Memes and Motifs: Living Memories

PLEASE NOTE: In modern Internet usage, "meme" has become a word for cartoons and similar iconic imagery. Although this will probably be the meaning the word will retain in future, it is not the original meaning of the term. And I don't see that the new meaning has relevance for textual criticism. This article is about the original memes -- self-propagating memories -- which I think are an idea textual critics should think about.

One of the trickiest tasks of a textual critic is to figure out why scribes do what they do -- that is, why they make the conscious and unconscious changes they make. We don't really have an answer. This article can't give a definitive answer, either -- but it may offer a new way of looking at the question.

In 1976, Richard Dawkins published a book, The Selfish Gene. Most of it is about genetics, and argues that genes, not populations, are the basis for species survival and behavior. This is to some extent controversial (beyond the relatively mundane controversy over evolution, which of course is not controversial in scientific circles), but one concept in the book --- the "meme" -- has developed a life quite beyond the community of those interested in biology.

Dawkins does not explicitly define the meme, but he gives a derivation: on page 192 of the revised (1989) edition of The Selfish Gene, he offers it as a shortened form of Greek-influenced "mimeme," from the root for imitation.

So what is a meme? In simplest terms, it is a self-replicating unit of culture -- anything which is passed on from person to person repeatedly by behavior rather than genetic influence. One might almost say a meme is a "brain virus," save that many memes are unquestionably positive. The analogy is in fact to genes -- what Dawkins calls "replicators." Just as genes reproduce so as to yield more copies of themselves in living organisms, memes reproduce so that they are remembered and transmitted by more people.

As examples of memes, Dawkins cites popular melodies, catch phrases (as an American youth will say "whatEVER" to mean "maybe, but the details don't matter to me and they shouldn't matter to you either"), and fashions in architecture or clothing or almost any other widely disseminated object. Fins on cars, by this definition, were a meme. Or shoes with pointed toes. Or "White Christmas." Or patterns on china plates.

We note, incidentally, that the word "meme" is a meme. Just a minor point for the set theorists out there. But it is a successful meme -- as we have seen, it is now a common word (although the common use is not the meaning Dawkins intended), and the term has been accepted into various dictionaries, and at least three books have been written about it by terms other than Dawkins. One of them, The Meme Machine by Susan Blackmore, points out the interesting fact that memes seem to be largely a human (or at least a primate) invention: Children imitate from a very early age. Few other animals directly imitate. They learn, certainly, and they learn by guidance, but not really by imitation (see Blackmore, pp. 3-4, or the example of the can opener below). It is humans who have developed the meming ability (and gone on to create various even more precise meme-preservation tools, such as books and computers). So strongly is the meme implanted in us that I have never heard any proposal for a society, or even an intelligent species, based on anything other than memes.

The key is that memes survive and spread. Some memes I find rather unconvincing; the fins on cars (not cited by Dawkins, I should note) were failed attempt to create a meme -- the equivalent of an extinct species. But the meme "Toyota" is going strong at the moment -- say "Toyota" and most hearers today will think of quality and fuel economy, despite the fact that Toyota, in addition to its small, efficient, high-reliability cars, makes small trucks with relatively poor gas mileage and reliability little better than their American-designed counterparts.

Some memes, in fact, are demonstrably false. Blackmore, p. 176, mentions the meme of Aliens Breeding With Us. Now it is possible that there are aliens among us -- unlikely, given the constraints of relativity theory, but possible. What is not possible is that they are interbreeding; if they can breed with us, they are us. That's what makes a species. So this isn't happening. But the story apparently is widespread enough to have a certain amount of medical literature about it. It's a contrary-to-fact meme -- and the fact that it's common would seem to demonstrate its memehood: It's breeding despite being absurd. Indeed, it's arguably increasing in popularity as meme transmission becomes more reliable. (Blackmore, p. 204, argues that memes, in their attempts to reproduce more accurately, are encouraging replicating mechanisms such as books, recordings, computers, the Internet. This I think does not follow -- it would be more effective to make people's memories better, and these devices in fact make their memories worse. But they do let memes spread faster, which is another desirable goal if you think memes have a "purpose.")

What makes memes memes is their transmission by imitation. Blackmore (pp. 44-45) notes how this is different from other sorts of learning. If you hear a story and retell it, it's a meme. If you hear it and forget it, it's not. Nor is it a meme if you learn it by conditioning, as one conditions a pet ("stand on your hind legs and I'll give you a doggy treat"). Passing on the behavior is the essence of the definition. On page 51, Blackmore notes three essential attributes of a meme: Heredity (the behavior is copied/imitated), variation (not every imitated version exactly matches the original version), and selection (not every copy survives; some are retained, some discarded). This surely will sound familiar to textual critics!

There are, of course, objections to the theory of memes. Some argue that they "don't exist." That is, you can't point to something and say, "that's a meme." It can be written down, but it, as a physical object, cannot be extracted from its context.

This is true, but not very meaningful. By this standard, a computer program doesn't exist, either -- it's just a pattern of magnetism on a hard disk or in a computer's RAM. You can't physically extract a program from the computer any more than you can extract the meme from the human. Sure, you can call up the program from the computer -- but you can call up the meme from the human, too: "Hey, pal, tell us the one about the banana, the handkerchief, and the railroad engineer."

Others object that memes aren't alive, even in the limited sense that viruses are alive, and so can't "reproduce." This is simply not true; thing which are not alive can reproduce. As counter-evidence we point to prions -- the chemicals which cause, among other things, mad cow disease. No one considers these alive, but they do reproduce -- and evolve. In a strange way, we seem to be seeing more and more of these non-living reproducing mechanisms -- probably because they're all parasitic, so they need life to let them exist. Unlike viruses and prions, at least memes aren't inherently "costly" to their hosts; it takes effort to maintain a brain, to be sure, but it's presumably no more difficult to store the meme of a popular song than an unpopular. It may even be easier, because something that "fits" the brain will need less storage space.

Perhaps this can be made clearer by noting an analogy of the brain-and-meme combination. The temptation may be to compare the brain to a computer, and a meme to a computer program. An even better analogy, though, might be to a computer program and a document -- say, a word processor and the file it creates. To open the file and make sense of it, you must have the program; the document alone is generally gibberish. And the size of the file is not always proportional to the actual amount of data. For instance, if a memo consists mostly of "boilerplate" text from a glossary, the file may be very small because all the parts are stored in other places. An efficient meme would presumably be one which can make use of other material already stored in the brain.

(There has been, to be sure, some "over-claiming" here, including an argument that memes are now driving evolution. Blackmore, in fact, argues that memes have created our notion that we have a "self," which she considers an illusion. This goes so far beyond the data as to be almost ridiculous. It is a curiosity that most of those who make extravagant claims for memes seem to be rather rabid atheists. I can't see why this should be so. In any case, these issues are for the scientists -- and the mystics -- to argue out. Our concern is not with how memes may have helped to create human society; rather, it's with how they might have affected the transmission of the Bible. Memes may or may not influence overall human behavior, but that's memes as a driving force, which is an extension of the original concept. We care about memes as survivors -- the ideas and phrases most likely to be preserved and passed on, whether they are Big Ideas or not.)

There are at least hints of some of the biology behind memes. The reason humans can imitate so much better than other animals is something called "mirror neurons," found in monkeys and apes but in much greater numbers in humans. (It has been speculated that the reasons humans have such big brains is so that the number of mirror neurons can increase.) What mirror neurons do is allow an observer to mentally imitate the actions of another person -- if you watch someone cutting a piece of paper with scissors, say, mirror neurons will play back your own actions so that you can "feel" the feeling of cutting the paper yourself. Your hand won't actually pump up and down, but they will tense up just a little as if preparing to do so.

It appears that mirror neurons, and the actions they imitate, can be started up by either sights or sounds. So you could either see or hear someone taking scissors to paper and still imitate the act. This, incidentally, includes facial expressions. Watching someone feel sad, your own muscles start twisting in sad ways -- which is how you can empathize with the sadness. (It may be that this is one of the problems of people with autism: This mechanism doesn't work as well, so they have a hard time imitating and a harder time understanding how others feel.)

The ability to imitate is vital. You've probably encountered dogs or cats which have learned the sound of a can opening, and scramble to come and get fed when it happens. But the dogs and cats don't try to learn how to open the cans; they just have a Pavlovian response: "I heard a phhs! sound -- that means food!" A human child, watching a can being opened, won't necessarily associate it with food -- but will go to the toy kitchen you bought two months ago and will play with the toy can opener. The dog or cat responds; the human practices. This is what appears to make memes a largely human thing.

(Note: Most of the material on mirror neurons comes from the program "Quirks and Quarks" on CBC radio, April 2, 2006. How much has been published and verified I do not know.)

A third objection is less to the concept of memes than to the (potential) science of mimetics. The objection is that there is no unit to measure memes. The standard example of this is Beethoven's fifth symphony and the opening line dit-dit-dit-daaaah. Is this a meme? Is the whole symphony a meme? The first phrase is the best-known part of the work, but the whole symphony is variations on a theme. Both are widely recognized, and repeated. Both seem to fit the definition of a meme. Does this mean that memes contain memes?

This can be argued in both ways. It is usually said that genes cannot be composed of other genes. On the other hand, if we look at information theory, a set can consist of other sets, and subsets of a set are still sets in their own right. (And sets, we note, have the same sort of non-existent existence as is claimed for memes.) The analogy to computer programs is also apt. Chances are that your computer has some sort of startup script, though the names vary from machine to machine. The startup script is a program, but it calls other programs. Those programs it calls are programs in their own right -- but are also part of the bigger program.

Even in biology, there are partial analogies -- notably in proteins. Certain complex proteins consist of assemblies of smaller proteins, which can perform some function independently but which perform another function when grouped. An example, cited by Jonathan Weiner in Time, Love, Memory, p. 194, is the process responsible for Huntington's chorea, involving two proteins, one produced by the huntington gene and another named glyceraldehyde-3 phosphate dehydrogenase, a common enzyme. The glyceraldehyde protein is used for many things (genes with multiple uses are so common that the term "pleiotropism" has been coined to describe their effects), so the DNA sequence that codes for it is clearly its own gene -- but it is also required by both the good and the bad versions of huntington (the reason for the disease is that the huntington protein grips the enzyme too tightly for it to perform its proper role). So we have a huntington gene and a glyceraldehyde gene -- but the extended gene which causes or prevents Huntington's Disease consists of the gene at the huntington spot, plus the glyceraldehyde gene. The glyceraldehyde is very much like a subroutine in a library of computer routines: It's used by many other body processes. Thus, there is no real reason why memes cannot be composed of other memes.

And while the lack of a unit of memes is a drawback, it is not necessarily fatal to the science. Set theory has no unit except sets. More to the point, a science can proceed without knowing its units. Darwin proposed the theory of evolution without even knowing that there were genetic laws (which, incidentally, made Darwin's original theory rather different from what we know now. Darwin didn't really propose a theory of evolution as such; the word wasn't even a regular part of his vocabulary. He proposed the theory of natural selection, which became the modern theory of evolution when combined with genetic theory). Somewhat later, when Gregor Mendel indirectly completed the theory of evolution by discovering his genetic laws, no one even knew that cell nuclei contained genetic information -- so while there were clearly things such as genes, they were just black boxes, with no known mechanism or location. Later, nucleic acids were discovered, but nobody knew what they did. Still later, Crick and Watson discovered the structure of DNA, but that didn't sequence the genome. Even now, there are plenty of genes whose functions we don't understand. Some sciences start from the bottom up -- but others start from the top down. Mimetics, if real, appears to be one of the latter: The big picture precedes the dirty details.

We do note at least one major difference between memes and genes: Memes -- at least, some memes -- have two ways to reproduce. The terms used by Blackmore (p. 63) are "copy-the-product" and "copy-the-instructions." The former reproduces by observation, the second by recipe. To take a very simple example, consider peeled carrots. Two people who have never eaten peeled carrots come to a party and see the peeled carrots and like them. One goes to the host and asks how they were prepared; he says, "Peel them with a potato peeler." The guest does so; that's copying the instructions. Following a reciple or an instruction manual is copying the instructions.

The other guest says to himself, "peeled carrots -- I can do that!" -- and peels the carrots with a knife. Same end result, different means. This is copying the product.

Genes reproduce by copy-the-instructions; they make reproductions of DNA. Most of the time, this produces exact replicas of the original genes. On the other hand, if there is an error in the instructions (e.g. if someone accidentally writes "peel the carrots with a potato grater" instead of "peel the carrots with a potato peeler") the result is nonsense and the outcome bollixed. Copy-the-product is much more subject to small variation (but also, indirectly, to improvements); my suspicion is that it is less likely to be rendered complete nonsense. This strikes me as rather similar to scribes who copy letter-by-letter versus phrase-by-phrase. The former may make an error on a particular letter, yielding nonsense but only in a very local way; the latter may make a more substantial change but almost certainly one that makes some sort of sense.

So why do memes matter to textual critics? Because the basic characteristic of memes is that they spread. Random ideas generally are not memes. Memes somehow fit into the shape of our brains, and to tend to intrude into other thinking. Think of how many phrases from Shakespeare or the King James Bible survive in English. Those are, emphatically, memes.

Now think about assimilation of parallels. Not all parallels get assimilated. Assimilation is toward the most familiar reading. In other words, it's not assimilation of parallels. It is, precisely, assimilation of memes.

Is this just terminology? I suppose you could say so -- but the concept is well worthy remembering. The key point about those assimilated phrases, like the memes themselves, is that they have survived and propagated. Why?

I can't help think of folklore motifs. For example, my three favorite works of fantasy are Tolkien's The Lord of the Rings, Ursula K. LeGuin's Earthsea books (at least the first three), and Lloyd Alexander's Chronicles of Prydain. These three stories do not have the same plot, but they have the same folklore motifs: The Quest, the Coming of the King, the Price (one person paying a high penalty to save many). The last of those, of course, is familiar to all Christians.

Why do these items appeal to me? Evidently there is some deep psychological "lock" to which they are the key. "The Price" appeals to billions, since Christianity survives to this day. The others aren't quite as popular, but they have appealed to millions -- e.g. the books I cited have between them inspired six movies that I know of, and possibly more, though the movies generally have been much inferior to the books.

Or consider the concept of "magic." Grant that ancient peoples had no scientific method; they could not explain lightning or earthquakes or hurricanes. But they understood that there was a natural order of things -- they had to, to follow the seasons! So why magic? The notion seems to be very widespread -- and yet it is most unlikely that any of them saw human beings perform magic. All the "unnatural" things they saw were without evident cause (the work of God, not a human being). There is no reason why a concept of "magic" should evolve, let alone be widespread. So why do so many cultures have it? Presumably because it suits our thought processes somehow.

What this says is that some things stick in the brain better than others. Whatever the reason, whatever their nature, they replicate and spread.

As far as I know, no one has set out to study what makes a meme -- that is, no one has done research on what sticks best in people's brains, or whether some people are more receptive than others. But there are hints. Sound patterns help -- rhymes, alliteration, metre. There are also indications that non-metrical patterns can help. The parallelism of Hebrew poetry, for instance. I recently saw an argument (I'm not sure where) that the reason Lincoln's Gettysburg Address is so memorable is its balanced phrasing: It repeatedly uses concepts in threes ("we cannot dedicate -- we cannot consecrate -- we cannot hallow"; compare the second inaugural address: "With malice toward none; with charity for all; with firmness in the right"). This may explain, e.g. why, when Matthew took out the "to hear" from Mark's "He who has ears to hear, let him hear," scribes repeatedly re-introduced it. Mark's form is "meme-able"; Matthew's probably less so. This is speculation in the absence of knowledge, but we have to start somewhere. We often say that Mark is assimilated to Matthew because the latter is the "stronger" gospel (whatever that means). But it may just be more mimetic -- after all, Mark sounds very Aramaic, while Matthew and (especially) Luke sound more Greek. To a Greek speaker, those Greek gospels must be easier to remember.

It should be stressed that memes are not necessarily good -- Blackmore, pp. 76-77, offers the suggestion that evolution favoured those who were the best at imitating successful people. In other words, you can get ahead by being personally successful -- or you can get ahead by aping or attaching yourself to someone who is a good leader. The result, perhaps, is a tendency toward fads -- some of them good, many neutral, some mildly bad (think, say, the 2005 trend toward pointed shoes for women, which are ugly, uncomfortable, and hard on the feet; or the current trend toward eyeglasses which are too small to allow decent peripheral vision), and some incredibly dangerous (the Nazi party, say, or the Bolshevik version of Communism). Memetics serves yet again as confirmation that what is popular is not necessarily what is right -- something very important in assessing, say, the Majority Text.

There has been, to my knowledge, no exploration of memes with connection to textual criticism. But it seems to me that it is an area that should be examined closely. If we can learn which stories and phrases stick in scribes' heads, we can much more easily guess how they will change the texts before them.

Some other characteristics of memes: They come in clusters. This again is like biology, where genes are grouped in chromosomes and a full set of chromosomes make up a genetic code. Words and music of a song can be separated but also go together (and reinforce each other). We have bins in our brains for Bible quotes or Shakespeare quotes. Some of these can be of different types -- we can't quote all of Hamlet, but can outline the plot while quoting "To be or not to be" and "The play's the thing" and the like. Is there a pattern to these clusters? And might it affect scribes?

Anybody want a nice interdisciplinary thesis?