in Drosophila // Cold Spring Harbor Symp. Quant. Biol. 1985. Vol. 50. pp. 155‑164. Maderson P.F.A., Alberch P., Goodwin B.C. et al. The role of development in macroevolutionary change // Evolution and Development / Ed. Bonner J.T. Berlin: Springer, 1982. pp. 279‑312. McGinnis W., Krumlauf R. Homeobox genes and axial patterning // Cell. 1992. Vol. 68. pp. 283‑302. Quiring R., Walldorf U., Kloter U., Gehring W.J. Homology of the eyeless gene of Drosophila to the Small eye gene in mice and Aniridia in humans // Science. 1994. Vol. 265. pp. 785‑789. Roux W. Gesammelte Abhandlungen über Entwicklungsmechanik der Organismen. Leipzig: Engelmann, 1896. Vol. 1. 816 p.; Vol. 2. 1075 p. Salthe S.N. Development and evolution. Cambridge (Mass.); L.: MIT Press, 1993. 357 p. Shishkin M.A. The development of the evolutionary theory. Some lessons // Rivista di Biologia. 1989. Vol. 82. pp. 323‑325. Shishkin M.A. Evolution as a maintenance of ontogenetic stability // Acta Zool. Fennica. 1992. Vol. 191. pp. 37². Thomson K.S. Morphogenesis and evolution. New York: Oxford University Press, 1988. 149 p. Vogt W. Entwicklungsmechanik und Gewebezüchtung // Arch. Exp. Zellforsch. 1934. Vol. 15. No. 2‑4. pp. 269‑280. Vrba E.S., Eldredge N. Individuals, hierarchies and processes: Towards a more complete evolutionary theory // Paleobiology. 1984. Vol. 10. pp. 146‑171. Waddington C.H. The Strategy of the genes: a Discussion on some aspects of theoretical biology. L.: Allen and Unwin, 1957. 262 p. Waddington C.N. Principles of development and differentiation. New York: Macmillan, 1966. 115 p. Wake D.B., Roth G., Wake M.M. On the problem of stasis in organismal evolution // J. Theor. Biol. 1983. Vol. 101. pp. 211‑224. Weismann A. Das Keimplasma. Eine Theorie der Vererbung. Jena: G. Fischer, 1892. 628 p. Development and Lessons of Evolutionism. M. A. Shishkin Paleontological Institute, Russian Academy of Sciences, ul. Profsoyuznaya 123, Moscow, 117997 Russia E‑mail: shishkin@paleo.ru Received August 30, 2005; in final form, December 14, 2005 **Abstract**—The present crisis of evolutionism was predictable initially, since the pre‑formational model of development expressed in the idea of discrete heredity contradicts the systemic properties of ontogenesis. Correspondingly, the principle of selection of inherited factors cannot explain evolution. The synthetic theory based on this principle contains insoluble contradictions in its key notions. According to the alternative epigenetic theory based on the integrity of living organization, heredity is a product of selection and expresses a teleonomic direction of development toward a stable final state. Unification of the genetic concept of evolution with recognition of the integrity of development is principally impossible. The cause of the dominance of genetic views on evolution lies in the correspondence of the mechanistic tradition of the 18‑19th centuries, rather than in their logical substantiation. For the same reason, biology as a whole is characterized by the identification of specific linear dependencies with the laws of evolution. Following this path in search of a “new evolutionary synthesis” invites a priori its failure. Evolutionary interpretation of genetic generalizations is only possible on the basis of their description in terms of development. **Key words**: system of development, heredity, evolution, stability, regulation, reductionism.