Ontological Foundations for Evolutionary Economics:
A Darwinian Social Ontology
J.W. Stoelhorst
Amsterdam Business School
University of Amsterdam
Roetersstraat 11
1018 WT Amsterdam
The Netherlands
Paper to be presented at the 2008 EAEPE meeting in Rome
Research Area V: Ontological Foundations of Evolutionary Economics
Key words: Generalized Darwinism, Ontology, Multi-level selection, Institutions
Ontological Foundations for Evolutionary Economics:
A Darwinian social ontology
The purpose of this paper is to further the project of generalized Darwinism by developing a social ontology on the basis of a combined commitment to ontological continuity and ontological commonality. Three issues that are central to the development of a social ontology are addressed: (1) the specification of units of analysis above the level of the individual, (2) the specification of the relationship between self-organization, developmental processes, and environmental selection as drivers of socio-economic evolution, (3) the nature and the role of agency in socio-economic change. It is shown how the explanatory logic and ontological commitments of generalized Darwinism help resolve these issues. Specifically, the paper builds, first, on the distinction between ‘units of selection of’ and ‘units of selection for’ that is central to the explanatory logic of generalized Darwinism, and second, on the distinction between within-group competition and between-group competition that is central to generalized Darwinism’s ontological commitment to multi-level selection processes. This leads to a social ontology in which (1) only individuals and groups of individuals are ‘units of selection of’, (2) agency is limited to individuals, (3) game-theoretic interactions between individuals establish the social rules that allow groups to self-organize into stable social structures, (4) these social rules become ‘units of selection for’ in between-group competition. The resultant view of economic change is of a bottom-up process in which individual action is the only driver of economic change, but in which both individuals and groups are seen as essential carriers of productive knowledge.
1. Introduction
The project of generalized Darwinism is to specify the general nature of an evolutionary explanation by uncovering ontological commonalities across evolutionary processes in biology, on the one hand, and cultural and economic change, on the other (Aldrich et al. 2008; Hodgson, 2002; Stoelhorst, 2005). The purpose of this project is to derive an explanatory framework that can serve as a meta-theory for evolutionary explanations in the social sciences in general, and economics in particular. The principle method for deriving this framework has been to start with the current understanding of the nature of biological evolution, as originally captured in Darwin’s theory of evolution and since developed in biology, and to build on ongoing work in the philosophy of science to generalize Darwin’s explanation of evolution by abstracting from causal mechanisms that are specific to evolution in nature. This project has led to some progress in specifying the nature of a Darwinian meta-theory that can inform theory development in evolutionary economics (Hodgson and Knudsen, 2006a; Stoelhorst, forthcoming).
So far, the primary focus of the project of generalized Darwinism has been on deriving the general causal logic of an evolutionary explanation. While the project of generalized Darwinism has been positioned as an ontological endeavor (Hodgson, 2002; Stoelhorst, forthcoming) it has not as yet developed an explicit social ontology. The purpose of this paper is to embark on this task. The term (social) ontology is here used in the sense of a specification of the main building blocks of (social) reality and the characteristics that these building blocks share (cf. Lawson, 2004). The intended contribution of the paper is to propose a social ontology that is consistent with the basic tenets of generalized Darwinism, and in doing so dispel some of the criticism of the project of generalized Darwinism.
The viability of the project of generalized Darwinism has been contested on several grounds, but mainly on the argument that cultural evolution, and by extension economic change, may simply be too different from biological evolution to warrant a search for ontological commonalities across the natural and cultural domains. The consequences that the critics (Foster 1997, Cordes 2006, Witt 2004) and skeptics (Lawson 2003, Nelson 2006, Vromen 2004) derive from their arguments are somewhat different. For instance, Nelson is skeptical of ontological commonalities and content to continue using Darwinian concepts in an analogical way. Vromen allows for the possibility that ontological commonalities can be derived but is skeptical of the value of such commonalities when explaining cultural and economic evolution. Cordes and Witt deny the possibility of ontological commonalities and propose the idea of ontological continuity as an alternative view of making connections between the theory of biological evolution and theories of economic change.
This paper argues (1) that a satisfactory social ontology can only be developed when the notions of ontological continuity and ontological commonality are combined, (2) that this combination brings the agency-structure problem into focus, and (3) that the explanatory framework of generalized Darwinism suggests a social ontology that allows us to deal with this problem. This social ontology emphasizes (1) individual agents and groups as the only interactors that are subject to environmental selection, (2) individual agency as the only source of variation in evolutionary processes, and (3) both individuals and groups as carriers of productive knowledge.
2. Ontology in evolutionary economics
The term (social) ontology is here used in the sense of a specification of the main building blocks of (social) reality and the characteristics that these building blocks share (cf. Lawson, 2004). There is currently much interest in the ontological foundations of economics and a number of authors have recently suggested ontological foundations (Lawson 2003, 2004, Dopfer and Potts 2004, Herrmann-Pillath 2001). The general search is for an ontology that does justice to the dynamic, open ended, and self-transformational nature of social reality to inform theory development in evolutionary economics, if not heterodox economics in general (Lawson, 2006). The general purpose of the ontology proposed in this paper is similar, but its specific focus is somewhat different. Whereas Dopfer and Potts and Herrmann-Pillath are primarily concerned with developing an ontology that is sufficiently general to encompass all of reality, the focus of this paper is more specifically on developing a social ontology. This focus on social reality is similar to Lawson’s project. In fact, his view of social reality as ‘an emergent, open-ended, structured, transformational process in motion, in which the parts are constituted in and through their (changing) relations to each other’ (Lawson, 2004, p.19) is here accepted as a good approximation of the nature of social phenomena. In contrast to Lawson, however, whose project has so far largely focused on the general nature of social reality (or, what he refers to as ‘social philosophical ontology’), the focus in this paper is rather more on the specification of the social entities that are of interest in the development of evolutionary economic theory (or what Lawson refers to as ‘social scientific ontology’).
This paper proceeds on three presuppositions about the role of ontology. First, the development of an ontology is a ground-clearing project to allow for better theory development (Lawson, 2004). Second, the development of an ontology consists of developing heuristics that lead to conjectures about reality (Vromen, 2004), which is to say that, rather than engaging in prima philosophia, the purpose is simply to make progress in articulating a view of reality that can lead to better theories. Third, the ultimate goal of ontology is to inform explanatory projects, which is to say that it is not sufficient to specify the basic building blocks of reality and their common characteristics. To further explanatory projects, an adequate ontology must also be able to point towards the causal mechanisms by which basic entities relate to each other. On this view, there are three questions that a social ontology must answer:
-What are the basic entities of social reality?
-What are the common characteristics of the basic entities of social reality?
-How do the basic entities of social reality relate to each other?
In the remainder of the paper, the method for deriving a social ontology that will be employed is to consider social reality in Darwinian terms. Generalized Darwinism comes with a number of ontological commitments that are entirely compatible with the view of social reality as a dynamic, open-ended, self-transformational system that is widely shared among evolutionary economists (Stoelhorst, forthcoming). In addition, generalized Darwinism also offers a powerful heuristic for the specification of relevant social entities and their common characteristics. And, crucially, it offers an analytical framework that allows us to specify how the basic entities of social reality are causally related to each other.
3. Ontological commonality: The explanatory logic of generalized Darwinism
To see how generalized Darwinism can help us answer the three questions formulated above, we need to consider its explanatory logic and ontological commitments. This section will first consider generalized Darwinism’s explanatory logic and the related commitment to ontological commonality, which helps answer the first two questions. The next section will turn to generalized Darwinism’s commitment to ontological continuity and its implications for answering the third question.
Generalized Darwinism derives its explanatory power from a recursive causal logic that is best understood in terms of its explananda, explanantia, and units of analysis:
The explananda of generalized Darwinism
The explananda of generalized Darwinism are threefold (Dennett, 1995). First and foremost, generalized Darwinism explains how open, complex systems achieve adaptive fit. Systems are open when they require resources from their environment to be viable, and systems are complex when they consist of interacting elements. In addition, generalized Darwinism explains how the same process that allows open complex systems to become adapted to their environment leads to variety from common origins and the accumulation of adaptive complexity.
The explanantia of generalized Darwinism
The explanantia of generalized Darwinism are similarly threefold. They are the well-known mechanisms of variation, selection, and retention (Campbell, 1960, 1965). Their interplay establishes an algorithm that necessarily leads to evolution (Dennett, 1995). In other words, when open, complex systems have a mechanism to introduce variations in the way they interact with their environment, are subject to a consistent selection mechanism, and have a mechanism to retain the variations that increase their success in securing resources from the environment, they will evolve adaptations to their environment.
The units of analysis of generalized Darwinism
The units of analysis of generalized Darwinism are twofold (Dawkins,1976; Sober, 1984; Hull, 1988). The algorithmic nature of the Darwinian variation, selection, retention triumvirate depends on the establishment of a feedback loop that ensures that the success, or lack thereof, of the interactions of an entity with its environment biases future interactions with the environment in the direction of what worked in the past. In addition to the system itself, this feedback loop requires an additional unit of analysis. Something needs to code for the nature of the environmental interactions of the system. In biology, this second unit of analysis is the genotype, in which the relative success of phenotypic interactions with the environment is encoded. Different labels have been suggested for units of analysis that can serve an explanatory function that is logically equivalent to the genotype-phenotype distinction in biology. We will here refer to ‘interactor’ (Hull, 1988) or ‘unit of selection of’ (Sober, 1984) for the system itself, and ‘codex’ (Wilkins, 2001) or ‘unit of selection for’ (Sober, 1984) for the underlying unit of analysis that codes for the nature of the system’s interactions with its environment.
On the basis of this specification of the explanatory logic of generalized Darwinism we can easily derive answers to the first two questions specified earlier. Given its commitment to ontological commonality, or the notion that all evolutionary processes share common features, the relevant entities in social reality are open complex systems. Moreover, these systems share the common characteristics that they are able to vary the nature of their interactions with the environment and are able to feed back the success of these interactions to a codex that biases their future interactions in the direction of what worked in the past.
As yet, these answers are at the level of philosophical ontology, and answers at the level of scientific ontology require some further analysis. It is to this analysis that we turn next.
4. Ontological continuity: The multi-level nature of generalized Darwinism
There is a widespread commitment to a multi-layered ontology among evolutionary economists. Cultural change and economic growth are seen as the result of the interaction of evolutionary processes taking place at multiple levels of analysis. The additional ontological commitment of generalized Darwinism is that all these processes share the common features discussed above. However, this commitment to ontological commonality is not enough to derive a satisfactory social scientific ontology. It is not just that the Darwinian logic applies at multiple levels of analysis. Cultural evolution and economic growth need to be understood in terms of the relationships between these levels of analysis. This brings us to the question how the basic entities of social reality relate to each other. Answering this question, in turn, brings us to the notion of ontological continuity.
A combination of ontological commonality and ontological continuity is entirely consistent with the explanatory logic of generalized Darwinism (cf. Stoelhorst, forthcoming). In fact, given generalized Darwinism’s insistence on causal explanation (Hodgson, 2004; Vromen, 2004), its claim that the explanatory logic of Darwinism also applies to cultural and economic evolution can only be upheld when ontological continuity is accepted as part and parcel of the project of generalizing Darwinism. Because Darwinism does not allow for uncaused causes or skyhooks (Dennett, 1995), because cultural evolution and economic growth are the products of human action, and because humans are a biological species, it follows that cultural evolution and economic growth must be understood as proceeding on a biological substrate. This biological substrate consists of the genes that code for our behavioral dispositions, handed down by natural selection. Ontological continuity, of course, is the notion that these behavioral dispositions are what both enables and constrains cultural evolution (Witt, 2003).
However, the notion of ontological continuity also throws up a fundamental question for any attempt to derive a social ontology for evolutionary economics: where do social phenomena come from in the first place? The Darwinian variation, selection, retention algorithm can only reward behaviors that are in the interest of the individual organism, so how do we get from the individual level of analysis to collective phenomena? How, in particular, can the cooperative behaviors between individuals that are the defining characteristic of social species arise at all? By itself, the explanatory logic of Darwinism is unable to answer these questions. However, when the Darwinian algorithm is embedded in multi-level selection logic it does become possible to answer these questions.
Multi-level selection logic is concerned with the interaction of selection processes at multiple levels of analysis (Price1970, 1972; Sober and Wilson, 1998; Maynard-Smith and Szathmary, 1997). It sees evolutionary outcomes as the joint effect of within-group competition and between-group competition. That is to say that both individuals and groups are seen as possible interactors. In keeping with the Darwinian logic, individuals are the primary interactors that compete with each other for scarce resources from the environment. However, groups of cooperating individuals can also become interactors. This is only possible if groups bring an advantage in securing scarce resources from the environment to their members, and if they find ways to keep within-group competition among their members from undermining cooperation. According to multi-level selection logic, such an outcome can emerge when there is competition between groups. Between-group competition allows groups to evolve cooperative structures that can withstand invasion of free riders.
There are three things to note about multi-level selection logic in relation to the ontological foundations of evolutionary economics. First of all, when we take ontological continuity seriously, we need multi-level selection logic. The human ability to sustain cooperation in large groups of non-kin can only be explained in terms of multi-level selection logic. In other words, our behavioral dispositions are the result of multi-level selection processes. Without multi-level selection logic, we are simply unable to explain where the higher levels of adaptive complexity that are the hallmark of cultural evolution come from. Second, multi-level selection logic is premised on the notion of ontological commonality. Its central idea is that the Darwinian algorithm simultaneously operates at different levels of analysis. Third, multi-level selection logic entails an additional ontological commitment to simultaneous within-group and between-group competition as the driver of evolutionary processes. The corollary of this commitment is that all open complex systems need to be understood in terms of how they keep competition among their elemental building blocks from undermining their viability.
5. A social ontology for evolutionary economics
So far, the discussion has been at the level of social philosophical ontology and accordingly has proceeded in suitably abstract terms. It is now time to turn to social scientific ontology and translate the ontological commitments of generalized Darwinism into a specification of the basic building blocks of social reality that can further theory development in evolutionary economics. What generalized Darwinism contributes to the search for a satisfactory social ontology is a well-developed view of the accumulation of adaptive complexity as the result of the interaction between within-group competition and between-group competition. On the basis of this view, the following social ontology is proposed:
What are the basic entities of social reality?
The basic entities in social reality are individuals and groups. Individuals are here defined as members of the human species. Groups are defined as collectives of individuals that can be distinguished on the basis of a boundary condition that separates members of the group from outsiders. Both individuals and groups are open complex systems. As open complex systems, they are subject to selection pressures. These selection pressures result from the individuals’ and groups’ needs to secure scarce resources from their environment.