If you prefer to listen rather than read, this blog is available as a podcast here. Or if you want to listen to just this post:
A few posts ago I talked about memetic evolution. As a result of this post one of my readers, Mark, and I had an in depth discussion about what mechanism, exactly, I was trying to describe and whether there really is such a thing as memetic evolution. Mark is a scientist specializing in oncology research (he also has a blog, which you should check out) and he pointed out that evolution is exceptionally complicated and that many people use the term to describe lots of things that aren’t actually evolution by natural selection. Particularly when they’re trying to use it by way of analogy which I was. As part of our discussion a lot of things were clarified for me, and I think I’ve tightened up the analogy and hopefully gotten rid of most of the issues Mark pointed out. This post is about sharing the additional insights which came out of that discussion.
I.
Mark was, of course, correct, there are in fact lots of pitfalls involved in the discussion of evolution and selection, and even if you manage to avoid making any big mistakes there are still numerous specifics that can trip you up as well. For example, most people don’t realize that there are two competing theories regarding the rate at which evolution occurs. And the difference between these two theories turns out to be very important. Not only in general but also for the point I want to make.
The first theory, and the one initially put forward by Darwin, is phyletic gradualism. Under this theory the creation of new species happens very gradually, almost imperceptibly as small changes accumulate over tens of thousands of years. Because of how gradual this process is, you might not end up with a clear line where you can say that one species has changed into another, and, insofar as a layman thinks about evolution with any rigor, they probably envision it working something like this.
The second theory, which was proposed only in 1972, by Niles Eldredge and Stephen Jay Gould, is called punctuated equilibrium. This theory holds that species appear fairly suddenly in response to some rare and geologically rapid event (the punctuation) and that once a species appears that it ends up being relatively stable (the equilibrium). To reiterate, I’m no expert, but it’s my impression that this theory has the most support among scientists, particularly when you’re talking about the big evolutionary events, like speciation. To be clear both kinds of evolution, gradual and punctuated, appear to be taking place, but the latter is more impactful, and more important, particularly when the survival of a given species is really in question.
Having, hopefully, grounded our understanding and discussion of evolution on a somewhat firmer footing, we are still left with the question of how much of that understanding and discussion maps cleanly to the topic of cultural evolution, and beyond that to the more speculative topic of memetic evolution. For instance, insofar as cultural evolution is adaptive, is this adaptation gradual? Or does it operate more along the lines of the punctuated equilibrium model? I’m not entirely sure what would count as hard data when considering these questions, but at the level of anecdote, I’m inclined to believe that the situation is similar to genetic evolution, both forms occur, but that the cultural selection which occurs gradually ends up being less impactful than cultural selection which happens at times of rapid change and extreme crisis.
As I said this is mostly at the level of anecdote, but consider the example of Germany. It’s hard to argue that Germany didn’t have a long martial tradition, starting with their first appearance in the records of the Roman Empire and continuing down through the centuries to the two World Wars. Would you say they still have that culture today? I think most people would agree that they don’t, and that it all changed during the extreme crisis at the end of World War II. Sure there have been many gradual changes to German culture over the years, but the fact that there’s also numerous long-standing stereotypes about Germans would seem to indicate that a relatively stable equilibrium existed as well. From where I sit, this example has all the elements you’d expect if cultural evolution also happened according to the punctuated equilibrium model.
Another example would be the creation of the United States of America. Evolution through natural selection concerns itself with the creation of new species. The parallel in cultural evolution would be the creation of a new culture or nation, and this is an example of exactly that. And, once again, it happened over the course of a few years where things were rapidly changing under crisis conditions. Additionally what resulted was not some incremental change in English culture (though there are obvious connections) but an entirely new culture forged in the fires of the Revolutionary War and the many debates over governments and rights.
The more I consider the question the more I am convinced that there are numerous examples of punctuated equilibrium with respect to cultural evolution. I suspect all of the examples of nations in crisis given by Jared Diamond in his recent book Upheaval (see my review), would end up being examples of the punctuated equilibrium model of cultural evolution as well. And of course these are successful “mutations”, if cultural evolution is anything like biological evolution most mutations are going to end in failure. Is that perhaps the best way of describing communism and fascism?
Obviously not all cultural changes are so large, as I said, I’m sure that things also change gradually, but we would appear to have less to fear from those changes. Sure the vast majority will fail just like all “mutations”, but that failure should be much easier to recover from. Much less disastrous than the analogous “speciation” of adopting something like communism.
II.
If you’ve followed me this far and you accept (even if only for the sake of argument) that punctuated equilibrium applies not only to biological evolution, but to cultural evolution as well, then we’re finally ready to revisit memetic evolution, though right off the bat I’m going to dump the word “evolution”. One of Mark’s bigger contributions in the discussion we ended up having was to point out that once we’ve reached this point that things have been stretched so far that using the term evolution conceals more than it reveals, particularly if we’re more interested in the punctuated equilibrium model of evolution. So we need a new term, but before we get to that what exactly are we talking about? And, in what sense are we talking about something separate and interesting?
When considering the punctuated equilibrium model most of the attention ends up on the punctuation part, but what’s happening during the equilibrium part? Here I’m going to quote liberally from Mark:
[The punctuated equilibrium model] posits [that] major selection events might be somewhat uncommon. As such, we would expect to see accumulations of multiple different mutations, all present in a species’ gene pool simultaneously. The longer the period of time free of selection, the greater the potential for diverse new mutations within the species. Since anything directly lethal is going to weed itself out fairly quickly, this enriches for potentially-beneficial mutations. With all these mutations lying around, it’s possible for individuals to even have two or more traits that might not be adaptive on their own but that function very well together. This period of stability can be thought of as ‘good’ in that is allows for much greater variability to enter the population.
Along comes the selection event – the filter, removing anything that can’t pass through a particular challenge – and most of that diversity disappears. However, since the population experienced a long period of growth and mutation without being subject to a filter, it’s possible that the adaptation that made it through the filter is more complex – is a bigger change – than the kind of single-mutation adaptation you would see from a series of rapid filters. Populations that instead pass through serial filtering events will only be able to select based on single-mutation traits.
….We expect to have multiple possible pro-adaptive traits at any given time, waiting to pass through the next, unexpected, filter and join future generations. Thus, memetic evolution is simply a sub-process of cultural evolution. It would be as meaningless to speak of it in isolation as it would be to talk about accumulating mutations prior to selection events (filters) when speaking of biological evolution.
…Memetic ‘evolution’ is simply another name for cultural evolution prior to selecting events.
Some of this is obviously speculative, but on the whole Mark’s comments were fantastic, and really helped me to understand something that had previously eluded me, and I agree with everything he said, with one exception… I don’t think it’s “meaningless to speak of it in isolation”. I think “it” is very important to talk about. What is “it”? What is this thing that’s worth discussing, but which is not evolution? I’m going to call it “memetic accumulation”.
III.
For most of history the rate of accumulation for genetic mutations has probably been fairly static. I assume that during periods of greater radiation (if any) that it might have increased, or perhaps the greater the variety of life the greater the space for mutations to occur and perhaps there are other factors as well, but I don’t see any evidence that there were periods where it was significantly faster or slower. There is the Cambrian Explosion, but remember we’re talking about the rate of accumulation, not the rate of evolution or of speciation, and while it was an “explosion” for many things, I don’t think it was an explosion in the accumulation of mutations. In other words I think the rate of mutation accumulation with natural evolution has been pretty constant.
Even when humans entered the scene and started the selective breeding of domesticated animals, this didn’t change the mutation rate, even for the animals in question. (CRISPR, however may be another matter.) We just introduced a lot more filters and selection events. So, if mutations are relatively constant in natural evolution, what about cultural evolution? Has that rate also been constant? I would argue that it hasn’t, and this, more than anything else, is why it’s worth discussing. I suppose, given the fact that humans can introduce new ideas, new potential memes into the space of culture whenever they feel like it, that there are a great many things which could affect the speed at which memetic accumulation occurs. But certainly technology and progress has to have a large impact on that speed, and almost exclusively in the direction of speeding it up. In fact, “something which speeds up the rate of memetic accumulation” is not a half bad definition of progress. But beyond that, might technology and progress have any other effect than generating ideas quickly?
With the advent of global communication and social media, we are moving ever more rapidly in the direction of creating a single ecosystem for ideas, and I don’t think we’ve fully come to terms with what that means or how it will play out. Certainly ideas propagate faster, and I would also say we end up with a handful of “apex ideas” similar to the idea of an apex predator. Which is to say that we’re in a space where a memetically fit idea is able to very quickly outcompete all the other ideas among people susceptible to that idea. (Notice the increase in the number of people who believe in conspiracy theories.) Leading to a stratification at the level of ideas rather than at the level of a community or nation. Basically, social media and global communication have allowed invasive species/ideas to go everywhere.
On top of all this there’s one final thing which needs to be pointed out, humans are more removed from issues of actual survival than ever before. Toss all of this together and we have rapid memetic generation, but which results in a relatively barren collection of a few dominant memes/ideologies, none of which are likely to have anything to do with actual survival. Now I’m aware that this is something of an oversimplification, culture is still complex and varied, and people still worry about survival, but we have nevertheless lost an awful lot of both those qualities.
Finally, if I’ve convinced you that memetic accumulation is speeding up, then even if you disagree with me about everything else, you might at least want to examine what the potential consequences of that are with respect to cultural evolution.
IV.
Having examined what the modern state of memetic accumulation is within the equilibrium part of the model, what does all of this mean for the eventual “punctuation”? How does our rapid, barren and superficial method of memetic accumulation play out when we actually run into a selection event? Into rapidly changing crisis conditions? Well that’s hard to say, though none of those elements would appear to be positive.
Just by itself, the rapid part isn’t necessarily bad. Perhaps if culture is moving rapidly, then, by the time the eventual crisis rolls around, we will be in some location uniquely well suited for surviving that crisis, a location we would not have reached had we not been moving so quickly. And certainly if there were a bunch of cultures all speeding off towards their own unique locations we might have some expectation that at least one of these locations would be exactly the spot they should be in, but this is where the lack of variety comes into play, we’re not all choosing different locations where we can survive the potential crisis, we seem to all be journeying as quickly as we can towards a small handful of locations, and the rapid bit means if it’s not the right place we will have gone an awfully long distance in the wrong direction. Furthermore, what do these locations look like? If we were really concerned about survival, they would hopefully be strongholds, but if we don’t factor in survival I would think they’re more likely to end up looking like expensive penthouses. Dwellings which look really nice and are great for entertaining, but also the last location you’d want to be in when the zombie apocalypse starts. There’s obviously still a lot of variety in these dwellings, but can anyone honestly tell me we’re not building a lot more penthouses than strongholds these days?
There also seems to be significant effort being spent on getting people to abandon locations which proved to be strongholds in the past. I think I’ve already said this, but it bears repeating, there are essentially three ways to choose a “location” we can choose them randomly, which is essentially what natural evolution is doing. We can choose one based on whether it sounds good or not, but in this sense, as I already pointed out, we’re probably not choosing a stronghold so much as a nice place to live. Or we can choose one based on what’s worked in the past. Any option where we choose is going to be better than random (one would hope) but it’s not clear to me that “sounds good” is definitely better than “worked in the past” (in fact, I strongly suspect it’s worse) and in any event it’s probably best to have cultures in both types of locations.
To be clear, we don’t know which location will best withstand the eventual crisis, because we don’t know what that crisis will look like, but you could certainly see how changing the way in which memetic accumulation happens could change the likelihood of being in the right location. And I hope we can agree on this, even if you don’t agree with me on exactly how memetic accumulation has changed
But beyond all this, there’s probably more bad news, particularly if you believe Nassim Nicholas Taleb’s contention, that in addition to changing to speed of memetic accumulation, that progress and technology has also changed the nature of potential crises as well. That we have made them less frequent, but in the process we’ve also made them larger. As a real world example, lots of people feel that there is no safe location (both figuratively and literally) if the crisis ends up being full scale nuclear war or runaway climate change (I disagree, but I’ve already covered that in past posts). Both crises that have only been made possible recently.
I will freely admit that I’ve followed a long chain of assumptions to get to this point, but strip all that away and I would contend that the two initial ideas, 1) that cultural evolution also follows a pattern of punctuated equilibrium, and 2) that technology and progress can change the rate at which cultural mutations/memes accumulate, are both pretty solid. And both of those together should be enough to introduce serious uncertainty into any claims that conditions are following some long-term, unstoppable, positive trend.
A couple of final things to think about, which I leave as an exercise for the reader:
Are we at a point of “punctuation” right now? If so how’s it looking?
Could memetic accumulation get so out of whack that it actually causes the crisis?
With this post I’ve gone a long way down a pretty obscure road. It you like that sort of thing consider donating. If you don’t like that sort of thing you should also consider donating, but I think it will make less of a difference.
I’m not quite clear what we mean by ‘genetic mutations accumulating’. Accumulating where? If you choose some arbitrary point then two things can happen. Genetics will change from that point or genetics will be highly conserved. Show dogs would be an example of the latter. Since their breeding is always highly selective for ‘purity’, their traits will not vary much from whatever baseline you declare to be ‘the original’. On the other hand a show dog sold to a guy who likes to breed dogs but doesn’t really care about purity will pretty rapidly start showing dramatic changes in his descendants.
The baseline, though, seems totally relative and arbitrary. There also seems to be little in the way of ‘accumulation’. Dogs, show or not, will have roughly the same amount of DNA as their ancestors. “Accumulating” sounds like books in your private library. As you get older and keep buying books, your library gets bigger and bigger thereby increasing the odds that at some dinner party where the Internet goes down, you are likely to find a book to settle some obscure argument going on.
But species do not just add DNA except under some exceptional events. Clearly there’s a carrying capacity in always trying to make copies of longer and longer DNA so new code discards old code. It’s like a private library where you’re only allowed to keep 100 books. Maybe all your books will be brand new or maybe many old but you are either ditching old or new books to keep the library constant.
Communications may be better than before but there is a finite limit to ‘culture memes’. You can, for example, only read so many books in a year (or lifetime). If you read 100 blogs today, you are reading fewer books. Watching more TV means fewer movies and vice versa. You can socialize with a lot of people lightly or a few people intensely. Time remains limited for humans hence serves as a constant check on any culture that can be produced or consumed.
You have to think in terms of the genetic diversity of the population, not of the individuals. If you took purebred beagles and selected one new mating pair from each litter every generation to establish the next generation (adopting out the rest – though in the lab with mice we just kill them, or ‘sacrifice’ them) you’d get genetic drift within the mating pairs just due to normal genetic mechanisms and random mutations. In your example above, both the purist method and the enthusiast method will result in the same quality of genetic drift over time.
Your sense that it doesn’t really matter what your selection criteria are, you still won’t get any kind of ‘accumulation’ is correct. Because that’s not what we mean when we talk of accumulation. If your selection criterion was height, you’d expect to see the average dog height grow over time. But each time you select a breeding pair, you also discard the mutations in all the other pups in the litter you didn’t allow into the breeding pool. Each selection event – whether artificial or natural – reduces the total number of unique genetic mutations in the population.
However, if you took the same pair of beagles to a massive prairie and let them all interbreed for 10,000 generations you’d end up with a huge number of drifting genetic lines all interacting in dynamic ways. Dogs could be born from two separate lineages that developed multiple random mutations independently. In essence, mutation now happen both in serial and in parallel. This is because you’re keeping many of the new mutations produced by each successive generation, instead of discarding them through selection.
If your criterion was still based on height, you’d have only one selection event you’re working from, but a richer diversity of genetically exchanged material randomly distributed through the canine population. Your resultant dog would likely not be as tall as the dog subjected to serial selection events, and the genetic strategy it uses to achieve greater height is likely to be different as well.
This is an example where I’d expect to see genetic drift excel, compared to punctuated equilibrium, at achieving a targeted outcome. It’s the classic story of the giraffe’s neck (though there’s a complex biophysiology side of the giraffe story the drift case doesn’t address, but that’s a different debate).
That’s because drift works great for incrementally escalating selection events. Think of any biological arms race in continuous competition for resources. It works less well for more abrupt, complex events (i.e. major disruptions, and nearly all biochemical pathways).
Indeed, part of the allure of the theory is that it begins to make evolutionary theory relevant again to anyone working at a biochemical level. Of course natural selection and common descent are pivotal concepts used in everyday lab work. But when you work with nothing but very complex interacting pathways, the vanilla evolution explanation you read about in high school textbooks stops making any sense. It certainly doesn’t do any explanatory work.
It’s possible for a single pair of dogs to produce single mutations, or even the rare paired mutation. But very few biochemical pathways resemble anything that could be produced through that kind of process. Your body isn’t an accumulation of hundreds of thousands of single elements doing their thing; it’s massively complex pathways that are interdependent, multi-step, and indirect. You can’t get there through gradualism, so some mechanism has to explain the complexity (actually multiple mechanisms, but I won’t go there).
Much of this requires selection against a larger pool with a more diverse set of mutations that have the benefit of developing together and intermixing throughout the population in lots of different ways. Once a selection event reduces the population size to only those organisms that carried the rare trait (potentially a combination of genes), you’re once again reduced to working from a more limited genetic pool. Hence, it will take time to generate the genetic diversity and interaction you enjoyed before the selection event, but now all organisms within the community share the trait that provided survival benefit.
TL;DR – Accumulation of genetic mutations is a function of the size of the population and the amount of time since the last selection event. It is meaningless at the level of the individual.
*quantity, not quality.
I’m wondering how much of this applies to ‘memetic evolution’ or if we have mined an analogy dry?
I think this is at the heart of what the OP is trying to say:
– Insistence on cultural conformity is selecting on arbitrary criteria and potentially reduces adaptivity down the road.
– Cultural development in a digital forum may be a different type of memetic accumulation that crowds out real-life cultural development
In other words, it’s at the population level that we may be creating dangers as technology creates a different dynamic. And you can’t know whether this is a problem until a selection event, at which point it’s too late to go back and develop the diversity of culture that might have contained the cultural adaptation we needed to survive. Very close to the theme of the blog in general.
An analogy I have heard used with evolution is a search method. You are on an alien landscape and want to be as high as possible. One method might be to adopt an algorithm along the lines of:
Walk in a random direction. Are you going up? Then walk again in a random direction but increase the weight of the previously successful direction. Are you going down? Then do the reverse. If you are in a landscape with one tall peak, this will almost certainly get you to that peak. However if you are in a landscape of many hills, you can easily end up on a hill that is not the tallest but the only way to find a taller peak would be to sustain a long period of going down.
We don’t know the full geography, though. If there’s one peak then you’re going to get cultural convergence. For example, if you feel capitalism is the only really viable economic system, then over the long course of history all economies have to converge to capitalism. If equality is a vastly important value, then all societies will converge upon it including, say, SSM.
I think conservative minded thinkers are in a bit of a bind here. You can talk about the “laboratory of democracy” meme to praise diverse experiments in social policy but labs converge upon standards. There are thousands of medical labs yet if you sent your blood to them to be typed you would not be happy with a ‘diversity of responses’ the way you are ok with a diversity of blogs on the Internet.
I think we can accept a ‘hilly landscape’ when it comes to aesthetics. We can accept the greatest pieces of classical music and the greatest pieces of jazz are very different but there’s no easy to way to rank one higher than the other. We may end up on top of different hills depending upon random factors like where on the map we started our random walks. But I’m not convinced conservatives or even liberals are open to accepting this in terms of larger cultural evolution. If there is a highest peak out there, we are all destined to get closer to it and milling about different spots in the middle is a failure, not a feature.