How darwinian is population thinking? (applied to culture)

I am reading Tim Lewens’ new book Cultural Evolution (I am still halfway through it, so perhaps other posts will follow). One aspect I found interesting – and I imagine this will not come as a surprise to readers of this blog – is a peculiar conception of what identifies an approach to culture as “evolutionary”.

Lewens attacks the problem from a familiar perspective (see, for example, here), defining a taxonomy of possible evolutionary commitments for cultural explanations. As usual, some of them are too broad (Lewens calls them “historical” approaches), while others are too restrictive (he is skeptical of replication-based accounts of cultural evolution): things are more interesting in the middle.

A first difference in respect to what is, at least for me, a standard view (see for example this recent paper), is that Lewens considers the approach of Boyd, Richerson and colleagues as “not-selectionist”. In any case, this is not the topic of this post, and Lewens might have good arguments to defend the assertion that “claims about selection are only of secondary importance to them, given their primary motivation for taking an evolutionary perspective on culture” (p. 16). So what is this motivation? As Boyd and Richerson asserted in several places, biology is fundamentally darwinian because its explanations are rooted in population thinking, and cultural explanations are darwinian as long as they are rooted in population thinking too. This seems to be a quite uncontroversial position. Sperber and colleagues, which disagree on various points with Boyd and Richerson, nonetheless recently wrote (here):

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]).

They continue giving a somehow standard definition of what population thinking is, namely:

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.

Starting by saying that I am not an expert, the concept of population thinking (as used in cultural evolution) has actually always puzzled me, and reading this book perhaps helped me to understand why. As the story goes for biology, population thinking represented a potentially revolutionary, counter-intuitive, position against a typological/essentialist view of natural history. But what about social science? This is how Lewens defines population thinking (p.17):

the assumption that […] large scale patterns extending across time could be explained in terms of the aggregated effects of many small-scale events that occur in the lives of individual organisms.

While there are countless ways a social scientist can be against this assertion, clearly this has not much to do with “evolution”, and in fact Lewens draws an analogy with the kinetic theory of gases, where population-level features (e.g. temperature) are the aggregate outcomes of individual-level interactions, concluding that “there is no particular reason to call a populational approach ‘Darwinian’: we could just as well describe it as a kinetic theory of culture”, which is what he does indeed.

So what does remain of cultural evolution, if one is only committed with population thinking?  While Levens can call these approaches “kinetic”, a big all-capitals “Cultural Evolution” is on the cover of the book.  His idea is that these approaches combine population thinking with an adaptationist stance. In the case of Boyd and Richerson, this is the fact that social learning (and the general-domain mechanisms that implement it) is itself an adaptation, while in the case of Sperber it is a more general commitment to ‘standard’ evolutionary psychology to explain human cognition.

While this picture is mostly correct, it is, at least from my perspective, slightly intellectually unsatisfying. In my view, an evolutionary approach to culture should be committed to the idea of the process of cultural change being (broadly) darwinian. What are the options then?

  1. We accept that views of cultural change that are not committed to more than a general idea of population thinking are not, in fact, evolutionary.
  2. There is a way to define population thinking which is both coherent with what we know about cultural change and enough restrictive to justify the term cultural “evolution”.
  3. There is a way to define “evolution” interestingly, that is not related to population thinking, and that is well suited for culture.
  4. We adopt a view of cultural change in which selection, for example, has a strong enough role (not necessarily as strong as in biological evolution) to, again, justify the usage of the term cultural “evolution”.
  5. We use “evolution” in a broad sense (i.e. not implying analogies in the process)

As I mentioned, (5) is not very satisfying for me, but perhaps there are good pragmatic reasons to support something similar – and perhaps there are good pragmatic reason to not support (1). I’d be happy to know more about (2) and (3). I defended, indirectly, a version of (4) in this paper, and, if someone does not convince me of the contrary, I will probably remain on this position!

9 thoughts on “How darwinian is population thinking? (applied to culture)

  1. Quick comment here, Alberto.

    First, what really hangs on the term ‘evolution’ for social scientists like yourself, Mesoudi, Boyd, et. al? There are so many processes involved in culture (whatever that is) and cultural change (however that happens), that I struggle to make sense of your commitment to the idea that the processes of cultural change are in some way Darwinian.

    I take it that Lewens’ point is to be relatively persnickety about the term ‘evolution’ and ‘selection’, partly because so many cultural evolutionary theorists use the term in a relatively loose manner. If I understand Lewens’ point, its that that for a process of evolution by natural selection, one needs retention of heritable variants through differential reproduction — and that it’s less than clear whether many of the phenomena cultural evolutionary theorists talk about have this ‘selectionist’ character.

    1. Hi Andrew, thank you for your comment. I think it is perfectly fine being persnickety (new word of the day for me!). So you would say that, according to Lewens, the views on culture of Mesoudi, Boyd, Sperber, etc. (and, well, myself…) can not be really defined as “evolutionary” (as long as they do not commit with a strong-enoguh selectionist stance)? This would be basically my option (1) above.

      1. Not necessarily. Lewens does point to areas where Boyd and Richerson (2005) discuss phenomena that are properly selectionist: some cultural phenomena (lets say behaviours) may be associated with individuals with higher biological fitness, and spread as a direct consequence of such fitness. On the flipside, some behaviours may fail to spread because of the low biological fitness of its bearers. So there are clear ways in which cultural evolution may intersect with selectionist models — Henrich’s model of the loss of the Tasmanian tool-kit may be one such example, Sterelny’s (2012) discussion of the extinction of the Neanderthals, another.

        But I think Lewens is right that many models of Boyd, Mesoudi, Henrich and all that bunch are not selectionist in this way. Does this mean they are not evolutionary though? Only if one narrowly defines evolution in terms of natural selection. This is why, I take it, Lewens is careful to use the term ‘selectionist’ rather than ‘evolutionary’. And maybe this is part of the confusion between you and me Alberto: it seems like in the discussion above, you may be identifying ‘evolution’ with ‘natural selection’, is that right? As I see it, Lewens’ discussion doesn’t really weigh in on any of the options you’ve put forward, because he isn’t committed to claims about ‘evolution’ — only claims about natural selection or ‘selectionism’ and their presence in models of culture.

  2. Thank you for your clarifications Andrew.
    Regarding the “properly selectionist” phenomena in Boyd & Richerson: when I talk about “selection” – here and elsewhere – I am not talking about the effect of natural selection on traits’ bearers, but on *cultural* selection i.e. whether one assumes that cultural traits exhibit variation, heritability and fitness differences. This seems to me the standard way to use the term in cultural evolution, and it is also how I interpret Lewens usage of the term (see, for example, pp.14-15: “A cultural selectionist […] is someone whose evolutionary approach to culture is justified on the grounds that cultural entities are engaged in a form of competitive struggle”).
    This does not change anyway the fact that, as you write, “Boyd and Richerson (2005) discuss phenomena that are properly selectionist” and that, more generally, some cultural “evolution” models are selectionist and other are not (keeping in mind my different usage of the term selection), so I can also expand on the second part of your comment:
    No, I do not identify ‘evolution’ and ‘natural selection’. The main point of the post is in fact how to identify ‘evolution’ (in cultural evolution).
    You are right that Lewens does not weigh in on any of the options I’ve put forward: this is my question. If we agree that most models in cultural evolution are not-selectionist (let’s not worry now about the definition of “selection”) shall *I* call them “evolutionary”? Is it enough a commitment to population thinking (especially as Lewens defines it, i.e. with the “kinetic” analogy)?

    1. Hey Alberto! Thanks for your post.

      Okay, I think I understand your position a little better now.

      I think the trouble, one that memeticists, cultural evolutionists, and the like have struggled with is to make explicit what it means for culture to have heritable differences in fitnesses. It runs head on into the problem of how to differentiate cultural units, what it means to be heritable, and what it the term fitness means. This issues remain even without weighing in on whether there is some limited resource being competed over (as Alex Mesoudi seems to suggest).

      I think the position you struck in the Acerbi and Mesoui (2015) article is a reasonable one: there might not be a privileged level or way of understanding the processes of cultural change. Maybe some phenomena will appear more ‘selectionist’ than other when viewed at different levels of spatiotemporal grain — thus the safe position might be one of ‘cultural evolutionary pluralism’ (maybe a sixth option to add to your list above?), where one allows for multiple, even competing accounts of evolution to be applicable to the same phenomena, granted they provide some useful empirical means of latching on to the phenomena.

      At least in my understanding of the field, this is what we confront. There are some instances (like template copying) where cultural items seem to be clear Darwinian individuals. There are many other instances where cultural items seem to be tested and trialled by humans, with only some aspects of them picked up and passed along. There is also an aspect of guided variation, or creative intelligence at work in some behaviours. So I think the problem isn’t just one of what evolution means in the cultural evolutionary context, since there are likely to be several different kinds of evolution, but where these different definitions might be applicable, and thus, what kinds of models and tools one should bring to bear as a result.

      And just as a coda, I think that Lewens’ kinetic analogy really only works for the models of Aoki, Feldman, Laland, Boyd, Richerson, Henrich et al. who use formal mathematical models to derive results in a way akin to population genetics (which is still a sizeable portion of the cultural evolutionary market). Lewens would also think that the kinetic analogy would be applicable in population genetics, and he wouldn’t be the first to argue as much. R.A. Fisher developed his fundamental theorem at least partially by analogy with the kinetic theory of gases. So there is a clear precedent in what Lewens is talking about.

      1. Good, I think then we are more or less on the same page here! Just allow me to add a coda to your coda…I do not have problems with Lewens’ kinetic analogy, and, as I wrote in the original post, I think it actually helped me to clarify my problems with the usage of “population thinking” concept in cultural evolution. My point is, if we accept that the kinetic analogy is a good analogy for population thinking (and I might be happy with this, but I am not an expert), then there is no reason, as Lewens notes, to call “evolutionary” an approach to culture that is only committed to population thinking (and that is not selectionist, etc.).

Leave a Reply

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

WordPress.com Logo

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

Twitter picture

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

Facebook photo

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

Google+ photo

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

Connecting to %s