Can Darwinism be “Generalized” and of what use would this be?

Abstract

It has been suggested that, by generalizing Darwinian principles, a common foundation can be derived for all scientific disciplines dealing with evolutionary processes, especially for evolutionary economics. We show, however, that in the development of evolutionary biology, the abstract principles of so-called “Generalized Darwinism” have not been crucial for distinguishing Darwinian from non-Darwinian approaches and, hence, cannot be considered genuinely Darwinian. Moreover, we wonder what can be gained by invoking the abstract principles of Generalized Darwinism given that they do not suffice to substantiate an explanation of actual evolutionary processes. To that end, specific hypotheses are required. They neither follow from the suggested abstract principles, nor are they more easily found on that basis. Accordingly, we find little evidence in the literature for the claim that generalized Darwinian principles enhance the explanatory power of an evolutionary approach to economics.

1 Introduction

In reflecting on evolutionary phenomena in their respective domains, many disciplines borrow conceptual tools from Darwinian evolutionary biology. Evolutionary economics is no exception. It is popular, for example, to construct analogies to natural selection, particularly in the neo-Schumpeterian literature (see, e.g., Nelson and Winter 1982; Metcalfe 1998; Windrum 2007; Andersen 2009, chap 12). More recently, it has been argued, however, that it is not only possible but also desirable to go beyond metaphors and analogies. It is claimed that general principles of evolution can be derived through isolating abstraction from the Darwinian theory that explain alike evolution in the biological domain and non-biological domains, e.g. that of economics (Hodgson and Knudsen 2006a, 2010). The claim that “Darwinism involves a general theory of all open, complex systems” (Hodgson 2002) ties in with an ongoing debate in the philosophy of science originally triggered by Dawkin’s (1983) vision of “Universal Darwinism”. In that debate, variation, selection, and inheritance (for the latter also the terms “replication” or “retention” are sometimes used) have been postulated by some writers as domain-independent general principles (see Campbell 1965; Dawkins 1983; Hallpike 1985; Hallpike 1986; Wilkins 1998; Wimsatt 1999; Hull 2001; Crozier 2008). These principles also form the core of “Generalized Darwinism”, an approach propagated in a series of works by Hodgson and Knudsen (2004, 2006a, b, 2010) with special reference to evolutionary economics.

With their recent manifesto “In defense of Generalized Darwinism” in this journal (Aldrich et al. 2008), several prominent scholars have endorsed such an encompassing approach to evolutionary economics. Its tacit presumption is a fundamental homology between evolution in nature and evolution in the economy—“social evolution is Darwinian”, as Hodgson and Knudsen (2006a) have put it. In practice, the homology assumption provides a basis for adopting still other domain-specific, biological concepts in an abstract disguise. For example, the biological notions of genotype and phenotype are transformed into an abstract “replicator–interactor” dichotomy which is then suggested as a general analytic tool in evolutionary theories (Hull 1988; Hodgson and Knudsen 2004, 2006b).

The claims for Generalized Darwinism have stirred some opposition in evolutionary economics. Buenstorf (2006), for example, argues that the concept of selection based on a replicator–interactor distinction contributes little to understanding industrial evolution: “The homology between biological inheritance and the replication of routines is achieved at the price of abstraction from economically relevant details.” On the basis of his “ontological continuity hypothesis”, Witt (2003 chap 1; 2004, 2008) claims that the Darwinian theory is relevant for understanding economic evolution, albeit not in the form of abstract one-fits-all principles but rather as a meta-theory about how man-made evolution could emerge from, and is molded in, evolution in nature. The notable difference here is that, as a result of the intervention of human intelligence, intentionality, creativity, and knowledge accumulation, the rules of man-made evolution are likely to differ from those of biological evolution. Supporting this view, Cordes (2006) points to the fact that biological evolution has created a human “behavioral repertoire” which forms the bedrock of cultural evolution and gives rise to specific adaptation dynamics for the understanding of which the abstract principles of variation, selection, and inheritance contribute little, if anything. Vromen (2008) notes that “Hodgson and Knudsen do not counter this critique by dismissing the extra substance that Witt’s continuity hypothesis adds..., but by diminishing the ontological substance of their own Generalized Darwinism.... The price they have to pay for this is that it leaves their Darwinian principles with virtually no content.”

With this background, the present paper sets out to do two things. First, it will go beyond the criticism launched against Generalized Darwinism so far by showing that its very core—the identification of Darwinism with the principles of variation, selection, and inheritance—is highly problematic (an argument that can also be leveled against Dawkin’s Universal Darwinism). Conceptualizing evolution by means of these abstract principles not only comes at the price of abstracting from economically relevant details, as Buenstorf (2006) notes, but also at the price of abstracting from crucial biological “details”. In biology, it depends on the specific features of the processes by which genetic information is varied and inherited whether different theories about variation, selection, and inheritance qualify as Darwinian, non-Darwinian, or anti-Darwinian (see Levit et al. 2008a). These features require specific hypotheses and addenda (left out by the suggested abstraction) which ultimately decide about the fit with the logic of Darwinism and the empirical evidence that has been gathered for it. The three principles that remain as allegedly unifying framework after the isolating abstraction has been carried out are but an empty hull that, as Vromen has suggested, lacks explanatory substance.

Second, and in recognizing the history of evolutionary theorizing in biology, the paper will examine more closely the very idea inspiring Generalized Darwinism. This is the claim that the abstract principles identified by Generalized Darwinism are helpful in, or even necessary for, building up an encompassing evolutionary theory in economics. To invoke abstract principles “top-down” to construct a theory of evolution, e.g., for the economic domain, means, of course, that hypotheses on the “details” of the particular domain must be found that account for the specific conditions of, say, economic processes. However, such specific hypotheses can neither be logically derived from the abstract principles, nor do the abstract principles seem particularly useful or even necessary for arriving at specific hypotheses by which explanations for economic evolution can be provided. Doubts may therefore be raised as to whether Generalized Darwinism offers any advantage in explaining economic evolution. It seems worth noting in this context that evolutionary biology, the very discipline from which Generalized Darwinism borrows its principles, has never taken resort to the research strategy which the proponents of Generalized Darwinism suggest for evolutionary economics.

Accordingly, the plan of this paper is as follows. Sections 2 to 4 give an overview of three different phases in the development of Darwinism. Each of these phases is indicative of significant changes in the central tenets that the Darwinians were forced to accept in view of the problems emerging from the specificities discovered in the various interconnected research activities, reaching from the paleontological record to the molecular processes governing life and reproduction. In this process, the principles of variation, selection, and inheritance did play a role. But they were neither the only general principles considered relevant, nor did they serve as the criterion distinguishing Darwinian from non-Darwinian interpretations. In the light of these insights, Section 5 turns to a discussion of the prospects and limitations of the top-down approach suggested by the proponents of Generalized Darwinism for the development of evolutionary economics. Section 6 presents the conclusions.

2 Darwin and “Classical Darwinism”

Darwinism is a dynamic and complex theoretical system consisting of several necessary, interconnected postulates and numerous theoretical addenda and specific hypotheses. A difficulty that arises in defining Darwinism results from the fact that the theory of natural selection achieved its logical consistency and conceptual maturity only decades after Darwin’s death. For that reason, an appeal to Darwin’s own writings is not necessarily the best way to determine whether a particular concept in question is, or is not, “Darwinian” in character. To account for this complication, it is necessary to distinguish different, historically contingent, phases of “Darwinism” (and of anti-Darwinian concepts) in biology. The majority of historians of biology agree that there are three major phases in the growth of Darwinian thought (Reif 2000; Reif et al. 2000).

The first phase is Classical Darwinism, represented by Darwin’s The Origin of Species by Means of Natural Selection (1859), subsequent works, and their immediate reception among biologists. Pushing the idea of organic evolution and common descent, Darwin’s contemporaries grappled with how strictly to interpret the relative importance of the principle of natural selection, the concept of inheritance combining both hard and soft (‘Lamarckian’) mechanisms, and the postulate of gradualism. In fact, Darwin had proposed a theoretical system which implied four evolutionary mechanisms: natural selection, sexual selection, direct effects of the environment, and inheritance of acquired characteristics. This pluralism of evolutionary mechanisms had led by the end of the 19th century to a second phase that culminated in the establishment of two major schools of Darwinism.

The neo-Darwinians, in particular August Weismann (1885), eliminated Lamarckian mechanisms and Darwin’s pangenesis, thus opening the avenue to a consequent selectionism. In details, however, Weismann’s approach was hardly compatible with the mature version of Darwinism (i.e. the synthetic theory of evolution).¹ The interpretation of such “details” of the various principles and their relative importance also gave rise to alternative (anti-Darwinian) and rival theories of evolution. Incoherences within the Darwinian theories and unresolved puzzles left by classical Darwinism and neo-Darwinism even resulted in what became a temporary eclipse of selectionism in evolutionary biology (Bowler 1983).

Only with the third phase of Darwinian thought could the essential difficulties be overcome to a certain extent. This phase started with the break-through of the Synthetic Theory of Evolution in the late 1930s, the so-called Modern Synthesis prepared by the rise of Mendelism and population genetics. It was one more step to Darwinism as it is now: a logically coherent and empirically applicable research program. In this form, it is able to integrate classical genetics, population genetics and molecular genetics, systematics, evolutionary morphology, developmental biology, palaeontology etc. on a selectionist basis while, at the same time, acknowledging the importance of non-selectionist factors of evolution, especially isolation, chance events, and population size. This means that “... selection is regarded as important, but only as one of several evolutionary factors” (Reif et al. 2000).

Integral to Classical Darwinism—the starting point of the long development—was Darwin’s and Wallace’s introduction of the “principle of natural selection” as a hypothetical mechanism of biological evolution (Darwin and Wallace 1858; Hossfeld and Olsson 2009). Darwin’s influential German disciple, Ernst Haeckel,² argued that a scientific theory of transmutation (evolution) had already been formulated by Lamarck (a theory for which Haeckel coined the very term “Lamarckism” as opposed to Cuvierism, a theory demanding constancy of species). According to Haeckel, this was a general theory of descent aiming at a complete and harmonic picture of evolution by reducing all its phenomena to “the only physiological process of nature, the transmutations of species” (Haeckel 1866, p. 167). The theory of natural selection, by contrast, reveals the exact machinery, the “mechanical causes” of transmutation (evolution), and explains its directionality. Accordingly, Lamarck should be appreciated, Haeckel claimed, for promoting general evolutionism, while Darwin must be praised for his causal explanation of the theory of descent. Yet what Haeckel considered “Darwinism” differed crucially from Weisman’s “Darwinism” and the differences rooted, first of all, in their interpretation of variation and inheritance (see Hossfeld 2010).

Darwin read the proof-sheet of Haeckel’s opus magnum and approved Haeckel’s interpretation of his own theory, albeit expressed concerns about Haeckel’s boldness in stating the theory.³ Haeckel had indeed made some strong claims. One was related to the progressive nature of evolution, where he argued that natural selection gradually, but permanently, works to perfectionate the organisms’ organization. It is well known that Darwin to some extent supported the idea of evolutionary progress. On the other hand, he was extremely cautious in formulating this idea and stated that the naturalists were in disagreement about “what is meant by an advance in organisation”. Haeckel, who was influenced not only by Darwin but also by natural philosophy, knew these doubts but did not take them seriously. Another of his bold statements related to the idea of internal and external constraints in evolution, echoing the idea of orthogenesis.⁴ Darwin modestly supported the idea of “constraints”, but again, Haeckel made it a “strong argument”.

With his elaborations, Haeckel contributed, in a sense, to the sharpening of controversial points arising from Darwin’s original concept. Haeckel, as other early Darwinians, was aware of the fact that selection, variation, and retention played a role (though at the time they were not spelled out as general principles), but the exact mechanisms underlying them still needed to be specified. In trying to do so, he remained speculative, supporting ideas of natural selection combined with the concept of environmental influences on the organism’s heredity, evolutionary constraints, and the idea of progressive evolution—all compatible with the abstract principles. Thus, it was not by these principles that the further fate of selectionism was decided upon.

3 Darwinism eclipsed: alternative evolutionary theories and the spac of logical possibilities

Due to the incomplete and sometimes even contradictory data of paleontology, anatomy/morphology, biogeography, systematics, and genetics, the reconstruction of evolutionary history and evolutionary mechanisms still had to be provisional and even speculative (Gould 1977, 2002; Bowler 1983, 1992). Failure to answer convincingly the questions concerning heredity, variation, evolutionary progress, and orthogenetic series indicated that Darwinism as a theoretical system was still incomplete. This weakness meant water on the mills of theories of evolution that were considered rivals of Darwinism. Indeed, in the heydays of these rivals in the first decades of the 20th century, a situation prevailed in evolutionary biology in which the Darwinian theory of natural selection became just one of an entire set of more or less plausible theories of how evolution proceeds. In the majority of cases, these were complex theoretical systems that combined the elements of selectionism with non-selectionist ideas. The variety of theories was so large that it was judged to almost exhaust the “space of logical possibilities” for explaining phylogenetic history (Zavarzin 1979).

One of the theories was neo-Lamarckism, often seen as a major alternative to selectionism. By claiming that acquired characteristics are heritable, neo-Lamarckians suggested a logically consistent explanation for the high rate of evolutionary change and the appearance of complex adaptive structures. (The Darwinians had difficulties with explaining these phenomena on the basis of their core assumption of variation being random.) However, the majority of the 20th century Lamarckians did not reject completely the idea of natural selection and other evolutionary mechanisms. To the contrary, they often combined the inheritance of acquired characteristics with the Darwinian idea of natural selection, orthogenesis, mutationism, etc. Furthermore, neo-Lamarckism itself never represented a monolithic theory, and the notion of the inheritance of acquired characteristics is not more than a kind of ‘umbrella-concept’ for several different explanations of the evolutionary mechanism.⁵

A second theory that became especially popular in the first half of the 20th century, and that was partly rivaling Darwinism, was based on the idea of directed evolution (orthogenesis). It came in more than twenty different variants (Levit and Olsson 2006). Although some of them included finalistic ideas (see Mayr 1982, p. 959), the common denominator of all of them was the idea of morphological, molecular, or other constraints supposed to guide phylogeny in such a way that the mechanism of natural selection becomes superfluous either for explaining evolution as a whole or for explaining certain periods of phylogenetic history. The majority of advocates of directed evolution proceeded from the assumption that organisms are predisposed to vary in certain directions and that such biases determine major transitions in evolution. They were convinced that these directions are empirically observable in the paleontological record and clearly definable. Accordingly, it was assumed that evolutionary events follow certain clearly defined laws and restrictions, and that evolution proceeds to a significant or predominant extent in a non-adaptive fashion. In other words, orthogenesis was strongly coupled with the idea of non-adaptive trends in evolution.⁶

A third major anti-Darwinian theory in evolutionary biology was saltationism—which is still influential today. The advocates of saltationism (e.g. Bateson, Goldschmidt, Schindewolf) rejected the Darwinian idea of a slowly and gradually growing divergence of characteristics as the only source of evolutionary change. They did not necessarily deny gradual variation or natural selection any relevance, but claimed that major changes in the “body plans” come into being as a result of sudden, discontinuous and unique changes, e.g., in a series of macro mutations. The latter were held responsible for the sudden appearance of new higher taxa, while the small variations resulting from natural selection were supposed to be responsible for the adaptations below the species level, i.e. the “fine tuning” adjusting organisms to their environment. Hence, they were considered of secondary importance.⁷

Considering Darwinism and all its rivals of the time, an impression can be gained of how rich are the logical possibilities for making sense of evolutionary change in the biological domain. There is no logical restriction implied by the three “Darwinian” principles—as can be seen from the fact that they are shared by all these variants. The relationships between the various explanatory approaches and their mutual (in-) compatibility are determined by a plethora of other doctrines. Thus, pure neo-Lamarckism and strict selectionism are not mutually compatible, because the inheritance of acquired characteristics already guarantees adaptive changes and natural selection loses its shaping role (in the presence of adaptive mutations, natural selection becomes a secondary evolutionary mechanism—raising the question of what could be assumed to guide the inheritance of acquired variation). Saltationism and mutationism are incompatible with all historical forms of Darwinism, because evolution cannot be gradual and saltational at the same time. The latter two approaches are incompatible with orthogenesis, because evolutionary changes cannot be random and directed at the same time. Orthogenesis, in turn, is the more difficult to align with Darwinism, the more constrained the random process of variation is imagined to be.

4 Modern Darwinism: the synthetic theory and the expanded synthesis

In the first third of the 20th century, Darwinism and all its scientific rivals appeared to be on equal terms regarding their plausibility. Accordingly, many outstanding evolutionary biologists of those times subscribed to some form of a “synthetic” theory, combining various selectionist and non-selectionist arguments in a way that tried to explain the diversity of the living world by some major mechanisms of evolution combined with various auxiliary mechanisms. To some extent, this was in accord with Darwin’s own appeal to additional hypotheses, but made the resulting theoretical systems awkward, sluggish, and sometimes even amorphous. In the understanding of the early “synthesizers”, certain principles (for example, the principle of natural selection) were valid for certain domains or levels of biological reality, but not for others.

Before this background the selectionist turn of evolutionary biologists worldwide from the mid 1930s onwards came suddenly and was remarkably radical (as if all of them were passionate Kuhnian revolutionaries). All of a sudden it was recognized that the nature of inheritance and variation could be re-interpreted in the light of a new genetics. Following a re-interpretation, it was realized that biological systematics could be incorporated into a general theory of evolution involving the fields of palaeontology and evolutionary morphology. The paleontological data fit the new explanatory paradigm developed in molecular genetics and microsystematics. Moreover, it became clear that the theory not only covered the findings at the population level, but could also explain many macroevolutionary phenomena without appealing to saltationism, orthogenesis, or neo-Lamarckian mechanisms.⁸

The “synthetic theory of evolution” was based on three crucial assumptions: (1) natural selection is the major direction-giving factor in evolution; (2) random mutations and recombinations cause variation and therefore imply the incorporation of genetics into the theory; (3) geographic isolation is the most important mechanism separating populations (Junker and Hossfeld 2009). The original “Origin-of-Species-Darwinism” (Darwin 1859, 1871) served as a motivation and a paradigmatic framework, but the scope and content of the newly established theoretical system went far beyond the historical sources and corrected them in many respects. In forming an alliance with genetics, the synthetic theory of evolution not only enriched itself, but also restricted the panoply of concepts that previously appeared potentially or actually acceptable. The concepts of inheritance of acquired characteristics, orthogenesis, and saltationism were rejected as incompatible with what established itself now as mainstream “Darwinism” in the form of the synthetic theory. Thus it were the architects of the “second Darwinian revolution” (Mayr 1991) who, for the first time, drew an ultimate and clearly defined line between “Darwinism” and its actual and potential rivals.

However, the synthetic theory did not solve all problems and retrospectively appears to have been too simplistic. There were still a whole range of phenomena which remained unexplained. For example, the “phylogenetic explosions” that had been observed by Schindewolf and Beurlen in the paleontological record (Beurlen 1930) remained a puzzle. Eldredge and Gould (1972) took this problem up again and proposed the well known concept of punctuated equilibria as a solution. This was first thought to represent an alternative to the gradualism of the synthetic theory of evolution, since the authors initially intended to demonstrate that there are domains of evolution inaccessible for gradualism. However, recent accounts of the problem have shown that evolution is both gradual and punctuated, since even accelerated evolutionary transitions are gradual processes at the micro-evolutionary level (Bokma 2002; Gould 2002; Kutschera and Nicklas 2004). What is still an issue, though, is the exact role of species selection and group selection from the empirical and the theoretical points of view (Reif et al. 2000).

Yet another problem posed to the synthetic theory that requires further clarification and possibilities for further expansions is the whole complex of issues around developmental biology and its relations with other biological disciplines (Hall et al. 2003). The most recent synthetic movement of an “ecological developmental biology” tries to integrate the whole range of relevant disciplinary domains in a further extension of the classic synthetic theory of evolution (see, e.g., Gilbert and Bolker 2003). In that light, for example, allelic variation in the structural genes is no longer the only source of variation. There are two other sources of variation accessible for natural selection, namely allelic variation in the regulatory regions of genes and “developmentally plastic variation” (Gilbert and Epel 2009, p. 318). The latter can be crucial for macro-evolutionary processes producing the variations associated with phyla and classes of species. This example shows that the occurrence of variation as such does not mean much as to whether it is to be considered a Darwinian principle or not. Depending on how the nature of variation is interpreted, the results of its combination with the principle of selection will also vary. The “devil is in the details” (Vromen 2007) indeed. Amendments to a general concept such as variation that seemed of only technical character turn out to make a radical difference with respect to whether biological evolution is explained in a Darwinian or a non-Darwinian way.⁹

It is not our purpose here to give a full account of the directions in which the synthetic theory currently expands.¹⁰ The examples are rather given to demonstrate that contemporary evolutionary theory development continues to be driven by bottom-up modifications, syntheses, and enrichments. It expands partly by accounting for elements that were previously considered anti-Darwinian, but that can now be made sense of by the discovery of yet other evolutionary mechanisms consistent with the general framework of modern Darwinism. This is not to say, though, that “anything goes” provided only the principles of variation, selection, and retention are complied with. It is the specific combination of the particular evolutionary mechanisms and their “details” that matter for whether or not an extension of the complex theoretical system is “Darwinian” in nature.

5 How useful can generalized Darwinism be for evolutionary economics?

The brief review of the developments in evolutionary biology in the previous sections shows that different variants of the theory of variation, selection, and inheritance have played a role in almost all approaches that have been influential over the past 150 years in biology, whether Darwinian, non-Darwinian, or even anti-Darwinian. For this very reason we have argued that variation, selection, and inheritance as such are no particularly “Darwinian”. However, the sketch of the history of evolutionary biology also shows that explanatory progress in this discipline was driven by the inquiries into details of evolutionary processes that were not captured by abstract principles. These inquiries evoked more general conclusions which were scrutinized and eventually modified or rejected by further work on “the details”—what we have referred to as a bottom-up research strategy. In comparison, Generalized Darwinism proposes an entirely different research strategy for evolutionary economics. Starting from the abstract principles of variation, selection, and inheritance, it suggests deploying them in the search for auxiliary hypotheses that explain the details of evolutionary processes in the economic domain. This proposal raises several questions which, in our view, warrant reflections that lead up to a critical assessment of the usefulness of Generalized Darwinism for evolutionary economics.

A first point that, we think, should be taken into consideration is the complete absence of something comparable to the top-down strategy of Generalized Darwinism from the history of evolutionary biology. The lesson that can be learned from the leading evolutionary science is that evolution can be expected to be an extremely complex process the explanation of which requires a similarly complex network of specific hypotheses. In biology, this emerged from an extremely broad, rich, and controversial discourse running the whole gamut from the details of the paleontological record to the findings of molecular biology. It was the attempt to give the complex network of specific hypotheses a coherent causal and functional structure that eventually led to the formulation of general principles—not the other way round. One may wonder what evolutionary biology would look like today had it proceeded in a top-down fashion, starting just from the abstract principles of variation, selection, and inheritance. Why should what appears a strange idea with regard to evolutionary biology be a promising idea when it comes to other domains such as that of evolutionary economics?

A second point arises from the fact that Generalized Darwinism is based on the claim that the three abstract principles derived by isolating abstraction from findings in the biological domain govern “all complex population systems (in the biological and social worlds)” (Hodgson 2009). Hence, evolution is conceived of as a homologous process in different disciplinary domains—among them, in particular, the domain of evolutionary economics. However, the homology postulate is an empirical claim. It is far from being self-evident, since an abstraction arising from a bottom-up discourse in one particular discipline is influenced by the domain-specific (ontological) conditions of that discipline which are not necessarily the same elsewhere. Hence, the claim needs to be supported by empirical evidence for each disciplinary domain for which its validity is asserted.

As far as evolutionary economics is concerned, a proof is lacking and, as Nelson (2006) has argued, doubts may be raised as to whether it can be expected to ever be delivered. The best way to confirm the asserted homology with evolutionary biology would be the demonstration that the principles of variation, selection, and inheritance emerge identically from isolating abstraction in a “bottom-up” discourse reflecting specific research results on evolutionary processes in the economy.¹¹ However, such a bottom-up discourse is precisely the opposite of what the proponents of Generalized Darwinism have in mind. Their idea is to invoke variation, selection, and inheritance in a top-down fashion as principles governing economic evolution and, assuming their validity, to search for suitable “auxiliary hypotheses” by which the abstract scheme can be filled with economic explanatory substance (see Aldrich et al. 2008).

A third problem we see relates to the dichotomy of abstract principles, on the one hand, and “auxiliary hypotheses” on the other. (To recall, the latter are needed because, by its very nature, Generalized Darwinism lacks disciplinary substance.) By taking the validity of its abstract principles for granted, Generalized Darwinism factually turns into a search heuristic designed to find something equivalent to, or reminiscent of, variation, selection, and inheritance at the disciplinary level. On purely logical grounds, this is a possible way to proceed—as much as it is possible to start from any other general principles and to look for suitable auxiliary hypotheses on “the details”. But there is a non-negligible risk involved in such a top-down approach of inducing systematic heuristic biases. The interpretation of reality can be seriously biased by selectively recognizing only those elements that seem to fit the logic of the chosen principles and ignoring all others, or by constructing disciplinary correlates that actually misrepresent the economic conditions.

In view of the criticism raised against Generalized Darwinism, particularly its top-down research strategy, it may be asked what evidence for fruitful applications there is in works so far published on Generalized Darwinism. Stoelhorst (2008) has argued that the three principles of Generalized Darwinism can be seen as part of an (abstract) explanans of an evolutionary explanation that can be put to work in the economic domain only if it is clear what the explanandum is going to be. One may add that it is precisely in specifying the explanandum that domain-specific “details” have to be entered and elaborated via “auxiliary hypotheses”. But, as Stoelhorst notes, this has not been clarified in the writings on Generalized Darwinism. Indeed, the already quite numerous publications on Generalized Darwinism offer little on the crucial “auxiliary hypotheses”.¹²

In their manifesto in defense of Generalized Darwinism, Aldrich et al. (2008) do not mention any specific implications of their approach for explaining evolution in the economy. They posit that domain-specific substance can be developed on the basis of the reflections about the unit of selection. In that context, they refer to the replicator–interactor dichotomy and discuss it at an abstract level. An interactor is defined as in Hull (1988, p. 408) as “an entity that directly interacts as a cohesive whole with its environment in such a way that this interaction causes replication to be differential”. Replicators are defined as entities that are replicated in a way that “involves a causal relationship between two or more entities, where there is substantial similarity between the original and the replicated entities, and where information concerning adaptive solutions to survival problems is passed from one set of entities to another”. But what are the replicators and interactors at the concrete economic level and how precisely are they varied, selected, and inherited, and how does all this contribute to understanding man-made evolution?

In the extensive discussions of interactors and replicators in their most recent book, Hodgson and Knudsen (2010) do not explain in any detail how replication and interaction work and drive evolutionary processes in the economy. In terms of concrete applications, their reflections basically come down to suggesting certain pairings of interactors and replicators.¹³ It is left open, however, how the developed categorizations lead to explanations of concrete economic evolutionary phenomena. In a paper inquiring more deeply into the role of firms as interactors, Hodgson and Knudsen (2004) argue that habits and routines, which are more or less specific to individual firms, replicate under the influence of selection forces. But again, neither the selection forces nor the assumed selection and replication mechanisms are explained in detail. The authors mention the Price equation (Price 1970; see also Frank 1995)—another very abstract concept. Yet, they do not demonstrate what causal explanations can be derived regarding specific evolutionary processes in economics. Hodgson and Knudsen repeatedly claim a broad agreement of their approach with the argumentation and formalizations in Nelson and Winter’s (1982) seminal work on the evolution of firms and industries. (See, however, the reservations Nelson 2006 makes regarding this claim.) Yet, they do not show what explanatory conjectures their Generalized Darwinism is actually able to contribute beyond those obtained already by Nelson and Winter on the basis of only a loose analogy to the notion of natural selection.

In view of the problems that have been pointed out in this section, the rather thin outcome of the top-down approach of Generalized Darwinism in terms of concrete, new explanations offered for evolutionary phenomena in economics may not come as a surprise. By comparison with the successful bottom-up research practice in evolutionary biology, one may infer that the reason for this outcome lies in the nature of heuristically proceeding top-down from abstract principles with uncertain validity for economics to suitable auxiliary hypotheses representing the domain-specific, economic explanatory substance. If so, the recommendation for evolutionary economics would be to focus on analyzing the huge variety of specific evolutionary processes in the economy at a concrete level, and only when explanatory progress has been made at that concrete level to engage in a (bottom-up) discourse of how the complex set of specific hypotheses can be organized into a more coherent causal and functional structure.

Such a discourse may eventually lead to the formulation of general principles. Since all of the specific economic processes are instances of man-made evolution, it may turn out that general principles akin to economic evolution relate to the core features of human adaptiveness. These are individual and collective learning, experimentation, insight, inventiveness, and the restless motivations driving these human forms of adaptations. If so, to understand and explain how these features shape the specific, evolutionary processes in the economy may be more important than demonstrating that the abstract hull of Generalized Darwinism can be stretched to fit them.

6 Conclusions

In a recent series of publications, “Generalized Darwinism” has been proposed as a new overarching research strategy that is based on the assumption of a fundamental homology between evolution in nature and the evolution of the economy. The principles of variation, selection, and retention that have been distilled from evolutionary biology by isolating abstraction are claimed to be generally valid. It is suggested to apply these abstract principles as a unifying framework for all evolutionary theories. By a brief reconstruction of the different historical forms of Darwinism, we have shown that the identification of these abstract principles with Darwinism is misleading. Moreover, on a priori grounds, other principles—non-Darwinian or even anti-Darwinian ones such as, e.g., orthogenesis, saltationism, or neo-Lamarckism—could claim a similar plausibility in explaining economic evolution.

The crux with such allegedly unifying abstract principles derived by isolating abstraction from findings in other domains is that they provide but an abstract hull. In order to become a useful heuristic device, they need to arrive at domain-specific explanation which, in turn, would require adding substance by hypotheses on the disciplinary “details” of actual evolutionary processes, e.g. in the economy. This is, of course, what is done in the first place in a bottom-up research strategy as it has fruitfully been practiced in the development of Darwinism in evolutionary biology. In the final section, doubts have therefore been raised as to whether the top-down approach propagated by Generalized Darwinism can overcome its inherent limitations and develop into a fruitful research strategy for evolutionary economics.