This is a list of criticisms of memetics, associated with my 2011 "Memetics" book - which is now available. This page is also associated with the Memetics FAQ.
In Freedom Evolves, Daniel Dennett (1995) said:
Most of the arguments that have been deployed against a science of memetics have been misguided and misinformed, and they betray a distinct whiff of disingenuousness or desperation.
Practically everyone seems to have a criticism of memetics. Also, there are a wide range of different criticisms. However, many of them appear to be based on misunderstandings. Here we present both criticisms of memetics and rebuttals of them.
To stop the more ridiculous criticisms from distracting the reader, the most pointless ones are given their own section at the end of this document.
- Criticisms of cultural evolution.
- Criticisms of memetics.
- Memes are not like genes because...
- Memes are not like viruses because...
- Particularly pointless or daft criticisms.
- Criticisms of cultural evolution.
- Adaptations cannot accumulate in the face of culture's high mutation rate.
- Directed mutations are prohibited by evolutionary theory.
- Directed mutations will swamp evolved adaptations.
- Lamarckian inheritance has been disproved.
- Long-isolated cultures can sill interbreed.
- We are too ignorant to say that cultural evolution is Darwinian.
- Cultural evolution is too different - we should start again.
- There are no cultural lineages.
- Evolutionists should present a united front.
- Cultural evolution exhibits progress.
- Culture is designed - not evolved.
- Deleterious cultural traits can't evolve adaptations.
- Criticisms of memetics.
- Memetics is a pseudoscience.
- Memetics has never taken off.
- The existence of memes seems questionable.
- Culture is not subdivided into discrete units.
- Memetics ignores developmental processes.
- Copying and selection may not explain culture.
- Memes are not copied but recreated.
- Meme transmission is tangled up with development.
- Memes are totally transformed during transmission.
- Culture is more complicated than that.
- Memetics does not explain meme fitnesses.
- Memetics makes no predictions and is unfalsifiable.
- Memetics makes weak predictions.
- Memetics has yet to find its Mendel.
- Memetics is a dangerous idea.
- Memetics is nothing new.
- Memetics terminology is pointless.
- Memetics represents an unpalatable truth.
- Naked memes are problematical.
- Cultural evolution exhibits complex developmental tangles.
- Memetics is full of "just so" stories.
- Memeticis is not socially acceptable.
- Memetics hasn't produced anything original.
- Memes are not necessarily good for you.
- Meme fitnesses are not determined by meme attributes.
- Critique from semiotics.
- Memetics violates Occam's razor.
- Memetics is too negative.
- Memeticists can't agree on what a meme is.
- Memetics is “mind-blind”.
- Memetic linkage is too strong.
- Memes are not "quasi-autonomous bots".
- Memes are not like genes because...
- ...memes are so intangible.
- ...there is no memetic code.
- ...memes do not have "loci".
- ...memes are not necessarily transmitted with high fidelity.
- ...genes replicate - while cultural evolution is not based on replicators.
- ...culture exhibits blending, while gene transmission is descrete and particulate.
- ...memes are sometimes analog.
- ...memes undergo continuous variation.
- ...memes are not part of a functionally integrated structure.
- ...meme mutations affect adaptive evolution.
- Memes are not like viruses because...
- ...memes are often beneficial, whereas viruses are deleterious.
- ...memes can be transmitted vertically from parent to offspring.
- ...memes interact inside their host more than viruses do.
- ...memes are engineered - but they are not like computer viruses.
- Particularly pointless or daft criticisms.
- Culture exhibits insufficient variation.
- Memetics denies a role for chance processes.
- Memes - unlike genes - are not copied
- Evolution requires storage in stable molecules.
- Memes do not self-replicate.
- Memes cannot represent complex cultural traits.
Criticisms and responses
To look up references from this document, see here.
- Criticisms of cultural evolution.
Adaptations cannot accumulate in the face of culture's high mutation rate.
- Here is Kim Sterenly (2007) expressing this point:
Social learning is common, but only humans routinely learn new abilities by imitation (Tomasello, 1999; 1999; Whiten, 2000; Heyes, 2001). Yet these very abilities undermine the fidelity of the intergenerational flow of information. For they increase the extent to which an agent’s beliefs and behaviour are sensitive to his or her own experience, and they decrease the extent to which these beliefs and behaviours are the result of blindly copying those of the previous generation. Patterns of behavioural similarity will not be transmitted deeply through the generations.Alas for this, adaptations can and do accumulate. The evidence shows that complex adaptations (e.g. as found in engine designs) have accumulated historically. Copying fidelity really is high enough to support this. The Holy Bible was transmitted down the generations with reasonable fidelity, for example. This was achieved by using error detection and correction technologies. These days, high-fidelity transmission is becoming more common - and so there are fewer accidental mutations than there ever have been before. Further cultural evolution uses much the same tricks as organic evolution to improve copying fidelity: natural selection and sexual recombination. If you make lots of copies, throw out the bad ones, and recombine from the good ones, it improves the overall copying fidelity.
Directed mutations are prohibited by evolutionary theory.
- No, they are not. Most formulations of evolutionary theory are pretty agnostic about the source of variation. Directed mutation is just another type of mutation to add to an extensive existing list which includes inversions, point mutations, duplications and deletions.
Directed mutations will swamp evolved adaptations.
- Not necessarily. Selection between cultural products is a common and important process. In VHS vs Betamax both sides were memetically engineered, but selection still played an important role in determining what came to dominate.
Lamarckian inheritance has been disproved.
- In memetics, inheritance is mostly Weismannian - but occasionally Lamarckian. Lamarckian inheritance really requires a split between inherited information and phenotype, reverse-engineering the phenotype and storing the results in the genotype. That doesn't happen too much in the organic realm, but it can sometimes happen in the realm of culture. So empirically, Lamarckian inheritance does not happen to a significant extent in organic evolution - but it can happen in cultural evolution. This isn't evidence against evolutionary dynamics applying to culture. Rather it represents evidence that the dismissal of Lamarckian inheritance was premature.
Long-isolated cultures can sill interbreed.
- Some think the relative lack of isolation in cultural systems represents some sort of problem. Here is Orr, H. Allen (1996a) on this topic:
memes and genes differ in other fundamental ways. Species, once isolated, almost never exchange genes, while exchange between long-isolated cultures is immensely important in the history of ideas.
Stephen J. Gould agrees, saying:
The basic topologies of biological and cultural change are completely different. Biological evolution is a system of constant divergence without subsequent joining of branches. Lineages, once distinct, are separate forever. In human history, transmission across lineages is, perhaps, the major source of cultural change.
There is a difference between organic and cultural evolution in this area - cultural isolation is less extreme. However, isolation in organic evolution is less than perfect - viruses sometimes drag genes around from one lineage to the next. Cultural evolution does exhibit a range of types of partial isolation: for example, COBOL programs interbreed little with origami designs. So, this seems to be more like a quantitative difference than a qualitative one. Even if there was no cultural equivalent of a species, culture would still evolve. Much of population genetics and evolutionary theory would still apply in the cultural domain.
We are too ignorant to say that cultural evolution is Darwinian.
- Here is Orr, H. Allen (1996a) on this topic:
it is far from clear that Darwinism can account for the percolation of ideas, styles, and songs through culture. In fact, there is a basic problem with any such claim - we are very ignorant of how humans hold ideas in their heads and of how the ideas in your head influence the ideas in my head. So how can we possibly conclude that the process "must be" Darwinian?
This may sound like humility, but it goes too far. The discovery of Darwinian evolution preceded the discovery of DNA by almost a century. We don't know exactly how ideas are represented in brains, but the lack of a cultural Watson and Crick should not cause scholars to get stuck in a pre-Darwinian era.
Cultural evolution is too different - we should start again.
There are no cultural lineages.
- Tim Lewens (2006) writes:
Genes form lineages; cultural units do not
Assuming that we are talking about genetic lineages, it is simply false that memes don't form lineages. It turns out that Tim is actually complaining that not all cultural paths from ancestor to descendant form lineages. He says:
In principle, I could look into my genome and say (for most of my genes, at least), which came from my father and which from my mother. Each gene is derived from a single individual, in such a way that we might trace a lineage back through time. Can we do this for cultural items? Not always.However, you can't "always" do it in genetic evolution either. Lineages can be - and frequently are - interupted by crossovers.
Kim Sterelny (2007) expresses a similar objection:
Sperber points to a real difference between ideas and genes, for genes really are copied. Except those few arising newly by mutation, every gene in me has a distinct, identifiable ancestor in one or another of my parents. Genes form true lineages of ancestor-descendant pairs. But my ideas are not links in true lineages. Even if my ideas about Bradman are content-identical to those of my father, they did not arise from a single copying episode. They will have been constructed from multiple episodes of hearing about Bradman from my father (and others). My Bradman-conception and my father’s Bradman-conception are not two links in a lineage. This is a very important objection to meme-theory proper, i.e. to a conception of memes that see them as genuine lineages, with fitness values to call their own, coevolving with humans.
Cultural transmission does frequently form lineages. However, Kim is simply mistaken when he says "Except those few arising newly by mutation, every gene in me has a distinct, identifiable ancestor in one or another of my parents". Some genes are present in neither parent - since a crossover occurred in the middle of them during meiosis. Crossovers are pretty common - and classifying crossovers as "mutations" would be rather unorthodox.
Evolutionists should present a united front.
- Some are cautious about controversial evolutionary ideas, and wonder if it would be better to present a united front against the creationists.
In some countries Darwinists battle against biblical creationism, and there is a history of the creationists using Darwinian controversies on the edges of the science to make out that evolution is somehow not settled science, and is still controversial. For an example, see Robert Wright's 2001 article, "The Accidental Creationist". Since memetics is controversial, some say that perhaps scientists should hold back, so we don't show our underbellies to the creationists. Here's Richard Dawkins in the "Evolutionary Perspectives" panel discussion.
I know that there's a certain amount of hostility "out there" to people who think that Darwinism is being too aggressively promoted - and I am sensitive to that and I am sufficiently anxious to promote the fundamental form of Darwinism which is the one that Darwin himself and the explains all of life - I mean that's quite a big thing in the first place I think it would be rather a shame if we lost the battle to get people to understand this really very fundamental fact through being a little bit too over-enthusiastic to push the Darwinian line on all sorts of other things which some sensitive people regard as their own territory. So sometimes for tactical reasons I shrink back from venturing into other fields that others have pushed Darwinism into - but I think my heart's with them.
I don't think caution on this front is necessary when it comes to culture, which plainly evolves. It is not necessary to keep quiet about memetics "for tactical reasons". Indeed, if anything memetics is an important tool for understanding how religions can still spread even though they present false accounts of the world, and may not necessarily be good for those who are involved with them.
Cultural evolution exhibits progress.
- In a discussion relating to David Hull's ideas about the evolution of scientific theories, Michael Ruse wrote:
It makes good sense to say the Mendel was ahead of his predecessors, just as Watson and Crick were ahead of their predecessors. Yet as Darwinian evolutionists are perpetually telling us, biological evolution is not progressive (Williams 1966). Appearances to the contrary, it is a rather slow process, going nowhere.
This seems fairly straight-forwards. In a sufficiently large and benign environment, evolution is indeed progressive. The progress made by cultural evolution is, in fact, built directly on top of the cumulative progress made by organic evolution. Not all evolution is progressive. For example, if there was a nuclear war, we would probably see a lot of extinction and devolution, neither of which are progressive. If meteorite strikes on the Earth were more frequent, evolution here might not be progressive. However, meteorite strikes are relatively rare - and so evolution is progressive.
It is popularly claimed that organic evolution is not progressive - since evolution is more like a bush than a ladder. Rather, adaptation takes place in the context of local conditions - and complexity levels, intelligence and size can all increase or decrease. In this limited sense organic evolution is not progressive - and neither is cultural evolution. However, despite occasional setbacks, overall progress is completely obvious in both organic and cultural evolution. This view is unpopular - since it has previously been associated with social Darwinism and "ethnic cleansing". However: political correctness is one thing and science is another. Denying evolutionary progress is a completely crazy and indefensible stance arising from Marxism. The position should be widely denounced by real scientists as delusional nonsense. For more about this issue, see here.
Culture is designed - not evolved.
- In 2009, Steven Pinker said:
Design without a designer is essential for biological evolution - but it is perverse for cultural evolution: there really is a designer - the human brain - and there's nothing mystical or mysterious about saying that.That some aspects of culture are designed does not mean that culture doesn't evolve, or that we don't need a selection-based model to model how it changes. There are plenty of places where memes compete with each other and undergo selection. That happens in science, technology, marketplaces, languages and charities. That is what you need a Darwinian model of culture for.
Memetics does not deny a role for intelligent design by the human mind. In memetics, intelligent design is a source of mutations. Mutations are a basic element of all evolutionary theories - and there is no rule saying that mutations have to be random.
However, it is worth noting that intelligence itself contains evolutionary processes. Intelligence is not magic. Minds constantly attempt to maintain a compact and parsimonious world model, and their model of the world evolves. A considerable amount of trial-and-error testing processes go on in minds. Thoughts compete with other thoughts, possible actions compete with other possible actions and synapses compete with other synapses.
Most optimisation processes work using trials, variation, and information that persists via copying. You really have to go back to random search to find an optimisation process that doesn't even feature a type of inheritance.
Deleterious cultural traits can't evolve adaptations.
- David Burbridge (2003) claims that the deleterious side of culture is problematical because:
Biological traits are usually adaptive for the individuals who possess them, in the sense that possession of the trait enhances their reproductive fitness. Genes producing traits that impair reproductive fitness will be eliminated by natural selection. In contrast, there is no reason to suppose that cultural traits (with some important exceptions, such as economic competition in a free market) are usually beneficial in any sense to the individuals or groups that possess them.
Here, Burbridge is looking for benefits to the individual humans. However in memetics the fundamental idea is that cultural entities benefit. When a wave of copy-cat suicides takes place, the hosts don't benefit, it is the suicide meme that gets greater circulation. If you compare with viruses, you see that cold viruses are not advantageous for their hosts, either. When a human sneezes, one does not ask how this benefits them. The benefit accrues not to the sneezing human, but rather to the virus they are infected with. Similarly, cultural adaptations, don't necessarily benefit their human hosts. The benefits accrue either to the cultural entities themselves - or to other humans that engineered the memes in order to manipulate others.
- Criticisms of memetics.
Memetics is a pseudoscience.
Memetics has never taken off.
- This, alas, is true. Steven Pinker - in a 2009 Harvard lecture - said:
For one thing, just empirically, the idea of memetics, of a science of cultural change based on a close analogy with natural selection, it is just a fact: it's never taken off. It's thirty-five years old almost at this point. Every five years a paper appears that heralds the final development that we have all been waiting for of a science of memetics - and nothing ever happens.
Compare this to other sciences that have just flourished since 1976: neural networks, cognitive neuroscience, evolutionary psychology - there are conferences and journals and textbooks - we don't have a science of memetics - despite the constant promise that it is just around the corner - and I think that there is a good reason why we don't that there is something deeply flawed with the idea.
What we do have is a science of gene-culture coevolution. This is a branch of population genetics - and on close inspection, it is closely isomorphic to memetics - a fact which both students of memetics (Blackmore 2006f) and people from the population genetics side (Kendal and Laland, 2000) have previously pointed out. This is essentially memetics without the "m"-word - and it has all the features that Pinker objects to. Pinker doesn't make any mention of this.
The other main problem is with social scientists dragging their feet when it comes to embracing the principles of Darwinian evolution. Darwin and humans just don't mix for them, it seems. That memetics has not taken off seems to be a rather embarrassing fact for rational humans. The correct response is not to look for holes in memetics, but rather to set to work constructing the science - since it is better if it is built now than later.
There's a page about this criticism here.
The existence of memes seems questionable.
- Most definitions of the term "meme" result in words being memes. So: memes have the same existential status as words.
Culture is not subdivided into discrete units.
- If you check with the definition of 'meme' in a dictionary there is no mention of 'discreteness'. However, this is a common criticism of memetics - so common that there's now a whole article discussing the issue. Please see that for more details.
Memetics ignores developmental processes.
- Population genetics - and to some extent genetics itself - typically treats development as a black box. Those who perform meme frequency analysis do exactly the same thing. Of course, development is important, but there is much that can be said without getting into the details of how it works. Development is mostly another topic.
Copying and selection may not explain culture.
- This is one of the complaints of Dan Sperber. Here he is in 2000:
For memetics to be a reasonable research programme, it should be the case that copying, and differential success in causing the multiplication of copies, overwhelmingly plays the major role in shaping all or at least most of the contents of culture.It is probably true that new culture arises largely from existing culture - but memetics doesn't depend on that being true. For example, most theories of cultural evolution and memetics would still apply if mutational forces outweighed selective ones. In that case the resemblance between cultural and organic evolution would be reduced - and there would not be cumulative cultural evolution - but evolutionary theory would still apply.
Some other factors - besides copying and selection - that affect culture are: learning, recombination, interpolation and extrapolation. Sperber might think that such things invalidate memetics - but they just don't.
Memes are not copied but recreated.
- Andrew Brown (2009) writes:
What we know about the transmission of meaning, like that of memory, is that it involves continuous recreation rather than simple copying.
Since most cultural information has gone digital these days, it is indeed reproduced largely by "simple copying" - for example on peer-to-peer networks, or via "retweeting". However, even back in the stone age, when that wasn't true, cumulative cultural evolution still took place. Definitions of evolution do not say anything about "simple copying". Evolution involves the transmission of heritable information from one generation to the next. Exactly how that happens is an implementation detail, not a fundamental feature.
Meme transmission is tangled up with development.
- Jablonka and Lamb (2005, p.209) write:
Since heritable variations in behaviour and ideas (memes) are reconstructed by individuals and groups (vehicles) through learning, it is impossible to think about the transmission of memes in isolation from their development and function.
In many cases this isn't true - plenty of memes are digitised and copied with no significant developmental process being involved. However, some memes are reverse-engineered from meme products that are the product of developmental processes. That doesn't happen in organic evolution, and it represents an interesting difference from cultural evolution - though not a problematical one. Fortunately, there's a neat conceptual way of "filtering out" development and functional issues while analysing memes - just look at the Shannon mutual information between parent and offspring. That shows what information has been transmitted between them - irrespective of any developmental issues.
Memes are totally transformed during transmission.
- Some say that memes are so mangled in minds as to become practically unrecognisable. Here is Maurice Bloch in 1995:
Sperber, Levi-Strauss and most of the ir colleagues - as well as myself - accept the fundamental criticisms formulated by the American consistency theorists against the diffusionists: criticisms which apply with equal force against memeticists. Agreement is focused on the fact that the transmission of culture is not a matter of passing on "bits of culture" as though they were a rugby ball being thrown from player to player. Nothing is passed on; rather a communication link is established which then requires an act of re-creation on the part of the receiver. This means that, even if we grant that was was communicated was a distinct unit at the time of communication, the recreation it stimulates transforms totally this original stimulus and integrates it into a different mental universe so that it loses its identity and specificity. In sum, the culture of an individual, or of a group is not a collection of bits, traits or memes, acquired from here and there, any more than a squirrel is a collection of hazelnuts.Here, the idea that "nothing is passed on" is simply wrong. Shannon information is passed on. The donor and recipient of the meme come to share Shannon mutual information. Sometimes what is passed on is indeed faithfully copied - as by biblical scribes. Sometimes what is passed on is not faithfully copied - as we see in Chinese whispers. The former process leads to long term adaptive evolution, while the latter one often leads to an error catastrophe. Memetics - just like genetics - handles both cases just fine. Memetics is just - by definition - the study of cultural heredity, and variation.
Culture is more complicated than that.
- Memetics is a beautiful, simple idea. However many resist simple explanations of biological phenomena, saying something like: "it is more complicated than that". One example of this is provided by Maurice Bloch (2005):
British anthropologists see culture as existing on many levels, learnt implicitly or explicitly in a great variety of ways (e.g. Leach 1954; Bloch 1998). It is not a library of propositions or memes. This type of argument is principally intended as a criticism of American cultural anthropology, which (as we saw) was itself a criticism of diffusionism. But clearly it also applies to the simple diffusionist idea that culture is made up of "bits of information" that spread unproblematically by "transmission" where transmission is understood as a unitary type of phenomenon. British anthropologists, including myself, would argue that knowledge is extremely complex, of many different kinds, and impossible to locate, as though it was a single type.
So, the idea of memetics is that memes are made of information. Information is complex - in the sense that it can be used to represent or describe anything else, including complex objects - but it could be described as being of a "single type". Information can be measured in bits. Memes themselves can be complex - even though the idea of a meme is essentially simple.
The vision of "a library of propositions or memes" conjours up an image which seems obviously wrong. However, nobody thinks that memes are somehow neatly stacked up in brains in the first place. It is well understood that brains are messy tangles. Also, something like riding a bike exists partly in muscle memory. Some of that information is stored in the spinal cord - and not in the brain. Cells where viruses reproduce are not neat and orderly either - yet we can still have an epidemiology for viruses. Similarly, despite the brain's tangles, we can still usefully have an epidemiology of ideas.
If there is a problem with diffusionism, it is more that the word "diffusion" implies a rather passive process. Memetics rejects the the diffusion terminology (from early cultural anthropology) as being misleading - and replaces it with concepts derived from epidemiology - infection, vector, plague, transmission, pandemic, etc. That in turn promotes the "memes are negative" misunderstanding, but at least it successfully gets across the important idea that culture is composed of living, reproducing, evolving, active agents.
Memetics does not explain meme fitnesses.
- The longest criticism of memetics I have seen comes from Maria Kronfeldner (2007). Her critique runs to over 300 pages - and has now been turned into a book. In her summary (p.290) she writes:
Given that there is no independence of meme diffusion from human individuals, the explanatory units of selection analogy ends up in an explanatory dilemma: Either the analogy is heuristically trivial, because it loses its main claim, namely that memetics presents an alternative to the traditional explanation, which is given in terms of properties and interests of humans, or the explanatory units of selection analogy is trivial in explanatory terms, because it is tautological – it does not explain anything, since it merely states that those memes that have a high actual survival are those memes that have a high propensity for survival, without explaining where this high fitness emerges from.
Genetics doesn't explain much about the fitness of genes either - it is more about the nuts and bolts of how genes combine. However, if you look a little further afield, to evolution and ecology, there is a wealth on information about why some genes are fitter than others. It is much the same with memes - that isn't really the domain of memetics, but we really do have lots of information about how and why some ideas spread, and others do not. Ideas vary in their truth, how memorable they are, how short they are, whether they activate humour, desire, fear or boredom in those exposed to them - and so on. We don't know everything about why meme fitnesses vary - but we don't know everything about how gene fitnesses vary either.
Since this is a common criticism, there's now a whole page all about it, titled:
Can memetics predict meme fitnesses?
Memetics makes no predictions and is unfalsifiable.
- Massimo Pigliucci (2007) goes so far as to claim that memetics is unfalsifiable.
Popper’s objection remains valid for memetics: the only way to tell which memes are going to be successful, which tunes are going to stick in your mind, or which religions are going to become popular is by waiting and seeing what happens. That is, memeticists completely lack a functional ecological theory of memes. Without it, the whole enterprise is scientifically empty.Of course, this accusation is totally untrue. Scientists know a lot about what spreads and what does not. It is totally false to say that the only way to know if something will be successful is to try it and see. For instance, these days we have a mountain of information about what kinds of status updates get "retweeted" on social networks. Dan Zarrella's 2009 paper "The Science of Retweets" has lots of details about that.
Memetics makes weak predictions.
- John Maynard Smith (1995) writes:
The explanatory power of evolutionary theory rests largely on three assumptions: that mutation is non-adaptive, that acquired characters are not inherited, and that inheritance is Mendelian - that is, it is atomic, and we inherit the atoms, or genes, equally from our two parents, and from no one else. In the cultural analogy, none of these things is true. This must severely limit the ability of a theory of cultural inheritance to say what can happen and, more importantly, what cannot happen.
Maynard-Smith is correct in stating that cultural evolution permits a wider range of possibilities than is available to in organic evolution. However, the possibilities are not so broad that predictions based on the theory become useless. Mutations may not always be non-adaptive - but they are usually small - so culture forms a branching network similar to that found in organic evolution - which itself constrains the resulting possibilities. The occasional cases of inheritance of acquired characteristics, and analog inheritance don't make that much difference. The increased range of possibilities available in cultural evolution represents an interesting feature - but not a terribly problematical one.
Memetics has yet to find its Mendel.
- John Maynard Smith (1995) writes:
My uneasiness with the notion of memes arises because we do not know the rules whereby they are transmitted. A science of population genetics is possible because the laws of transmission - Mendel’s laws - are known. Dennett would agree that no comparable science of memetics is as yet possible.
It is hardly surprising that we know more about organic gene transmission as meme transmission. Organic gene recombination is a relatively simple matter - while for memes, most recombination currently takes place in the tremendously complex environment of the human mind. However, we certainly do have predictive theories about meme transmission and recombination. Psychology and anthropology have a wealth of knowledge on these topics.
A full understanding of memetics may have to wait until we have cracked the problem of constructing artificial intelligence - and can simulate the human mind reasonably well. However, much progress can be made - and has been made - by treating the brain as a black box - examining what goes into it, and what comes out of it - and then inferring things about the underlying dynamics. This is the approach taken by those performing meme frequency analysis.
Memetics is a dangerous idea.
- One commonly-cited critical paper is entitled: "Memetics: a dangerous idea" by Luis Benites-Bribiesca (2001). Part of the abstract reads:
Memetics is a dangerous idea that poses a threat to the serious study of consciousness and cultural evolution.
Inflammatory material - though perhaps Dennett would agree that this material is "dangerous". Another who would raise the alarm is Allen Orr (1996a):
Except to a handful of biologists, there is little "dangerous" in the idea that Biology is Engineering. There is something dangerous, though, in the idea that Darwinism transcends biology, undermining our views of culture, consciousness, and morality.
OK, then: if that is dangerous, then memetics is indeed dangerous. Those attached to rigid and inaccurate conceptions of the scope of Darwin's theory of evolution should take special care with it.
Memetics is nothing new.
- One complaint is that memetics fails to offer anything new. The exact complaints vary. Here is Maurice Bloch (2005) on the topic:
I noted above that, in many ways, Dawkins' work on memes - and that of other writers who have followed him, such as Dennett - is a good, accessible introduction to what is intrinsic in social and cultural anthropology. This fact, however, will not necessarily endear memetics to anthropologists. At a general level, Dawkins and Dennett make very similar, if not identical, points to those which anthropologists have always made about human culture.
However, memetics certainly isn't just a rehashing of what anthropologists have always thought.
Michael Ruse (2008) has a slightly different complaint:
one is really just taking regular language and putting it in fancy terms. No new insights. No new predictions.
This isn't right - if the theory of Darwinian evolution covers all of human culture, that is a big deal and would represent a major revolution in the social sciences. Massimo Pigliucci (2007) says:
memetics - at least for now - doesn’t seem to add anything to the standard view of gene-culture co-evolution that was developed well before Dawkins put down his ideas in The Selfish Gene.
Before The Selfish Gene (Dawkins 1976), the field of cultural evolution was much more of a disorganised mess than it is today. Today, we have lots of scientific papers on the topic, but back then, academic study of the area was a mere trickle. Dawkins mostly contributed useful terminology - and drew attention to this important and sadly-neglected field.
Another complaint is this (anonymous) one:
The idea that culture, and ideas, are evolving is not new. Popper, with his "evolutionary epistemology" (Popper, 1969), was already advancing the idea that scientific ideas were evolving a little like gene; the most fit had more chance to survive longer, while poor ideas were doomed to extinction.
It is true that we have had theories of cultural evolution dating back over a century. However, Dawkins did a fine thing by drawing attention to the concept in his 1976 book, and the associated New Scientist article. He popularised the idea - much as he also popularised the work of Bill Hamilton in the earlier chapters of the same book. Popularisation is useful and important work.
Memetics terminology is pointless.
- Andrew Brown (2009) complains:
But why call these studies "memetic"? If you're going to use "meme" as a synonym for "idea"; what's the point? What is added to our understanding?
Meme is not a synonym for "idea", though. Memes as usually defined are transmitted from one person to the next, while ideas may not be. Memes may also include "muscle memories" - which are not usually classified as being "ideas".
Martin Gardner (2000) wrote:
The point is that the notion of a meme is much too broad to be useful in explaining human thinking and behaviour. A meme is little more than a peculiar terminology for saying the obvious. Who can deny that cultures change in ways independent of genetics, ways involving information that is spread throughout society mainly by spoken and written words?
Memetics can seem rather obvious - once you understand it. The problem is that only a tiny fraction of people seems to understand the idea. Gardner continues with:
To critics, who at the moment far outnumber true believers, memetics is no more than a cumbersome terminology for saying what everybody knows and that can be more usefully said in the dull terminology of information transfer.
By contrast, using a term for contagious ideas that evokes the biological concept of a gene seems to be a fine piece of terminological engineering to me. I think these days we can just point at the enormous scale of the modern usage of the term "meme" to illustrate that there was indeed a need for such a term before Richard Dawkins came up with it. The beauty of the meme terminology is that it lets you access most of the terminology of genetics, while neatly keeping tabs on whether you are talking about the genetic or memetic realms.
Memetics represents an unpalatable truth.
- Some apparently dislike memetics because it doesn't seem to present a very nice picture of human nature. Daniel Dennett (1991, p. 202) describes some of his first impressions of the idea:
I don’t know about you, but I’m not initially attracted by the idea of my brain as a sort of dung heap in which the larvae of other people’s ideas renew themselves, before sending out copies of themselves in an informational Diaspora. It does seem to rob my mind of its importance as both author and critic.
Conventionally, a human is the product of genetic and environmental influences. Memetics points out that some of the environment consists of memes. People are familiar with the idea that the environment contains pathogens and other organisms - who don't have their best interests at heart, and often attempt to manipulate them. One of the memetics insights is that much of the rest of the cultural ecosystem is like that too. Things that people didn't previously consider to be alive are actually out to manipulate them, in much the same way that pathogens are. People understand that other people may try to manipulate them through books, television and other media. However, memetics says that culture tends to behave like this - even if there are no manipulative human beings involved.
Some might feel liberated and empowered by this knowledge - a better understanding of the forces at work helps to deal with them. However, others don't like the idea that they are - in part - constructed my memes. People often don't like the idea that their behaviours are being controlled by DNA genes - and the idea that they are being controlled by memes is an additional insult. Humans are built by evolution to believe that they are powerful masters of their own fate. The idea that their actions are largely the product of a swarm of copied fragments of information seems to be an affront to their self-mastery.
One reviewer put it this way:
Some years ago, Richard Dawkins published “The Selfish Gene”, explaining how gene survival was fundamental in natural selection. He also coined the term “meme” to explain the dissemination of ideas across societies. Almost immediately, there was a strident chorus of objection, based on the theme of “you can’t say that about humans!”
Richard Barbrook put it this way in his "Memesis Critique":
The real crime of the Memesis statement is the way that it willfully obscures the process of human innovation and creativity under a mass of dodgy biological metaphors. In contrast, we must celebrate the Promethean power of humans to create - and recreate - themselves. It is precisely our refusal to accept our biological destiny which makes us more than insects. Unlike our fellow species, we can transform ourselves through thought and action.
This kind of thing is an appeal to the human ego. Human's cherish their sense of self-mastery. They don't like to hear that they are products of genes and memes - they often seem to find that insulting and degrading. Lastly, here's Dennett (2009b).
Finally, one of the most persistent sources of discomfort about memes is the dreaded suspicion that an account of human minds in terms of brains being parasitized by memes will undermine the precious traditions of human creativity.
People had a hard time adjusting to the first Darwinian revolution. The second wave of Darwinism is causing at least as many problems for people. It was bad enough learning that you are a product of natural selection - but to learn that many of your thoughts and ideas are themselves subject to mutation and selection as well seems to be too much for some people.
Naked memes are problematical.
- Some claim that memotypes (cultural genotypes) acting as meme phenotypes is a problem for memetics. For example, here is Jonathan Marks (2004):
In microevolutionary theory we distinguish between genes, the hereditary units themselves, and phenotypes, their expression and interface with the outside world. Memes are commonly used in both senses, suggesting that the analogy breaks down very quickly. Obviously if there is no distinction between the units of replication (genes) and the units of interaction (phenotypes), then it immediately becomes unclear just what value lies in modelling cultural processes as if they mimicked genetics.
Massimo Pigliucci has much the same complaint:
To begin with, unlike the case of genes, there doesn’t seem to be any distinction between memes themselves and the phenotypes they produce. Genes in some sense “encode” proteins, and proteins have a variety of effects that indirectly contribute to the fitness of the organism carrying those genes. In Dawkins’s own terms, there is a distinction between “replicators” (the genes) and “interactors” (the organisms themselves). But, in the case of memes, the replicating “unit,” for example, an annoying tune that gets into your head, forcing you to whistle it, and thereby gets stuck into somebody else’s head, is both replicator and interactor.
Willem Drees (2011) writes, in the Times Higher Education supplement:
In the theory of memes, the distinction between the genotype and the phenotype, in other words the one that exists between the recipe (genes) and the complete organism, is simply not there.This confusion should be easily dissolved. Culture does have the split between gene products and meme products, just as the organic world has the gene/phene split. Memes are the cultural equivalent of genes. It is true that there are some cases where the memes and their phenotypes are represented by the same physical structures. However, this is sometimes true in organic world as well. There are what are known as "naked genes". Prions are probably the most famous of these. Prions are able to spread without being contained within a cell wall, or a viral sheath. Naked DNA exists in the organic realm - where DNA can leak out of punctured bacteria, before being assimilated - when sometimes the original genes are reused. Cairns-Smith's clay mineral genes are also examples of naked genes. Naked genes are perhaps more common in cultural evolution than they are in organic evolution - but that doesn't seem to be a particularly big deal.
Cultural evolution exhibits complex developmental tangles.
- Some say that the tangles of development thwart a memetic analysis. Wimsatt (1999, p. 288) writes:
In genetics, and in evolutionary biology, the structure of our theories supposes that we can separate out processes of heredity, development, and selection - if not physically, then at least analytically, each from the other, in our models of the evolutionary process. But these three dimensions of the evolutionary process are inextricably fused and confounded in the process of cultural evolution. This happens because of the pivotal role development and the life cycle assume in cultural transmission.
From my perspective, this is the opposite of the "naked memes" objection (above). In that, it was claimed that cultural developmental processes are so simple as to be non-existent. Here cultural developmental processes are being claimed to be far too complex for an analysis of inheritance to succeed.
Development is incredibly tangled and complex in the organic world as well. Population genetics just treats development as a black box, and completely bypasses the details of development in its analysis. Population memetics does exactly the same thing with a similar level of effectiveness. It turns out that there are an awful lot of questions that can be answered using this kind of approach. Now, I can easily imagine how ignoring development might irritate developmental biologists - but it really does make a lot of science to have a science of inheritance and a science of development. These really are pretty different subject areas.
Memetics is full of "just so" stories.
- Memetics sometimes gets criticised for being a collection of "just so" stories. Howard Klepper (2000) suggested that Blackmore's The Meme Machine went too far in that direction:
Blackmore applauds sociobiology as a science that has made great progress, and sees its critics as Luddites who are afraid of having their superstitions about human nature dispelled. But she never engages the serious criticism of sociobiology (and its close relative, evolutionary psychology) by those who find it lacking in the elements of hard science. Those critics call sociobiology a collection of "just-so" stories, which confuse teleology with causation and lack any evidentiary basis in genetics.
Mark Rosenfeldter (2001) offers a similar criticism of Aaron Lynch's book Thought Contagion:
Lynch's book is deeply disappointing; what memetics has chiefly generated is a new way to blather about society, sex, and politics, without rigor and without the slightest need to make sense. Lynch worries a bit in the introduction about this reaction; he pleads for a little patience - he wants to show what sort of exciting ideas memetics can come up with, not get bogged down in factual nitpicking. What he doesn't see is that the real problem is not just that he gets facts wrong. It's that he's developed memetics into a scheme for generating factless scenarios. By excluding rigorous analysis, testing, and verification from his methodology, he's simply refined an ability to tell just-so stories about social behaviours.
This certainly seems to be a reasonable criticism of Thought Contagion. That was a relatively early book, but it was full of speculation, with very little examination of the facts.
However, "just so" stories afflict genetics too. These days, there are plenty of people doing real science in the field - not just telling appealing stories. A few people telling what might turn out to be speculative "just so" stories does not seem to be a good reason for writing off the whole field.
Memeticis is not socially acceptable.
- Kate Distin (2010, p.233) suggests that memetics is not socially acceptable:
For now, at least, even though in my view memetics has established that it is quite theoretically respectable, in practice it is not yet quite socially acceptable.
There are certainly some areas where plenty of scorn is poured on memetics. Distin's solution was to simply drop the "meme" terminology. However, that solution has some pretty serious problems - you lose access to the meme term. Also, it isn't just the term "meme", it is all the rest of the genetics-based terminology that comes along with it. Being able to say "meme pool", "meme expression" and "meme flow" is nice - whereas repeatedly saying things like "cultural variant pool" soon becomes unnacceptably long winded.
So: political correctness be damned. If some people don't get along with meme terminology, that is their loss. Memetics is the correct theory of cultural evolution - or so close to it that it can be easily bent into shape. The serious students of cultural evolution within academia have converged on something very close to memetics - and memetics has the best and most popular terminology.
Memetics hasn't produced anything original.
- Some claim that memetics has not resulted in anything original. For example, Adam Kuper (2000) says:
And that is my deal objection to the whole memes industry: it has yet to deliver a single original and plausible analysis of any cultural or social process.
Cultural evolution is an old idea, but memetics has stimulated many people to consider it who would probably not otherwise have done so - and they have produced an original body of work. Blackmore's idea that memes promoted human ultrasociality has mostly cashed out in the last decade. Her idea that memes are responsible for the enlarged human cranium is both original and plausible, in my opinion. Memetics was first to apply epidemiological ideas to cultural phenomena (though see Cloak, 1975), and now this idea is much more widespread. Peter Richerson (2011f), an expert in cultural evolution, concedes that Dawkins got this analysis right:
I think it is near to undeniable that cultural variants are sometimes selected to become selfish pathogens along the lines that Dawkins suggested.
Also, in the mean time, ideas such as "viral marketing" and the "epidemic threshold" have become basic concepts in social media marketing (Zarella 2011b).
Memes are not necessarily good for you.
- Tim Lewens (2006) writes:
In organic evolution, the swift spread of some variation through a population typically indicates that the variation in question confers high reproductive success on its bearers. Things are more complicated at the cultural level. We cannot infer from the swift spread of a tune through a population that the tune has features than make it likely to hop from mind to mind. The tune may not be especially ‘contagious’ or ‘catchy’ at all; the tune’s producers may just be powerful enough to make it ubiquitous, hence more likely to be learned than far catchier, but more poorly-funded, competitors.
It is a straight-forwards misconception that "the swift spread of some variation through a population typically indicates that the variation in question confers high reproductive success on its bearers". To give a counter-example, cold and flu viruses spread rapidly through populations of humans. Yet they do not confer high reproductive success on their bearers.
Meme fitnesses are not determined by meme attributes.
- Tim Lewens (2006) writes:
This underlines an important limitation for memetics. In organic evolution, the swift spread of some variation through a population typically indicates that the variation in question confers high reproductive success on its bearers. Things are more complicated at the cultural level. We cannot infer from the swift spread of a tune through a population that the tune has features than make it likely to hop from mind to mind. The tune may not be especially ‘contagious’ or ‘catchy’ at all; the tune’s producers may just be powerful enough to make it ubiquitous, hence more likely to be learned than far catchier, but more poorly-funded, competitors. Once again, it is important that our cultural evolutionary theories are rich enough to document the diverse reasons why an idea may spread, and the memetic theory, by drawing a very close analogy between organic and cultural evolution, threatens to obscure the important distinction between contagious and power-assisted spread.
Alas, this supposed 'limitation for memetics' is a simple misunderstandiong of how biology works. Genes get forcibly distributed, just as memes are forcibly distributed. Traumatic insemination is one common example of forcible distribution of genes. A wasp planting an egg inside a caterpillar is forcibly distributing her genes. The idea that 'power-assisted spread of heritable material is confined to the cultural realm is simply a ridiculous one.'
Critique from semiotics.
- Some theorists from semiotics don't seem to like memes. Deacon (1999) says:
The theory of memetics is not the answer to a theory of social and psychological evolution, but reinterpreted it may suggest some bridging concepts that can lead to a unifying methodology for the semiotic sciences.
...while Krull (2000) says:
meme is a degenerate sign in which only its ability of being copied is remained.
The semiotics students are correct to say that memes are signs - provided we use the definition of the term "sign" from within semiotics - which is - pretty confusingly - different from the usual English usage of the term. However, they don't seems to have any kind of coherent case against memetics. It seems to be more that they would prefer to use the terminology of semiotics to describe memes - even though the evolutionary biologists got there decades before them, and have a far more relevant discipline behind them.
Memetics violates Occam's razor.
- Mary Midgley (2004, p.70) thinks memetics is too complicated - and that its story of parasites and mutualists is unnecessary:
As William of Occam observed, varieties of entities should not be multiplied beyond necessity. When human beings think and act, no extra entities need to be present in them besides themselves.
What Occam said was that you should not multiply entities beyond necessity. A certain amount of explanatory material is often needed is one is to actually explain the facts. Memes are, in fact, a pretty neat and beautiful explanation of human culture. Memetics essentially says that culture is part of biology, and fits neatly into a Darwinian framework. Where previously there were independent explanation for phenomena in the cultural and organic realms, memetics explains a lot of the phenomena in both realms with a single theory, Darwinian evolutionary theory. The motive is to explain many facts with a single simple, general theory. The numerous symbiotes and parasites of memetics arise naturally from the dynamics of evolutionary theory, and should not be counted as being additional assumptions. This is surely a project that gets the "thumbs up" signal from Occam.
Memetics is too negative.
- Some say that memeticists exaggerate the negative side of memes:
The meme, like any apparatus of evolution, lacks intent or purpose. Dennett, however, subverts this claim by describing the meme as a vicious and aggressive "parasite." Meme replication, in Dennett's language, sounds like a dangerous plague attacking humanity.
This is one of the complaints Peter Richerson (2010f) raises about memes as well:
One of the problems with the meme concept as it evolved is that users of the term focused far too heavily on the selfish potential of memes.Historically, it appears that there's some truth to this accusation of negativity. However, some of this is understandable. It is important to bear in mind that the cases where memes act as parasites (where meme interests and gene interests conflict) are the very ones where it is easiest to distinguish between the meme-based hypothesis, and the "inherited treasures" model of cultural evolution, and the idea that benefits must ultimately accrue to DNA genes. Also, cases where memes benefit at the expense of humans are often the cases of most interest - since many people want to know what to avoid. Tales of conflict often make the best stories as well - and then selective retelling results in the conflict stories being encountered more.
So, yes, some meme enthusiasts have emphasised the negative side of memes - but there are reasonable reasons for that.
Memeticists can't agree on what a meme is.
- Mary Midgley (2004) wrote:
Unless some clear picture emerges, showing what kind of entity memes are supported to be, the parallel between them and genes surely vanishes, and the claim to scientific status with it.
No. We understood that organic world evolved long before the mechanisms of inheritance were uncovered and the structure of DNA was revealed. Darwin didn't know what genes were or how they were implemented. A lot of useful scientific work can be done by treating the brain as a black box, considering its inputs and outputs.
We won't know the mechanisms by which memes evolve in as much detail as we know how nuclear genes evolve until we crack a lot of problems in artificial intelligence and neuroscience. However, there are still plenty of things we can do in the mean time. Population memetics does not depend on having such a detailed understanding of the microscopic nature of memes.
Memetics is “mind-blind”.
- Atran (2004) describes memetics as being “mind-blind”. Similarly, Pinker (2009) criticised Dennett's presentation of memetics as lacking details about cognitive processes, emotions and motives.
Population memetics treats the mind as a black box. Since most meme mutations take place inside minds, that means that most memetic mutational processes are also inside the black box. Such models are very useful for modelling cultural change - and they have the virtue of being simple enough to be tractable - but they make no claims to being complete models.
Of course, any complete theory of memetics would have to model the insides of the black box. However, we don't fully understand all the details of that just yet. So, I will plead guilty on the part of population memetics to being "mind-blind". Black-boxing complex and poorly-understood elements is a standard part of scientific modelling. Hopefully - as time passes - we will be able to fill in the details of the box, and develop more complex models of memetic recombination, mutation and de novo creativity in the process - though these will probably complement rather than replace the simpler population memetic models.
However, in the mean time, black-boxing the mind has proved to be a masterful move. Most of the successes in the field to date are due to using population memetics. Also, most of the peer-reviewed science in the area has arisen out of this approach as well.
Memetic linkage is too strong.
Memes are not "quasi-autonomous bots".
- William Benzon wrote in 2010:
She has misstated my problem with memes, asserting that it is the TERM I do not like. Not so. I think the term is brilliant, which is why I use it. What I object to is the USE of the term to indicate quasi-autonomous bots that go hopping about from brain to brain commandeering neural realestate in competition with other one another.
Memes aren't "quasi-autonomous bots". It does seem fairly accurate to describe them as pieces of heritable information of "quasi-autonomous bots", though. Many of the cultural creatures whose genomes are made of memes might be compared to symbiotic viruses or bacteria - due to them frequently having small genomes and short generation times. Academia apparently can't handle cultural creatures yet either. For details see the articles: Why no cultural creatures in academia? and Cultural creatures.
- Memes are not like genes because...
...memes are so intangible.
- Memes can be represented in any information storage medium. However, that is not so different from genes - which are frequently stored in databases these days. The simplest perspective that unites the cultural ands organic realms is to say that both memes and genes are forms of heritable information - and information can be stored in a range of different storage media.
...there is no memetic code.
- Some critics claim that there is no memetic code. For example, here is John Wilkins:
There is no code, no digital structure, and no conception like a locus, in genetics, where alternative genes compete under the neo-Darwinian conception.There are memetic codes that translate between memes and meme products. There are lots of them: BASIC, Cobol, Java, English, etc. What critics should be saying is that there is no universal memetic code. That is true - but: so what? Even with DNA the genetic code is not universal and variants on it exist. There is no universal genetic code for DNA either.
...memes do not have "loci".
- Memetics does have the concept of a "locus". "Locus" just means "place" and memes must be physicallly instantiated - and so they must necessarily have physical locations - which could reasonably be called their "locus".
The concept of "place" often makes use of some kind of an addressing scheme - and that is often true in memetics.
Addressing schemes allow memeticists to make statements such as, for example: the memes "red", "rouge", "rot" and "vermelho" compete for the same locus inside human minds.
...memes are not necessarily transmitted with high fidelity.
- If you check with the definition of 'meme' in a dictionary there is no mention of 'high fidelity transmission'. As with genes, memes can be exposed to any mutation rate their environment generates. High-fidelity transmission is common and useful, but it is not required. Adaptive evolution requires at least some high fidelity information transmission - so that some signals from ancestors reach descendants more-or-less intact - but that is a bit of a different issue.
...genes replicate - while cultural evolution is not based on replicators.
...culture exhibits blending, while gene transmission is descrete and particulate.
- For example, Alex Mesoudi (2011) writes:
There is also evidence that cultural traits blend when transmitted. During language acquisition, children appear to blend the speech sounds of their parents and their peers resulting in a shift toward the average pronunciation of several people.Memes can recombine using averaging and interpolation - because the environment in which they recombine is the human mind, which allows for more sophisticated types of recombintion that the simple copy-and-paste operations that cells can perform.
Remember that memetics does not claim that memes are like genes in every single respect - just that there's a substantial list of similarities as a result of both being the heritable basis of a Darwinian system. The details of how recombination happens does exhibit some differences - because recombination inside brains makes use of multiple variants and uses better tools - compared to what is available inside cells.
...memes are sometimes analog.
- According to John Maynard Smith (1999), the analog nature of some memes prohibits adaptive evolution:
Two features necessary for any genetic system that is to support adaptive evolution: The system should be digital and should not permit the ‘inheritance of acquired characters’.However, this claim is simply wrong. Analog systems can support adaptive evolution - according to basic information theory considerations.
Many memes have analog elements. Brains themselves contain numerous analog components, and - although brains make ubiquitous use of thresholding to cut down the noise levels - not much inside them is pure-digital.
Of course, if you do have low-fidelity transmission, then the inherited cultural information typically fails to result in adaptive evolution, and is often rapidly lost. Similarly with analog information. Analog information typically has terrible copying fidelity, and attempts to transmit it down the generations soon results in outcomes similar to those found in games of Chinese whispers.
So: analog and low-fidelity transmission mostly represent rather uninteresting corner cases of cultural evolution. However, it seems best for theories of cultural evolution not to explicitly exclude analog information. Mathematical models of cultural evolution are mostly indifferent to whether the inherited information is analog or digital.
...memes undergo continuous variation.
- Michael Doebeli (2011) writes:
This idea goes of course at least as far back as Richard Dawkins’ notion of memes (Dawkins 1976), but this is a loaded term that seems to necessitate the existence of cultural “units” of reproduction. Instead, the notion of cultural content is more flexible in general, and can for example easily accommodate replication of continuously varying cultural properties (Henrich et al 2008), as well as complex cultural traits, such as cooperation, or components of technologies and ideologies.Michael suggests that culture undergoes continuously variation - but memes are defined in such a way that they can only vary discontinuously. This idea is bizarre. Where is Michael getting this curious idea about memes from? It isn't clear - since he doesn't say. Perhaps he is making it up.
...memes are not part of a functionally integrated structure.
Specific genes are functionally organized to replicate organisms of a specific type, and specific types of organisms are functionally organized to replicate specific genes. The relation between memes and human organisms constitutes no such functionally integrated structure in a replicative process. Specific ideas and cultural practices (memes) might or might not contribute to the inclusive fitness of individual human organisms, but human organisms are not functionally organized to replicate specific memes.
Here the problem lies in looking for the benefits to humans. In memetics, benefits can also accrue to the cultural entities themselves. For example, humans have few adaptations for spreading Catholicism - but Catholicism has lots of Catholicism-spreading adaptations. The idea is that human genes benefit humans in the same way that Catholicism's memes benefit Catholicism. Humans are meme hosts. They don't necessarily benefit from their memes any more than they benefit from cold and flu viruses they carry.
...meme mutations affect adaptive evolution.
- Boyd and Richerson say:
genes can also be transformed by spontaneous changes called mutations. But genetic mutations are rare, occurring about once every million replications, and as a result their effect usually can be ignored when thinking about adaptations. If mutations occurred more often - say, every 10 replications - they would have a significant effect on which genes were most common. We think this situation is exactly what occurs with ideas, which can transform rapidly as they spread from one person to the next. If we are right, cultural change will be understood only if the effects of transformation and natural selection are combined.
This is not unreasonable. Meme mutations are probably more common in many cases - resulting in larger mutational pressures and a greater influence of mutational forces on the trajectory of adaptive evolution. The extent of these differences is sometimes exaggerated by a failure to consider copying and evolution processes within the mind. Memes - like parasites - evolve within their hosts - as well as between them. Once this is accounted for the mutation rate of memes is reduced somewhat. While this is an interesting difference between the way evolution works in the cultural and biological realms, it has little impact on the relationship between genes and memes.
- Memes are not like viruses because...
...memes are often beneficial, whereas viruses are deleterious.
...memes can be transmitted vertically from parent to offspring.
- This is one of the complaints made by J. Storrs-Hall in his book Beyond AI. He wrote:
but the virus analogy is way overblown; we think of a virus as a disease because it is foreign to our organism, opportunistic and antithetical to our well being. Conventional religions, in memetic terms are much more like chunks of our native genome - a memetic chromosome. We get them from our parents, typically only once in a lifetime. While they manipulate us for their benefit, so do our genes: the whole thesis of The Selfish Gene is that the genes have their own interests at heart, not ours. The superstructures of belief and behaviour - the memetic phenotype, if you will - that religions build are broadly beneficial. Their major ill effects are side effects of their struggles with their direct competitors - other religions and science. On a proper view, memes form the substance of our minds, rather than being diseases. Few religions are only transmitted down the generations vertically. Most religions have some potential for evangelism and horizontal transmission. It is true that some people do compare memes to viruses - regardless of whether they are beneficial. There are beneficial viruses, and they can be used to justify the practice. However, there is the word "symbiont" - which lacks the connotations of being deleterious to the host and so is probably more technically correct. Most memes are a lot like the heritable information of symbionts - and not much like native chromosomes - since their life cycle involves flitting between hosts. Thinking of memes as being like chunks of native memetic chromosomes is the fallacy of the extended genotype - a very bad way of thinking about how culture evolves. The symbiosis perspective is simply superior to the perspective of the extended genotype.
...memes interact inside their host more than viruses do.
- Here is Wimsatt (2010):
Suppose that whether you could catch a given virus, and how it was expressed, was a complex function of what other viruses you had caught and in what order, and you were forced to deal with complex interactions of tens of thousands of virus types per individual What would epidemiology look like then?However we do see interactions between virus-borne pathologies in the organic world too. Infection with one virus results in antibodies that sometimes offer subsequent protection against other strains of the same virues. A classic example of this occurred in the manufacture of the smallpox vaccine. In 1796 Dr. Edward Jenner was visited by a milkmaid who was infected with Cowpox. He infected his gardener's son with Cowpox. He contracted the disease - and recovered within a few days. Weeks later, Dr. Jenner exposed the boy to smallpox. Amazingly, the boy did not contract the deadly disease. This experiment led to the development of a smallpox vaccine. These days it is common for weakened forms of a virus to be given to people as vaccines - allowing them to produce antibodies against the real disease.
This is broadly similar to the way that exposure of one pyramid marketing scheme helps to produce immunity to other ones.
There are also cases where an organic infection results in other infections. Being sick puts a load on the immune system, which makes it more vulnerable to opportunistic infections. Persistent viral infections often make the host more susceptible to opportunistic infections by decreasing the production of some types of interferons. Some pathogens (e.g. HIV) sabotage the immune system, which can lead to infection with other pathogens. This is broadly similar to the way in which the idea of getting on the internet makes you more vulnerable to many other memes.
Memes probably do interact with one another inside hosts more than viruses do. However, the differences in this area seem to be more like differences in degree than differences in kind.
...memes are engineered - but they are not like computer viruses.
- Bill Benzon claims that memes are not like computer viruses here:
Dawkins has suggested computer viruses as an analogy (Viruses of the Mind, 1991). But those viruses do not spontaneously arise in an ecology of communicating computers. They’re designed by programmers living outside the ecology, programmers who understand how computers work and who use their design knowledge to design and code viruses which they then insert, from ‘above’ (in effect using Dennettian skyhook cranes), into the computer ecology. In the case of memes there is no such designer outside the system, unless, perhaps, you were to designate a god as the designer. And that is exactly what Dawkins does not want to do. The computer virus analogy is thus an utter failure.
I think this is silly. Memes are like computer viruses because they spread between hosts, behave rather like computer software when they are inside them and are engineered. The precise location of the memetic engineer seems like a side issue. Computer viruses are engineered by kids and criminals. Memetic engineers work for advertising companies and the military. That is a difference - but not one with a significant negative impact on the analogy. Indeed computer virus makers are memetic engineers - since computer viruses are part of human culture - and are a special kind of high-tech meme.
- Particularly pointless or daft criticisms.
Culture exhibits insufficient variation.
- Culture doesn't have enough variation to evolve. According to David Burbridge (2003):
In biology, most organisms have the capacity to produce many offspring, and there is considerable variance in reproductive success. This is a prerequisite for natural selection to operate. In culture, by contrast, even if social groups may sometimes in a loose sense reproduce (e.g. by forming colonies), the rate of ‘reproduction’ is very low, and has little variance. For example, there are nearly 200 recognised independent countries in the world, but it is doubtful if any of them can be said to have ‘reproduced’ during the last century (unless you count the breakup of the USSR and Yugoslavia as ‘reproduction’). Yet there has been immense cultural change in all of those countries during that period.
Here Burbridge is just giving a cherry-pickled example. Consider instead the billions of copies of "The Holy Bible", vs the small number of copies of "The Little Cyanide Cookbook". That is a massive difference in reproductive success. Or consider "The Beatles" vs "Vashti Bunyan". Again, a massive difference in reproductive success. Reproductive success exhibits huge variance in cultural evolution just as it does in organic evolution.
Memetics denies a role for chance processes.
Memes - unlike genes - are not copied
Evolution requires storage in stable molecules.
- Critic Luis Benites-Bribiesca (2001) says:
For evolution and selection to take place, genetic information has to be stored in a relatively stable molecule such as DNA in what Schrodinger referred as a "code script".
This is not technically correct. Evolutionary theory simply doesn't require that genetic information has to be stored in "stable molecules". It makes no mention of the medium of inheritance. There are other possibilities - for example, the information could be stored in streams of photons or electrons. This kind of thing actually happens fairly frequently - when memes move around in the modern world.
Memes do not self-replicate.
- In a section entitled "The Myth of Self-replication" Wimsatt (2010) complains that memes are not self-replicators, as supposedly claimed by Dawkins in 1976. His objection is that they need a lot of complex equipment in order to make copies of themselves, and so are copied, rather than being true self-replicating entities - which would make copies of themselves.
Memes do not literally make copies of themselves unassisted. Nor do DNA-genes, which also depend on a lot of equipment before copies are produced. I don't think anyone ever claimed otherwise. Memetics certainly doesn't depend on the memes making copies of themselves without assistance. There is a sense in which such entities "self-replicate" - if you consider the rest of the organism they are part of to be part of their environment. All things that copy themselves exist in some kind of environment, and depend on features of that environment to operate. However, this whole point about self-replication seems too mundane and trivial to bother having much of an argument about. That the thing that is copied must also do the copying is not a tenet of genetics - or memetics - in the first place.
The critique of Wimsatt (2010) is revealing in another way. The author acknowledges Dual Inheritance Theory, cites Boyd and Richerson - and agrees that culture evolves. In other words, the guts of the meme theory are accepted. He even postulates his own copied cultural elements - which he calls "Meme-Like-Things" (MLTs). The main difference between these and memes seems to be that his MLTs don't "self-replicate", whereas memes - supposedly - do.
I find this sort of thing rather exasperating. Call them "memes" already! The battle for the terminology to describe small bits of information in cultural evolution is surely now over. The meme term is fine. We don't need every author on the topic making up their own terminology in the hope of coining a new term.
Memes cannot represent complex cultural traits.
- Michael Doebeli (2011) writes:
This idea goes of course at least as far back as Richard Dawkins’ notion of memes (Dawkins 1976), but this is a loaded term that seems to necessitate the existence of cultural “units” of reproduction. Instead, the notion of cultural content is more flexible in general, and can for example easily accommodate replication of continuously varying cultural properties (Henrich et al 2008), as well as complex cultural traits, such as cooperation, or components of technologies and ideologies.Michael suggests that memes cannot represent "complex cultural traits". That's what memeplexes are for. The idea is that complex cultural traits may be represented by memes by using multiple memes. This is precisely analogous to the way in which complex entities may be represented in digital media by using multiple bits.