Is memetics valid?

[the tree]

For the sake of this debate, "Memetics is an approach to evolutionary models of information transfer based on the concept of the meme." and "The term "meme" (IPA: [miːm]), refers to any unit of cultural information, such as a cultural practice or idea or concept, which one mind transmits (verbally or by repeated action) to another mind"¹

[...] I believe the concept [of memetics] is utterly useless and pseudoscientific. It's an interesting subject' date=' and IMO it's worth a debate.[/quote'']Start a thread [...]

I am currently in niether camp, I do however see sociology and psychology as perfectly valid fields of study so it seems posible that memetics is a valid branch of those.

I don't know enough about either to have an opinoin yet.

 

_1. Quotes from respective Wikipedia articles.

[bascule]

I'll go ahead and state that I myself see memetics as a protoscience, but one I have a great deal of faith in (more so than Dawkins himself, it would seem) because it's so intuitively obvious to me.

 

I think perhaps the most rousing argument for it came in the form of Susan Blackmore's The Meme Machine. I could try to restate the ideas from that book in this thread, but what I'd rather know is what aspects of memetics people feel are particularly unevidenced or otherwise unsubstantiated. (I'm guessing the likely answer I'm going to get back on that is "All of it")

[PhDP]

Psychology, sociology and anthropology are sciences. They can be tested and have descriptive and predictive power. I also don't believe the study of cultural evolution is worthless, or that evolution shouldn't be applied to human psychology, evolutionary psychology does a fairly good job in this area. Memetics, however, is pseudoscience. It claims to be science, but it doesn't help us to understand anything, it adds new words (with vague meanings) and it can't predict anything (except extremely trivial things). It has been more than 20 years, still, no clear mechanism of transmissions were discovered.

 

To simplify a little, evolution is caused by three great mechanisms (+ some minor ones); mutations, drift, selection. Contrary to what Dawkins said (1986), we have serious reasons to believe that mutations, or drift, might be more important than selection. Mutationism is far from being dead (Nei, 2005), and neutralism and near-neutralism either (Ohta and Gillespie, 1996). Dawkins and Dennett has been particulary arrogant toward those criticisms. For them, selection is the answer (Dawkins, 1976, 1986; Dennett, 1995). While most biologists acknowledge that adaptation is an important part of evolution, Dennett and Dawkins defend a vision were adaptation IS nearly 100% of the "important" part of evolution, calling neutral modifications "boring" (Dawkins, 1986). Natural selection is the differential survival of genes. Without genetics, it wasn't very helpful, but with the arrival of population genetics and mathematical ecology, natural selection has been able to predict very interesting events. And we are currently developing very good models to predict age at maturity, the effect of environmental factors on phenotypic evolution, neofunctionalisation... But, natural selection acts with constraints.

 

When facing the idea of constraints, Dennett defend the very strange idea that constraints are adaptation in themselves (Orr, 1996). It's this kind of excess that have generated strange researches, like one about rape. The conclusion; it's an adaptation, but a careful study of this "adaptation" show it's a poor one (Coyne, 2000). Memetics is the consequence of this vision of evolution driven by natural selection and adaptation, and were "by-products" and "constraints" often don't even figure in the list of possibilities (Dennett nearly deny their existence).

 

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. Memetics is far more than that. First, it's an analogy to genetics, and analogy that cannot be justified. Sure, meme have the power of "self-replication", but they blend. Memes are not discrete in the way genes are, and population genetics can't work without discrete units. Because meme can blend, something that nearly killed "natural selection" as a valid mechanism before the arrival of Mendel's genetics, how can we create a science of population memetics ? And if they have nothing in common with genes, with not just calling it "cultural evolution" ?

 

Second, what's the use of this analogy ? Memeticists have created a whole new set of words, meme, memetics, memeplex, but can you use memetics to create knowledge ? Sure you can use memetics to make naive predictions like evolutionary biology used to do in its early years, the "memetics" of religion is a good example. Did we really need a new science to tell us that a religion unable to convince anyone would eventually die ? Or that a religion using the period in life were children are vulnerable (from a psychological point of view) will have good success ? Read an article about memetics, eliminate everything related to memetics, it still makes sense (in fact, a lot more). It's not normal. It's even more surprising because Dawkins advocate a new scientific language, devoid of jargon, I agree with him, but then why creating a new terminology (because memetics is everything but a novel idea) that cannot generated novel research projects ?

 

For example, what about religion. Memetics, or even cultural evolution is unlikely to be the ultimate cause of existence for religion, which seems to be a by-product of human evolution (Persinger, 1987; Atran, 2002). After you know that, saying that the religion with the best replicating value survived is trivial. Sure, if Islam is more widespead than Judaism, it's probably because Judaism is not as good as Islam to "replicate" their values. Did we need memetics to know that ? Do we need a "science" based on a faulty analogy to understand that ? Was memetics able to create new knowledge. No. According to Edmonds (2005), not only did memetics failed, but the scientific community is even loosing interest in it. It is surprising ?

 

About the current status of memetics. I know only one interesting research in evolutionary science where memetics is mentioned, it's an article by Lynch and Baker (1994), and I'm not sure Lynch is still supporting memetics. It's about the cultural evolution of songs in some species of birds. But because there's no real "mechanism of memetics", I fail to see where the concept of memetics is necessary. Also, Edmonds (2002) ask "Three Challenges for the Survival of Memetics" in the Journal of Memetics. Three years later (Edmonds, 2005), he concluded that memetics has failed, and he show that memetics, in the scientific community at least, is starting the fade. I don't consider memetics a protoscience because of its lack of nuance about natural selection, the absence of clear mechanisms of memetics and, more importantly, because protosciences have to make some progress if they want to be considered science one day. It has been 30 years, and nothing conclusive.

 

I'll quote Dennett; "My refusal to play ball with my colleagues is deliberate, of course, since I view the standard philosophical terminology as worse than useless--a major obstacle to progress since it consists of so many errors". It's exactly my view of memetics; a new terminology, useless, and its an obstacle to progress because it puts an emphasis on something that doesn't seems to exist; a relation between the mechanisms of population genetics and the evolution of ideas, and a belief that nearly everything is an adaptation.

 

P.S.: I have all the articles I've quoted in PDF if somebody wants them.

 

Atran, S. 2002. In Gods We Trust : The Evolutionary Landscape of Religion. Oxford University Press

 

Coyne, J.A. and Berry, A. 2000. Rape as an adaptation (A review of "A Natural History of Rape"). Nature 404, 121-122.

 

Dawkins, R. 1976. The Selfish Gene. Oxford University Press.

 

Dawkins, R. 1986. The Blind Watchmaker. W.W. Norton & Company.

 

Dennett, D. 1995. Darwin's Dangerous Idea. Simon & Schuster

 

Edmonds, B. 2002. Three Challenges for the Survival of Memetics. Journal of Memetics 6.

 

Edmonds, B. 2005. The revealed poverty of the gene-meme analogy – why memetics per se has failed to produce substantive results. Journal of Memetics 9.

 

Lynch, A. and Baker, A.J. 1994. A Population Memetics Approach to Cultural Evolution in Chaffinch Song: Differentiation among Populations. Evolution 48(2), 351-359.

 

Nei, M. 2005. Selectionism and Neutralism in Molecular Evolution. Molecular Biology and Evolution, 22(12):2318–2342.

 

Ohta, T. and Gillespie, J.H. 1996. Development of Neutral and Nearly Neutral Theories. Theoretical Population Biology, 49, 128-142.

 

Orr, A.H. 1996. Boston Review.

 

Persinger, M.A. 1987. Neuropsychological Bases of God Beliefs. Praeger Publishers

 

Popper K. 1969. Conjectures and Refutations: the Growth of Scientific Knowledge.

[bascule]

Phil, you certainly seem well versed in the subject, but it seems to me you're merely attacking the novelty of the idea and glossing over the new concepts it brings to the table and unifies into a higher level system. Really it's no different than the arguments that I hear against the novelty of Kurzweil's Law of Accelerating Returns (itself an extrapolation based upon sociotechnological evolution), or for that matter, the Singularity concept itself. The idea of the Singularity as machine surpassing man was pervasive throughout the '50s, seen in places like Fredrick Brown's "Answer" as early as 1954.

 

The truly novel idea behind memetics, as far as I'm concerned, is that the type of relationships we see developing via the collective evolutionary process of all life on earth are repeated throughout human cultural evolution. Perhaps the best example of this, and the easiest to parallel to the biological world, is language, because individuals overwhelmingly tend to reproduce within the same language group, which is a part of the overall power of a common language to unify a particular group. Languages appear to speciate as we see various groups who originally spoke the same language becoming geographically isolated, and through random drift we see the languages eventually diverge to the point where they become incompatible.

 

We see the same thing with religion. Judaism, Christianity, and Islam all share a common ancestry as Abrahamic Religions, but have diverged to the point of incompatibility. We see speciation playing itself out within these groups as well. For example, among Christians we see the Catholics, Mormons, the Jehovah's Witnesses, Seventh Day Adventists, and a whole assortment of Protestants too numerous to name. Judaism has fragmented into Kaballah and Messianic Judaism, and Islam into the Shia, Sunnis, Wahhabists, and other groups which have diverged in the precepts of their beliefs.

 

Do you see these sorts of parallels as simply irrelevant, or already encompassed by some other area of study?

[PhDP]

Phil, you certainly seem well versed in the subject, but it seems to me you're merely attacking the novelty of the idea and glossing over the new concepts it brings to the table and unifies into a higher level system.

 

It unifies ? There's no "population memetics", no mechanism of the evolution of memes. Memetics is based on an analogy it hasn't justified, how could that be unifying ? I'm all for searching general laws, I'm the first to get angry at biologists for not trying hard enough, but you can't draw an analogy without a justification, just because it seems to unifies. It's making me think about the analogy between the solar system and atoms, it's elegant, but it's downright false and unjustified. If you believe memetics is valid, you have to justify the need for this analogy, as memetics is based on an analogy. You certainly don't think memetics has invented the study of cultural evolution ?

 

About language,

 

From what I know, the study of the evolution of language is booming, but it's not based on memetics. Searls (2003) is building trees for languages, much like phylogenetic tree we use to draw the tree of life. Is this tree a good tool to understand the evolution of human population and dating languages, he thinks so. In this very short (2 pages) articles, he draws a lot of parallels to evolutionary biology, going as far as invoking Margulis' "endosymbiotic theory" to explain the invasion of the English language by Germanic elements. But he don't only draws parallels, he use mathematical and statistical tools from evolutionary biology, his "trees" are not trivial, they are an interesting contributions to our understanding of human evolution. Memetics, in 30 years, has done nothing of the sort. So Why is memetics necessary to study language ? Why do we need an analogy between genes and ideas ? It's not because the evolution of languages and religions "looks like" the evolution of species that we can make a science out of this resemblance.

 

About religion,

 

I still fail to see in your parallel between the evolution of religion and evolutionary biology how the concept of "meme" is useful. Perhaps the link you made between religions and species is novel, but what does it explains ? We all know religions are not species, and while I'll agree some parallels can be made, I fail to see, from a scientific point of view, their use, and I fail to see how you can call memetics "protoscience" with such parallels. At best, it's simply a metaphor, an analogy, a pedagogical tool maybe, but not science, it doesn't give us anything new. When it does present new "ideas", they're not very good, like the idea that memes are behind the enlargement of the human brain (Blackmore, 1999).

 

Blackmore, S. 1999. The Meme Machine. Oxford University Press

 

Searls, D.B. 2003. Trees of life and of language. Nature 426:391-392.

[bascule]

Phil, the real problem is that memetics is supposed to be a unified study of how information evolves in collective intelligence systems, but the modus operandi of consciousness is not yet understood.

 

You sound rather well versed in Dennett so I will assume you understand his multiple drafts model, which I would equate to neocortical columns employing a selection mechanism to extract phenomenological objects ("phenoms" in Dennett's language) from the global workspace of consciousness, believed to be the thalamus, and apply transformations which they resubmit back to the global workspace of consciousness, where they can in turn be selected by other neocortical columns. In this way the most selected objects become the most developed, and this in turn parallels the alleged memetic approach to the way in which collective intelligence operates.

 

Consciousness itself (if you ascribe to something similar to Dennett's model) provides the selection engine by which memetic theory operates, and therefore memetics cannot and will not have any sort of scientific footing until we have a detailed, empirical theory of consciousness.

 

Without an empirical theory of consciousness (although Dennett asserts multiple drafts is "an Empirical Theory of Mind" and tries his best to substantiate it with real-world experiments) I admit that memetics must be taken on faith.

[PhDP]

Well, Bascule, I think you are dodging the real issue here. I'm aware of Dennett's theory, but although I'm skeptical of his understandings of evolution since I've read the "debate" he had with Allen Orr, I don't know enough about consciousness to say his theory is plausible or not. But this has nothing to do with memetics, for some reasons, you're making memetics' ambitions shrink, dramatically ! It's supposed to be an unified theory of how "information" evolves. It has been "used" to understand the evolution of birds' songs, and we're still waiting for "population memetics". But if population memetics ought to be used for the evolution of information, from birds' songs to computer viruses, how could it depends on an empirical theory of consciousness, which has nothing to do with birds, or, for that matters, nothing to do with at least 99% of the evolution of information on earth ?

 

Also if memetics; has nothing to justify the relation between "genes" and "memes", if it predicts nothing new and has to be taken on faith, but still claims being a protoscience/science, then it's clearly fitting into the definition of pseudoscience given by wikipedia; "Pseudoscience is a term commonly applied to any body of knowledge, methodology, or practice that is portrayed as scientific but diverges substantially from the required standards for scientific work or is unsupported by scientific research".

[ydoaPs]

interesting article

[bascule]

Phil,

 

Call it a protoscience or pseudoscience if you want...

 

I guess I should add that my concept of memetics comes just as much from James Burke as it does from Dawkins, Dennett, and Blackmore. Burke's obsession was tracking the origins of novelty and innovation and showing how they spread throughout a population. He loved to show how one person influenced another, and how the presence of (technological) objects within one's environment affects one's behavior.

 

I really include memetics as part of my own philosophy of how progressive evolution permeates the universe, and I suppose its exact workings, as I envision, deviate significantly from those of Dawkins, Blackmore, Dennett, etc. As I stated earlier, to me, it's really about how phenomenological objects evolve in collective intelligence environments... and it's a pattern I see everywhere, especially in the field in which I work: social networking and folksonomies

 

If your argument is simply that memetics isn't a science, well, then I'm forced to agree, and have already conceeded this.

 

If not, what is your point? It seems like otherwise this is just a boring semantic argument

[Genecks]

***message deleted***

[PhDP]

Bascule,

 

Memeticists (and you) are using memetics to explain religion, politics, morality; if you concende it's not science, why are you using it to explains X and Y ? What's the difference between saying religion is god's gift to humanity and that it's caused by memes ? Both being nonscientific, you can't choose rationally between those two options. I really prefer looking at what science (evolutionary anthropology/psychology) has to say.

 

I think it makes no sense to defend evolutionary biology on one hand and then to use something as speculative and pseudoscientific as memetics. Memetics is using and distorting the language of evolution ("genetics", "population memetics", "natural selection"), how could I not care ? We're fighting creationists and we told them that natural selection is important, it's not tautological, it can be used to make predictions. And then, we have a bunch of memeticists, using natural selection for trivialities.

 

If the pattern you're talking about was more than superficial, we could build a predictive science out of it. The problem is that cultural evolution may looks like the evolution of genes, but it's totally different.

 

It seems like otherwise this is just a boring semantic argument

 

Memetics is all about semantic.

[bascule]

Phil, I readily admit that memetics has languished of late. The most pertainent problem was the internalist versus externalist debate. The externalists were the ones really trying to make memetics into a science, because memes-as-cultural-artifacts means you have empirical evidence to work with.

 

I don't think externalist memetics works though. Cultural artifacts are not the base replicating unit of memetics. A meme is a phenomenological object, an association pattern that exists within consciousness.

 

I really think the languishing we've seen in the memetics movement is due to the fact that externalists have realized that their approach simply doesn't work and internalists know that it will be at least a decade before we can even begin to conceive of studying phenomenological objects empirically.

 

I don't see how you can look on the patterns as being "superficial" at all. Perhaps I've been exposed to some different material. One of my favorite ones expressing very meme-like ideas was James Burke's Connections: An Alternative View of Change series, where he traced the evolution of ideas backwards in time, showing the bizarre ways in which ideas influence each other, and using a number of case instances drawing some general conclusions about the way information evolves. One of the most poignant that he brought up was that the rate at which change occurs depends upon the rate at which we communicate, which has continued to increase throughout history.

 

Anyway, certainly nothing scientific, but I found your comparison to electrons orbiting a nucleus to planetary orbits as an indication that you're rather averse to the concept altogether, and I'm surprised that, given your obvious facination with these sorts of ideas ideas, that you have rejected the concept outright. Are you rejecting it for any other reason that it's not empirically based and that the concept doesn't necessarily explain the sorts of predictions it makes?

[DV8 2XL]

My two cents: Memetics is just approaching the Mendelian level of understanding as it were, (to compare with genetics) we are still a far cry from the Crick and Watson stage.

[jasper7777]

I'm a graphic designer and am just reading through this forum -- trying to get my head around memetics.

 

Quote Phil "Memeticists (and you) are using memetics to explain religion, politics, morality; if you concende it's not science, why are you using it to explains X and Y ?"

 

Are gestalt theory and semiotics considered to be a science? To me it seems like although memetics isn't a science it has explainitory power in the same way. And also predicts-- Through memetics I can predict that if I design a advertising campaigne featuring anorexic models-- then people will not only copy the fashion of clothes-- they will adopt the idea that being thin is beautiful, so the anorexia meme will spread. But I guess what Phil is saying is that there are other theories that explain this better.

 

As a designer what Bascule wrote about language- following an evolutionary model exhibiting drift and speciation is fascinating to me.

I see the icons, typefaces, symbols, graphic designers make as visual language that also could be studied through memetics. For instance when graphic designers appropriate cultural symbols like the Maori Koru, into logos and branding systems- they eventually create variations that compete with the origninal symbol and change it into something else over time. In New Zealand the Koru no longer means Maori -- it means New Zealand. And becuase so many government institutions use it-- the Koru may eventually mean government institution. Anyway, I am finding memetics works well to explain the evolution of graphic design.

[PhDP]

Anyway, certainly nothing scientific, but I found your comparison to electrons orbiting a nucleus to planetary orbits as an indication that you're rather averse to the concept altogether, and I'm surprised that, given your obvious facination with these sorts of ideas ideas, that you have rejected the concept outright.

 

What kind of scientist would accept a theory which add new words, doesn't add anything to our knowledge of a subject, and lead to misleading reasonings ? It's catchy, but it's not science. Scientists are studying how culture is evolving (Nowak's papers are quite fascinating), but they don't need to base their researches on an analogy with a certain perception of genetics.

[bascule]

What kind of scientist would accept a theory which add new words, doesn't add anything to our knowledge of a subject

 

I'll be the first to admit that I don't adhere completely to belief in scientific knowledge. How can you?

 

and lead to misleading reasonings ?

 

Like what?

 

It's catchy, but it's not science. Scientists are studying how culture is evolving (Nowak's papers are quite fascinating), but they don't need to base their researches on an analogy with a certain perception of genetics.

 

Again, no one is claiming memetics is a science. Utilitarianism isn't a science. Does that render it invalid?

[PhDP]

Bascule,

 

Again, no one is claiming memetics is a science. Utilitarianism isn't a science. Does that render it invalid?

 

What is it then ? Blackmore, and many memeticists promote memetics as an scientific approach to cultural evolution. The journal of memetics was there to promote memetics as a science, nothing less.

 

Like what?

 

The point made by Dawkins was that we should expect all replicators to show a similar "selfishness" than genes. Recently, in a book devoted to selfish elements, "Genes and Conflict" (2006, Burt & Trivers), the authors, right at the start, say that most genes aren't selfish. Generally, authors are starting their books by saying how dramatically important the subject is to science, but in this case, they knew they had to break down the misconception that all genes are selfish. So, the vision of genetics Dawkins sought to expand to all replicators isn't even true for genes. Blackmore's hypothesis that brain-size has evolved because of memes is a good example of the kind of misconception you can get with memetics. The human brain stopped enlarging half a million years ago, while the supposed meme cambrian-like explosion occured 30 000 years ago, yet the enlargement, she said, is caused by memes.

 

So, we have, on one hand, plenty of trivial comparaisons and predictions, and one the other hand, confusion generated by memetics. It doesn't seem unfair to say memetics is useless.

[bascule]

What is it then ?

 

I'd view it as a study of information evolution as an emergent effect of the collective action of cognitive systems. Until we have a scientific description of how those cognitive systems function, we can't begin to engage in a scientific study of memes (as you can probably tell, I'm an internalist).

 

Blackmore, and many memeticists promote memetics as an scientific approach to cultural evolution. The journal of memetics was there to promote memetics as a science, nothing less.

 

Externalism just doesn't work. Memes are information replicators within cognitive systems. We have to understand how those systems work first. Otherwise memetics is like trying to understand genetics without first having knowledge of chemistry.

 

The point made by Dawkins was that we should expect all replicators to show a similar "selfishness" than genes.

 

I think radical fundamentalist religion, particularly Islamicism, is an excellent example of this. Suicide bombers who die in a glorious attack do an excellent job of replicating the fundamental units of information which underlie their radical religious furor, even though the cognitive system itself died in the process.

 

Recently, in a book devoted to selfish elements, "Genes and Conflict" (2006, Burt & Trivers), the authors, right at the start, say that most genes aren't selfish. Generally, authors are starting their books by saying how dramatically important the subject is to science, but in this case, they knew they had to break down the misconception that all genes are selfish. So, the vision of genetics Dawkins sought to expand to all replicators isn't even true for genes.

 

You have it here folks: the final nail has been hammered into the coffin of selfish gene theory.

 

But that aside: I don't think all memes are selfish either. So what's your point?

 

So, we have, on one hand, plenty of trivial comparaisons and predictions, and one the other hand, confusion generated by memetics. It doesn't seem unfair to say memetics is useless.

 

So what sort of label would you put on James Burke's connection-oriented historicism and the way in which ideas replicate between individuals who improve them over time through a combination of stochastic processes and directed action?