How Darwinian is cultural evolution?

Dual-Evolution-Systemby Scientia Salon

This is a paper published by Nicolas Claidière, Thomas C. Scott-Phillips and Dan Sperber in the Philosophical Transactions of the Royal Society (Series B, Biological Sciences), in 2014 (here is the open access link to the full article).

The abstract states:

Darwin-inspired population thinking suggests approaching culture as a population of items of different types, whose relative frequencies may change over time. Three nested subtypes of populational models can be distinguished: evolutionary, selectional and replicative. Substantial progress has been made in the study of cultural evolution by modelling it within the selectional frame. This progress has involved idealizing away from phenomena that may be critical to an adequate understanding of culture and cultural evolution, particularly the constructive aspect of the mechanisms of cultural transmission. Taking these aspects into account, we describe cultural evolution in terms of cultural attraction, which is populational and evolutionary, but only selectional under certain circumstances. As such, in order to model cultural evolution, we must not simply adjust existing replicative or selectional models but we should rather generalize them, so that, just as replicator-based selection is one form that Darwinian selection can take, selection itself is one of several different forms that attraction can take. We present an elementary formalization of the idea of cultural attraction.

I actually disagree with the authors when they wish to treat cultural evolution as “Darwinian,” and in fact I think they unwittingly make a good argument that it is not, or at the least not in great part. Still, the excerpts below give an idea of a number of interesting notions discussed in the paper.

Of course, the analogy with biological evolution is not perfect. For example, variations in human cultural evolution are often intentionally produced in the pursuit of specific goals and hence are much less random than in the bio- logical case.

In this paper, we argue that there are important aspects of cultural evolution that do not fit even within a selectional model, and which are better explained and modelled as part of the broader population and evolutionary frames. Specifically, we argue that cultural evolution is best described in terms of a process called cultural attraction, which is populational and evolutionary, but only selectional under certain circumstances.

While some cultural items may indeed be propagated by imitation and other forms of copying, it is clear that a large number are not. In particular, many are also (re-)constructed. For example, a student taking notes in a lecture does not simply copy any spelling error that the lecturer happens to write down, but will in fact, in her own notes, correct the error and in doing so re-construct the correct spelling. As such, cultural propagation is partly preservative, but also partly (re-)constructive, to different degrees in each particular case. As such, it is not only a matter of inheritance, as is gener- ally the case for biology, but also of reconstruction. Whichever of these is more important in any given case is an empirical question, but either way, the direct analogy with biological evolution is considerably weakened by this fact.


How deep is the analogy between biological and cultural evolution? Memetics assumes that it is deep indeed; that the main relevant details of the biological case have direct equivalents in the cultural case, such that there is, for example, a cultural phenotype, which achieves a certain level of (inclusive) fitness, which will in turn determine the phenotype’s relative success in the population. Selectionist approaches loosen the analogy somewhat, moving from a replicative frame to the more general selectional frame. We have argued that the analogy should be loos ened further: cultural evolution is broadly Darwinian, in the sense that it is a population-level evolutionary phenomenon, but there is no empirical reason to think that it sits entirely or even in general within the selectional frame.

Another important disanalogy between biological and cultural evolution is the mechanisms by which traits propagate through a population. In biology, the mechanisms of transmission are in general only preservative. In the cultural case, however, the mechanisms of transmission are many and varied, and include both preservative and constructive sub-mechanisms.

Both attractors and the process of attraction are statistical notions. They do not denote a type of causal process or the outcome of a specific such process, and as such they do not pro- vide explanations of cultural phenomena. Rather, they provide relevant descriptions of what is to be explained. Attraction should instead be explained in terms of factors of attraction. Factors of attraction in an epidemiological population will generally be partitionable into two classes: relevant properties of the individual members of the host population (such as the psychological and biological susceptibilities of humans); and relevant properties of the environment of these individuals, including the demographic properties of the host population itself.

Darwinian selection leads to the maximization of inclusive fitness, and this explains the appearance of design in the natural world. Is there an analogous result for cultural attraction? As selection is a special case of attraction, design is possible and in some cases explicable in standard Darwinian terms. Having said that, such explanations will not apply generally, and may not even apply commonly.

A general, formal statement of what cultural attraction leads to does not presently exist, and we see the development of such a statement as a major goal for future modelling work.

61 thoughts on “How Darwinian is cultural evolution?

  1. There are clear differences between cultural and biological evolution. That is beyond doubt.

    But I think that the analogy to Darwinian evolution is helpful, because the two are not completely different either. Pre-Darwin, one might have appealed only to intentional design (as well as some randomness, I guess) as a force for shaping cultural evolution.

    But it is obvious to me that selection must play some role — for instance religions that encourage proselytisation (e.g. evangelical Christianity) obviously have a survival advantage compared to those that don’t (e.g. Zoroastrianism), so that if we choose a religious person at random we can expect that their religion is more likely to derive from the former category than the latter. The point of the analogy is just to acknowledge this fact.

    Similarly, Yusuf al-Qaradawi, a prominent Sunni leader in the Muslim brotherhood, has been quoted as saying “If they had gotten rid of the punishment for apostasy, Islam would not exist today.” Whether or not this is true, it is an interesting statement to consider. It would normally be interpreted as a teleological rationale or justification for punishing apostasy, but it can also be interpreted as a historical explanation for why Islam exists and why it has some of the features it does. The point of the analogy to Darwinian evolution is to acknowledge that both perspectives can shed light on the matter. As in biology, we can talk either of the purpose of a thing (teleology) or why we should expect to see it in surviving populations (history) — the difference being that there is no mind to generate purposes in biology.

    That said, there are clear barriers to making a scientific study of this aspect of cultural evolution, because the effects and the entities are very vague and difficult to define. It seems more a sort of philosophical principle to bear in mind rather than anything that could ever approach the kind of scientific rigour we see in study of biological evolution. As such, I’m pretty skeptical of the science of memetics even though I think the principle is reasonably sound.

    Liked by 1 person

  2. cultural evolution is broadly Darwinian, in the sense that it is a population-level evolutionary phenomenon, but there is no empirical reason to think that it sits entirely or even in general within the selectional frame.

    I would say the same about biological evolution, and I think many biologists would agree.

    On the other hand, it is also a mistake to deny that selection has a role. Something new can be introduced into culture by invention. But if an invention is not easy to learn, then it will change (or evolve) toward something that is easier to learn and easier to pass on to future generations. And that’s a variety of selection.

    Liked by 1 person

  3. DM, Neil, even if “selection” does play a role in cultural evolution, this doesn’t make the process Darwinian, since selection is a necessary but not sufficient condition for Darwinism. And of course since cultural selection is neither “natural” nor “sexual” in the Darwinian sense, we’d also have to parse exactly what we mean by selection in this context.

    Liked by 2 people

  4. Hi Massimo,

    Two things.

    Firstly, I didn’t actually say that cultural evolution was Darwinian. I said that the analogy of cultural evolution to Darwinian evolution is useful and informative.

    Secondly, what exactly it means for a process to be actually Darwinian rather than just similar to Darwinian evolution is not in my opinion as cut and dried as you present it. Darwin was talking about the evolution of species, so you could say that any form of evolution of anything other than species is not Darwinian. On the other hand someone else would in my view be equally entitled to interpret Darwinian evolution as any process where selection of any kind is applied to a varying population of any kind.

    For this reason, I don’t think that the question of whether it actually is Darwinian is particularly interesting. This is a pseudoquestion about how we each prefer to define “Darwinian” rather than about how the world works. The meaningful question to my mind is how similar is cultural evolution to biological evolution: what does it have in common and in what respects is it different?

    > And of course since cultural selection is neither “natural” nor “sexual” in the Darwinian sense

    This objection is a little strange on two counts.

    First, “natural”. What is natural? Is the evolution of the behaviour of ants natural? I would say so, but then I would ask why is the evolution of the behaviour of humans any different? Human culture, like ant behaviour, is entirely a product of the interaction of natural organisms, which are natural creatures. The idea that there is a clear divide between the natural and the artificial or the cultural is perhaps just an artifact of our particular perspective — we see ourselves as above nature though we are as much a part of it as anything else. As such I think I would reject “natural” as a meaningful category. This means that I regard artificial selection (of cattle say) as just a special case of natural selection, where human intentions (via human behaviour) form part of the natural environment determining the fitness function which directs evolution.

    This is certainly how we would regard it if we were studying a symbiotic relationship between any other species, (and human relationship with cattle actually is in my view genuinely symbiotic — there are many more cattle in the world than there could be otherwise because of our protection, veterinary care, provision of land and so on). Some species of ants for instance have much the same relationship to certain kinds of aphids as we do to cattle, and this relationship has probably shaped both species significantly.

    Second, on “sexual”. Why would you bring sex into it? Darwinian evolution doesn’t need sex — asexually reproducing organisms can evolve just fine (if slower). I think you may have meant to say something about reproduction in general?

    Liked by 1 person

  5. Hi Massimo,

    As a biologist, I agree with Neil, that biological evolution is not Darwinian just in the same sense that cultural evolution is not, given all non-adaptive evolutionary processes. Of course, the difference between then lies in how important is selection for each case. I rather to think that both process have a contribution of, but neither are, completely Darwinian.

    Although the precise definition of what means Darwinism is a complex question per si, mainly for historical issues, I think selection is the most important element of anything called a Darwinian process. I usually think of Darwinian evolution as adaptive evolution through selection. What elements you would consider necessary and sufficient, besides selection, for cultural evolution to it be Darwinian, at least in the cases where selection is operating?

    Liked by 1 person

  6. Surely, one may sensibly use the label “Darwinian” in several, narrower or broader senses. This is what we did in the article, where – adapting a suggestion of Peter Gordfrey-Smith – we distinguished four nested explanatory frames (populational, evolutionary, selectional, and replicators), all of which could usefully be labelled “Darwinian”, depending on one’s purposes. We also argued that selection is not the main force of cultural evolution, and in so doing implied that cultural evolution is not Darwinian in this narrow sense of the term. We argued instead that the main force of cultural evolution is cultural attraction, of which selection is a special case. If we are right, this makes cultural evolution Darwinian in the wider, populational sense of the term. If one thinks that this wider sense doesn’t deserve the label “Darwinian,” so be it. Is it worth doing more than agreeing to disagree on this terminological issue? We would welcome a discussion of the substance.

    Liked by 1 person

  7. While I am not entirely sure what “Darwinian” means, it is possible that cultural evolution is in fact entirely Darwinian. The process does not know where it is, nor where it is going, nor obviously does it care. The disasters facing the human specie can be explained in terms of evolutionary dynamics. Population in excess of what parts of the earth can sustain, climate change, nuclear war threats, these are things which evolutionary dynamics may posit as being beyond the ability of humans to control.

    Liked by 1 person

  8. I do not feel that what constitutes “culture” has been adequately addressed here. The point is well made that culture constitutes a kind of second replicator. Humans interact with their physical environment directly, and also indirectly through their cultures, which encompass economies, institutions (like marriage, religions etc) and ideologies. All three of these are integrated in each cultural system on the planet. The intersubjective, symbolic communication between human beings is primarily about their cooperative activities within cultures, while systems of ideas are rationalizations – or systems of explanation – for these activities.

    To make matters more interesting, there is a discontinuity between the biological and the cultural, for individuals can learn more than one language and culture during their lifetimes. If a culture that is not assimilating changes fast enough to keep up with changed environmental conditions (including domination by some other cultural ecology) can lose it’s biological hosts to the more successful culture, or even fragment and scatter.

    Cultures are not designed by human host populations, but rather represent a replicator best compared to the kind of quorum sensing mechanisms that exist in bacteria and insects, an analogy to other species’ behavioural algorithms also suggested by E.O. Wilson.

    Cultures, despite their capacity for rapid adaptation, do persist and replicate themselves over time, through the profound effects of belief systems and paradigms – in other words, they do have fidelity. They can also generate a counterpart to genetic heterogeneity, for each new generation finds new expressions and innovations, some of which flare as briefly as top songs on a hit parade. As long as the economic system remains sustainable, the tune might change, but people will sing about similar things.

    Liked by 2 people

  9. It’s a long article; it’ll take time to read carefully; but browsing it, I confess myself unhappy with it.

    “In other words, attraction can also result in design as an indirect (proximate) effect of the natural selection of factors of attraction. (…) one real world example is the gaze of sitters in portraiture: humans are particularly sensitive to direct eye gaze, and in cultures where a portrait sitters’ gaze direction is left free to vary, the culturally preferred form tends, over time, to move in the direction of direct eye gaze, and away from averted gaze.”

    I tried to hunt down the source of the reference, a paper by Olivier Morin (“Cognitive attraction in cultural evolution”), but could only find a precise ( Its own point of reference is the development of portraiture in the Renaissance.

    In the 1500s, art was making the rather profound historical transition from being designed to propagate Christian theology, to providing a pleasurable means of secular self-reflection. From his precise, Morin seems not to account for the pressures of the market at the time – artists in the 1500s depended on the express desires of their patrons in order to survive. “Those artists show a preference for direct eye-gaze portraits as soon as they start painting, suggesting that they acquired the new style in the years of their apprenticeship,” writes Morin. But more likely, those artists reflected a preference of their patrons, deriving from the greater interest in themselves (liberated from religious purity), allowed by the affluence acquired from the massive expansion of commerce during the Renaissance. (BTW, one reason the Renaissance artists could develop the secular gaze in their work, is because the Renaissance style is a direct development of new techniques capable of greater precision.)

    Of course, they had a precedents. No, not simply in the work of the masters to whom they were apprenticed. I mean the surviving portraiture from ancient Rome, the overwhelming majority of which were done with direct outward gaze at the audience. For the claim of Darwinian cultural evolutionary theory to be true, we would need to suppose that human cultural evolution fully developed such “factors of attraction” in the arts some 2000 years ago. Then we would need an evolutionary analogy that might explain how these got forgotten for a thousand years.

    Otherwise, this sounds pretty much like cultural history, which has been studied for many decades, in various disciplines.

    It should be remembered here that when the very idea of evolution began its development and dissemination, it involved introducing the notion of history into studies that were believed to be non-historical – geology, biology, the ways in which humans relate to other animals. But cultural history has always been historical, and it has its own history. What precisely do we gain by redefining such studies in terms of biology and evolution? And what do we lose?

    Liked by 5 people

  10. For the theory offered in this article to be true, it would need to be argued for most human cultures currently existing, and readily traced back historically without breaks. This is just about impossible; similar problems have led cultural theorists in various disciplines to accept any cultural determination as profoundly contingent and contextually embedded. Thanks largely to globalization, pretty much all cultures today share in the same world history; but while all cultures have changed in adaptation to globalization, not all cultures have adapted in the same way. Without an account of this, a Darwinian evolutionary theory of culture seems trivial where true, distraction where poorly grounded.

    I latched onto the remark on ‘the gaze’ in the article, because of my background in semiotics. Cultural criticism and semiotics have well developed theories and critical language concerning the phenomenon of the gaze, that are sociologically well-grounded, historically informed, and politically nuanced *. Again, it is unclear why any of this should be re-inscribed as somehow biologically driven.

    There are considerable risks in reducing culture to biology, even analogically. I noted in my previous comment that the changes in art during the Renaissance had largely to do with the economics of the Renaissance, which liberated art from its previous religious-pedagogical function, and also allowed people some liberty to reflect on themselves in a secular, visually pleasing manner. But who were these ‘people’? Certainly not the serfs. They were the rich and powerful of their day. Their portraits were as much an assertion of their power as it was a (somewhat narcissistic) contemplation of themselves.

    If we reduce culture to biology, what gets lost is politics.

    There’s a brief remark on two-party politics, in the article. In part: “The two-party system: whatever the initial frequencies, the system tends towards an equal frequency of the two types because of the positive impact that each type has on the other.” This tells us nothing other than that the parties exchange dominance over time. There’s no reflection here how the two actual parties in America have changed dramatically over history; how frustrated many voters feel with the system; how the system is maintained by media focus, empowered by advertising dollars; or other factors that seem to have left that system both moribund and yet still in control.

    Similarly, the section in the article discussing how the pronunciation of the word ‘data’ has shifted between ‘dahta’ and ‘dayta,’ evidences no consideration of the fact that Americans tend to pronounce ‘a’ nasally, so ‘dayta’ would likely be the pronunciation in America; and America happens to be the globally dominant cultural model currently, to which other cultures must adapt, or resist.

    Without a compelling account of the politics of culture, in large matters and small, in the contemporary era and throughout history, a theory reducing culture to biology (even loosely analogically) is going to remain unconvincing.

    * See for instance Daniel Chandler’s “Notes on the Gaze,”

    Liked by 3 people

  11. I read the original paper. It fights the notion that culture is “Darwinian” (so it somewhat counters the implied meaning of its own title!) I agree with some of what it says, and I have used some similar arguments for years.
    From the conclusion:

    How deep is the analogy between biological and cultural evolution? Memetics assumes that it is deep indeed; that the main relevant details of the biological case have direct equivalents in the cultural case, such that there is, for example, a cultural phenotype, which achieves a certain level of (inclusive) fitness, which will in turn determine the phenotype’s relative success in the population. Selectionist approaches loosen the analogy somewhat, moving from a replicative frame to the more general selectional frame … We have argued that the analogy should be loosened further: cultural evolution is broadly Darwinian, in the sense that it is a population-level evolutionary phenomenon, but there is no empirical reason to think that it sits entirely or even in general within the selectional frame.

    Another important disanalogy between biological and cultural evolution is the mechanisms by which traits propagate through a population. In biology, the mechanisms of transmission are in general only preservative. In the cultural case, however, the mechanisms of transmission are many and varied, and include both preservative and constructive sub-mechanisms. Constructive sub-mechanisms are common and, because they are often shared within a population, they often transform cultural traits in systematic ways, such that they converge upon particular types, which we call attractors.”

    I wrote a comment to SS. My comment argued that biological evolution is not just “Darwinian”. And it’s not just about genes. Thus, if cultural evolution is in analogy with biological evolution, it cannot be just about “memes” (as the paper concludes)… for these reasons alone.

    I also argued that scientific teams (I am involved with one) are busy trying to demonstrate much more advanced biological evolution mechanisms which are somewhat teleological (definitively not just Darwinian). So an analogy culture-biology may soon have to involve teleology on a biological basis alone.

    As far as I am concerned, it is obvious that all cultural traits are designed, or have been designed, intelligently by human beings. Agreed, it was that intelligent design was then and there (it may well be entirely obsolete now). I re-used the apostasy example which DM was allowed to evoke.

    The paper uses the semantics of “attractor”, well known in dynamical systems. This is an equivalent notion to teleology: a system is determined by its ends (a notion found in Aristotle, Hegel, etc.).

    I do not disagree with the paper. Instead of saying outright that so-called “memetics” has no standing, it tip-toes around that critique by pointing out that neurologies (as found in human children) “re-construct” and “correct” cultural elements rather than mechanically duplicating them (as memes guys have it).

    An expanded version of my rejected comment can be found at:

    Liked by 1 person

  12. Darwin sought to explain the multiplicity of the “tree of life,” arising from a lessor number of forebears.

    Culture, on the other hand, is the exact opposite process. How diverse input coalesces into a single or common community or culture.

    Now if we are to consider some much larger dynamic, that would include both the tendency toward variegation and toward congregation, then Darwinism might not be the most effective mold to fit it into.

    Liked by 2 people

  13. I think that the similarities between cultural transmission and biological evolution are very superficial indeed and the extent to which such an analogy may be useful, this is heavily outweighed by the misleading aspects.

    When Richard Dawkins proposed the idea of a meme he had in mind that a meme would have a definite physical signature in the brain, analogous to the physical nature of a gene. He only made this explicit in endnotes to later editions of the Selfish Gene. Had he been more forthright about this aspect from the start and in the main text I am pretty sure it would have sunk the idea sooner.

    He said he felt emboldened to say this because he had, he thought, some support for the idea from a neuroscientist. I remember rather rolling my eyes at the idea when I first read it, some decades ago.

    As Dawkins himself says, every idea deserves a chance and there was an attempt by others to start up Memetics as a science. But I think Dawkins would agree that this idea, at least in the form he presented it, has failed.

    I think the problem is that in science today there is an emotional investment in the name ‘Darwin’ and you can ‘sex up’ an idea by calling it Darwinian. But I don’t think that it is very fruitful in this case.

    I think that the only things we can currently call ‘Darwinian’ are biological evolution, and perhaps the mathematical abstractions from it and some uses to which this might be put.

    Liked by 7 people

  14. @ejwinner If I understood Olivier Morin’s argument correctly, even your alternative explanation (the clients not the painters) would fall under the same process of cultural attraction. The clients were more inclined to find direct gaze figures more pleasurable (interesting?etc) than non-direct gaze. And again you provide a very good hypothesis for why direct gaze was perhaps. not forgotten, but better said suppressed: modesty – which could have swayed either paints, or clients, or both. (I believe Olivier suggests that sheer complexity of technique is a better explanation).


  15. It is interesting how the discussion immediately gets bogged down in trying to figure out what any of the relevant terms mean. If everybody has a slightly different understanding of Darwinian, evolution, and culture, then every take on this issue becomes plausible. One wonders if it is even a fruitful question to ask under those circumstances.

    One thought that occurs to me when considering the question to what degree culture is shaped by selection is that most people may underestimate its importance because they take the products of stabilising selection as so axiomatic of culture that they don’t see the selection. Yes, whether our language has a ‘th’ sound or not, or whether adult women are supposed to cover their hair or not, or whether our national cuisine includes potatoes or not, all those aspects of culture are probably mostly down to a random walk or other non-selective constraints.

    But when we imagine a culture with very different values, for example one that considers stealing, lying, murder, cowardice, or perhaps laziness to be virtuous, or one whose national food is gravel, or one that exclusively sends its young women off to die in war, it should become obvious that there are indeed some rather stringent selective pressures on human societies. And the same goes for organisms: a few ecologically and developmentally important proteins are under strong selection but 90% of the genome is undergoing a random walk.

    In that sense, there is something of a parallel where using biological evolution as a metaphor or analogy might be useful. But the differences are otherwise so big that the metaphor is mostly misleading, starting with the fact that most biological lineages have to acquire a given innovation independently whereas cultures can easily copy each other’s innovations. Biological evolutionary history is 99% tree-shaped (with rare horizontal gene transfer accounting for the rest), but human cultural history is mostly network-shaped with lots of ‘horizontal meme transfer’.


  16. Daniel Dennett led a workshop* around the content of this paper (ECMs – Evolutionary Causal Matrices). Seems like there was disagreement here about the the degree of overlap of Darwinian and cultural evolution.



  17. Radu Umbres,

    I was aware that my first comment would be open to such criticism when I posted it; but I would need more space than this format allows, for a larger, more engaged argument to make my point firmly. But really, why try to find a single explanation, or a reduced set of explanations, for cultural behaviors and events with rich histories and complex contexts?

    So in my second comment, I tried to explain my two primary issues with Morin and with Sperber et al, and with the the whole endeavor of developing a biologically grounded theory of culture.

    1. We already have satisfactory cultural theories and histories that give more than adequate account of the phenomena such a theory would need to address.

    2. The attempt to redefine culture in biologically evolutionary terms, even by loose analogy, risks eliding politics.

    Many thinkers in various fields have tried to find some way to discuss culture without addressing politics. It can’t be done. Even in a stagnant culture with a rigid social structure everyone in it assumes to be a product of nature (and hence fail to recognize the politics involved), there are political relationships demanding political observance. Neither the lord of the manor nor the serf recognized their relationship as political, but it was, and this could readily be discerned whenever the two communicated.

    It is notable, that at roughly the time painters were developing new techniques to please the rich and powerful, Machiavelli wrote “The Prince,” instructing those wealthy patrons of the arts how to lie to the people and betray their friends, for “it is better to be feared than loved.” This was not in anyway coincidental.

    The problem with politics is that it is all about self-interest, both the self-interest of the individual, and the self-interest of the collective. The trouble is, this self-interest is defined by the person or the collective as whatever most interests them. So if they think that what is most important to them is their skin color, or their sex organs, or their money, or their religion, or even their teeth, that’s what will motivate their political choices.

    When fluoride was first introduced into American water supplies, the extreme right wing went ballistic, convinced that it was a commie plot to poison young patriots, making them weak and infertile. Where did this come from? What would it mean for a social theorist to remark this as reaction by a fringe political movement? What are the historical implications of the election of the son of a fringe anti-fluoridian to the office of Vice-President (Dan Quayle)?

    I don’t think any ‘Darwinian’ theory of culture is likely to provide satisfying consideration of such issues; but time will tell. The real problem is that those proposing such a theory haven’t even admitted that such issues are pertinent. I admit myself somewhat annoyed with this.

    Liked by 3 people

  18. In his/her second post, ejwinner writes : “If we reduce culture to biology, what gets lost is politics.” What gets lost if we reduce culture to biology, I would say, isn’t just politics, it is culture itself. In our article, we certainly didn’t propose reducing culture to biology; on the contrary, our approach is blatantly non-reductionist. On the other hand, I would say that, if you ignore the role of biology in culture, what gets lost is a great amount of intelligibility.

    In his/her third post, ejwinner evokes a variety of interesting social sciences and humanities issues and concludes: “The real problem is that those proposing such a theory [ours] haven’t even admitted that such issues are pertinent. I admit myself somewhat annoyed with this.” Not only do I admit that such issues are pertinent, but I have spent quite a bit of time working and writing on some of them. Our article is on a related but rather different kind of issue: How to improve on current ‘Darwinian’ formal models of cultural evolution? We are aware that, for most social scientists, this is not a pertinent issue at all. C’est la vie! I am not even annoyed.

    Liked by 1 person

  19. Just a little remark: Many of the comments above are cogent and interesting. Culture is “Darwinian” in a trivial way.

    However, many of the commenters seem not to have read the original Royal Society paper in depth. If one does, one realizes that the authors actually comes to the conclusion that the “Darwinian”, and especially “memetic” models are either too small, or even completely irrelevant (Darwinian), and, or, false (memetic). However the authors are apparently so preoccupied with not upsetting the Darwinian-Memetic partisans that they use a lot of what Wikipedia would call “weasel words”.

    In one sense culture can be drastically Darwinian, as when the enemy comes and wipes out a culture. As the Conquistadores did with Aztecs and Mayas.

    When a culture is founded on wrong, or inferior metaprinciples (as the Celtic culture was), it is outcompeted by higher principles (fostered by the Greco-Romans, in the case of the Celts) within its own foundation, and can disappear peacefully, but thoroughly (as happened to the Celtic culture and languages: very little is left; entire languages have been lost without a trace).

    But of course history, military or not, sociology, anthropology, linguistics and the like have pondered these questions ever since they exist. Thucydides, in his 24 centuries old treatise, famously reduces the catastrophic, thirty years Peloponnesian war to a psychological anxiety the Spartans had, regarding Athens’ success.

    Liked by 1 person

  20. At this point in the discussion, it may be worth saying a few words to clarify the objectives of our paper.

    Many commentators, perhaps guided by our title, appear to have read our main objectives as being terminological. This is not the case. We are instead most concerned with the matter of how to best think about and model cultural evolution. We do however believe that distinguishing between different uses of the term “Darwinian” can make a positive contribution to that discussion.

    Let me expand. The title of our paper was a question: “How Darwinian is cultural evolution?”. You could gloss our answer as: It depends what you mean by “Darwinian”. As Dan Sperber’s summarised in his first comment above, we distinguished four different senses of the term, and argued that cultural evolution is not Darwinian in the two narrowest senses, but it is Darwinian in the two broader senses. Some might object that these broader senses don’t really merit the label “Darwinian”. If so, we don’t object. As I say, we are not so concerned about these terminological issues for their own sake.

    The main contribution that we are aiming for with this paper is instead to make a positive suggestion about how one might go about modelling the evolution of culture. We provide a number of arguments in the paper that the best way to do this is within a framework of “cultural attraction”. Our paper gives arguments for and reviews data that support this conclusion, and we present a simple model of cultural attraction, for the purposes of exposition. The perspective we are arguing for is a departure from much present work on cultural evolution, which works within a narrower frame of reference (as we explain in the article). We would like to believe that these positive contributions are of more serious interest than any terminological issue.

    Finally, I would also like to clarify that there is no attempt whatsoever in our article to reduce culture to biology, as some commentators have suggested/intimated.

    Liked by 1 person

  21. I know next to nothing about this subject, but do appreciate the comments from two of the authors as to their purpose, which according to Dan Sperber is “How to improve on current ‘Darwinian’ formal models of cultural evolution.”

    As a layperson in such matters, the heavily qualified statements (there are countless examples) made from the onset only make me wonder what has been accomplished since I don’t even know how well-regarded a “Darwinian” approach to a model of cultural evolution is as opposed to competing theories. For starters, I don’t even understand a preference for the term “cultural evolution” instead of, say, “cultural change and/or absorbtion.” And, rather than being put off by any attempt to engage in reductionism, I found the attempt to broaden the application of “Darwinism” to be disconcerting. What seems assumed in this scholarly paper is that the ‘Darwinian’ model of culture can yield fruit, but should be directed toward exploring a notion of cultural “attraction,” despite the stated reservation that “A general, formal statement of what cultural attraction leads to does not presently exist, and we see the development of such a statement as a major goal for future modelling work.”

    Perhaps, in part, the generally negative commentary, then, simply fails to take into account the intended audience for the paper, that is, those who believe some kind of “Darwinian’ model can be employed over time to provide a robust model of what is meant by cultural evolution or cultural change.

    Liked by 2 people

  22. Dan Sperber,

    After reading your comments, and those of Patrice Ayme, I recognized that I had missed something in my initial, and frankly rapid, reading of the article. So I have spent the afternoon re-reading your article more carefully and thinking it through.

    I now admit that I largely misread the article, even reading some paragraphs pretty much the to opposite of their intent. For instance, I didn’t really recognize what you and your colleagues were doing with the problem of idealization, which becomes important again towards the end of the article. Nor did I recognize the anti-reductionist thematic of the paper. So allow me to apologize.

    Discussing the misreading might be of some benefit. I had in mind, while reading, a recent discussion here at Scientia Salon, concerning Evolutionary Psychology. As you probably know, Evolutionary Psychologists seem to take the Darwinian paradigm very seriously as causal explanation of human behavior. The justification for this seems pretty weak; so I guess I was applying the bias I had developed during that discussion to your paper, which was uncharitable of me.

    This also raises a larger question, however. You are apparently fairly comfortable using certain terms analogously in the loosest way possible. For instance, a key term you deploy for research is in developing an ‘epidemiology’ of cultural transmission. Although I recognize its value as a counter to ‘Memetics,’ I confess myself uneasy with this terminology, not because I hope humans are some how more precious than microbes (they may not be), but because it does invite a reductionist reading of the research.

    When some Evolutionary Psychologists talk about human behavior directed by criteria of ‘fitness,’ say, in child-raising, they are drawing a very strong analogy from biological evolution – to the point where it doesn’t read as analogy at all. I think some of the issues raised by some readers here arise from differing understandings of the degrees of analogical equivalence in the general discussions concerning the applicability of Darwinism to cultural behaviors.

    Viewed in another direction along the same path: when do analogies become so loose that their application loses clarity? And at what point does such analogical terminology get stretched to the point of simple metaphor? The problem is that metaphors sometimes get unnecessarily hypostasized into perceived idealizations of the kind which I now think you are arguing against.

    As to the general argument I made against reduction of culture to biology, I still stand by it. But I am sorry for having misread your project as a part of that. I still see problems with some of the cases the article addresses. But I think the general theory may hold promise. However, I would need to see it addressed to further cases; so I take it I have some further reading to do.

    Liked by 4 people

  23. I enjoyed the paper and the discussion, it was very interesting! 🙂

    I have two questions for the authors.

    The first is a general motivational question — what do you expect to achieve by describing the cultural dynamics in terms of Darwinian-like evolution? Does it introduce any level of clarity about this dynamics that was not already there, or some of its mechanisms that were previously unrecognized? From your article, I don’t quite see cultural dynamics being *explained* by your Darwinian model, but rather merely *described*. In other words, I see it as a new language for the same underlying thing. Am I missing something?

    The second question is technical — why do you assume that the causal evolution of cultures is linear? I feel that using the evolutionary causal matrices is an oversimplification, and way too crude to describe the actual evolution. A priori, I don’t see that interaction of cultures is always proportional to the number/amount of “individual elements” representing any particular aspect of a given culture. The linear approximation implies that “the more instances there are, the better opportunities for survival” may work in evolutionary biology, but if you apply it to culture, it seems that you lose all those interesting things that happen because of a single creative moment (sometimes produced by an individual human) that revolutionizes an entire culture. Things like paradigm shifts, creative insights, etc. For example, would the English language be the same today in an alternative history where one William Shakespeare never existed? The ECMs seem to ignore these kinds of effects on the evolution of a culture. So, is the linear approximation just there for the sake of simplicity, or is there some argument that claims its dominance in some way?

    Liked by 3 people

  24. I found it brilliant and thought-inspiring; I’m afraid my remarks may make you think I’m overinspired or giddy.

    The question was “*HOW* Darwinian is Cultural Evolution?” — hard not to ask unless we abolish the phrase cultural *evolution* — and a nuanced answer was given, with a good account of how Darwinian evolution must be modified to come up with cultural evolution.

    I will mention some thoughts that occurred while reading it:

    * “Culture as epidemiological phenomenon”: In a sense, cultural ideas are “viral”, but as with viral videos, this is not a terrible thing, so is there an epidemiology of some mix of nasty things with benign or positive symbiots. Indeed, since people are incomplete (unthinkable really) without some culture, they are necessary, like some bacterial fauna in our intestinal tracts, though far, far more visibly completing our essential nature.

    * Chain letters aren’t the only replicative texts. Many religions are a sort of chain letter. “Give copies to 10 people and you will prosper. A curse on you if you don’t”.

    * Dawkins’ Selfish gene idea was something like a Fourier transform on evolution. Should we think of a gene as a discrete snippet of DNA? Or rather as the structure defining that snippet. Would the latter view be “structuralist”? It also resembles somewhat the technique, in _Drawing on the Right Side of the Brain_, or turning a photograph upside down to draw it. Some of the iconicity of the eyes, mouth, etc tends to vanish and you get much better at attending to the actual pattern of light and dark that is there.

    * In the view of the gene as a structure, or sort of mathematical or Platonic object, we could be asking what mods would lead to more instantiations. Trouble is, thinking of the gene in isolation makes no sense as it needs to play as part of a team.

    * Evolution calls for “almost perfect” replication. DNA as a sort of digital medium, with redundancy like check-sums in a computer program is a good candidate. The insistence on a canonical spelling has a sort of check-sum effect, cutting down on drift. So does poetic form, favored by illiterate or minimally literate cultures with sacred texts to preserve and propagate.

    * Is the source of culture a kind of exuberant generation of far more raw proto-meanings than are needed? Some is lost, some makes its way into sacred narratives, and cultural facts and norms; much of the excess becomes secular art, song, dance, fashion. This resembles brainstorming (rule: no idea is discouraged) followed by critical decision making. I believe it also resembles some processes of early brain formation which has been called “evolutionary” by some very imprecisely I guess.


  25. The article asks an intriguing question. I suppose I ended up more distracted than helped by the use of biological properties/functions to dicuss sociological properties/functions. As a proponent of emergent properties I tend to think it’s better to resist using language and models from other levels. Sometimes that can be handy, but sometimes it invites confusion.

    In this case, if I were to consider cultural evolution (which i agree exists) in relation to biological concepts I would not think of Darwin. The way cultures change over time and give rise to new cultures which often hold exaggerated positions from their ‘parent’ culture, I’d say they were decidely Lamarckian.

    In fact the attractors concept sounds rather like a Lamarckian example of the naturally elongating neck of the giraffe (driven by internal and external pressures).

    Also, many cultural changes involve features similar to horizontal gene-transfer in bacteria (which i believe someone already mentioned). It is hard to view evolution within populations of some prokaryotes as entirely Darwinian in nature, given alternatives provided by horizontal transfer and symbiosis, and that to me seems the closest biological analogs to sociological level entities… if we decide to make such comparisons.

    Liked by 2 people

  26. I am interested in how a model for cultural evolution is tested.

    The problem I see is that our observations of cultural evolution are through the prism of our own cultural assumptions.

    The problems here are twofold 1) Our cultural assumptions skew the observations of cultural evolution (i have given a few examples of this in the past) and 2) this skewing is in itself a driver of cultural evolution by producing variations..

    Models of cultural evolution themselves can become causes of this skewing and also a unit of cultural transmission in itself.

    I am sure the authors are aware of the problem, I would be interested in how they avoid the problem.

    Liked by 5 people

  27. It may be useful for this discussion to keep in mind Richard Lewontin’s famous criteria for evolution by natural selection (i.e., “Darwinian” evolution), as spelled out in his classic “The Units of Selection.” Annual Review of Ecology and Systematics 1 (1970): 1-18. From the paper:

    “As seen by present-day evolutionists, Darwin’s scheme embodies three principles…

    1. Different individuals in a population have different morphologies, physiologies, and behaviors (phenotypic variation).

    2. Different phenotypes have different rates of survival and reproduction in different environments (differential fitness)

    3. There is a correlation between parents and offspring in the contribution of each to future generations (fitness is heritable).

    These three principles embody the principle of evolution by natural selection. While they hold, a population will undergo evolutionary change.”

    Given the above, my qualms in calling cultural evolution “Darwinian” arise from: i) There is no clear definition of differential fitness; ii) there is no clear theory of what causes differential fitness (without which the definition of evolution becomes notoriously tautological: the most fit survive, those who survive are the most fit…); iii) there is no clear concept of heritability.

    Liked by 3 people

  28. Dan Sperber says:

    Surely, one may sensibly use the label “Darwinian” in several, narrower or broader senses.

    Is this such a case, though?

    I say no, on historical as well as biological and sociological grounds.

    While not saying that the authors of this paper have such a stance, “cultural Darwinism” sounds close to “social Darwinism.” I’d think the authors would, even if they stand behind their ideas, be wary of such a phrase. Related to that, and contra thomscottphillips, even if your primary goal is not terminological, terminology still matters. What if I had a paper talking about “Lamarckian cultural evolution”? EJ covers some of this in his latest comment.

    Beyond that, it sounds about as strong, as far as an actual case being made, as some of D.S. Wilson’s claims for the top level of multi-level selection.

    Per Robin Luethe, yes, such a thing is “possible.” But, unless and until we know that culture evolves in a Darwinian way, it’s inaccurate to apply such a modifier.

    EJ gets at this when he notes that what may not be new in this paper has long been discussed under other terms.

    To move beyond EJ, Massimo has himself written about emergent phenomena in relation to the development of individual consciousness and related phenomena. Surely we have something like that in spades here.

    And, per Massimo’s comment, claims to have a quasi-Darwinist model for cultural evolution run aground on the shoal of “memes,” which of course, at least in a strong form, have been abandoned by all original proponents.

    To otherwise carry Massimo’s analogy further, new cultures often emerge from a mix of old ones via “blending” or similar. Given that in biological evolution, no such thing happens, this is another good reason to reject not just the word “Darwinian” but any idea of Darwinian or quasi-Darwinian modeling.

    At least we haven’t gone down the lane of Dan Dennett to claim that cultural evolution, because it can be modeled Darwinianly, is also algorithmic. (Insert eye roll here.).

    Liked by 3 people

  29. Maybe just taking a few propositions from evolutionary game theory, and avoiding any strong analogues with biological evolution would avoid some of the pitfalls pointed out in the paper.


  30. SocraticGadfly: “At least we haven’t gone down the lane of Dan Dennett to claim that cultural evolution, because it can be modeled Darwinianly, is also algorithmic. (Insert eye roll here.).”

    I believe Daniel Dennett is applying “algorithmic” correctly with respect to cultural evolution. I don’t know how “Darwinianly” it is though.


  31. Ejwinner,

    I am glad that, after all, our perspectives may not be as incompatible as you initially thought. Just a couple of points that, I hope, may make the compatibility even more plausible (with the risk, of course, that they will have the opposite effect).

    Epidemiology: The idea that ideas, fashions, styles, and so on may spread like epidemics is an old one in the social sciences (think of Gabriel Tarde). Actually, one of the first use of the Greek term epidemics is in Plato discussing the spread and success of Gorgias’ rhetoric style in Thessaly ( see Monique Canto « Un cas de style contagieux. L’épidémie du rhéteur Gorgias en Thessalie », Traverses, , septembre 1984, n° 32). Epidemiology is about processes of propagation in a population. Nothing in the very idea imposes that it be the propagation of a pathology. So, I don’t think our use of the term is a very loose analogy. More a moderate and not particularly original extension of the term in its narrow modern sense.

    Evolutionary psychology: A great many things are done under this label, some excellent, some not. Judging an approach – any approach – by its less successful outcomes isn’t really helpful. Anyhow, the initial impetus, when, in particular Leda Cosmides and John Tooby started developing a program for an evolutionary psychology in the 1980s, was against directly explaining behaviour in term of the genes (as sociobiology tended to do), arguing that this way of doing things missed a crucial link: psychological mechanisms, which are more directly under selective pressure than behaviour (without ever denying or even just ignoring the truism that the development of these mechanisms crucially depends on the environment).


  32. With this comment I will respond to a couple of the above comments directly. These responses are mine, but I do not believe that my co-authors would disagree on any matters of substance (hence my use of the first-person plural).


    Massimo: “It may be useful for this discussion to keep in mind Richard Lewontin’s famous criteria for evolution by natural selection”.

    We agree, and we also agree that cultural evolution does not satisfy these criteria, at least not in general (it may do in certain special cases). Nevertheless, we do not think that this disqualifies it from the label “Darwinian”. This is because we do not read “Darwinian” as synonymous with “natural selection”. Natural selection is the idea Darwin is most renowned for, but it was not his only insight worthy of eponymisation. In particular, population thinking is arguably as important. Population thinking involves looking at a system as a population of relatively autonomous items of different types with the frequency of types changing over time. (This can be contrasted with an essentialist approach, in which populations are defined by some supposed essential quality.) As we said in the article, the notions of population thinking and natural selection “are [both] Darwinian, but in different ways that are worth spelling out”.

    Correspondingly, a central claim of our paper is that while cultural evolution does not in general proceed according to natural selection, it is useful to approach it in populational terms, with the distribution of the different types changing over time. It is in this sense that cultural evolution can be said to be Darwinian. In a previous comment you suggested that we “unwittingly” made a case that cultural evolution is not Darwinian. If by “Darwinian” you meant “proceeds according to natural selection”, then we did not make the case unwittingly at all – it was, on the contrary, one of the main points we wished to make!


    Marko Vojinovic: “What do you expect to achieve by describing the cultural dynamics in terms of Darwinian-like evolution?”

    Our agenda was not to describe cultural dynamics in Darwinian terms, simply for the sake of doing so. As you intimate, it is not clear what purpose such an agenda would serve.

    The aim was instead to make a contribution to the development of a naturalistic theory of culture. Over the past 20 or so years, substantial progress towards such a theory has been made under the label “epidemiology of representations”. This literature has several points of contact with the cultural evolution literature. One aim of our paper was to clarify the similarities and differences between these two approaches. One point of agreement of particular relevance to your question is the view that formal modelling of cultural dynamics can be, as it is in the biological case, a source of useful and novel insights about how and why some cultural items become commonplace in a population, while others do not.


  33. A specific aspect of culture relevant not just to describing it but also to explaining several of its features is its well-recognized ‘epidemiological’ character

    This is something that raises questions for me. It is unsurprising that cultural items should spread in a similar pattern to that of disease. It was, after all, the basis of the Catholic Church’s strategy for controlling heresy for many centuries (perhaps the modern idea has evolved from the older one).

    But I think that this is a case of torturing an analogy. There being similar spread patterns is not, to my mind, a warrant for calling us a “host” for cultural items and I can see many ways in which this terminology can become misleading.


    When we say that P evolves, we mean two things: (i) that the frequency of types in P changes over time and (ii) that the frequency of types at a given time step is a function of their frequencies at earlier time steps.

    This seems to leave out a good deal of what we mean by evolution in a biological sense, perhaps even the most important part, which is the stepwise change of the “P” in question.

    Also, I am not sure that the causal matrices are enough to describe the spread of an idea, it seems that we need conditional probabilities rather than plain probabilities (or some sort of Bayesian matrix algebra). For example the probabilities in the ‘data’ matrix seem to be dependent on the cultural background of the speakers and in particular the way they already pronounce related words. The matrix might be different for a British person than it would be for an American.

    On the other hand I can see the introduction of this might make the model so cumbersome as to be unusable.


  34. I can see the value for biology of introducing culture as an element — given that culture affects the environment in which human beings live, it will be relevant to biological evolution.

    I don’t, however, see the value for our understanding of culture of “biologizing” culture. I don’t see how it adds to our understanding of cultural transmission across cultures — what of use would biology tell me, for example, about the way in which Durer served as a conduit of Italian Renaissance methods and tropes into the Northern Renaissance? — and I don’t see how it adds to our understanding of cultural transmission, within a culture over time. What am I going to learn from a biologist about this that I wouldn’t learn a thousand times better from an art historian, cultural historian, or historian of ideas?

    From the article:

    “We agree with Richerson & Boyd [8] that the overall general framework for the study and modelling of cultural evolution should be that of ‘population thinking’ (so named by Ernst Mayr, who described it as one of Darwin’s most ‘fundamental revolutions in biological thinking’ [10]). Population thinking involves looking at a system (such as culture) as a population of relatively autonomous items of different types with the frequency of types changing over time. The types themselves are not defined by their ‘essence’ but as historical subpopulations, features of which may change over time.”

    This sort of analysis can only strip away everything that is important and interesting and necessary for a real understanding of culture. I came away from this article feeling like I’d eaten a meringue. I had the illusion of chewing on something substantive, but afterwards, the whole experience just kind of evaporated, and I was left feeling like I hadn’t actually consumed anything. I certainly don’t feel like I know a single substantial thing about culture that I didn’t know before, something that I would never say after reading a Kristeller, a Panofsky, or a Gombrich.

    To be fair, however, this really isn’t a critique of the authors, but rather of the sort of thing they’re doing, which, I gather from the references, is relatively common fare. I have just never seen much benefit from this sort of importation of hard-scientific categories and methods into the study of culture. (Approaches like those one finds in cultural anthropology are a somewhat different matter, although even there, it is tempting to think that one is “doing science,” much more than one really is, as opposed to, say, engaging in historical and other forms of narrative description.

    Liked by 1 person

  35. I am somewhat puzzled by the hostility that some commenters here display towards the concept of memes. Admittedly I haven’t read the original book or the attempts by some people to build a formal science of memes, but the concept of a meme as a replicator undergoing duplication, mutation and selection in the ecosystem of human cultures seems sound.

    Differential fitness: some memes get taken up by more people than others. What causes differential fitness: complex, but no more so than in the case of biology. Some memes (ideas, beliefs, stories, pop cultural knowledge, techniques) bestow an advantage on the person who accepts and uses them, others are merely attractive to the human mind although they may even be harmful. Heritability: well, being able to transmit memes to other members of one’s generation and to the next generation is pretty much the definition of having a culture in the first place.

    The major problem is how to define a meme. Then again, that can only seem exceptionally problematic to somebody who has never given serious consideration to the complexity and multitude of replicators in biology: individual organisms, super-organisms, viruses, protein coding genes, genes coding other stuff, pseudogenes, transposons, and so on.


  36. @EJWInner:

    I am the author of the paper about portraits that made you “unhappy”. I don’t know why you could not trace the reference back. It was published in 2013 in Evolution & Human Behavior and can also be downloaded from my site. (

    I wish you had read it before you criticised. You would have seen, for instance, that you and I agree on the importance of changes in market mechanisms to explain what happened to direct gaze portraits. I think other interrogations of yours are already answered in the paper, but in case they are not, or for participants who may not want to read it, let me clarify a few things, in my next comment


  37. Depictions of direct gaze faces are almost ubiquitous in the world’s cultures, not just in the Greco-Roman portraits you alluded to, but in masks, paintings and engravings all over the world. The peculiar attraction of the direct gaze (which attracts attention but can also be interpreted as a sign of aggressiveness) also means that it is frequently surrounded by various beliefs about its powers, and sometimes its depiction is prohibited. All this is, of course, acknowledged in my 2013 paper. Its point was *not* to dwell once more on the power and significance of the gaze. It was more precise.

    I predicted that, where and when the two types of portraits, direct-gaze and averted-gaze, are in competition, direct-gaze portraits should enjoy a comparative advantage, at least when they are rare enough, and ultimately displace averted-gaze portraits as the majority style. I tested this prediction in two portrait traditions, the second one being Korean portraiture.

    In the course of testing that hypothesis, I realized that the rise of direct gaze portraits in the European Renaissance obeyed a peculiar mechanism: there seems to be a critical period during which the proportion of direct-gaze portraits in a painter’s style gets fixated, so to speak. Individual painters do not change their style as time goes by. Their collective style does, because individual painters come and go. I do not see how you can explain this pattern if you assume, as you so confidently assert, that patrons were responsible for every change in portraiture style.

    I am acquainted, of course, with the view that the rise of individual portraiture in the Renaissance was a manifestation of modern individualism in the elites. I have read my Burckhardt just like you. In some people’s view, everything about Renaissance portraits should be viewed in this light, including, of course, direct gaze. I disagree.

    First, countless cultural historians have challenged the Burckhardtian dogma. The version of it that you defend is (forgive me) a collection of clichés. The view that Renaissance elites were “liberated from religious purity” (your words) does not take religious history seriously. Neither does the view that “Renaissance liberated art from its religious-pedagogical function” (what about Baroque Catholicism?). Social historians of XVIth art in, for instance, today’s Netherlands, have extensively shown that portraits were becoming a mass-market bourgeois commodity, and were no longer reserved for the very rich and famous. As for “serfs”, a humanist like you should know that they were rare and getting rarer in XVIth century Western Europe.

    Second, my findings also obtain in Korea (I suppose Korean elites were contaminated by individualism as well). Lastly, the rise of direct gaze portraits does not in any way fit the usual narrative about rising individualism. The rise of direct-gaze portraits happens much later than the rise of individual portraiture. It happens quite independently of Reformation or Counter-reformation movements. It happens in wealthy, commercial Flanders, but also in conservative Italy or war-torn France.

    Liked by 1 person

  38. Lewontin’s three characteristics can easily be extended to include many behaviours

    “1. Different individuals in a population have different morphologies, physiologies, and behaviors (phenotypic variation).”

    Some individuals say day-ta, and others da-ta; some individuals drive on the left, others on the right; some B cells have higher affinity for a particular antigen than others

    “2. Different phenotypes have different rates of survival and reproduction in different environments (differential fitness)”

    Pronunciations most in keeping with that of other words are more likely to be retained ; individuals choosing the less popular side of the road are more likely to be run over ; B cells capture more antigen and are permitted to proliferate

    “3. There is a correlation between parents and offspring in the contribution of each to future generations (fitness is heritable).”

    Children follow their parents’ pronunciation OR peer pronunciation OR an individual’s pronunciation is relatively constant from t1 to t2 ; ditto ; genetic transmission to daughter cells.

    I think the point is whether it generates useful mechanistic understanding of population dynamics, as per Marko’s point. I don’t think it is sufficient to point out the process is relatively short-lived eg lymphocyte clonal selection, given that the Lewontin description doesn’t explicitly mention mutation/recombination (ie novel variation) at all.


  39. Hi Robin,

    When Richard Dawkins proposed the idea of a meme he had in mind that a meme would have a definite physical signature in the brain, analogous to the physical nature of a gene. He only made this explicit in endnotes to later editions of the Selfish Gene.

    A fairer summary would be that his original intent was to clarify understanding of the logic of genetic evolution by discussing alternative replicators. Then, later, he occasionally discussed how strongly the idea could be interpreted.

    You seem to dismiss the idea of a physical instantiation in the brain, but any and all ideas that you currently hold must be physically instantiated as neural-network patterns.

    On the central issue of how similar cultural and genetic evolution are, the answer seems to be that there are some clear parallels, which are well worth pointing out, but only limited ones, and that the big differences mean that the comparison can’t be pushed very far.

    An example of where such explanations are useful is the fact that the meme “God exists” is so often accompanied by the meme “believing in God is a virtuous act”. If an idea such as “God exists” can’t proliferate owing to external evidence for it, then the only way it can propagate is on its own internal merits. Coupling it to the meme “belief in God is virtuous” gives a reason to believe it in the absence of actual evidence for it, and thus the coupled meme propagates much better. And that’s why it is widely considered to be virtuous to believe in a god.

    Such explanations are worth pointing out, but they are rather trite and limited.

    Hi Massimo,

    … without which the definition of evolution becomes notoriously tautological: the most fit survive, those who survive are the most fit …

    I still don’t understand your objection to accepting that “survival of the fittest” is tautological. As you say, Lewontin summarised Darwinism as (1) phenotypes vary, (2) different phenotypes give different fitness/survival-rates, (3) phenotypes and fitness are heritable.

    None of those three statements is tautological, and from those alone we get Darwinian evolution. Given that “fitness” means “best at surviving”, the phrase “survival of the fittest” is indeed tautological, and indeed SotF follows tautologically from the above three statements. So what? The information-content of the theory, the non-tautological bit, is those three statements. Adding a tautological commentary to them doesn’t invalidate them, but can help people to realise the implications of them.

    Further, since “survival of the best at surviving” is tautologically true, any definition of “fitness” in which SotF is *not* tautological would then be false (or, at best, only approximately true).

    Which is why I thought that biologists had long adopted the “yes it’s tautological, so what?” position on SotF. The fact that SotF isn’t an axiom of Darwinism, but just a commentary about Darwinism, is shown from that fact that the first edition of OofS didn’t use the phrase, and thus Darwinism can be straightforwardly stated without it, as in Lewontin’s three points. So, I’m still baffled by why you reject this.


  40. There seems to be some debate as to whether the authors actually mean to show cultural evolution can be understood in terms of Darwinian biological evolution, or whether they are actually trying to show it isn’t.
    Given this is a short essay, addressed to a general audience and not a select group of their peers, it would seem the reasonable conclusion would be that it is what it purports to be, that culture can be modeled in those terms.
    Otherwise they run the risk of one of the examples they use, that of the student correcting the professor’s language to what she assumed he meant. As the old saying goes; Understand your audience, if you want them to understand you.


  41. How cultural evolution goes beyond (or extends) a Darwinian evolution model is the approach of Cultural Algorithms. On the difference between Cultural Algorithms and Genetic Algorithms in this framework:

    Evolution takes place at both the cultural level (the belief space) and the population level (for each individual). The belief space is the knowledge that is shared amongst the agents in the population. This model of dual-inheritance is the key feature of Cultural Algorithms, as it allows for a two-way system of learning and adaptation to take place. In a dual-inheritance system, the fit population members, as selected by an acceptance function on the basis of a fitness value, add their knowledge and experience to the belief space, thereby sharing it with all other agents in the environment. The belief space knowledge in turn helps guide the agents in the population from one generation to another. Other evolutionary approaches, such as Genetic Algorithms, allow for evolution to take place only at the individual (or population) level, i.e., they do not support a dual-inheritance system.

    A cultural algorithm framework allows for the knowledge of each individual agent community to be shared by the overall population, and also allow for the global knowledge to determine the evolution of that population. Such a formulation allows for the addition of various social, economic and biological factors into the population and observe and track the eventual evolution as a result of these additions.

    “A Survey on the Use of Cultural Algorithms in Multi-Agent Systems”
    “Cultural Algorithms: A Tutorial”
    “Cultural algorithm”


  42. Hi Thomscottphillips, given your reply to Massimo regarding a focus on population level changes, I am forced to ask again (now more forcefully) why Lamarckian is not an apt description?

    Just because it was not correct for biology does not mean it is not correct for cultural evolution.

    As I said, your attractor concept looks pretty similar, and since you are leaving out one of the main planks of Darwinian evolution while discussing something similar to Lamarcke sort of raises the question.

    Liked by 1 person

  43. Dan Sperber,

    Thank you for your clarification. That is not the common understanding of ‘epidemiology,’ but in the context it is a justified usage. As to Evolutionary Psychology, I remain skeptical, but am still reading on the matter, so we’ll see.

    Olivier Morin,

    Let me offer an apology to you, for taking the all too brief abstract I was able to find and reading too much into it.

    Thank you for the link to your paper, and have read it, and suggest that others here do as well. I found it well written, accessible, well researched and well argued. It was also more inclusive of interests in history than the abstract led me to expect.

    One of the mistakes I made in my original comments was failing to remember that research in one area, whether broad or narrow, by no means excludes research in other areas, and can be complementary to other forms of research.

    My own interest in the Gaze phenomena derives from semiotics, not Burckhardt. In semiotics the question is always ‘what does this signify?’ The question can only be answered by study of social context considering the expectations of interpreters of the sign. However, fuller elaboration of that in reference to the original article was not possible here; so, yes, I did indulge somewhat in cliche, in order to open up the possibility of viewing the matter from a different perspective. Certainly, adequate understanding of the innovations in the arts during the Renaissance cannot be reduced to development of individualism or the loosening of religious social constraints. *

    So understand that what I wrote was compacted due to space and time limitations. I’m well aware that cultural changes during the Renaissance were more complex than I could account for here. “Rich and powerful” is a thumbnail sketch; which I hoped to broaden by reference to the “explosion of commerce” during the era – but apparently this was ineffective. The “rich and powerful” of the Renaissance became so through the economics giving rise to a bourgeois culture that also allowed for the greater dissemination of the arts, which needs accounting for. Even painters pursuing religious-pedagogical purposes were pressed to sensualize and secularize their work to appeal to the appearance of a more sophisticated audience desiring pleasure as much as guidance.

    However, I believe that the research and argument of your paper can be useful in developing such a larger account. So again, I apologize for being too quick to react – and for not stating my own case well.

    Please bear in mind that comments here are rather more like “letters to the editor” than programmatic “responses” in scholarly journals. Most of us are writing fairly quickly using only knowledge we have ready to hand.

    That said, some good came from my misreading, as it brought forth comments from yourself and Dan Sperber, leading to greater clarity; as well as the link to your interesting paper.
    (* As to Korea, I cannot say, being unfamiliar with that culture.)
    (Fifth/final comment.)

    Liked by 1 person

  44. Philip,

    “I believe Daniel Dennett is applying “algorithmic” correctly with respect to cultural evolution”

    I’m not sure what a cultural algorithm is. If they existed, wouldn’t people have solved the problem of how to make something go viral at will?


    “we do not read “Darwinian” as synonymous with “natural selection”.”

    Right, but you should. It was that insight that made Darwin’s theory (common descent was already an accepted notion). And natural selection still is the only explanation for adaptation. The term “population thinking” was introduced by Mayr in the context of the Modern Synthesis, and at any rate is not what makes Darwinian theory Darwinian, since one can think in terms of Lamarckian populations as well.


    “I am somewhat puzzled by the hostility that some commenters here display towards the concept of memes.”

    Well, it’s essentially a dead field, as recounted in this recent article:


    “I still don’t understand your objection to accepting that “survival of the fittest” is tautological.”

    It isn’t just me. The theory as stated is tautological. Which means it is true by definition. Which means it isn’t that interesting until one fleshes out the fitness bit, for which one needs a theory of functional ecology. Which exists in biology, but not in cultural studies.

    “since “survival of the best at surviving” is tautologically true, any definition of “fitness” in which SotF is *not* tautological would then be false”

    Really? So you think any scientific theory that is not tautological is false? Or are you instead suggesting that all scientific theories are tautological?

    Liked by 1 person

  45. To my opinion there should be made differentiation between cultural phenomena that is not transformable to other group, because it has no universal tools of communication, and cultural phenomena that can out reach to other social groups by universal communication means.
    The first type of culture distributes itself only within the unique social group, and propagates by missionary persuasion, sometimes peacefully, but many times using force on individuals or social groups to join the group. The basic atom of this culture is usually faith. Such a culture, if it has higher propagation rate, tends to annihilate other cultures, and usually replaces them. It is usually very intolerant to other cultures.
    The second type of cultural phenomena is rather supplementary cultural phenomena, that can live side by side of any cultural phenomena. It propagates by its own attractiveness, like junk food culture, clothing, generally pop culture. Usually this kind of cultural phenomena is harshly opposed by the first type of cultural phenomena, because it sees is as penetrative to its purified cultural values. (The Communistic opposition to jeans or western pop music )
    As to the money, it is a very unique and interesting cultural phenomena. It is on one side voluntarily adopted cultural value, but still it has the power to enforce itself upon the members of any cultural membership. Money has universal power of penetrating into any cultural type of social grouping.


  46. Massimo: “I’m not sure what a cultural algorithm is. If they existed, wouldn’t people have solved the problem of how to make something go viral at will?”

    I gave references to cultural algorithms (CAs) above. It is a subject that’s been around for years in the AI world. Here’s another reference:

    “Knowledge Sharing Through Agent Migration with Multi-Population Cultural Algorithm”
    Proceedings of the Twenty-Sixth International Florida Artificial Intelligence Research Society

    To answer your question” Perhaps they (CA researchers) are a bit closer to solving that.


  47. Hi thomscottphillips,

    I would agree with Massimo that defining “Darwinian” as broadly as you do is unhelpful, since what is distinctive about Darwin’s explanation is natural selection.

    Hi Massimo,

    The theory as stated is tautological.

    I don’t regard “survival of the fittest” as a theory, rather it is a tautological commentary, an aside, a pointing out of a tautology to help people understand the implications of Darwinism. The theory is the three points by Lewontin that you listed.

    Really? So you think any scientific theory that is not tautological is false?

    Not at all, it’s that I don’t regard “survival of the fittest” as the theory.

    But the point is this, we agree that: “Survival of the best at surviving” is correct, tautologically so. Thus “survival of the X” would be correct (tautologically so) if X is “best at surviving”. That means that if X is anything else, then it is less correct, at most it is only partially true. Thus, for example, “survival of the fastest runner” would be partially true, but only one aspect of the truth.

    Thus if “fittest” means anything else than “best at surviving” then it is only partially true, and indeed is only true to the extent that it means “best at surviving”.

    Of course one can properly examine what characteristics make for good survivors in any given niche, and there is nothing tautological in listing those characteristics, but the only way of listing those “fitness” characteristics is by discerning what characteristics lead to survival, which means that “fitness” has to be defined that way and thus SofT has to be tautological.

    Liked by 1 person

  48. Massimo,

    I am aware of that, but that doesn’t explain the hostility towards the concept. There are many terms that don’t have an official, quantitative science and a journal attached to it but that are still useful to describe aspects of reality and don’t meet with angry dismissal by so many people.


  49. Philip I know what an algorithm, is. My eye roll is at the idea that evolution can be reduced to an algorithm, and my bigger eye roll is that, per “Darwin’s Dangerous Idea,” all Darwin-like, or quasi-Darwinian evolution, is algorithmic. It’s exactly why I also do an eye roll whenever Dennett talks about “greedy reductionism” and fails to include himself.

    I had halfway, or more, in earlier readings in modern philosophy, agreed with what Dennett had to say up to that book. He lost me after that.

    Specific to this issue, I hit the Wikipedia link and grokked the start of the tutorial link.

    The main basic issue/problem I see is the one I mentioned as part of my first comment. I don’t see culture as having discrete units, whether memes, or something else. I noted this under this issue of “blending.” Such blending and development is not necessarily — and usually is not — a Hegelian process.

    Wikipedia kind of gets at this itself. It notes that cultural algorithms are part of evolutionary computation. At that link, among other things, I find two paragraphs which I connect:

    In computer science, evolutionary computation (a.k.a. evolutionary computing) is a subfield of artificial intelligence (more particularly computational intelligence) that can be defined by the type of algorithms it is concerned with. These algorithms, called evolutionary algorithms, are based on adopting Darwinian principles, hence the name. …

    Evolutionary algorithms form a subset of evolutionary computation in that they generally only involve techniques implementing mechanisms inspired bybiological evolution such ss reproduction, mutation, recombination, natural selection and survival of the fittest.

    Dunno about others, but to me, it seems this is a bit of the cart before the horse. One assumes that cuture evolves Darwinistically, and one then and thus derives cultural algorithms.

    Another problem I see is that no individual is a product of, or inhabitor of, one cultural space. When A interacts with B in cultural space F, A also has bits of cultural spaces 1, 2, 3 and B has bits of cultural spaces 4, 5, 6. At a minimum, the Darwinian analogy to culture would have to include an analogue to epigenetics as well as genetics.

    Anyway, I’ll read further as I have time, but I right now stand by former Sen. Arlen Specter’s noting that Scots criminal law has a third verdict option of “not proven.”

    Otherwise, on memes? Sorry, Alex, but I don’t see them useful in any non-trivial way.


  50. To Sperber et al
    Tho I have much praise of the article, not everything was great. The full frontal gaze bit was confusing IMHO — a case of where my creative writing teachers would say “You need either more or less of this”. Really develop it or take it out (though leaving it in had the unexpected benefit of bringing Olivier Morin into the conversation).

    I’m still struggling with the distinction between selectional and replicative. “a population that *renews itself through the reproduction of its members is subjected to Darwinian *selection* …” “a population is replicative if heritability is secured by some form of replication”. OK “renews itself through … reproduction …” vs “heritability is secured by some form of replication”. I think you should have stopped right there to clarify that distinctions, and frankly every time I think I’m close to getting it, I don’t after all.

    I really appreciate the authors staying with the conversation. Some others haven’t participated at all, and agree with him or not, Sperber is, in my time at least, the most prominent guy we’ve had in this forum.

    Also appreciate the modest with which the “ECM” concept was presented. It’s quite the fashion these days to put computer simulations in social epistemology (Goldman faction) papers, and presenters rarely bother to situate them so appropriately in the scheme of things, and they’re usually much more “toy” constructs than this one.

    For the first time I think I got some intuitive grasp of the “attractor” concept and its potential breadth of application. Based on the following, someone may tell my I didn’t really get it but:

    1) Orthodox spelling is an attractor which tends to limit the drift of text
    2) The color wheel is an attractor
    3) Poetic form is an attractor (favored by illiterate and marginally literate peoples to boost reproducibility of important texts)
    4) 1 & 3 are somewhat analogous to check-sums in computer data, and 1-3 could help instantiate a class of deviation reducing attractors.
    5) The set of phonemes of a language might be considered “digitalization attractors”. These actually actually permit conversion between the binary (any alphabetic regime is equivalent to binary or digital) and the continuously varying. A computer modem is a simpler version of the same thing — at least the input side is an attractor.
    6) The redundancy and the “alphabetic” nature of DNA could also be part of the deviation reducing class of attractors, and might be viewed as helping instantiate alphabetic regime attractors (no continuously varying anything to translate to/from).
    7) Ray Jackendoff (Foundations of Language) suggests that what can be salvaged of Universal Grammar can be best seen as an attractor.


Comments are closed.