Thursday, September 9, 2010

Memetic Sophistry

Over at the Psychology Today blog complex, Joseph Carroll is taking Norman Holland to task on remarks that Holland made concerning the relationship between the reader of a literary text and the text itself. Though I disagree with Carroll on many matters, I agree with him on this one particular issue. Beyond that, I think his critique of Holland can also be applied to Susan Blackmore’s equivocations on memes. Here’s what Carroll says about Holland:
This whole way of thinking is a form of scholastic sophistry, useless and sterile. It produces verbal arguments that consist only in fabricated and unnecessary confusions, confusions like that which you produce as your conclusion in the passage you cited from your book: “the reader constructs everything” (p. 176). This conclusion seems plausible because it slyly blends two separate meanings of the word “constructs.” One meaning is that our brains assemble percepts into mental images. That meaning is correct. The other meaning is that our brains assemble percepts that are not radically constrained by the signals produced in the book. That meaning is incorrect. Once you have this kind of ambiguity at work for you, you can shuffle back and forth between the two meanings, sometimes suggesting the quite radical notion that books don’t “impose” any constraints—any meanings—on readers; and sometimes retreating into the safety of the correct meaning: that our brains assemble percepts.
Blackmore equivocates in a similar fashion on the question of whether or not memes are active agents. Here’s a snippet from a TED talk she gave last year:
The way to think about memes, though, is to think, why do they spread? They’re selfish information, they get copied if they can. But some of them will be copied because they’re good, or true, or useful, or beautiful. Some of them will be copied even though they’re not. Some, it’s quite hard to tell why.
Here she talks of memes as though they are agents of some kind, they’re selfish and they try to get copied. A bit later she says:
So think of it this way. Imagine a world full of brains and far more memes than can possibly find homes. The memes are trying to get copied, trying, in inverted commas, i.e., that’s the shorthand for, if they can get copied they will. They’re using you and me as their propagating copying machinery, and we are the meme machines.
Here memes are using us as machines for propagating themselves. And then we have this passage where she talks about a war between memes and genes:
So you get an arms race between the genes which are trying to get the humans to have small economical brains and not waste their time copying all this stuff, and the memes themselves, like the sounds that people made and copied – in other words, what turned out to be language – competing to get the brains to get bigger and bigger. So the big brain on this theory of driven by the memes.
The term "meme," as we know, was coined by Richard Dawkins, who is also responsible for anthropomorphizing genes as selfish agents in biological evolution. Dawkins knows perfectly well that genes aren’t agents, and is quite capable of explicating that selfishness in terms that eliminate the anthropomorphism, which is but a useful shorthand, albeit a shorthand that has caused a great deal of mischief.

And one of the most pernicious bits of mischief is providing an example that Dawkins and others, such as Dan Dennett and Susan Blackmore, have to use to authorize talk about memes as agents. The problem here is that no one in memetics, as this pseudo-science brands itself, has been able to explicate memetic processes in a way that is both satisfyingly detailed in its treatment of mind and brain and that eliminates the anthropomorphic agency figuratively conferred on memes.

I called Blackmore on this point in a comment on her recent NYTimes post, The Third Replicator (in which she posits temes in addition to genes and memes). She’d referred to some of my work in a misleading way. Here’s part of my response:
Blackmore 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 real estate in competition with other one another. I see no evidence that such entities exist and find no theoretical value in positing their existence.
While her Times post was relatively free of talk about memes-as-agents, that post was also linked to her TED talk, which is why I’ve quoted it here. Here is how she responded to my remark:
I apologise for misstating your problem. I understand the difference. However, I do not think of memes as “quasi-autonomous bots” hopping from brain to brain but rather more simplistically as all the information we come across, day after day, hour after hour, and have to deal with. Some we use and forget, some we effectively copy on to other people, most we do neither. This includes stories in the morning paper, adverts, emails, songs, jokes, theories, designs for buildings or tea pots, images, dance steps, and countless other things.
At this point she’s clearly talking about humans as agents disposing of memes as they will. She continues:
“Commandeering neural real estate” makes it sound as though the neurons ought to be doing something else, but no, they are doing what they are designed to do, process all this stuff. Inevitably there is competition because there is just too much stuff for any single brain (piece of “neural real estate”) to cope with.
Note that neurons have no become agents, agents ‘designed’ to ‘process all this stuff.’ Given that neurons are living cells, they have such agency as is appropriate to cells. They ‘seek’ nutrients and ‘eliminate’ waste. And somehow ‘all this stuff’ gets processed, perhaps as some emergent property of billions of interlinked neurons in a nervous system ‘designed’ to process stuff. She concludes her reply to my criticism with memes as agents, competing to get copied:
You say you find no evidence that these entities exist but surely you cannot mean that songs, stories, theories, and stories in newspapers don’t exist. In my view, these are the memes. They exist and they inevitably compete to be copied and thus to continue existing.
And so it goes. As long as memetics is built on such sophistry, it will get nowhere, which is where it’s been ever since people tried to turn Dawkins’s casual suggestion in The Selfish Gene into a tool for studying culture.
X-posted at A Replicated Typo.


  1. I use the term meme with some reluctance, I Must confess I don't go out of my way to read anything by Dennet, Dawkins or Blackmore on the subject.

    Meme's are highly usefull things in oral story telling, the little series of hooks contained in narrative that allow you to remember the story after only one telling.

    A somewhat crucial thing for pre-historic, non-literate socities as it means they can remember the fact they have a culture. Its a highly economic way of storing information.

    Personaly I find some perspectives highly disreputable and I seriously question the motives that lurk behind such responses.

  2. Meme's are highly usefull things in oral story telling, the little series of hooks contained in narrative that allow you to remember the story after only one [hearing?].

    FWIW that's not how DD&B use the term, though just what they mean, either intensionally or extensionally, is not at all clear even to them. I don't know what term would be good for your purpose.

  3. Dennet seems too with regard to the talking tree, at least in oral performance in lectures where he uses memes partly to entertain and make his own story memorable.

    Striped of the belief that it is viral the basic definition of a meme complex, seems to fit rather well with this feature of oral story-telling.

    Telling = hearing, Scottish term. My use of terms is generaly terrible. A failing I need to correct.

    Dennet is fond of the talking tree, my favorite that I learnt in one telling and has entertained all my kids is the tale of the enchanted turd.

    I think DD&B have conjured and study just such an object.

  4. n.b. The enchanted turd subverts the normal conventions and outcomes of a standard narrative.

    Jack hears an enchanting voice singing from a locked tower in the woods and swears he will marry the owner of such a voice. He has not been carefull with what he wishes for.

    His rash haste and failure to properly identify the object of his desire results in his meeting something other than the beautiful princess.

  5. OK, the term 'meme' has a history and that history includes a popular use and a whole lot of fruitless quasi-serious debate. Dawkins coined the term fairly casually as the cultural analog to the biological gene. He figured imitation was the key to cultural evolution.

    The term went into popular circulation where it has a loose meaning of an idea or practice that's readily passed around. In the popular sphere the term may or may not be associated with evolution.

    At the same time various thinkers have attempted to make serious intellectual use of the term in the study of cultural evolution. I adopted the term back in the mid-90s when looking for a way to talk about cultural analogs to genes. When I discovered that Dawkins had coined a term for that use I adopted the term, not realizing that a lot of loose thinking had been built up around that term. If I'd known that, I might not have adopted the term. Having adopted it, I've attempted to assimilate it into ongoing theoretical work in a reasonable and useful way. I've written a bunch of blog posts about that and used the term in my book on music.

    However, when analyzing specific cultural objects and practices apart from trying to develop that more general theory, I don't use the term. On the one hand, it doesn't carry any conceptual content that I can't get otherwise and, on the other hand, it would likely to be distracting. Within literature, for example, one can talk of tropes, topoi, themes, and motifs, all of which are in use, and none of which bring along any of the baggage associated with "meme," which one can't control.

    As far as I can tell DD&B haven't provided the term with an interesting and useful theoretical context, nor as far as I can tell, has anyone else flying under the memetics banner.

    So, you say "Striped of the belief that it is viral . . . " And that's the problem. The popular use pretty much incorporates that meaning, and so does the use by DD&B. Stripping the term of that belief is more easily said than done. If you just use the term for its convenience, well, it will keep that meaning anyhow, because that's how your reader will interpret it. It will be a term that means "viral" but it will not link up to any detailed and useful model or theory.

  6. Using the term "neurons" with "memes" is kind of odd. Neurons are physical cells. Memes (if we use that word at all) are information. Information may be expressed in physical things, but information itself has no mass nor volume. It can't me measured in physical units. Therefore I would say that memes - or replicable units of information - compete for attention. And by "compete" I mean there are more replicable units of information than I can attend to, and the ones I do attend to have an advantage over those I don't, and the ones I copy "win."

    So, what is attention?

  7. Information and physicality is a tricky business. There is the informal slash common sense notion of information, which is mostly what people mean, even in the most intellectual of contexts. Mostly. But there's also information in a technical sense as defined in, say, information theory. Information in that sense is physical, perhaps in the same way that 'circle,' as the shape of a wheel, is physical. Physical things have shapes, and 'circle' is a kind of shape.

    Information is a rather more abstract physical property. But it IS a PHYSICAL property. A phone line can transmit so many bits per second -- that's the problem that lead Shannon to create a technical concept of information. If you want to play in this conceptual universe, then attention is a function of your capacity to process information. How many bits per second can your nervous system handle?

  8. A phone line can transmit so many bits per second --

    That's a constraint of the particular physical medium of the information - electrons, photons - not the information itself. Yes, information needs a physical medium to communicate, but information itself is not physical. Memes, if they exist, are not physical. Genes are molecules, memes are not.

  9. That's the standard doctrine, but I'm not buying. Real information, information that is active in the world, is always physical. If it weren't then it wouldn't have any effects in the world. What's peculiar and special about information is that it's possible to construct physical systems that preserve slash conserve information from one embodiment to another. It is also possible to have a physical system that allows one 'broadcast' from one embodiment to many other embodiments in one operation.

    That's why patterns of information are not 'rivalrous' goods. And that, in turn, is why copying is fun.

    Oh, and I'd be careful about saying that genes are molecules. Genes are properties of DNA molecules, which is not quite the same thing.

  10. Oh right, sorry. DNA is a molecule, gene is some variable unit. Hmm.

  11. Re: "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 real estate in competition with other one another."

    Are genes "quasi-autonomous bots"? Not really. Memes are the cultural equivalent of genes.

  12. In the organic world, it's phenotypes that act. Similarly with memetics.

    Religious memes *themselves* don't act - it is the bodies of those that they infect that bang on people's doors and are responsible for their copying and spread.

    So - it is usually not memes = robots - it is: memes = robot-making-instructions and phenotypes = robots.

    I hope that helps to clear up this muddle.

  13. Memes are the cultural equivalent of genes.

    Well, yeah, sure. It's easy to say that, but difficult to deliver on it in a useful way. Dawkins and Dennet haven't delivered, not even close. And Blackmore has a different conception, one where memes are like viruses, not genes. Dawkins sometimes talks that way as well.

    I've spent a fair amount of effect attempting to deliver on that notion — that memes are the cultural equivalent of genes — and, while I like what I've done (see, e.g. these notes), it's not clear where that line of thinking is going or whether it will pay off. But I don't try to get any conceptual mileage out of talking of memes as agents, while DD&B, and others, really can't get anywhere without such talk. It's the irreducible rock-bottom last-turtle-down foundation of their thinking about cultural evolution. They may pretend otherwise, but meme-as-agent talk isn't simply a convenient metaphor for a technical account available elsewhere. These folks have no technical account.

  14. Looks like you got it bad, Tim. Tell me, in detail, just how memes 'infect' people. And who or what is it the builds the robots using those meme instructions?

  15. Re: "meme-as-agent talk"

    You sound as though you are one of those who objects to "selfish genes" as well.

    If you can handle "selfish genes", then you should be able to handle memes being selfish as well - the metaphor is the same - and if you can't handle "selfish genes", you should probably get with the program.

  16. Re: "Tell me, in detail, just how memes 'infect' people."

    That's standard terminology in memetics. If you don't get what is supposed to mean, perhaps go read one of the books that explains it: "Virus of the mind" - or even the dreaded "Thought Contagion".

  17. Re: "who or what is it the builds the robots using those meme instructions"

    Who or what builds phenotypes out of *genes*? The laws of physics, seem to be ultimately responsible. Memes are just the equivalent of genes in cultural evolution - so it is the same with memes.

  18. Re: are memes supposed to be akin to genes - or viruses?

    "What's a virus? A virus is a string of nucleic acid with attitude." - Dan Dennett.

  19. I think i will just stick with motif. However I think some folks will identify them as viral memes anyway.

  20. Yeah, but at least you wont' be responsible for whatever nonsense they read-in via virality.