2 min read

13

"All models are wrong, but some are useful" — George E. P. Box

As a student of linguistics, I’ve run into the idea of a meme quite a lot. I’ve even looked into some of the proposed mathematical models for how they transmit across generations.

And it certainly is a compelling idea, not least because the potential for modeling cultural evolution alone is incredible. But while I was researching the idea (and admittedly, this was some time ago; I could well be out of date) I never once saw a test of the model. Oh, there were several proposed applications, and a few people were playing around with models borrowed from population genetics, but I saw no proof of concept.

This became more of a problem when I tried to make the idea pay rent. I don’t think anyone disputes that ideas, behaviors, etc. are transmitted across and within generations, or that these ideas, behaviors, etc. change over time. As I understand it, though, memetics argues that these ideas and behaviors change over time in a pattern analogous to the way that genes change.

The most obvious problem with this is that genes can be broken down into discrete units. What’s the fundamental unit of an idea? Of course, in a sense, we could think of the idea as discrete, if we look at the neural pattern it’s being stored as. This exact pattern is not necessarily transmitted through whatever channel(s) you’re using to communicate it — the pattern that forms in someone else’s brain could be different. But having a mechanism of reproduction isn’t so important as showing a pattern to the results of that reproduction: after all, Darwin had no mechanism, and yet we think of him as one of the key figures in discovering evolution.

But I haven’t seen evidence for the assertion that memes change through time like genes. I have seen anecdotes and examples of ideas and behaviors that have spread through a culture, but no evidence that the pattern is the same. I haven’t even seen a clear way of identifying a meme, observing it’s reproduction, or tracking its offspring. Not so much as a study on the change of frequency of memes in an isolated population. Memetics today has less evidence than Darwin did when he started out; at least Darwin could point to discrete entities that were changing.

Without this sort of evidence, all the concept of a meme gives me is that ideas and behaviors can get transmitted, and that they can change. And I don’t need a new concept for that. Every now and then I’ll run a search on memetics just to see if anyone’s tried to address these problems — after all, a model describing how the frequency of ideas change in a population could be extremely useful to me — but so far I’ve seen nothing, and I don’t usually have the time to run a truly thorough search.

If any of you have, and if you know of evidence for the concept, please send me a link.

New Comment
49 comments, sorted by Click to highlight new comments since:
[-]ewang190

If I recall correctly, the concept of "memes" was invented to illustrate the universality of Darwinian evolution in non-genetic systems, rather than as an attempt to explain results or make predictions.

Many others quickly rectified that. The man who coined the term later joined in - e.g. see the chapter about memes in The God Delusion.

The Journal of Memetics shut down. That is a bad sign.

Edit to add: the final issues has some discussion of the issues and is very interesting. The Neglected Virtue of Scholarship!

Final issue: http://cfpm.org/jom-emit/2005/vol9/

An article on why it failed: http://cfpm.org/jom-emit/2005/vol9/edmonds_b.html

Referring to an article on why it was in danger (3 years earlier): http://xn--jomemit-706c.cfpm.org/2002/vol6/edmonds_b_letter.html

That Edmonds article is largely based on a statistical mistake - as explained here.

Good catch, thanks! That plot is indeed very weak.

But if we're charitable enough to ignore the plot and related argument, the article still makes a good argument. Given that the journal shut down for lack of quality submissions, I think it's fair to say that if the field is not dead, it's definitely resting.

Ok, more seriously: people are using memetics-like ideas, sometimes under different names, in other fields and publishing in other journals.

Given that the journal shut down for lack of quality submissions, I think it's fair to say that if the field is not dead, it's definitely resting.

It was a freebie. Not everyone wants to write for free. I don't think a free journal not getting material says terribly much.

IMO, the editor was reading his trend lines all wrong. In fact the intenet meme explosion didn't really get going until 2011.

people are using memetics-like ideas, sometimes under different names, in other fields and publishing in other journals.

Yes, like here: http://on-memetics.blogspot.com/2011/10/memetics-journals.html

[-]asr30

It was a freebie. Not everyone wants to write for free. I don't think a free journal not getting material says terribly much.

Academic journals (of which JoM‑EMIT was an example) generally don't pay for contributions. Even so, there are lots of low-impact low-quality journals out there with plenty of submissions. If a journal goes under from lack of interest, that suggests something was really wrong with the field.

There's been a flood of articles and books on the topic since then, though. The field is flourishing - albeit under different names.

Again thanks for the link. Just a note that "Memetic Computing", while it might accept articles on memetics proper, is mostly about "memetic algorithms" which, as Edmonds says, have little to do with memetics.

Memetic algorithms have "little to do with memetics"?!? Huh? Of course they do. Memetic algorithms are cultural evolution in silico. As I say in my book:

Memetic algorithms will probably prove to be the most important application of memetics in the long term.

It sounds odd, but it's true. So-called memetic algorithms are just genetic algorithms with extra local search, and some variations. The difference between genetic algorithms and memetic algorithms doesn't capture any of the difference between genes and memes. That is my opinion and the opinion of quite a few researchers I know who work in evolutionary computation and, obviously, Edmonds.

The link to memetics comes directly from the field's founder, in the paper that started it all:

While Genetic Algorithms have been inspired in trying to emulate biological evolution, Memetic Algorithms (MA) would try to mimic cultural evolution.

The paper discusses at some length how Dawkins and memes inspired the approach in section 5 ("Towards Memetic Algorithms"). I don't really know what you are talking about by apparently denying the link.

Memetic algorithms are important - since the process that leads to superintelligent machines will inevitably work more like cultrural evolution than plain organic evolution. Memetic computing is one of the main areas which studies such algorithms.

I've read some memetic algorithms papers, but I never read this one. You did me a favour by pointing to the right section among the 68 pages :)

Now, I don't deny that Moscato and other MA people were inspired by memetics and Dawkins. What I'm saying is the difference between GAs and MAs fails to correspond to the difference between genes and memes. I'm not alone: p. 19 says "The GA community would like to say that MA are only a special kind of GA with local hill-climbing." That is what I said before.

The essential difference between MAs and GAs is the extra local search. This is supposed to be analogous to the ability of a martial arts master to make not random changes to his/her memes (genetic mutations are random), but directed changes. There are two problems with this:

  1. In MAs, the hill-climbing does, in all cases I have seen, boil down to using random mutations and discarding the bad ones. (How else would we do hill-climbing, on a black-box fitness function?) So the changes are not really directed any more than GA mutations are directed, when the GA uses selection.

  2. I think there is a lot more to the idea of memes than just directed changes. What about their non-particulate nature, in Dawkins' phrase? That has no analogue in MAs that is not already in GAs. What about their weird method of combination, which is more like compounding than crossover? That has no analogue in MAs, as far as I know.

Are there some other features that distinguish MAs from GAs? Moscato mentions non-"genetic" representations, meaning non-linear ones. That has been common in GAs for a long time. I use tree-based, grammatical, and real-valued representations all the time. Even if (using the example in the paper) two-dimensional representations were common in MAs and unheard of in GAs, memes are not more accurately represented by matrices of bits than they are by lists of bits. Neither representation is adequate for the almost unrepresentable space occupied by memes.

Any other distinguishing features? Cooperative versus competitive coevolution (p 20)? Common in GAs, and more importantly, common in real-world genetic evolution.

EDIT: added to second-last paragraph.

The essential difference between MAs and GAs is the extra local search. This is supposed to be analogous to the ability of a martial arts master to make not random changes to his/her memes (genetic mutations are random), but directed changes. There are two problems with this:

In MAs, the hill-climbing does, in all cases I have seen, boil down to using random mutations and discarding the bad ones. (How else would we do hill-climbing, on a black-box fitness function?)

That is an easy question. A "black-box fitness function" doesn't mean the function is completely unknown. One is allowed to presume Occam's razor. That means that a range of techniques are likely to work when designing the next generation of trials: linear interpolation, extrapolation, fourier analysis of the fitness landscape - and so on. You can also keep historical records of notable past successes and failures - to help guide your search, use inductive inference, and take advantage of the rest of standard scientific toolkit.

I think there is a lot more to the idea of memes than just directed changes. What about their non-particulate nature, in Dawkins' phrase? That has no analogue in MAs that is not already in GAs.

Well that's because there are already analog genetic algorithms - or at least real-valued ones - which are about as "non-particulate" as you can get while remaining inside a digital computer.

To simulate cultural evolution some of the more important things you need are individual learning and social learning in a population. Much depends on how good your learning algorithms are. Yes, there are other aspects of cultural evolution - but if we knew how to reproduce them all in machines, we would have advanced machine intelligence by now. Today's memetic algorithms are necessarily a work in progress. However, the goal of simulating cultural evolution - and taking advantage of its obvious power - was the aim from the very beginning.

WTH? Who claims that an obvious decade-long trend has ended and reversed based on one or two years' data?

WTH? Who claims that an obvious decade-long trend has ended and reversed based on one or two years' data?

I would, given the right kind of data. A year or two is a lot, 10 years isn't an excessive amount more.

"Meme" is not a model, it's a reference class of models, many of which are informal. In order to talk about testing it, you must first zoom in.

Could you elaborate on that?

Memetics consists of terminology and framework for cultural evolution. Cultural evolution covers a lot of hypotheses. For instance there's the hypothesis that the human brain swelled up to accommodate memes or the hypothesis that memes made humans sociable - since they need social contact between their hosts in order for them to reproduce - or they hypothesis that memes were implicated in the high frequency of speciation among our ancestors - just as songbirds speciate frequently. There's quite a lot of associated hypotheses - no doubt some are correct and some are not.

[-]TimS00

Consider the question: "Where did that viral video come from - why that video and not another video?" - That's the colloquial use of meme, but it's not a very useful question because the answer often "Random chance at the intersection of timing and relevance."

Consider the different question "When did it become unacceptable for males to have intimate friendships with other men?" Looking at the source of that, including with other ideas supported its creation and continuance - is essentially the only useful aspect of analysis of memes.

Consider the different question "When did it become unacceptable for males to have intimate friendships with other men?" Looking at the source of that, including with other ideas supported its creation and continuance - is essentially the only useful aspect of analysis of memes.

As it stands, this looks like an impoverished view of evolutionary change. Do you think the same thing about genes? What about things like their destruction and recombination? How about how microevolution leads to macroevolution? What about kin selection and group selection? I could go on...

I wrote a book on memetics last year. There's an associated web site. There's various material online, including a FAQ, my references, and a memetics timeline, illustrating the history of the idea - and a blog. Chapter 3 in the book covers the evidence for the idea. The nearest similar free online resource I am aware of is:

There's a lot of confusion about cultural evolution and universal Darwinism - but the basics are fairly simple - and there are plenty of books and resources available to help these days.

It's true that there are some meme synonyms. However, "meme" is by far the most popular.

However, "meme" is by far the most popular.

How about "idea"?

"Idea" is not really a meme synonym. Ideas are not necessarily socially transmitted. The concept bundles individual learning and social learning together. Memes, by most conventional definitions, are to do with social learning. It is socially learned ideas that are most likely to result in open-ended adaptive evolution - since they have potentially large lifespans.

Why don't you publish an ebook version?

The reason is here.

[-]Shmi20

Is there a mathematical model of meme propagation?

[-][anonymous]10

There's certainly mathematical models of rumors, which is a similar enough but not quite the same concept.

From memory, they model similar to epidemics, which I'm not sure how that related to genetic drift and selection.

I seem to remember more elaborate techniques that I think were trying to capture genetic drift and selection, but I can't find them at the moment.

A quick google along the lines of "mathematical model meme propagation" does tend to pop up quite a few models. Here are two that seemed interesting: http://cogprints.org/531/1/mav.htm and http://cfpm.org/jom-emit/2000/vol4/kendal_jr&laland_kn.html

So, perhaps start with my references.

Interesting models started in the 1970s, and there were three books on the topic in the early 1980s:

  • Lumsden, C. and Wilson, Edward O. (1981) Genes, Mind, and Culture.
  • Cavalli-Sforza, L. L. & Feldman, M. W. (1981) Cultural transmission and evolution: A quantitative approach.
  • Richerson, Peter J. and Boyd, Robert (1985) Culture and the evolutionary process.

Since then the field has exploded.

The article Mathematical Models for Memetics explains how this material relates to memetics.

I don't know for sure about scholarly work on memes, but language changes over time. Language is used to convey ideas, and in turn is changed, shaped by which ideas need to be represented in different cultures and different times. Maybe on some level you could consider components of language (words, grammatical rules) to be that discrete unit.

Languages definitely have been classified phylogenetically/cladistically. Drift and change eventually leads to different languages stemming from a common ancestor. See for instance Indo-European languages or even Nostratic languages.

For instance, on a more granular level, examples of this phenomenon would be the more and more common confusion between "it's" and "its" or "your" and "you're" as well as "they're" and "their". It's a minor drift, a permutation which seems almost a neutral mutation at a very low level of language. But I see people using it more often, even educated people who seem to have otherwise perfect command of English (I couldn't find scholarly work on this topic though).

Language is used to convey ideas, and in turn is changed, shaped by which ideas need to be represented in different cultures and different times. Maybe on some level you could consider components of language (words, grammatical rules) to be that discrete unit.

Not unless you want to confine memes to language. Memes are intended as general units of cultural inheritance - and are not specific to language.

My two cents on the subject may be found in the article: Are memes discrete?.

Traits were recognized as inherited before genes were discovered. Ideas don't change over time in a manner analogous to genes- they change in a manner analogous to traits.

The current generation of ideas consists of those ideas which were passed down from the previous generation, plus those which arose spontaneously- just like the current generation of other inheritable traits.

Ideas don't change over time in a manner analogous to genes- they change in a manner analogous to traits.

That sounds rather dogmatic to me, I'm afraid. Ideas are associated with heritable cultural information, which many call "memes". Similarly, DNA strands have associated heritable cultural [*] information, which some - following G. C. Williams - call "genes". So, memes are the cultural equivalent of DNA genes - in this precise sense.

The term "meme" was coined to sound like "gene". It was intended to be the cultural equivalent of a gene. If you are claiming otherwise, you simply aren't using the word in the way in which it was originally intended.

You are using a different definition of 'cultural information' than everyone else in the world. Genes code for proteins; proteins combined with other environmental factors cause traits to be exhibited.

Genes are the underlying mechanism by which genetic traits are transferable, just as 'memes' (original definition) are the mechanism by which ideas and ways of thought are transferable.

However, a meme (such as 'expected results inform present decisions') by itself is about as meaningful as a strand of RNA without a cell. An idea (such as Bayesian Rationality), on the other hand, is comparable to a protein (e.g. procollagen), in that the memes determine how the idea is expressed.

You are using a different definition of 'cultural information' than everyone else in the world.

Not so. Dawkins listed "ideas" in 1976 - in his examples. Most in the area accept that ideas qualify as memes.

Examples of memes are tunes, ideas, catch-phrases, clothes fashions, ways of making pots or of building arches. Just as genes propagate themselves in the gene pool by leaping from body to body via sperms or eggs, so memes propagate themselves in the meme pool by leaping from brain to brain via a process which, in the broad sense, can be called imitation." (Dawkins, 1976, p 192).

DNA strands, on the other hand, do not contain cultural information in any serious sense, any more than adrenaline or other hormones do.

Oops! I should apologise for my writing mistake. I meant to write:

Similarly, DNA strands have associated heritable information [...]

...rather than:

Similarly, DNA strands have associated heritable cultural information [...]

Obviously, cultural information is in bibles, DVDs, brains and hard disc drives - not DNA.

This is the internet. You, personally, directly see and hear memetics at work every day. Disbelieving in memetics is not even like disbelieving in an obvious-but-opposed-by-biases thing like evolution, it's like disbelieving that animals can move!

Nope. Memetics is not simply the claim that ideas change over time; it is a more specific analogy to evolution, which is almost certainly false. At least to me, the fact that memes can be deliberately (ie non-randomly and intelligently) created and changed is enough to disprove the analogy. Of course, empirical evidence would also be nice.

This seem to be a connotations problem. The words simply mean different things to us, and quite possibly yet a third unrelated one to the OP.

My connotation might be approximately summed up along the lines of "Ideas can change incrementally without a central authority in control"

FWIW, evolution is compatible with a central authority, IMO. The essential elements of evolution are copying, variation and differential survival - following Lewontin, 1970: "THE UNITS OF SELECTION". There's no mention of the presence - or absence - of a central authority. It's the same with textbooks on the subject of evolution. If there was a central authority dictating the contents of the ideosphere, memetics would still have something to say about how those ideas evolved.

At least to me, the fact that memes can be deliberately (ie non-randomly and intelligently) created and changed is enough to disprove the analogy.

Why is that? Genes can be deliberately engineered as well, albeit with greater difficulty.

Can was probably the wrong word.

The "natural" method of change for genes is random change, with some selection acting afterwards.

Changes do occur randomly to memes, but when approaching an arbitrary meme, I generally feel >95% confident that it was, at some point, extensively and deliberately designed, and >75% confident that the deliberate design was the dominant influence on it. In addition, basically everybody who is passing on a meme tries to communicate it in the most effective way possible: so even when looking at a change, my prior is that it is almost always at least partially deliberate (a result of somebody reasoning that they should change it in a certain way) instead of accidental.

In addition, genetic evolution is much more advanced than memetic evolution: the most competitive organisms in the wild are all randomly evolved, yet almost all effective memes were artificially designed.

I don't think there's much debate about this. Everyone agrees that there's more design in cultural evolution than there has been in organic evolution so far - though this may well change, as organic evolution tries to catch up. Of course some designers at work doesn't obviate the need for Darwinian evolutionary theory. The products of design are naturally selected, just like everything else. For example, in VHS vs Beetamax, consumer selection played a critical role in determining what we see.

One other thing to be aware of is that there's copying, selection and variation within the brain too (e.g. see: William H. Calvin, "The brain as a Darwin Machine" 1987). It's rather like how viruses multiply within bodies - as well as having a transmission phase where they spread between bodies. Within the mind, there's an awful lot of trial-and-error - taking place at a low level. If you only look at the results of that process, then it seems smart. However at root, the process takes place at the level of neurons - where there isn't too much scope for an intelligent designer.

So: even if you think memes are intelligently designed, they were still produced by an evolutionary process - one going on inside an individual mind.

Can you taboo "design?" As I understand the concept, I don't really see how it's something that can be applied to ideas that one formulates and adopts oneself.

I don't think that the analogy is intended to be that close. It is a bit hard to imagine that Dawkins failed to notice that memes can be deliberately created and changed. So whatever he intended, it must have been a sufficiently abstract analogy to allow for that.

Memetics is not simply the claim that ideas change over time; it is a more specific analogy to evolution, which is almost certainly false. At least to me, the fact that memes can be deliberately (ie non-randomly and intelligently) created and changed is enough to disprove the analogy.

Memetic engineering seems generally similar to genetic engineering. Memetic engineering doesn't disprove Darwinian cultural evolution any more that genetic engineering disproves Darwinian organic evolution.