Chapter 2



This chapter outlines the basic argument and explains what is meant by generalizing Darwinian principles. Instead of the vague, extremely broad, and ambiguous word “evolution” we start from the basic ontology of the general kind of system that we are addressing. This ontology is described in section 2.1. Such worlds are given the name “complex population systems.”

Among its important features is a population of entities, each of which has the capacity to store and pass on information relevant for its survival. These entities face (at least) locally scarce resources, and have to struggle to survive and minimize degradation. We claim that this ontology applies to both biological and social reality: just as there are populations of organisms, in society there are populations of organizations.

We then interpret the Darwinian principles of variation, selection, and inheritance (synonymous with replication) as explanatory requirements: the facts of survival, variation and information transmission must be explained. Our next step is to argue that such explanations will inevitably involve a combination of generalities and attention to specific mechanisms. So explanations at the most abstract level are necessary but insufficient.

What would make this argument inapplicable? Critics could argue that the ontology of complex populations systems does not apply to social reality. Another objection would be that the argument does not help us very much. But we are not claiming that it is a complete explanation. It is more a meta-theory, or a way of organizing theories and raising questions that require further theories and explanations.


1. Is social reality a “complex population system” as described in Section 2.1?
2. What is meant by the Darwinian principles of variation, selection and inheritance?
3. What are the roles of general and more specific (auxiliary) theories in the explanation of complex phenomena?

32 thoughts on “Chapter 2

  1. len wallast

    Some penetrating questions with respect to Chapter 2:
    Is a particular population system also an interactor? Or is it the union of interactors and does that union also include the habitat?
    To implement the concept of selection we need to define the sample space (also called probability space in probability theory) holding the different states from and into which replicators are being selected during a particular time-interval. What is your sample space? What are the different states of selection? And what is your time-interval of selection? Concrete answers to these questions demarcate the (evolutionary) system we are considering. In principle you need to answer these questions for all potential systems that are object of study. Specific answers to these questions are needed in order to deal quantitatively (and sufficiently accurately) with input, output and net growth (negative or positive) of the system of evolution. Well, that can be done for economic systems [Wallast 2013, “Evolvodynamics- The Mathematical Theory of Economic Evolution”, Springer]. Unfortunately it is an illusion to achieve that in a general manner for all biological and social non-economic systems in a direct manner. No way to reconstruct the quantitative development of the population of the vampire bat and its habitat during the complete time-span of its evolution. We can only guess here and make it fit in the general qualitative framework of the Darwinian tale by the introduction of additional assumptions. The point is that those extra conditions to facilitate explanation hold in fact a set of auxiliary premises additional to the three Darwinian principles (selection, variation, inheritance). And a theory where we change and extend the basic premises dependent on what evolutionary system is to be explained has no right to claim the predicate of universality. Generalized Darwinism will then not be what its name pretends to be. In fact explanations where we need additional auxiliary conditions are in last resort tautological. Karl Popper (and many others) wrestled with this problem and it is worth recalling his keen argumentation:
    “Quite apart from evolutionary philosophies, the trouble about evolutionary theory is its tautological, or almost tautological, character: the difficulty that Darwinism and natural selection, though extremely important, explain evolution by the ‘survival of the fittest’ (a term due to Herbert Spencer). Yet there does not seem to be much difference, if any, between the assertion ‘those that survive are the fittest’ and the tautology ‘those that survive are those that survive’. For we have, I am afraid, no other criterion of fitness than actual survival, so that we conclude from the fact that some organisms have survived that they were the fittest, or those best adapted to the conditions of life. This shows that Darwinism, with all its great virtues, is by no means a perfect theory. It is in need of a restatement which makes it less vague ” [Popper 1979, p. 241, 242, Objective Knowledge, An Evolutionary Approach, revised edition, Oxford University Press].
    Well, I do not think Darwinism is in need of a restatement. Rather the three Darwinian principles are in need of a mathematical interpretation that clarifies how the three principles can be worked out quantitatively (there is only one unique way to do that) that provides for a general (non-tautological proof) of Darwinism. And the only domain of application where this can be done is the economic. Notwithstanding the fact that we can’t demarcate the variables of social (non-economic) and biological systems of evolution concretely, we can mathematically transcribe one by one all the concrete variables of economic evolutionary systems into their abstract counterparts of social (non-economic) or biological evolution. This implies that the proof for Generalized Darwinism follows from the mathematical proof of economic evolution.
    One final point: Popper’s remark stated above holds the indirect judgment that the concept of “fitness” cannot be given concrete significance to incorporate it within a general theory of Darwinian change. Attempts to lend it significance must therefore be avoided as it will not contribute to explaining the evolutionary process.

  2. Geoffrey M Hodgson Post author

    I agree with some but not all of Len’s points. Here is an emphatic point of agreement: as we explain in this chapter, any detailed application of Darwinian theory requires answers to specific and detailed questions. To some extent, mathematical modelling can also be very helpful.

    Len asks if a population can be an interactor. In response, we outline later in the book how single interactors can contain populations of interactors. For example, a national state (interactor) can contain a population of organizations (also interactors), just as our bodies (interactors) contain populations of bacteria (interactors).

    Popper shifted his position on whether the “survival of the fittest” is a tautology or not. He ended up with the view that it was not. So do Thorbjoern and myself. We can discuss this when we come to the tricky definition of “fitness” in a later chapter.

    1. len wallast

      Indeed in 1978 Popper reconsidered his ideas in a paper about the testability and logical status of the theory of natural selection [1978. “Natural Selection and the Emergence of Mind.” Dialectica, 32:339-355]. But he did that rather conditionally to avoid that it should conflict the impressive epistemology he had so scrupulously developed during his best years. In the paper he said “that natural selection may be so formulated that it is far from tautological”, immediately followed by the remark that “it turns out not to be universally true”. (emphases added by me). And Geof, that is exactly the main point I wished to make clear. Generalized Darwinism is no longer universally true if with each case to be explained we must resort to another ad hoc set of auxiliary assumptions. Moreover you should not forget that I share and appreciate the common idea with Thorbjorn and you that natural selection is not tautological, but – for my part – provided it is so formulated that auxiliary assumptions are avoided unless needed for demarcating the border lines of the evolutionary system considered.
      I still remain anxious to learn your views with regard to the other questions I posed. It is my conviction that we must follow the powerful rigorous rules of selection and probability theory that have so meticulously been developed and disclosed since nearly four centuries. That is where mathematics is for. Boltzmann’s entropy law, Claude Shannon’s theory of entropy, uncertainty and information (both so essential to understand the Darwinian mechanism of selection) are rigorously based on probability theory. We cannot disregard that fundamental piece of consistent mathematical reasoning as of only being helpful and replace it by a verbal ad hoc idea of the evolutionary selection process that is very likely to conflict with it if it is not grounded in it. Well Geof, perhaps the subject is too technical here to deal with it sufficiently, but we shall have to consider it at another occasion then.

      1. Geoffrey M Hodgson Post author

        Len writes “Generalized Darwinism is no longer universally true if with each case to be explained we must resort to another ad hoc set of auxiliary assumptions.” In response, we need to be clear what we mean by “Generalized Darwinism” (GD). For Thorbjoern and I, GD (1) outlines a very general (but not universal) ontology, (2) claims that this applies at an abstract level to both biological and social phenomena, and then (3) claims that variety, selection and information inheritance have to be explained. These are not claims of “universal” truth. Furthermore (3) is an explanatory requirement, rather than a truth claim about the world. The fact that we need auxiliary explanations is valid and important but irrelevant to claims (1), (2) and (3) and their validity. Len seems to think that GD is a grand theory or model that is attempting to explain everything. It is not.

        On Len’s other points, he is usefully pointing to mathematical tools that may be useful in our endeavours to understand evolution. Fine. But let’s focus first on what we mean by GD and what it is, and is not, claiming. Tools and models come later, as Len himself suggests.

        Finally, please note the correct spelling of Geoff.

    2. glens2013

      Our capability to contemplate “WHAT IF?” is a result of the massively parallel processing of associative recall by billions of dynamically, selectively, interconnected neurons in our brain.

      Similarly the biological capability to evolve results from the parallel processing of numerous individuals of a population in their environments.

      We see both processes as “CREATIVE” because they lead to unanticipated results. So much so, that SELECTION is necessary before the results are viable. We have evolved the “Stream-of-Consciousness” capability to become aware of what we need-to-know for practical purposes, selecting by consensus among the massive plurality of associative patterns. Phylogeny of living organisms is what we see from nature. We can describe both processes as a “parallel-to-serial-conversion”!

      We are unaware of most of what goes on in our brain – autonomic control systems, reflexive learned responses, and contemplation etc. Our awareness of THOUGHT has evolved in the context of language communication – from our need to see ourselves as others see us so we can help them understand what we say.

      Speech communication is a very narrow, a few syllables-per-second, serial data channel. We have evolved a very sophisticated capability to express and interpret ideas in the context of our story-telling, cause-and-effect, command-and-response interaction in a community of people.

      We jump to conclusions (parallel processing) but we need to “rationalize” – express them to ourselves in ways that others could understand – before we trust them. We are aware of ideas and have a vocabulary of logic in the context of serial communication. But that is an epigenetic result of the massively-parallel processing in our brain! Our awareness of thinking is what is selected by our stream-of-consciousness filter!

      The concepts ‘idea’ and ‘thing’ are part of our vocabulary to communicate our thoughts and they correspond to realities that ‘exist’, but in very different ways. A thing is material but an idea is a thought process. Our perception, and communication, of both are thought processes! An idea can represent a thing or a process but it’s our own creation with meaning that’s limited to the aspects of reality we choose to express. Reality exists in parallel but we perceive it serially, one-aspect-at-a-time.

      So philosophical confusions are everywhere! Plato and Descartes and Whitehead and Russell and Von Neumann offer clarifications but they are all limited by the serial nature of our awareness-of-thought. They express their insights in a vocabulary that was developed for elocution and political persuasion. And we must avoid jumping to conclusions about what they had in mind. The ‘idea’ of ‘infinity’ ‘exists’, an immaterial ‘thing’. But ‘infinity’ is some-‘thing’ larger than any-‘thing’! Words are ambiguous with meanings that depend on context.


      We must recognize that massively-parallel processing, in biological and social evolution and in our brain functions, has logical capabilities that we can’t expect to comprehend in the context of our serial communication vocabulary. We need to catch up with, and pass by, the millennia of verbal-logic philosophy and become aware of the many logical functions that are such a familiar fait-acompli that we take them for granted. For example unreliable systems can gain reliability through redundancy. And redundancy and unreliability facilitate creativity! Learning becomes practical in a neural net with adjustable interconnections!

      This effort, clarifying philosophical ideas about evolution, should be a step along that road.

      1. Geoffrey M Hodgson Post author

        I can see some connections between the points made by Glen and the themes of Chapter Two, and what Glen says is sensible:
        Comments that stray too far from the chapter in question may be heavily edited or removed.

  3. reasonableadventurer

    Despite the many hours of enjoyment and benefit I have received from Geoff and Thorbjorn’s writings, I am troubled by one aspect of chapter 2. I suspect these troubles emanate more from my own philosophical disposition than anything else, so I will elaborate on this where appropriate.

    Accepting the idea that humans are more adapted to their environs and therefore hold a survival advantage (p. 32) brushes over the ecological reason why this may be so. That is, humans more than most other species have a greater capacity to alter aspects of the selective and ecological environments upon which their survival depends. As a result, they may benefit from environmental selection as well as natural selection. However, the two processes are not the same and should be accounted for separately (see Brandon, 1990; 1996).

    To explain this idea in simply, preference can be given to elevating environmental selection over natural selection as a means for explaining the sustenance activities of firms. That is, rather than assuming that differences in firm fitness are derived from the sorting of firms in common environments, the actual heterogeneity of the environment is highlighted to enable firm adeptness to be viewed as a property-in-an-environment. As social scientists, a challenge we all face is that natural Selection requires the presence of 1) very similar types of entities who, 2) interact directly, in 3) a common environment. This is not always the case, as I will consider below.

    This issue becomes very problematic when we get to the issue of what is being selected and how (p. 35). “Through selection, a set of entities – a population – will gradually adapt in response to the criteria defined by an environmental factor”. Here, the process of niche construction must also be considered. A ‘population’ of beavers do not wait for the process of external selection to act upon those aspects of their individual physiology to produce the right set of such properties in a future population, they aggressively change the nature of environment/beaver interaction by carving out a niche building dams that provide instant protection from external threats.

    Further complicating the issue is determining what ‘in reality’ is a population. Debate exists (see Reiners and Lockwood, 2010) over what are populations, and the extent to which we (as humans) engage in a process of arbitrarily determining what we believe are populations (or species, communities etc.) for the ease of enabling our research to proceed. For example, in my past research, I was investigating the survival of restaurants in Australia and the UK and soon realized that it was not possible to study such entities as members of a population. Either they existed in isolation as the sole entity in a geographic location, or they were distinctly different from each other (i.e. a McDonalds outlet and a local fish and chip shop). In this instance, using the ecological ‘guild’ as the grouping of entities that exists as “a group of species that exploit the same class of environmental resources in a similar way” (Root, 1967, p. 335) made much more sense. It also meant that that the individual characteristics of the guild members needed to be understood to appreciate how they each behaved differently in their efforts to survive, yet collectively altered the environment in ways that may or may not have given a survival benefit to themselves and/or others.

    The point I am trying to make, is that once you lock yourself into an assumption that a) populations must exist if entities exist, and b) natural selection will primarily be responsible for shaping the composition of such a grouping of entities we potentially ignore a raft of other explanations of why individual entities survive. For example, consider Beatty’s (1984: 187) consideration of independent relationship between chance and natural selection.

    “But while Darwin considered the origin of variations a matter of chance, he did not consider the possibility that their evolutionary fates might also be a matter of chance. In particular, he attributed their evolutionary fates to natural selection in or against their favor. Natural selection took over where chance variation left off: those organisms that were by chance better equipped to survive and reproduce actually out survived and out reproduced the organisms that were by chance less well equipped, and thus advantageous variations increased in frequency from generation to generation. In Darwin’s scheme, in other words, chance and natural selection were “consecutive” rather than “alternative” stages of the evolutionary process. With the introduction of the concept of random drift, however, came the notion that the fate of a chance variation might itself be a matter of chance-i.e., the concept that chance and natural selection might be alternative rather than just consecutive stages of the evolutionary process.”

    I think what is missing here is a greater ontological concern for what exists, and subsequently how it can be known. My position here is consistent with autecology, the study of individual entities and their interaction with their environment. This is different from the more common synecology, which focuses upon relationships within and between communities of entities. The challenge I see for the social scientist is that while in the domain of the natural sciences, synecology grew out of autecology (to become the dominant approach), it did so with the knowledge and practices developed in autecology. Alternatively, in the social sciences we have tended to jumped straight into a synecological approach without first having cutting our teeth investigating how individual entities behave in relation to their environs.

    I summary, elevating the notion of a ‘population’ as a level at which judgments can be made about what has and hasn’t been selected introduces two specific problems. First it assumes the presence of a population when in reality no such grouping of identical entities in the same geographic area may exist. Second, it enables it far too easy to commit an ecological fallacy (Babbie, 2005: 102) where assumptions “that something learned about a particular ecological unit [i.e. a population] says something about the individuals making up the unit”.

    While all individual entities can be said to be the products of evolution, they also must also I believe, be afford the respect of creating much of the variation and circumstance that enables the process of evolution to occur in the first place. It is their existence that is the basic phenomenon we should be concerned with; for evolution would not have been possible without it (see Haukioja, 1982).

    Babbie, E. 2005. The Basics of Social Research. Sydney: Thomson Wadsworth.

    Beatty, J. 1984. Chance and Natural Selection. Philosophy of Science, Vol. 51 No. 2, pp. 183-211.

    Brandon, R. N. 1990. Adaptation and Environment. Princeton, NJ: Princeton University Press.

    Brandon, R. N. 1996. Concepts and Methods in Evolutionary Biology: Essays in Evolutionary Biology. New York: Cambridge University Press.

    Haukioja, E. Are Individuals Really Subordinated to Genes? A Theory of Living Entities. Journal of Theoretical Biology. Vol. 99, pp. 357-375.

    Reiners, W. A. and Lockwood, J. A. 2010. Philosophical Foundations for the Practices of Ecology. New York: Cambridge University Press.

    Root, R. B. 1967. The Niche Exploitation Pattern of the Blue-Gray Gnatcatcher. Ecological Monographs, Vol. 37 No. 4, pp. 317-350.

  4. Geoffrey M Hodgson Post author

    Thorbjoern and I use the general term “selection” rather than “natural selection”. We define selection in a later chapter so we should postpone the technical discussion of this. But let me make it clear that we emphasize that the selection environment is important, and that selecion and “fitness” are context dependent. Furthermore, we emphasize differences in environmental contexts and the importance of such phenomena as niches. I agree with Reasonable Adventurer that we leave out many things of importance in Chapter Two. But I would ask him to explain more clearly how the points he is making would contradict our claims in this chapter. If instead he is merely saying that the points he is making need to be taken into account, then that’s fine. Is he claiming that there is no such things as populations of entities?

  5. len wallast

    As to Geoff’s reply at 7.49 a.m. to mine, it is clarifying but not reassuring. It explains what Geoff and Thorbjorn understand by GD. But I think it is incompatible with Popper’s epistemology to which I subscribe: scientific theories are intended to offer general explanations in the sense of being independent of time and location (borders) of the considered domain of observation. One thing to clear up here immediately: I have nothing in common with “universal truths” and “a grand theory of everything”. Science is also an exercise in modesty.

  6. Geoffrey M Hodgson Post author

    The reply to Len on this is simple. GD is not a theory. It is a metatheoretical framework involving ontological claims and explanatory requirements. Put that way, it is entirely compatible with Popper.

    On a general point about science, it must involve some universal claims, as Popper himslef acknowledged. For me, one universal claim is that “every event has a cause” (where strochastic causation is a type of cause).

  7. reasonableadventurer

    Hi Geoff,

    Very much the former, I am suggesting that selection is a highly individualised process that requires many factors of explanation. But I am also saying that one’s philosophical position on these issues can greatly influence the identification of populations and ultimately the directions of causality we assume to occur between populations and individuals. I am not concerned about the presence of populations of entities. Rather my mind is always preoccupied by the simultaneous presence of downward and upward causation through which selection processes occur. I am not really disagreeing, more wanting to add more to the overall explanation.

  8. Geoffrey M Hodgson Post author

    Any differences now with Reasonable Adventurer seem tiny, if not invisible. I would underline that we are going to develop ideas about populations, and key concepts such as selection and inheritance depend on populations for their full meaning. This does not mean that the uniqueness and development of individual entities is unimportant – it is crucial. There is a famous “evo-devo” debate in biology where “evo” emphasises the role of population-level selection and “devo” the role of individual development. We emphasize both.

    Part of the problem with past debates about “evolution” in the social sciences has been the conflation of population-level phenomena with developmental processes. The difference between them has to be clearly understood. And we have to be clear that our ontology involves populations, and not just one developing entity.

  9. perttiaaltonen

    Dear Geoff

    Whereas other approaches fail, Darwinian evolutionary mechanism works well in a world where the external environment keeps changing in an unpredictable way so that agents in a more or less complex population system cannot fully anticipate future opportunities or risks or new designs for survival. For agents able to survive and reproduce for extended periods of time in an environment turning more and more alien they must bet, intentionally or not. Putting bets happens by variation. I would emphasize how Darwin’s approach explains the ability to cope with the unknown future aside with the complex population system; both give warrants to Darwinian explanation for long time spans.

    Higher levels of complexity in the population or in individual agents induce more variation, more diversity in betting. But being more complex does not spell being more competitive or fit for survival in general. Selection does not favor complexity as such. Simple agents tend to survive too. Although we humans now survive with what complex add-ons we have, the really marvelous story is that we survived without them, for aëons. As Gould noticed, collapses and re-emergencies of complex systems occur. It would be important to look at both short term competition and long term survival for deep historical understanding of what happens during different time spans.

    The concept of variation might deserve a chapter of its own. We know from biology is that there is no one-to-one mapping from genes onto phenotypes. There are also constraints in “Bauplan” or “correlations of growth” in variation. Anticipated by Vico, Nietzsche and Foucault in their genealogies, also economic or social structures are exapted to new functions. Following Turing and Gödel, one may ask: what is a code?

    Best regards, Pertti

  10. gmarletto

    Dear Geoff and Thorbjorn,

    I am reading your book in order to answer to the following question:
    “Can GD help in building a genuinely dynamic representation of economic change?”
    (while the representation of economic change delivered by most standard economics is – implicitly or explicitly – static).

    Chapter two of your book allows me to give two positive answers to such question.

    The following relevant features of economic change are consistent with a Darwinian metatheoretical framework (in brackets are the relevant features of a standard representation of economic change):
    – Non-ergodicity (ergodicity)
    – Strong uncertainty (perfect knowledge)
    – Bounded rationality (full rationality)
    – Adaptation (optimization)
    – Non-reversibility / historical time (reversibility / logical time)
    – Out-of-equilibrium dynamics (equilibrium comparative statics)

    The main Darwinian principles of variation, selection and replication are necessary to represent economic change.
    This assertion is the result of a more and more articulated research program which draws on evolutionary and institutional economics. This program started with the seminal works of Schumpeter on innovation and Veblen on institutions and was recently revived by the contributions of Nelson and Winter and (last) Douglas North. The current effort is to integrate the representation of techno-economic and institutional changes into a single Darwinian dynamics (Dosi, Nelson). The co-evolution of socio-technical systems (that is networks of – not only economic – organizations) and political institutions will probably be the next milestone of this research program (van den Bergh, Gowdy, Marletto?).

    Best wishes,

    1. len wallast

      Support for Gerardo’s list of features.
      The features listed by Geraldo are indeed very much the clues and guidelines to work out a general theory of evolution. I am glad that they are here stated so explicitly. The list of features I have been using to elaborate on a general theory of evolution (Wallast, 2013) is even longer and in some respects somewhat different. It may be worthwhile to list them here also for closer contemplation:
      -The evolutionary system is ‘Indeterministic’, which implies that statistical selection occurs at a micro level of events, like in statistical mechanics (Boltzmann) and in communication theory (Shannon).
      – The principles of ‘Continuity’ and ‘(physical) Homogeneity’ apply . Don’t confuse physical homogeneity here with economic homogeneity.
      – Selection reduces ‘Uncertainty’ (in the Shannon sense) so that selecting agents get their information simply (and solely) by the events of exchange (transactions).
      – The concept of rationality should be replaced by ‘Intentionality’ or ‘Purposefulness’ of selection, in accordance with Shannon’s existence theorem. Other forms of rationality (such as optimization behavior) do not play a role as they do not reduce Shannion uncertainty. So we arrive from another angle of incidence than the Popperian argument of tautology at the conclusion that maximization of fitness must be excluded as one of the tools by which we can explain the general process of evolution. (There is still another argument which I will keep in reserve for the moment).
      – ‘Adaptation’ is the process of reordering (rearranging) the order of input and the order of output after selecting input variations and output variations from/into the reservoirs of existing knowledge (information). The valuation of reordering is according to Shannon’s entropy formulas. The reordering will result in a net growth of entropy after the reservoirs have been readjusted for the net value of reordered output over input. This is all there is to explain net growth in a quantitative manner.
      – ‘Non-reversibility’ is a concept reducible to the property of entropy as manifest in the second law of thermodynamics. If we adopt the view that selection is concerned with the selection of Shannion entropy, this character of non-reversibility will generally apply without excursions to the theory of thermodynamics.
      – I know that equilibrium and disequilibrium are given much and even central significance in the dispute with neoclassicism and microeconomics, and also in complexity analysis and econophysics, but to discern between these opposing concepts is to no avail in a general theory of statistical selection. We do not need it at all. On the other hand ‘Time’ is nothing else than the successive order of events of selection. Both at the micro scale of selection and at the macro scale variables depend on time. Variables, parameters etcetera that do not depend on time do not exist within the evolutionary context. ‘Dynamics’ is a core concept for understanding the process of evolution. This is also why we need to define the time-interval of selection (See my comment in Ch.1).
      Moreover , the derived laws from the three Darwinian premises should explain (or consistently define) at the macro level:
      – The role of ‘Prices’, ‘Money’, the ‘Business cycle’, ‘Employment’ and ‘Productivity’ in economics
      – What has money to do with the Shannion concept of ‘Entropy Transmission’?
      – Why it is that input and output ‘must interact’ (i.e. must be statistically dependent) to furnish growth?
      – Why it is that the propensity to consume is usually around 70% in modern growth economies?
      – Why it is that a ‘Keynesian emphasis on investment’ will often and eventually stimulate economic growth?
      – What are the conditions to attain a large rate of positive growth and how sustainable is that?
      – Why it is that continued deflation is detrimental to economic growth and will eventually result in ruination?
      – Whether the begin (the alpha point) and the end (the omega point) of evolution are sufficiently ‘Unstable’ to allow evolution aptly to begin and just as well aptly to end. In other worlds, the fact that we see evolution all around us confirms that evolution is very likely to occur. Can we explain that? If the attainment of an alpha point is unlikely, we cannot explain what we see.

  11. Geoffrey M Hodgson Post author

    Gerardo, I think that your remarks are very much on target. You have outlined – in an upbeat fashion – the promise of our research program. But it is important now to deliver. Thorbjoern and I have outlined the basic approach: it involves a major shift of paradigm. The shift will not take place in the scientific community unless we get a multitude of studies, at all levels of analysis, down to the most mundane and empirical of applications, and detailed modelling where appropriate. There is a long (but very exciting) road ahead.

    The modern wave of “evolutionary economics”, prompted by the great work of Nelson and Winter (1982), is now over 30 years old. It is reasonable to ask if its core theory is sufficiently well-defined and scientifically innovative. If there is a deficit, can generalzed Darwinism help fill the gaps?

    1. gmarletto

      Geoff, one more point on the “road ahead”.

      Many scholars of the so-called ‘socio-technical’ approach (e.g. Frank Geels) use the Giddens’ notion of ‘structured action’ in order to represent the interaction between (social) agents and (social) structure: the structure enable and constrain agents’ behavior, that in turn influence the change in the structure.

      I think than Darwinian co-evolution would be a more powerful heuristics than ‘structured action’: the variation-selection-replication of agents (or – more exactly – of their routines) also depends on the variation-selection-replication of structural constituents (whatever they are – material or immaterial), and vice versa.

      I would also add that agents can be individual, groups and groups of groups; that is, a multilevel Darwinian framework is needed to represent social evolution.

      Inter alia, this implies that GD is able to accommodate both competitive (between individual and groups) and cooperative (within groups) forces. Thus – for once – both orthodox and heterodox economists can not complain…

      1. schmid2013

        The idea of structure enabling and constraining social agents makes sense to me and is similar to John R. Commons’ definition of an institution.

  12. Melissa Dennison

    Hi all, just catching up here. A very lively dialogue here! So to make an attempt at question 1) is social reality a “complex population system?” , well why wouldn’t it be? I guess that is my first thought. Do we need to define what we mean by a complex population system? Anyway for me the social world represents an evolving processual system, to use Whitehead’s words from 1929 ‘nature is a process’, a process of continual change.So for me questions around Generalized Darwinism relate to ways of understanding this process. I have a quote from a paper by Van der Ven and Poole which I think has relevance: “An organization, recognising the need for a strategic decision, starts in the identification phase and , once it has diagnosed its situation and the problems it faces, it then moves into a development phase, in which it searches for solutions and adapts them to its situation, following which it moves into a selection phase, in which it screens options, selects its best cause of action, and authorizes the organisation to proceed…an organization is always something in some particular state or phase of a process”. I thought this was pertinent as it discusses something of significance ie selection. How do organisations select? How does selection occur in a social system? Is this via communication between entities perhaps? as well as via decision making? Could it be partly an interaction between the internal (ie the mind) and the external (ie out there, ie ‘reality’ if you want to call it that). Could there be feedbacks between the two that are fundamental to the ongoing process of change. Could complex population systems be seen in some way as cognitive or thinking systems in the social world?

  13. Geoffrey M Hodgson Post author

    There is a lot to comment on here! So I’ll be very selective.

    First: “How do organizations select?” Organizations are primarily objects of selection (“selection of”) rather than agents of selection. When they recruit individuals they could be part of a selection process (of individuals), and thus possible agents of selection. Bankruptcy is part of a selection process of organizations in a population – in which case the organization itself is not necessarily an intentional agent leading to the outcome.

    Second: “Could complex population systems be seen in some way as cognitive or thinking systems in the social world?” The big (and very philosophical) problem here is bridging ideal and materail worlds in one (monist) social ontology. Many accounts of the “meme” (as ideas) fall down at this hurdle. Thorbjoern and I adopt the solution of the philosophical pragmatists to this problem, which they reconciled with a Darwinian evolutionary view. In this approach, ideas are expressions or emergent properties of material relations. So when we come to define the replicator, for example, we regard it as a materially-grounded informational mechanism.

    1. Melissa Dennison

      Hi. Amused by your reply where you state you will be ‘selective’. ha! So, you believe that selection acts at the macro level on populations of organizations? Ok, so how does this selection operate at this level? Is it not possible that habits and routines, behaviours etc within organisations have some influence on whether an organisation is ‘selected’ or not? I find your suggestion that ideas are expressions or emergent properties of material relations interesting. Particularly since I am reading about emergence, to Goldstein emergence is defined as something that arises as a result of novel coherent structures or patterns in the process of self organization within complex systems. These are seen as being dynamic and arise as the system evolves over time. Interestingly Goldstein argues that emergence is also on the macro level. Which is maybe in contrast with your view of this. I am not sure if emergence occurs at the micro or the macro level (or both), but it is an interesting concept.

      Another thought: do social systems reproduce themselves via language and communication? For me social systems differ from biological systems in that they are artificially constructed, or constructed by us, socially constructed!. There is a large amount of subjectivity involved, interpretation and generation of meaning. Internal realities influence external realities and vice versa. So to some extent social entities are conscious entities, as they are comprised of thinking individuals. How these thoughts result in actions, or how these actions relate to the environment surely has an impact on how the organisation varies and on it’s selection?

  14. reasonableadventurer

    It is useful to note that organizations (and the individuals within) are not only the objects of selection, but they also impact upon many environmental factors that relate directly to the selection process, i.e. the process of niche construction. This is one of the many issues that make so selection in the social world a very complex process.

    1. Geoffrey M Hodgson Post author

      Yes, of course, while organizations (and individuals) are objects of selection, they are often more than that. The term “interactor” is applied to these objects of selection, and it also connotes intercation with others and their environment. As Reasonable Adventurer rightly says “they also impact upon many environmental factors that relate directly to the selection process, i.e. the process of niche construction”. But such interactions are not unique to the social world – they are plentiful in biological evolution too.

  15. Adrià Aldomà Sió

    In the previous posts very interesting points have been raised, I will make some comments trying to stick with the topic of the chapter. (I guess topics about selection will be discussed later on). I apologize for the length of the commentary; I hope it will be interesting at least.

    1) From a theoretical point I would say yes, social reality is a complex population system. But reality is much more complex. The problem here is to pinpoint the entities of such population (in biology they have the same problem as the replicator-gene view is just a simplification). Ultimately there isn’t any entity that can be totally isolated as everything is related.

    But general Darwinism, especially in the specific case of such a complex stage, doesn’t need any clear entity. It only needs information (that can have any kind of content and physical form) to be inherited or copied with enough fidelity (plus variation and selection).

    As Geoff has previously mentioned populations of entities can also become emergent entities themselves (thus opening the door to multilevel selection) those emergent entities (such as institutions or language) have certain cohesive nature but I don’t think they can be regarded as entities in the sense that critics (specially those coming from biology) expect.

    The need for entities is usually assumed because in simple systems the mechanism that is in charge of copying the information needs simple inputs (thus recognizable entities) in order to perform correctly (such as the relatively simple process of gene transcription, or the Darwinian evolution of certain computer programs). If the entities doing the copying are humans (or networks of humans such as organizations) the complexity in the information and the nature of the copying process is so high that it is impossible in many cases to find such entities; although we have the certainty that information is being transmitted or copied.

    Generalized Darwinism, as any piece of knowledge, in interested in simplifying things and thus assumes the existence of entities. But then it is open to criticism when those entities are hard to find. As I tried to show it is possible to avoid such assumption but it makes argumentation more complex and difficult to grasp.

    3) I see your treatment of generalized Darwinism as being a <> very clever (as you say it enables the rebuttal of many criticisms such as lamarkianism). Generalized Darwinism materializes in many different circumstances (i.e. sexual vs. asexual organisms) and thus it needs additional theorizing to account for the specific circumstances.

    There is little I could add to this. The evolutionary process tends to get more complex with time, as there is a “meta-evolution” going on which improves the performance (capacity to innovate and adapt) of the evolutionary process (sexuality is an example of such). As I already argued, in human societies this complexity is so high that makes the simple principles of generalized Darwinism difficult to pinpoint. I hope your approach will also help to counter criticisms on this point.

    To end up, your comments in the end of the chapter about the reluctance of social sciences to accept Darwinian explanations are sadly of great importance. Many different explanations could be given to this fact. You mentioned reluctance on explanations and theories coming from biology; in the particular field of economics there is a great resistance to accept the flaws of the established neo-classical paradigm, especially if we take into account that our economic system is based on such “theories”. There is also a lot of resistance to any theory that poses an alternative (or parallel) explanation to individual freedom and intentionality being the creator of the social world.

    I would even mention another possible explanation; many academics of the social sciences cannot accept that such a simple thing as Darwinian evolution (which basically consists of three steps) could explain (in a general manner) the incredible complexity of human societies. It is so simple that it seems wrong.

    For these reasons I cannot avoid being pessimistic, although not evident on the surface, discussions in the social sciences are more determined by preconceived ideas, personal and organizational interests, and reluctance to change than by open and sincere rational discussions. Fortunately, now and then evidence triumphs over entrenchment.

  16. sinewtin

    If our blog community is the “complex population system”, it looks that interaction with environment, including specifically postal service, which is not fast with shipping me a book copy, forces me to face the problem of scarce resources, to understand finally how Darwinian approach fits the case.
    But based on my limited knowledge of just 1st chapter and following two discussions, I still have an impression that relational essence of social entities is missing with the use of the word “population”. Using it I think of entities as taken for granted, whereas social relations are in the constant process of reproduction through their activities or practicies. So can anyone help my mind to bridge this gap?

  17. danielcourgeau

    I would like to come back to Reasonable Adventurer question about what ‘in reality’ is a population (or species, communities, etc.). For Darwin the term species is ‘arbitrarily given for the sake of convenience to a set of individuals’, and ‘all true classification is genealogical’. So that such an explanation seems to me to be in agreement with the replication of habits, customs, rules and routines which you put for inheritance in social evolution. Those habits, etc. are common to a set of individuals, which will constitute a group of ‘statistical individuals’ on which we can constitute a scientific social discipline, as it is not possible to work on a single individual, while we observe only one realisation (the individual trajectory), leading to a non identifiable risk structure. However, the difference with a genealogy, where an individual has a unique genealogy, is that the same individual may be simultaneously a member of different groups. How do you solve this problem?

  18. John Cantwell

    Thankfully, and rewardingly, I found in Chapter 2 some of the answers to questions I had raised in relation to the Preface and Chapter 1. I suppose this is a good sign, and makes understandable the potential frustrations of the authors in following the commentaries of participants on this blog, knowing that these commentators may often merely be anticipating what is to come later in the book. In particular, in Chapter 2 I find a coherent defence of a generalized meta-framework for the analysis of evolution in complex natural and social systems when treated in common.

    This said, and despite the undoubted quality of this defence, I do still suspect that there are differences in the workings of the selection mechanisms between natural and social systems even at this level of abstraction, and not just in terms of the more detailed development of the argument in specific applications of the framework. The chapter briefly acknowledges some potential criticisms of GD, and I would not wish to side with those critics who apparently deny the existence of selection in the social world, or indeed the significance of all of the three elements of variation, inheritance or replication and selection (V-I-S). Rather, I think it may be helpful to distinguish at a general level between Darwinian natural selection and the processes that we might term as social selection in the workings of V-I-S.

    The concerns I have are somewhat related to the issues raised in Chapters 1 and 2 about the typical speed of evolutionary change, and the role of human intentionality, but neither of these really get to the crux of how I see the difference between natural selection and social selection. In natural selection, as reflected in a number of the arguments advanced in this chapter, entities adapt in various ways to their environment. The environment is not fixed or static, but adaptation requires that the environment changes more slowly than the characteristics of the entities which we are studying within that environment, and that the environment can be treated as functioning largely independently of those entities. Otherwise, the entities would never have long enough to adapt before the environment changed, and hence insufficient time to enhance their degree of fitness to that environment. On the condition that the environment essentially functions independently of the entities under investigation, it follows that survival depends on the munificence of relevant useful resources that can be found naturally in that external environment; and on the condition that the relevant features of the environment tend to change more slowly than the relevant characteristics of these entities, it follows that survival also depends on their own capacity through adaptation to successfully capture or appropriate from the environment the resources which meet their needs.

    In my view, neither of these conditions generally carries across from the natural world to the social world. Humans differ from other animals as producers, and hence human social systems are systems of production. Only in very early societies did humans mainly just react to an independently given environment, and consume by taking resources that were scarce in that environment through the means of hunting and gathering, thereby competing for the use of these scarce resources with other species and with one another. In later societies systems of production set out to deliberately change the environment in various ways (this is the intentionality part), and – this is the critical element – to create new resources not previously found in that environment, in a (more) useful form. Or, strictly speaking, social systems of production process and combine resources taken from the environment and previously created resources in ways that generate new resources for human consumption and investment. As Sraffa summarized the Ricardian or classical political economy perspective, a capitalist economy entails the production of commodities by means of commodities (the social replication aspect of the story is evident from this formulation too). The major classical theme was that such societies have the capacity to progress where the created resources include a surplus, over and above the immediate consumption needs of the human producers.

    So to return to the V-I-S framework, in social selection the process at work essentially occurs in the opposite direction to natural selection: the environment is transformed and adapted in response to the V-I-S of social structures. Human progress over the long term might be described – was described by Hegel and Marx – as a steady increase in the extent of human or social control over the natural environment. In today’s society, those of a ‘green’ disposition warn that this process may be going too far, the natural environment may be degraded or destroyed by social organization in ways that are ultimately counter-productive. Yet as in discussions over climate change, in principle at least it is possible for social organization to also act responsibly in its effects on the natural environment.

    As observed in Chapter 2, in natural selection scarcity plays a central role in the workings of V-I-S, since entities compete for access to and the use of resources that are scarce in their external or received environment. Conversely, in social selection, entities (social structures or organizations) compete in terms of their differentiated capabilities to generate new resources, and especially in terms of their capabilities to generate surplus resources which enable these social entities to expand the range of their activities and raise the living standards of their members. I.e. social selection depends on the capacity of entities to change their environment in favourable ways, rather than to adapt or respond to it. In this social context the scarcity of resources in the environment is rather besides the point, unless we imagine a hypothetical world in which everything we might want materially is available to us in an immediately usable form and at a low or zero cost at the press of a button, but I am sure we can agree that these kinds of contrasts with never-existing states of the world are unhelpful. I noticed with interest your attempt to rescue the centrality of scarcity for the analysis of social systems by distinguishing local and immediate scarcity (for specific organizations) from generalized scarcity. While of course the existence of such scarcity in the social world is beyond dispute, I am not convinced that such scarcity plays the same central role in social selection that it does in natural selection.

    I share your distaste of the orthodoxy in our mother discipline of Economics. But I fear that your focus on the scarcity of resources in the environment represents a residue of neoclassical thinking in your representation of the V-I-S framework, as reflected in your reference to Robbins’ interpretation of classical economic thought in footnote 3 in the chapter. Instead, in evolutionary approaches to innovation and growth in the classical tradition, competition or social selection occurs primarily on the basis of varying capabilities to create and expand resources over time through production, rather than on the ability to appropriate or exploit more intensively some fixed or exogenously given scarce resource base as a facet of the received external environment.

    One inference I would draw from the difference I see in the context of the operation of the V-I-S framework in the evolutionary processes of social vs. natural selection is that we can examine the relative success of different societies or social structures in terms of their capacity to transform the social and material environment of their day. Hence, e.g. we can describe the rise and fall of ancient Rome largely with reference to the changing internal dynamism or progressiveness of Roman society in its different phases of development or social evolution. When Rome was advancing, with very high V and strong I by the standards of the day, it was easily able to dominate other contemporary societies (social S), and thereby to transform their environment as well as its own. The most fiercesome warriors were no match for the technology and organization of the Roman legions. But once Rome fell into decline and its V and I mechanisms were undermined, barbarian forces were soon at the gate, ready to strip this society of its accumulated wealth and resources (de-selection). The effects of the V-I-S mechanisms of Rome on the environment of the societies of the time were humourously illustrated in the Monty Python sketch from the Life of Brian, ‘What have the Romans done for us?’ – ‘apart from the sanitation, the medicine, education, wine, public order, irrigation, roads, a fresh water system, and public health ….’!

  19. Geoffrey M Hodgson

    John makes a number of important points, but I confine my responses to two issues. The first is scarcity. Our concept of scarcity is different from Lionel Robbins, who writes of globally “limited” resources. By contrast, our concept of scarcity is local and immediate. Even an unlimited resource (like skill) can be locally and immediately scarce, in the sense that local time and effort are required to obtain it. There is surprisingly little dissection of the scarcity concept in neoclassical economics. We choose a specific concept of scarcity that differs from the one proposed by Robbins.

    Second, John raises possible differences between social and economic evolution. We acknowledge several in our book, including differences of evolutionary speed and interaction with the environment. The question is whether these can all fit within a Darwinian framework. Biological evolution is not all slow. Biological evolution also includes many cases where organisms create their own environments – niche construction, beavers with dams, etc.. Sexual selection involves strong, intra-species, interactive feedback loops. Darwinism can deal with all these, so the barriers to transferring general Darwinian concepts to social evolution do not seem insurmountable.

  20. Melissa Dennison

    Does the notion that biological evolution is slower than social evolution actually matter that much? Is the time issue really so significant? Whilst Darwin did promote the concept of gradualism, SJ Gould proposed the notion of punctuated equilibrium whereby there were periods of slow change and then periods of rapid change, (as found in some fossil assemblages). An interesting question might be does society and social organisations evolve slowly over time or more rapidly? As for the environment, well it is changing all the time, if you look back over climate history, even in the last 1000 years it has changed, as in medieval times there was a period of warming and then the climate cooled into the little ice age etc, It doesn’t stand still. The concept of gradualism maybe just one interpretation of the evidence from geological history. For example, after the last ice age the climate changed so quickly that within about 50 years the Mammoth became extinct. Therefore, how stable is the external environment and how quickly can it change? The social world is definitely not a static place, it is a dynamic environment and constantly changing (yes! evolving). Perturbations (maybe you could define these as emergent properties?) can be a beneficial thing as they enable new entities to come along. But I digress.

    Before I go, I have a question pertaining to an earlier reply: “Second: “Could complex population systems be seen in some way as cognitive or thinking systems in the social world?” The big (and very philosophical) problem here is bridging ideal and materail worlds in one (monist) social ontology. Many accounts of the “meme” (as ideas) fall down at this hurdle. Thorbjoern and I adopt the solution of the philosophical pragmatists to this problem, which they reconciled with a Darwinian evolutionary view. In this approach, ideas are expressions or emergent properties of material relations. So when we come to define the replicator, for example, we regard it as a materially-grounded informational mechanism.” My question being why is there a problem in bridging the ideal and the material worlds? Also do we need a single social ontology as such? You may think these are idiotic questions to ask, but social evolution is influenced by both ideal and material factors. The within (the subjective) and the without (objective reality). There is an interplay between these, so perhaps trying to bridge this ontological gap is worth considering?????

  21. Geoffrey M Hodgson

    Melissa asks “why is there a problem in bridging the ideal and the material worlds? Also do we need a single social ontology as such? You may think these are idiotic questions to ask, but social evolution is influenced by both ideal and material factors.”

    This is a very good (and definitely not an idiotic) question. Ontologies that completely separate the ideal from the material are known as dualist. And ontologies that see either matter or ideas as the fundamental stuff of reality are described as monist. Darwin was a materialist monist, and it seems to me that this is the most prominent position among philosophers today. It is found in the works of Mario Bunge or John Searle, for example. The problem for the materialist monist is to establish a position where ideas are still real, and make a difference. Bunge and Searle use concepts such as emergence to establish the reality of ideas, but argue that they are grounded upon, and are expressions of, material reality.

    Among the problems with dualism are the following. First, dualism cannot deal with the problem of describing how ideal and material reality interact. If ideas are “forces” but in a completely separate ontological sphere, how can they impinge upon material reality, which is separated from the world of ideas? Second, from the perspective of the evolution of species, how did an entirely new ontological sphere come into being when ideas first appeared in humans (or in primates or any other species for that matter)?

    Dualism was much more popular (even in biology) from about 1880 to 1950. It was while grappling with these problems that the philosophical pragmatists developed Darwinian ideas and key concepts such as habit (which help to bridge the realm of ideas with neurological reality, and from an evolutionary perspective). See the writings of William James, John Dewey and Thorstein Veblen, for example. Ulrich Witt has (rightly in my view) criticized much of social science and even “evolutionary economics” for being (at least potentially) dualist, by separating discussions of ideas or culture, on the one hand, from their evolved genetic and physiological foundation.

  22. Melissa Dennison

    Hi all

    Thanks to Geoff for providing a very good reply to what I thought was a silly question! I have scribbled this down and will go peruse the writings suggested. I do question dualistic thinking, as for me reality is both material and idealist, there is a reality out there but in the case of social reality I believe our thoughts and ideas shape this to possibly a large extent. So for me social evolution results from the interaction of human minds (individually) and collectively via the expression of thought (language) and external reality. It is problematic for me to distinguish between the outside and the within.

    I was thinking about evolution today and decided to rummage for an old text book by Futuyma (2005) which states that Darwinian natural selection is encapsulated by selection amongst individuals rather than selection amongst populations. As the good of the species has no affect on the course of selection amongst individuals. Selection among populations which is the only possible cause of evolution of a trait that is harmful to the individual whilst being beneficial to the population is a weaker force than selection among individuals. Why? Well in part, very simply there is a higher turnover of individuals than there is populations, so the rate of replacement is greater. The upshot is very few traits have evolved that benefit the population or species as a whole. Why do I mention this? Well, Darwinism is about populations thinking, but it is about the frequency of certain characteristics within the population and how these change over time. So how the individual interacts with other individuals and it’s environment is surely fundamental. For me one of the things GD needs to do is find a way of marrying up the abstract over-arching theory with what goes on on the ground, in the everyday thoughts, decisions, habits and routines that you find in organisations and society in general. But you may not agree, and all comments are welcome.


  23. Geoffrey M Hodgson

    Melissa writes “one of the things GD needs to do is find a way of marrying up the abstract over-arching theory with what goes on on the ground, in the everyday thoughts, decisions, habits and routines that you find in organisations and society in general”. I fully agree. The claimed value of GD is not what it manages itself to explain, but how it poses questions (like, what replicates?) and helps to organize possible theoretical explanations of particular phenomena.


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s