SUMMARY OF CHAPTER ONE
QUESTIONS AND COMMENTS ARE NOW INVITED ON THIS CHAPTER
The opening chapter of the book discussed the relationship between evolutionary biology and the social sciences and its history. A distinction must be drawn between “biological reductionism” (attempting to explain social phenomena entirely in terms of genes) and the application of abstract evolutionary principles to socio-economic change. Darwin himself (who knew nothing about genes) conjectured that his principles may apply to the evolution of entities at the social level. This idea was taken up by a number of writers – including Walter Bagehot and Thorstein Veblen – before the whole idea of using biological ideas in the social sciences became highly unpopular in the early decades of the twentieth century. The application of abstract Darwinian principles to social evolution was revived by Donald T. Campbell (1965) and others after the Second World War. But it was not until the last few years that abstract Darwinian evolutionary principles have been sufficiently refined to develop a theoretical framework that is applicable to socio-economic evolution.
KEY QUESTIONS
1. What is the difference between biological reductionism and the derivation of common principles applicable to both biological and social evolution?
2. Why did the idea of deriving abstract evolutionary principles for the social sciences fall out of favor for so long?
3. What is the difference between analogy and generalization?
Darwin's Conjecture 
When I was a young professor of ecology in the early 1970s our oral exams for PhD students had a mandatory topic “Principles of Ecology.” A favorite question for this topic was “Are there any principles of ecology?” The right answer was “not really.” At any rate there are no exceptionless law like generalizations on a par with the laws of physics. I think of Darwinian “principles” as more like a set of methods than a set of pat principles. Evolution cover a subject matter that is exceedingly complex and highly diverse.
I would agree. Rather than assuming principles can be laws or generalizable theories, we need to comprehend the overwhelming complexity and hetrogeniety by ensuring we focus on processes.
For Darwin’s principle of natural selection to apply to our investigation of social entities we need a few things. We need ‘similar’ entities ‘interacting’ in a ‘common’ environment.
All too often these three things are not present in studies that ascribe outcomes to the process of natural selection. The process of environment selection is more likely to be the culprit; but in the absence of any analysis of the environment, no such consideration could be given.
Reasonable Adventurer has a very good starting point. The key question that he raises is whether there is sufficient ontological communality between the social and the natural world for Darwinian ideas to apply. This question is raised in the second chapter of our book, as well as a discussion of what Darwinian “principles” might mean.
I suppose that most important idea of the Preface is that evolution deals with causality and complexity. But focusing on genes as holders of information we may loose the importance of causality. I like that you mentioned here the need for evolutionary thinkers to escape concepts of human intentionality and social structure as given elements. Anyway I expect (looking at earlier posts) ontological discussion to be the most promising.
The question that I raise on the Preface is whether the authors equate Darwinism with evolutionism or not?
Mikhail’s comment raises a number of important issues. First, “Darwin’s Conjecture” does not focus on genes. Genetic evolution is important. But the “conjecture” in the book is precisely that there are other “holders of information”, in particular at social levels, in addition to genes. We call these “social replicators” and claim that they exist in addition to genetic replicators. Of course they are very different and the detailed mechanisms of replication are dissimilar.
In answer to the question, we do not equate evolutionism with Darwinism. Darwin rarely used the word evolution. Evolutiom is a very vague and broad word. Darwinism refers to a specific type of evolutionary process that applies to a world where there are populations of entities with specific characteristics.
Am I right that authors separate (or at least relatively divide) biological evolution of humans and social evolution as well as causal mechanisms of biological replication (of humans) and causal mechanisms of social replication? But if so, what about the link. I mean not theoretical applicability to both realities, but kind of real interdependency between those two evolutionary processes. How does it work?
Since there is an interaction, but not just similarity of those processes isn’t it the reason for paying basic attention to methods of generalization?
Mikhail is right that both interaction (between genetic and socio-cultural levels) and generalization (establishing common principles that apply to both levels) are involved. Gene-culture interactiob has been investigated extensively by Robert Boyd, William Durham, and Peter Richerson (see his comment on this blog). Their work on this is invaluable. We refer to it in several places in our book (look up their names in the index).
Since evolution is just specific kind of change (i.e. causation), and on the other hand, you speak of just one sort of evolutionary processes, specifying it with focus on inheritance and selection, why you still oftenly use the term evolution, but not for instance selective replication or inherited selection? It looks that “vague and breoad word” appears to be more attractive than others. Even if there is a danger of misinterpretations.
Mikhail is right. We should have defined “evolution” in our book. The definition I would suggest now is that evolution refers either to (1) changes in an entity, or to (2) changes in a population of entities, possibily involving changes in the composition of the population as well as changes in the entities themselves. This is a very broad definition. “Darwinian evolution” would then be a specific version of the second option.
I also see the gene focus as being of little importance, Geoff is right to say that there are many ‘holders of information’, but to often we only move on towards routines and this can be potentially problematic as well. When we try to study the evolution of social structures we need to ensure we can focus upon that which can be observed. A quote from Haukioja (1982: 357-358) hopefully demonstrates my point. “let us imagine that we are descendants of the first organism which ever lived in the world. Our structure deviates radically from that of our ancestor. It is even possible that we do not share a single gene with the hypothetical organism. Here we instinctively notice change, evolution. The other side of the problem, what is the entity which changed, is easily pushed into the background; but this ‘existence of life’ is the basic phenomenon. Evolution would have not been possible without it”.
He goes on to say “Hence, it is not possible to find the surviving aspect from the structures but from functions which these structures make possible”.
I take from this that our focus needs to be upon that which is ‘life’ (or longevity). He goes on to say “Organisms are successful if they maintain themselves at the moment of evaluation”. I take from this that generalized Darwinism can be applied to the study of ‘functions’ in society, something that sits between the routines that give rise to the functions and the various forms of interactors that ultimately connect the entitiy to its environment. In a sense I am suggesting that if we can discern between an entity’s social structures and the functions performed, we can potentially consider exactly which aspects of an entity relate to external selection process.
That is, there are a myriad of ecological processes, all occuring in complex and unpredicatable ways that make evolution possible. Geoff and Thorbjørn say at the conclusion of chapter 1 (p. 29), “in the social sciences, innovation and novelty are amongst the most difficult phenomena to explain”. For the same reasons, I see the application of Generalized Darwinism as always facing the challenge of explaining what has evolved. So while selection can act at the level of the replicator or interactor, evolution may also be seen to have occured at the level of what functions have survived in society. Sorry if this has got a little off track.
Here are some short responses to Reasonable Adventurer. Genes are important. “Darwin’s Conjecture” focuses on other (social) levels. But this does not mean that genetic evolution can be ignored. To say “selection can act at the level of the replicator or interactor” is slightly misleading. Interactors are the OBJECTS of selection (we call this “selection of” interactors). The changing pool of replicators in a population is an OUTCOME of selection (we call this “selection for”). These ideas are visited in chapters four and five.
The charm and power of Generalized) Darwinism as initiated by Hodgson and Knudsen is the restriction to and simplicity of the premises (variation, selection and inheritance). All further elaboration and derivation should be based solely on these premises. That can be done as I have demonstrated. I think that by introducing auxiliary assumptions (See p. 5 lines 5-7, p. 23 lines 10-13) you change the rules of the game and you spoil the general nature of Generalized Darwinism. Then we fall in the same methodological trap that kept Ptolemaic astronomy captive for about 1300 years and mainstream economics now for about 3/4 of a century.
Len raises an interesting and important point. But I would suggest that the Darwinian mode of explanation generally requires auxiliary assumptions. On pages 4-5 we raise the following example. Darwin’s core theory (variation, selection, inheritance) cannot entirely account for the fact that some birds have colorful plumage and others are grey or brown. Some auxiliary explanations are required to explain these divergent outcomes. Darwin himself pointed to these, and some are special cases of his general principle of selection. Bright plumage is explained by the specific mechanism of sexual selection. Duller plumage is explained by the specific advantages of camouflage and the avoidance of predators. Darwinism relies on specific auxiliary theories or special cases in order to complete the explanation of the phenomena in question. This would apply to social evolution as well. This is one reason why Darwinian theory is quite different from theories in some other sciences, notably physics. Darwinism addresses highly complex and diverse phenomena and is not a simple axiomatic system. This does not mean that mathematical models cannot be used. They are useful in biology. But we shall never be able to model evolution in the manner that we can model the motion of the planets.
Evolution is the result of selective interaction of phenotype and habitat (which I call the evolutionary system). Both evolve. There are an enormous number of those interacting evolutionary systems. For the explanation of a particular system we need of course to demarcate the system of evolutionary explanation and look for the conditions and changes that demarcate this particular interaction domain of the individual and its environment in the course of time. If you consider this process of demarcation as a search for additional assumptions, ok.
But I am not pointing at assumptions that demarcate the particular evolutionary system of investigation. I am concerned with the volume of evolution and the growth of a specific evolutionary system, e.g. the Dutch economy and its inputs and outputs in the course of time. To explain that we have already cautiously demarcated our subject of study. Further assumptions other than demarcating the applicable phenotype and its habitat should be avoided, as they will degeneralize the Darwinian premises by which the system’s evolution is to be explained consistently.
Len, your suggestion of the Dutch economy as a specific evolutionary system, is that too broad a focus to contemplate the process of evolution? For instance, what would be the interactors? and how might you conceive the parameters of the environment in such a way as to investigate it with any precision?
In response to Len’s latest, I’d like to make some points about terminology. First, “evolution”, “evolutionary” and “evolve” are vague words, requiring definition. They precede Darwin and there is no strong reason for giving them a restrictive, Darwinian meaning. Second, the terms “phenotype” and “genotype” are typically associated with biology. Following David Hull, Robert Brandon and others we use the more general terms “interactor” and “replicator”, especially in regard to Darwinian processes in the social domain. Hence the interactor-replicator distinction is a generalization of the phenotype-genotype distinction, which is found in biology. But Len also makes a substantitive point to which others may wish to respond.
General theories must be stated as broad as we can. This would require that Darwin’s selection theory and concepts must be stated within the mathematically advanced general context of probability theory. This evokes necessarily questions like: what are the statistical experiments of selection, the sample spaces into and from which is selected, the stochastic events of selection, the units of selection (the replicators), the different states of selection (the interactors) required for selection (there must at least be 2, but there may be more), the probabilities of selection? I suggest that in the particular context of an economy the statistical experiments of selection allocate value to and withdraw value from sources (interactors) that reside in different states. E.g. spending for consumption or alternatively spending for investment is an example of a binary selection experiment I have considered in this regard. This interpretation sees the consumption sector as a state (the interactor #1) from/into which value is selected and the investment sector as an alternative environmental state (the interactor #2) from and into which value is selected. This can be worked out in a very general manner without introducing parameters. Much more can be said, but I think this will anticipate too much of the discussion of chapters ahead.
Against Principles II: The toolkit of models theory of theory.
In a previous post, I briefly criticized thinking of science as a quest for principles. Here I lay out an alternative conception that has a foundation in the fallibilist epistemology introduced by Popper and Campbell.
I don’t think the replicator concept works for cultural variation. It is easy to come up with inheritance schemes that do not involve literal replication. Classically, population genetics models come in two forms, those assuming replicators and those that treat inheritance as the transmission of quantitative characters. In the latter, offspring inherit their parental midpoint plus some random error. Nothing is literally replicated. We have always assumed that quantitative characters are “really” underlain by discrete characters, gene replicators. But of course, if enough genes each with small effects on phenotype underlay a trait, then you get to quantitative characters by finite approximation, as in the digital computer algorithms that mimic continuous time by making the finite time step sufficiently small. At this point, the “fact” that genes are ultimately replicators becomes a philosophical point of scant practical importance, which is what drove plant and animal breeders to quantitative character models long ago. I put “fact” in scare quotes because recent developments in genetics suggest that classic conceptions of the gene are gross oversimplifications. Epigenetic processes associated with developmental circuits and gene regulation include processes that can regulate gene expression quantitatively and at least some some of these factors can be transmitted across generations. The simple picture given by some classical mendelian loci of one gene one phenotype, complicated a bit by diploidy and dominance, like human eye color, is the limiting simple case. It certainly can’t be taken as a general model of how genes influence phenotype. When you take into account the peculiarities of culture the general utility of the replicator model is even more dubious. You can have a look at this paper on my web page for the case against the replicator concept for cultural traits: J. Henrich, R. Boyd and P.J. Richerson, Five misunderstandings about cultural evolution, Human Nature 19: 119-137. 2008.
I think that the complexity and diversity of evolutionary phenomena drive you to give up the quest for principles like the replicator-interactor concept. The alternative is what I call the the toolkit of models theory of theory. This idea goes back to a classic paper by the theoretical ecologist Richard Levins and has been developed by the philosopher Bill Wimsatt. It has eventually become, I think, the dominant concept of theory in ecology. See for example the statistics textbook of Burnham and Anderson. B&A’s starting point is that the dimensionality of evolutionary and ecological problems is tantamount to infinite. At the same time, the amount of data we have to evaluate any given problem is very much short of infinite, usually very, very, very much short. As the power of computers to fit models has gone up, it has become clear that it is often easy to “overfit” the data we have. If you make your models too complicated, you in effect fit the noise in the data and obscure whatever truth is in the data. B&A advocate using an information theoretic approach to data analysis which aims to extract all the data possible from your data. Your ability to do this is utterly dependent on choosing the right model or models to test. Hence you try to have a toolkit of models that have reasonably well fit similar problems in the past and have good theoretical motivation. You try fitting these candidate models to the data an see which fit the best. More than one may fit equally well. My wife has a manuscript in which it seems that four models derived from quite different theoretical lines of reasoning fit equally well. See Charles Efferson’s and my attempt to explain these and related concepts in a few words. Stay tuned for my colleague Richard McElreath’s forthcoming textbook written from a Bayesian perspective. If you write him you might talk him out of a look at the current draft.
In the Levins-Wimsatt conception, “principles” morph into “general models.” General models are ones that give you a high level abstract feel for the behavior of a whole class of problems even though they may not fit any particular case very well. Thus, the replicator-interactor and quantitative character models are both useful general models that capture different features of inheritance systems including genes and culture. You can think of others, but these are the most frequently used because they are mathematically tractable and have a sound theoretical basis. Indeed, because diploidy makes replicator-interactor models rather cumbersome we usually use haploid one-replicator, one interactor models of individuals in models of culture as, we have no reason to assume that culture has the analog of diploidy. General models are almost guaranteed to break down in most real world cases. Even so, in the interests of keeping dimensionality low, we often treat them as “good enough” if we want to complexify our model in other dimensions, for example in imagining that genes and culture are interacting or if we want to consider complexities in the phenotype-environment interaction you need to keep your model of inheritance simple. Again, when fitting models to data as well as when trying to do theory, we often use models that we know at the outset are false in the sense of being oversimplified.
In the toolkit theory of theory only models exist. Generalized models, specialized models, all in quite considerable diversity. And, as George Box is supposed to have said “All models are false. Some models are useful.” We hope to be able to cover whatever of the infinite dimensions of evolutionary and ecological problem Ma Nature and Pa Culture happen to throw at us. If we’re lucky, the problem is dominated by only a few dimensions, it is easy to get lots of good data, and our toolkit already has some good-enough models. Lacking any one of these conditions we have a hard problem on our hands. In a tantamount to infinite dimensional world we can be pretty sure that our toolkit is less complete than we’d like. Thinking back even a few years we can marvel at how naive we once were. At the beginning of my scientific life, the acceptance of seafloor spreading revolutionized geology. Toward the end here, the ‘omics revolution wrought by cheap gene sequencing is revolutionizing biology. We are no doubt still lacking tools to understand similar dimensions of nature and culture that are still over the scientific horizon.
Burnham, K. P. and D. Anderson. 2002. Model Selection and Multi-model Inference. Springer Verlag, Berlin.
Efferson, C. and P. J. Richerson. 2007. A prolegomenon to non-linear empiricism in the human sciences. Biology and Philosophy 22:1-33.
IMPORTANT ANNOUNCEMENT
Much as I greatly respect Peter and his work, his arguments here are posted on the “Chapter One” page of this blog, and his remarks are not addressed to the main themes of this chapter. Pete writes of “the replicator concept”. There is not one replicator concept in the literature, but several. I believe that the particular replicator concept that Thorbjoern and I develop in Darwin’s Conjecture is invulnerable to the criticisms that Pete and others have raised. Pete does not address our version of the replicator. Our definition of the replicator is first developed in Chapter Four and discussion of replicators should be posted on that page, when it appears in a few weeks.
Posts on a page that do not address the issues that are raised in the chapter that relates to that page are liable to be removed or severely edited.
The Missing Contributors?
One thing that seems obvious to me is that a large section of literature on the application of Darwinian ideas to the study of institutions and they’re environments is missing. The human ecologists, most notably at the Chicago School (i.e. Park, Burgess and McKenzie), but also others such as Bews, Quinn and Hawley are all missing. This despite their obvious influence on Aldrich, Freeman and Hannan et al.,
The period that is concluded to be ‘mostly dormant’ (p. 18) was amongst the most fruitful for human ecology. Given the direct connection between this early and innovative work and modern day use of Darwinian concepts by sociologists, I am puzzled by the omission.
I am grateful to Reasonable Adventurer for raising this possibility. We might have missed out some contributors from the 1930s and 1940s to the Darwininian theory of social evolution. Of course, in those decades, immense advances were being made in genetics, Darwinian evolutionary theory in biology, and in ecology. But we looked for people who applied the Darwinian ideas of selection, variation and inheritance to social, cultural, or technological entities or processes. If Reasonable Adventurer has an example of a human ecologist who did this, then we would be delighted to learn more detail about his or her contribution.
Whilst time unfortunately does not permit me to fully describe the development of human ecology, I believe a brief overview of the main players and their key contributions can evidence the notion that ‘Darwinian ideas of selection, variation and inheritance to social, cultural, or technological entities or processes’ were present in there work. Human ecology was developed most significantly by American sociologists in the 1920s to the 1950s, building explicitly upon the biological writings of the late 19th century of Darwin and his followers (Theodorson, 1961).
The first recognized human ecologist is widely considered to have been Robert Park. His most significant works being his Introduction to the Science of Sociology with Burgess in 1921. This tomb (1040 pages) drew heavily upon the established ecological / evolutionary theorists and pragmatists of the day, with chapter contributors including; Dewey; J. Arthur Thompson; William James; William Sumner; William Wheeler; Eugenius Warming; Emile Durkheim.
In his work in 1936, titled Human Ecology, Darwin’s idea of a struggle for existence was central to his arguments that competition was the ultimate mechanism of coordinating the frequency and distribution of entities. His ideas borrowed heavily from both Spencer and Elton in terms of what was viewed as ecology and evolution. Indeed, Alihan (1938: 31) states that Park also stipulates that “economic competition, as one meets it in human society, is the struggle for existence, as Darwin conceived it”.
The most orthodox Human Ecologist (in my humble opinion) was Roderick McKenzie. Within his various works, Darwinian ideas are easy to observe, especially as they were being used by bio-ecologists such as Charles Elton.
McKenzie defined human ecology (1924) “as the study of the spatial and temporal relations of human beings as effected by the selective, distributive, and accommodative forces of the environment. Human ecology is fundamentally interested in the effects of position, in both time and space, upon human institutions and human behavior … these spatial relationships of human beings are the products of competition and selection, and are continuously in process of change as new factors enter to disturb the competitive relations or to facilitate mobility”.
In this definition are Darwin’s ideas of selection and variation. His other major contribution to human ecology was his fascination with ecological succession within which he explicitly focused upon ecological inheritance and disruption across time. Indeed, he was a great promoter of what we would now call niche construction (or ecosystem engineering), another idea that Darwin himself championed in his work on vegetable moulds and earthworms.
He saw the essential difference between human ecology and plant ecology being that human have more freedom to choose the habitat they exploit and that they also have a greater capacity to control or modify the conditions of such habitats. That is, he saw humans as being able to alter the nature of selection they might ultimately be exposed to.
The more seminal work of Hawley (1950) built upon the foundations of Park, Burgess and Mckenzie and carried on the Darwin’s consideration of variation and its importance upon allowing the process of natural selection to proceed without fear or favor.
I will conclude this contribution with a quote from Gerald Young (1974: 31), one of the few contemporary human ecologists whose work is fascinating to read. With reference to Darwin and Malthus’s mutual concern of the struggle for existence, he observed that “ecology is really an extension of economics to the whole world of life” and “economics, therefore, is merely human ecology”.
I would recommend Young’s paper as a comprehensive overview of the development of the field of human ecology and the influence that Darwin and the other early ecologists/naturalists had on its development.
References
Alihan, M. A. (1938) Social Ecology: A Critical Analysis, New York: Columbia University Press.
McKenzie, R. (1924) The Ecological Approach to the Study of Human Community, American Journal of Sociology, 30/3: 287-301.
Park, R. E. (1936) Human Ecology, American Journal of Sociology, 42: 1-15.
Park, R.E. (1921) Introduction to the Science of Sociology, Chicago: University of Chicago Press.
Theodorson, G. A. (1961) Studies in Human Ecology, Illinois: Row, Peterson and Co.
Young, G. L. (1974) Human Ecology as an Interdisciplinary Concept: A Critical Inquiry, Advances in Ecological Research, 8: 1-105
Following on from the last few posts, I wonder if Peter’s comments relate to the different meanings attached to using words such as principle. From my reading of chapter one, and the wider Generalised Darwinist project, the concepts of variation-selection-retention represent an approach, which is the starting point for development of domain-specific theory (as outlined later in chapter 2). These generalised principles or generalised models represent abstracted conceptual building blocks, which of themselves are insufficient to describe or explain phenomena within say economics or organization studies.
When reading Pete’s comments about the toolkit of models, chapter 3 of Howard Aldrich’s Organizations Evolving text came to mind, where he presents the Generalised Darwinist approach (though he didn’t use that term at that time) alongside other generalized approaches (e.g. resource dependence approach, institutional approach etc). These competing approaches form the starting point for development of rival theories, (and indeed rival theories under the umbrella of each approach). What is central to the GD project is that variation-selection-retention is central to the development of these domain-specific theories, rather than seen as a bolt-on set of concepts used alongside some other theory, say of behavior or learning (see Greve’s 2002 chapter in Blackwells Companion to Organizations).
Dermot is absolutely right to raise the question of what we may mean by “principles”. When we address the Darwinian “principles” of variation, selection and inheritance (or replication or retention) in the book we argue that these “principles” are in large part “explanatory requirements”. They are not principally (say) axioms, models, or laws. This issue is addressed in chapter two. The Chapter Two page will open in about 12 days.
There are some very interesting comments here. Can I ask who or what would be a ‘holder of information’ at the social level?, Do you mean individuals within populations? Could a holder of information be an organisation within a wider population of organisations? Where the principles of Variation, Selection and Retention are concerned, isn’t a key issue seeing as we are dealing with the social world here, that social evolution involves variation of ideas, then selection of particular ideas? It is ideas that shape routines, behaviours and what form interactors take, or this is what I think.If so then the ideas we choose are at the heart of social evolution. Maybe this also involves social learning, or learning interactors? Don’t know if this is relevant! Just a thought.
Information can be held in many places in a complex evolving system. (Information is defined very broadly in the Shannon-Weaver sense.) Replicators (such as genes, habits, customs or routines) are special holders of copiable information – we can discuss this later when we come to chapters where where replicators are discussed.
Hi, thankyou for your reply. I find the concept of using principles taken from Darwinism and applied in the social sciences as intriguing. You are right that these are just principles and not laws or models as such. How information emerges or arises and how this influences change or evolution is interesting to me. We understand a great deal about how biological information emerges via the genes and mutations etc, but how information emerges and how it is selected in the social world seems more complex. I have been thinking about some of your questions (and dbreslin’s comments below about Darwinian principles dare not speak their name etc). One problem with discussing social change in terms of Darwinism is that the application of this theory to the social world, historically speaking has lead to some unpleasant occurrences including eugenics etc. So notions of favoured races or survival of the fittest are problematic when applied to society. However if we abstract certain principles and apply these precisely and carefully in certain areas of research then maybe we can avoid this problem. We need to think carefully about how we do so. I remember reading Evolution by D Futuyma who stated that biological evolution was not a model for society. This is why Darwinism has fallen out of favour isn’t it? in the social sciences?
I liked the sentence in chapter 1 ‘[Nelson and Winter] have inspired an entire generation of researchers deploying these core Darwinian principle in their work – many dare not speak their name’. This play on Oscar Wilde’s famous quote I think captures a reluctance of many to use the D-word (or GD- word in this case). I can think of a number of recent key papers/book chapters in the organization science literature that clearly develop the key concepts of variation-selection-retention and even genotype-phenotype (or replicator-interactor), with no mention of Darwin, or many of the early works referred to in chapter 1. These scholars seem to start the timeline with Campbell (1965; 1974).
The book by H.& K. addresses an issue with profound methodological implications.In general we must ask whether biological evolution can be considered a scientific paradigm (in the sense of Kuhn) or a more simple scientific approach. What is certain is that theories and models inspired by chaotic processes could offer some reasons to consider the theory of evolution as more than a scientific approach. However H.& K. undertake a first -and necessary- process that aims to indentify evolutionism in Darwinism, making of Darwinism a substantial scientific paradigm. Today, this is not a risky operation because in the context of the evolutionary paradigm, Darwinism has assumed a hegemonic role.H.& K. highlight how evolution cannot give rise to a level of predictability of evolutionary trajectories comparable to that of Newton’s mechanistic science. I do not consider this an element of “inferiority” of evolution with respect to Newton’s mechanical philosophy, if only because the analytical levels are different. Newton’s rational mechanics operate at the macro level, but we have known for a long time that physicists must compete with dynamics that refer to entities at the micro level and only probabilistically lead to predictabile trajectories.A biological evolutionary process can also be probabilistically predictable.
I found the study useful, although brief, of scholars in the field of social sciences who in the nineteenth century saw in Darwinism a useful conceptual framework of reference. Contrary to popular beliefs, even a scholar of Pareto’s calibre had the opportunity to ask some questions about the possibility of utilizing the darwinian schemes in economics, without ruling out the possibility of using such schemes in the near (for Pareto) future. In this regard, see my recent paper (Maurizio Mistri,”Vilfredo Pareto and the Evolutionary Approach to Political Economy”, in “Studi Economici”,1/2012).
For Pareto, the difficulty in using the Darwinian paradigm was mainly in the lack of a consistent and well-grounded theoretical framework for biology. In subsequent years, the Mendelian revolution was affirmed and along it, the modelling of evolutionary processes according to probabilistic schemes. To be noted is that at the end of the nineteenth century, Pareto’s ideas on Darwinism were drawn from the work that the great French zoologist Yves Delage publisherd in 1895. In 1909, a book by Delage and Goldsmith was published in which ample space is given to Mendelian scientific positions; it is a pity that Pareto did not read the second book.
Sections 1.4 and 1.5 outline the general framework of the complex relationship between Darwinism and economics, highlighting that the acceptance of evolutionary processes in economics has suffered fron some political oriented judgments against Darwinism. However, there is a field of economics and social sciences in general in which Darwinism, and in consequence evolutionism, has been gradually asserting iteself, namely in the analysis of institutions. Social rules are found not only in human soscieties but also in animal societies. In this regard. the biologists Maturana e Varela speak of the “autopoiesis” of biological forms, also incorporating social rule into their scheme.The concept of autopoiesis has not found much success in either sociology or economics. However this is still a field of research that deserves to be investigated, if only because it highlights the possibility of considering social rules as a complementary element to biological characteristics. A final consideration concerns the increasing use of game theory in biology. This is a terrain in which, in a sense, the relations between economics and biology are overturned. That is to say, it is no longer economics that draws inspiration from biology, bur biology thar draws inspiration from economics. (Maurizio Mistri)
After reading the first chapter my impression is that authors didn’t mention in “KEY QUESTIONS” (KQ) the problem, which I find decisive for their project. But it was recognized at page 25, specifically, first sentence: how to present social RELATIONS as the units of evolution (or Darwinian evolution). So far may be Luhmann did well in this direction with his double contingency etc (hope to discuss it in the 2nd chapter). Consequently my answere to the 2nd KQ is that social thought during the XXth century have ignored the central role of relations as ontological entities, moving to agency-structure dilemma. (Although there are many important social tendencies that explain this “fall out of favor”, I agree with Geoff and Thorbjorn, that they are nor essential for the books purposes).
3rd KQ. Not dogging in philosophical disputes on logical problem of analogy and generalization (with plenty of aspects), I see major difference (in lights of books goal) is that analogy appeares to be less certain gfeneralization. Therefore analogy between bioevolution and socioevolution means that our knowledge on biological reality as evolving onewas more certain, so thet we may spread some biotic features unknown in social sciencies on social reality (that was favorable at XIX century). Generalization is another kind of induction, when we equally well know similar features of both realities (both were measured up). But the problem is wether we really equally enough know those features of biotic and social realities that are to be generalized. I’ll be glad if the answere will show up in the following chapters.
In 1st KQ I would prefer to use not the terms “biological and social evolution” but “biological and social reality”. In first case it is easy to derive common principles of evolution from smth that we already accept as evolution. But the authors project succeeds if it’s about shared causal mechanisms. There is a need to find more sophisticated proof for social evolution, using sociological language. For instance, I suppose that at least the concept of socialization have to be grasped. Sociology matters.
Regarding the issue of better understanding ‘how to present social relations as the units of evolution’, you find the seminal work of Edward Haskell useful. In attempting to unite the biological and social sciences, he offered a theory of coaction. Haskell (1949: 46) claimed that in any activity occurring in society, ‘there are diversely powerful individuals which can be separated in two groups or classes, the weak and the strong’. Further, that ‘these two main classes can have nine, and only nine, qualitatively different relations towards each other’. Lastly, he noted that ‘the major properties of societies vary with coaction’.
His thinking was original in that he captured the fragments of everyone’s thinking up until that point in time and condensed it into one easy to comprehend framework. His ideas remain, unchanged, in every book on ecology that must always offer a way to explain the interactions between two or more species. This work potentially can add vaule to understanding what social units of selection exist and provides an alternative way to view the individual and/or institutions as social actors.
Haskell, E. F. (1949) A Clarification of Social Science, Main Currents in Modern Thought, 7: 45-51.
I have two initial observations.
The first relates to question 3 – I am doubtful that either analogy or generalisation are appropriate terms for what I understand the aims of the Darwin’s Conjecture project to be. Taking a lead from the Campbell quote on the final line of page 18, the project seems initially to be taxonomical, and therefore ontological, in nature. It is asking if it is possible to identify a genus of evolutionary systems that share sufficient common features to qualify for that classification. Specific instances of evolutionary systems, in biology, in human society, etc. can then be assessed for their degree of consonance with the features of the genus and included or rejected as appropriate. Reaching some tentative conclusions on the taxonomy, or ontology of these evolutionary systems, provides a point of departure for seeking a more detailed empirical understanding of how the common features, mechanisms, processes, etc actually work in each member of the genus.
Second, when we refer to genes, habits, customs or routines as examples of replicators I wonder if we are in danger of limiting our conception of replicators to our current state of knowledge? Genes were beyond the contemporary state of knowledge in Darwin’s time and it seems to me that we need to allow for the possibility that there may be replicator mechanisms in the social world beyond our current knowledge. One candidate that come to mind, for example, is Jung’s concept of the collective unconscious and its archetypes.
Unfortunately I’m not familiar with Haskell’s ideas. I’m just wondering if he saw environment of social relations as social reality too. Because if we treat environment as biological reality it is completely another story.
Thats a good question, I suspect his attempt to develop his ideas across both domains never got out of first gear in the social world. In the biological world the process of demarcation exists also, ensuring the man made views of what is an organism’s environment (community and population ecology) still exists. It was in theory supposed to capture the reality of each interacting entities payoffs (be they negative, nuetral or positive)
First I will ask you to excuse my non-native English.
In this first chapter you show that the mechanisms of evolution, as theorised by Darwin, are guided by a general framework consisting in the combination of random variation, inheritance and selection (pp. 4, 26, 27). I add here the term ‘random’ to ‘variation’ because I think that its introduction is the most important difference between Lamarck and Darwin: the Lamarckian model is deterministic, the Darwinian model is stochastic. I think more generally that the introduction of probability in social sciences is prior to Darwinism and permitted for example the development of political arithmetics during the XVIIth and XVIIIth centuries (see my 2012 book). It would have been interesting to contrast here the two theories of Lamarck and Darwin for evolution, as you will contrast them for social change in chapter 4. Why did not you do that in this introductory chapter? Is such a distinction between determinism and randomness not so important for you?
I think that the term ‘framework’ needs some more precise definition: is it in its conceptual sense that you use it? Do you consider Darwin’s principles as a paradigm, as an axiomatic theory, or as something else? In this last case it will be important to say it in this introduction.
I wonder also, as you are using the terms ‘mechanisms of evolution’, why in your references you did not cite the main contenders of mechanistic explanation: Abrahamson, Bechtel, Craver, Franck, Glennan, etc. As an example, I will take here the definition given by Bechtel and Abrahamson: a mechanism is a structure performing a function in virtue of its component parts, component operations, and their organization. Franck had shown that the processes of evolution, as theorised by Darwin, are guided by their functional architecture. Don’t you think that it would have been interesting to develop more this point in this chapter, and to try to see in the forthcoming ones its usefulness for economics and social science?
References:
Bechtel and Abrahamson (2005): Explanation: a mechanist alternative, Studies in the History and Philosophy of the Biological and Biomedical Sciences, 36: 421-441.
Courgeau (2012). Probability and social science. Springer.
Craver (2006): When mechanistic models explain, Synthese, 153: 355-376.
Franck ed. (2002): The explanatory power of models, Kluwer Academic Publishers.
Glennan (1996): Mechanisms and the nature of causation, Erkenntnis, 44: 49-71.
These are deep questions and some of them are to be discussed in later chapters. So I confine myself to one point. The term “deterministic” has multiple meanings. It requires definition. By some definitions, stochastic determination is a form of determinism. See Mario Bunge’s classic book on “Causality” (1959) and my article “Darwinism, Causality and the Social Sciences”, Journal of Economic Methodology,11(2), June 2004. There I discuss three possible meanings of “deteminism”:
(1) Determinism is sometimes defined as the epistemological doctrine that ‘any event can be rationally predicted, with any desired degree of precision, if we are given a sufficiently precise description of past events, together with all the laws of nature’ (Popper, 1982, pp. 1-2). We may call this idea ‘predictability determinism’.
(2) A different definition of determinism is the notion that, any given set of circumstances and state of the world must lead to a unique outcome: ‘given A, B must occur’ (Blanshard, 1958, p. 20). We may call this principle ‘regularity determinism’. It involves a denial of randomness and chance in the universe. This is an ontological rather than an epistemological notion: it says nothing about what we may be able to know or predict.
(3) Another definition of determinism is the notion that every event has a cause (Urmston, 1989). This is again an ontological statement about the world, otherwise known in philosophy as ‘the principle of universal causation’ or sometimes ‘ubiquity determinism’. Bunge (1959, p. 26) provided a more refined formulation. He described the ‘principle of determinacy’ amounting to: ‘Everything is determined in accordance with laws by something else’.
I consider version (3) to be acceptable, and I believe it was close to Darwin’s view.
Pingback: darwin’s conjecture, universal or general darwinism | orgtheory.net
Yes, I entirely agree with you that the terms ‘deterministic’ and ‘stochastic’ have multiple meanings. You consider here the causal meaning of the term ‘determinism’, and I prefer to speak in this case as Mario Bunge of ‘determination’: causal determination, in the commonly referred sense; quantitative self-determination; interaction; dialectical determination; structural determination; teleological determination and mechanical determination. However, in my comments, I was considering the mathematical meaning of deterministic and stochastic processes, which is entirely different. But I agree that it is erroneous to think that a phenomenon dependent on a stochastic process is not reproducible: it is perfectly statistically reproducible. However I don’t agree with the view you give in your paper on ‘Darwinism, causality and the social sciences’ of the mechanistic approach: this may be valuable for the XVIIth century mechanistic theory, but no more true for the new one developed by Abrahamson, Bechtel, Craver, Franck, Glennan, etc. But this may be discussed in later chapters.
With reference to your reconstruction of the use of evolutionary theory in social sciences, I was wondering whether one could consider the work of Norbert Elias and especially his book on “The Civilizing Process” as one of the best examples of an evolutionary approach in sociology? Elias discusses Darwin in “Towards a Theory of Social Processes: A Translation” (1977/1997), where he writes for example:
“Darwin was able to perceive the connection between a multitude of observable data which had previously been understood as either teleologically goal-directed or metaphysically as the working of mysterious vital powers, and to conceptualize them rather as blind, unplanned, purposeless and yet directional processes with no predetermined end. He was able further to discover the immanent dynamic of the processes which bring about unplanned and purposeless, but directional and structured changes of this sort and which, when one knows them, can explain such changes.” (378)
I think that this quote from Elias is vague and confused. Elias is right to note that Darwin gets away from teleleogical and vitalist explanations. But Elias then goes on with the notion of “blind, unplanned, purposeless” processes. Darwin does not use these descriptions. The notion of “directional” evolution is also vague and possibly inappropriate. Assuming that Elias originally wrote these words in 1977, by then there were several other postwar social scientists beginning to recognise the possibility of using Darwinian ideas. D. T. Campbell (1965) is much more sophisticated.
Thank you for your reply. Yes, this was most likely written in 1977. Similar ideas, but with only a single reference to Darwin, can be found in more detail and with an empirical analysis in “The Civilizing Process: Sociogenetic and Psychogenetic Investigations” (1939).
“A few days ago, Reasonable Adventurer suggested that work in ecology should have been included in our history of Darwinian approaches in social science. Ecology is a very important discipline and has been influenced by both Darwin and Spencer. But ecology tells only part of the Darwinian story. Typically ecology considers the interaction of populations among each other and with their environment, or single entities with their environment. Sometimes (as in Hannan and Freeman’s “organizational ecology”) selection is considered in populations, but it is generally confined to what we call “subset selection” and excludes “successor selection” (see chapter five of our book). Consequently, processes of informational transmission from entity to entity (in that population) are not addressed. We regard this informational transmission as vital to the Darwinian story.”
Doesn’t Darwinism (or shouldn’t Darwinism) include concepts from ecology, in particular different types of relationships beyond those which are competitive? After all the ecological relationships that we see are a result of variation and selection. It can be advantageous to be a symbiont or to be part of a mutualistic relationship or a wider network. Examples being mycorrhizal root interactions in woodlands, photosynthesis, oxidative respiration, lichens, corals etc. Darwinism is as much about types of relationship as it is about the processes which govern these relationships. Perhaps interpreting Darwinism purely in terms of variation, selection, inheritance and retention is only revealing part of the picture?
On to reductionism, in the social sciences I think that ideas of a cause leading to a single effect have fallen out of favour in recent years, along with positivism and the scientific method. Darwinism has been caught up with the latter. It seems that interpretive, hermeneutic explanations etc are more popular to describe social processes. Instead of simplifying there has been a trend towards complexifying (my word sorry!).
I have a question, or questions: How do we apply the principles of variation, selection or retention to the social world? What sort of knowledge or evidence is it that we are seeking? What other explanations are required to explain how these principles operate in complex systems? How does this variation emerge? What is selected? etc.
Finally, can evolution be seen as a self-reproducing system (or am I going off topic?)
Thanks
In a 1959 paper, “Darwinism and Darwinisticism” (accessible at http://www.jstor.org/stable/3825585), Morse Peckham wrote, “…Darwin did not have the word ‘ecology’. He was in fact an ecological thinker, and in ecology words like ‘competition’ and ‘cooperation’ are too inexact, too value-weighted, too metaphorical, and too anthropomorphic to be used at all.”
Peckham was a literature professor, and the paper amply reflects this, but he did edit a Variorum Edition of the Origin which is still in print today. One more quote from the paper: “[Darwin was] execrated for having led men to disbelieve the seemingly obvious truth that man’s proper, natural, and normal mode of behavior is cooperation, harmony, and love. Again, it is always being rediscovered that in natural selection there are cooperative as well as competitive mechanisms at work.”
There are lots of questions here. Hopefully most of them will be addressed when we move on to later chapters, when we discuss the meaning of selection, the objects of selection, etc. etc. Chapter two is where we discuss the nature of Darwinian theory, as generalized and applied to the social domain.
As we know devil is always in details. So why do you move from entities to population of entities? And if those entities are humans i.e. holders of social features the same way as biological features, social reality can be lost. Supposing so it is easier to generalize on the same ontological level. But if those entities are smth social (relations etc), what is the meaning of “population of social relations”? It looks unclear.
Darwinism is a populations thinking model as I remember it. The principles of variation, selection etc operate within populations and so change the composition of such populations over time. However I think it is interesting to consider the entities or individuals within the population and the decisions that they make. As it is these decisions made in daily life that influence the outcome. For example, on the Great Bear Stakeout on the Beeb a female bear was killed due to the fact she was in the wrong place at the wrong time, ie she encountered a pair of bears that were flirting on and off with one another. Neither bear needed to kill her, they just chose to. She was not unfit, just unfortunate. So some elements are hard to model, as they are seemingly random. Behaviour is complex, especially in social situations. What influences this and the choices that individuals make is surely relevant to understanding how populations change over time.
My guess is that thinking of society or social reality as just population is simplification that fraught with ontological incosistencies. Erlier in “Darwinism, Causality and Social Sciences” Geoff considered five Darwinian principles (detarminacy, continuity, cumulative causation, evolution, and consistency), as he wrote “even more fundamental to Darwinism” than variation, inheritance and selection, and I presume that those five principles are closer to play the role of general principles for social and biological realities. But the shift to 3 principles signifies sort of narrowing the initial project. I would investigate philosophical principles more carefully, before going to selection et al. But perhaps it goes later in the book.
Darwin’s (1859) ideas clearly relate to an individual, he talks about the “life of the individual” (p. 62), he talks of the struggle for existence as being between an individual and the phyiscal conditions of life (p. 63), and he talks about “every single organic being” (p. 66). So, Darwin clearly spoke of individuals struggling with what we now refer to as an environment.
That an individual may be connected to other like individuals through a ‘web of complex relations’ (i.e. its environment) does not assume that any indivdual must be. The processes of variation, selection and retention can be still be used to explain the existence of an individual within a given environment.
Equally, Darwin’s ideas also clearly relate to groups. This is evident in “The Descent of Man” (1871). In vol 1, chs 3-5 there is a notion of group selection, where group fitness is considered. But Darwin did not have the conceptual tools to develop a rigorous, multiple-level selection theory. Now we have those tools.
I agree Geoff,
The point I was trying to make was that when all is said and done, regardless of the level/s of selection, it is an individual entity that is selected for or against.
Later in the book we develop the concepts “selection of” and “selection for” (not “selected against”). For us, “selection of” refers to objects of selection. These objects can be groups as well as individuals (under specific conditions). “Selection for” refers to population level, replicator outcomes (e.g. genes, habits). We need to postpone discussion of this to Chapter Five, when we can go into depth.
agreed 🙂
look forward to that discussion, promises to get philosophical me thinks
Not sure that group selection is a concept that evolutionary biologists like these days. Isn’t that similar to kin selection? As far as I remember from reading this subject , individuals do not act in the best interest of the group, but rather in their own best interests. So, animals only display altruistic behaviour when it is in their best interests to do so. So animals display parenting behaviour as it is advantageous in terms of biological fitness, as this ensures they leave more offspring, or they are leaving more copies of their genes as Dawkins would say. As groups are made up of competing sets of genes then selection cannot act on them as a whole, but individually.
However, populations are comprised of individuals, and it is the frequency of individuals of a certain type, ie carrying certain genes that changes over time. Why this happens is fascinating I think. In human populations this could be down to culture and ideas. What do you think?
On the contrary, group selection has enjoyed a rehabilitation in biology. But this will be raised when we discuss chapter seven. Please hold back on this for now. In brief answer to the question, there is both cultural and genetic selection in human populations. Again, more of this in later chapters.
I enjoyed reading this. But I worry that insufficient attention has been paid to the justification of the need or the desirability of working through a generalized framework for the explanation of the evolution of both social and natural systems. Presumably this justification rests in part on a belief that what is common to social and natural evolution is – at a suitable level of abstraction – greater than what might keep them apart, and so a shared set of principles and organizing philosophy is mutually beneficial. Yet I’m not sure that this is adequately demonstrated, in this part of the book at least. We could take the Metcalfe quote at the start of Chapter 1 and turn it around in terms of the meaning that I suspect the authors had intended to convey: i.e. an evolutionary logic applied to social and economic phenomena can proceed perfectly well in its own right without any reference to evolutionary biology (whether through analogies or through the use of a common organizing framework). However, what I do find convincing in the argument is the idea that there is some commonality to be had in the analysis of complex systems, which points in the investigation of both social and natural systems to an emphasis on the explanation of the past rather than the precise prediction of the future, given the centrality of non-ergodic change, i.e. that the structure of relationships as well as the entities themselves evolve and change character over time.
The discussion of the history of thought on the applicability of Darwinian principles to social and economic evolution is nicely done, as one would expect knowing something of Geoff’s work (I’m less familiar with Thorbjorn’s). However, this discussion only hints from time to time as to possible answers to the broader question of why particular phases of history were characterized by different degrees of resistance to (or acceptance of) Darwinian principles and even evolutionary ideas more widely in the social realm. What is it about the wider social environment itself in each phase of history that tends to make attempts to link the evolution of natural and social systems either relatively appealing or unappealing? One thinks here of EH Carr’s contention that our interpretation of a period of social history (the evolution of a social system in the past) depends upon the contemporary context of the historian or analyst. Analyzing social and economic phenomena is even more subject to such contemporary influences than is the study of natural phenomena, as recognized in Geoff’s own account of the drive to supposedly make economics a more ‘objective’ discipline, briefly also alluded to in the Preface.
Given what I’ve said already, I struggled with the claim in the opening paragraph of the Preface that ‘there is no known alternative to Darwinism as a general framework with which to analyze the evolution of economic and social systems’. This claim seems to be to be at odds with Geoff’s observation in the blog and elsewhere that evolution is a general term that preceded Darwin, while Darwinism offers a specific form of evolution (evolutionary analysis). I would find the claim more credible had it been that Darwinism provides the only known general framework for connecting the analysis of the evolution of natural and social systems. More generally, I believe we need to think through more carefully the relationship between Darwinism and existing evolutionary theories of social and economic history. I can see that a relationship might be established at the level of abstraction presented here, based on the organizing principles of variation, selection and inheritance/retention/replication, although the application of these principles might be worked through very differently in the two realms.
Hegel and Marx derived an evolutionary theory of social history from the development of ideas that emerged in classical German philosophy, combined in Marx’s case with ideas taken from classical political economy. In this area the debate has often been whether the key drivers are to be found in processes located in the realm of the evolution of technology (the ‘forces of production’) or processes found in the realm of the evolution of institutions (the ‘relations of production’ and social superstructure), or equally well in both. In recent times GA Cohen restated Marx’s view that processes in the forces of production lie at the leading edge of evolutionary change (although the various parts of the system may co-evolve), while Deirdre McCloskey has been arguing the reverse. In connection with the streams of literature alluded to in Chapter 1, some of Dick Nelson’s recent work has emphasized the tensions between processes in the realm of technological change and those in the realm of institutional change, and how institutional change might constrain or hold back the drivers emanating from technological (and scientific) progress. As Dick argues, ideas along these lines have been developed by Chris Freeman and his co-authors. If we were to think in Chapter 1 terms, we might distinguish between replicators and interactors in each of these realms of analysis, of technology and institutions. What would be gained by doing so?
You raise some very interesting points, and I agree that we do need to think through the relationship that exists between Darwinism and existing evolutionary theories of social and economic history. One problem that I find with applying Darwinian theory to society is that (and this could be my personal interpretation) the principle of natural selection (or selection here) is discussed in terms that suggest that this is akin to an entity that resides outside of the system like some guiding hand, observing each individual within the system and behaving a bit like an accountant, weighing up the pros and cons of each individual and then selecting which survive and which do not. However, for me nothing exists outside of the system, you cannot stand outside of it, selection is enacted largely in every day life, through the thoughts, decisions and resulting actions of the individuals residing in the system. Thoughts shape events and there is a cumulative effect (seen when we look back, ie from a historical perspective) from these thoughts and decisions over time.
Perhaps in a way decisions can be seen as interactors or replicators?
We need to discuss Melissa’s points more fully when we come to chapter five. But – very briefly for now – selection does not necessarily mean a conscious selector, even in social evolution. Darwin worried that the term “selection” might connote someone or something doing the selecting. But his concept of selection refers generally to a process or processes where the membership of a population changes through time. This also applies to social evolution in populations. “Artificial selection” is a special case of selection in general where a human agent is purposively involved in changing the membership of a (biological or social) population. Contrary to J. R. Commons, it is a special case of, rather than a rival to, selection in general.
I find this forum very enriching and I look forward to discussing the following chapters of the book. I am an economist and I did my master’s thesis precisely on the subject of generalized Darwinism and its application to the social or cultural domain. As this first chapter is an introductory chapter I would like to make some general comments.
My view on the subject is very similar to the authors’ view so my comments will usually go in the way on how the argumentation could be strengthened.
I think it would be interesting to frame the discussion from a physical stance in which we recognize that all the things that we are talking about (genes, interactors, ideas, routines, institutions, etc) are physical entities. Firstly this allows us to build a link between all the entities we are talking about and also enables us to work on ontological solid ground.
Secondly and most important; this enables to bring on the stage the second law of thermodynamics which serves as a powerful argument. Simplifying the issue, the second law is merely a statistical law which simply states that what is likely to happen will tend to happen (tendency to disorder). Although obvious, this at first seems contradictory with the finely tuned (and thus unlikely) physical configurations of cells, organisms, computers or markets.
The physical world has certain ability to self-organize (as brilliantly argued by Stuart Kauffman in his book, Origins of Order), but we only know of one process capable of bringing such unlikely physical configurations into existence; Darwinian evolution, which as D. Dennett argues is a searching algorithm. This fact not only gives an edge to Darwinian explanations of social phenomena but also forces detractors to give an alternative explanation to the complexity (the physical unlikelihood) of such phenomena. Currently this complexity is just given for granted and requires no special attention.
Another comment I want to make is around the issue of the applicability of the Darwinian “triad” to the complex phenomenon of society and culture. Although the Darwinian “algorithm” is very simple, its manifestation as a physical phenomena (for example in biological evolution) is much more complex. This complexity is far greater in the cultural case as very elaborate entities such as human brains play a role in the process. But the features of the abstract simple “algorithm” are still present although they are not readily seen within such complex phenomena.
Related to the last point is a general comment about the need to make categories in order to understand the world. For example there seems to be a need to identify some clear entity as the replicator in order to corroborate the existence of Darwinian evolution. But the replicator can be a complex entity that cannot be pinpointed at; thus we need flexibility with categorization, some people interpret this as a weakness but it is actually a strength, as it takes into account the complexity of the world thus avoiding the error of fitting it into an oversimplified framework.