August 24, 2006

More talk about memes

I’ve just seen a bit of improvisational comedy.

Improv comedy purports to be something like on-the-spot writing – and maybe it is, in its way – but it feels like something else. Improv comedy tends to come off less like creative play than like the shuffling of pre-existing cultural molecules – what the kids today call “memes.”* Since the performers need to be in agreement during the unpremeditated performance, they’re forced to depend on strategies that they can be sure the others will “get” instantly – on concepts that are common property. Anyone watching improv is also there to appreciate the craft – the comedy itself is generally pretty pale when taken on its own merits – and for the most part that craft is the clever invocation and exchange of these existing notions. In this sense, memes, and the shared repository of memes that constitutes our common culture, are exposed in a particularly naked way during improv.

See, when I sat down to write this, all of the above felt like it would take about one sentence. But then I couldn’t figure out how to write that sentence and it turned into a whole paragraph. Very frustrating; the whole rhythm of my overall thought is thrown off; that was supposed to be the upswing but now it’s turned out to be the first movement, and I really don’t have the time to keep it up through what will apparently have to be a very long arc of speculation.** Let me try to just suggest the rest of what I wanted to say with a series of questions:

Is this (the shuffling of pre-existing cultural molecules) really any different from “real” creativity, or is it just a rougher-grained version of the same process? If these molecules (those that might be invoked during improv comedy) could be catalogued in a dictionary, approximately how many would there be? Do they really exist as discrete molecules in a way that such a dictionary is feasible, or are we tricked by our sense of recognition into thinking that they have Platonic existences outside of their specific usage, when in fact we are only recognizing rules of formulation (that is, recognizing individual snowflakes as being well-formed snowflakes and thinking this means we must have seen them before)?

If such a dictionary could be constructed – not only of comedic memes but of all dramatic (mimetic!) memes – could a grammar be specified to govern their usage? It certainly feels like such laws, or at least principles, exist – it is the elegance and charm with which these memes are deployed that we appreciate while watching improv… or reading a mystery… or taking in any form of art that does not attempt to disguise the fact that it is constructed from pre-existing molecules – and such judgments necessitate principles of taste, as well as underlying formal principles.

Listening to music, one often feels the same sense of continuously recognizing constituent gestures – “now he’s doing one of those; now he’s doing one of those, etc.” – but my efforts to isolate and catalogue these have failed instantly because the things are devilishly hard to disentangle from one another without them losing their essentials. Is that a case of the snowflake illusion, or is it just a difficulty arising from our poor ability to articulate the workings of music? Perhaps an attempt at isolating such things in spoken/dramatic culture is a more approachable precursor to determining the form of a musical equivalent.

Everybody likes concrete examples, so here’s a concrete example to finish. The other day an obscure movie that I haven’t seen was being described to me. In this movie, a man seeking to immigrate to the US tricks a woman into believing that he loves her so that he can marry her for his citizenship. Then, before they reach the border, events ensue such that he actually does fall in love with her. Then, at a crucial moment, while the man is elsewhere, a third party shocks the woman by revealing that he never loved her and was only using her – even though now he really does love her! When I was told this, I thought, “Well, sure, right. One of those.” The question is, one of what? Have I actually seen this snowflake before? My gut tells me I could make a lexicon of these and that such a lexicon is sitting in my brain right now. A commenter on this site once suggested starting a wiki of all memes, but before that process begins, a coherent theory of how they break down needs to be established.

* Long note on “meme”: I don’t like this word “meme” for what I’m talking about because the emphasis, as it was coined by Richard Dawkins, is on the fact that, in analogy with genetic information, such an idea propagates itself from person to person and is thus subject to principles of adaptation and evolution. What I want to talk about is the notion of a unitary cultural concept that is shared by many people, but without this emphasis on propagation, which is a limiting metaphor. After all, culture (and the molecules thereof) is not solely transmitted from person to person; concepts can, for example, lie dormant in books and films and whatever for years and then be picked up again by a new generation, now colored by all sorts of historical considerations. Or they can be disseminated to millions of people all at once on television, where in some ways they don’t really seem to their audience to have originated with humans at all. These kinds of events can no doubt find a place in a pseudo-genetic theory of human culture; my point is just that such theory shouldn’t provide the terminology for the culture itself. Meme essentially means “a unit of imitatable thought” when I want a word that means “a molecule of cultural convention.” Coin and suggest!

For your reference… Since my OED privileges have been wiped away, here’s the – ugh – American Heritage Dictionary on “meme”:

A unit of cultural information, such as a cultural practice or idea, that is transmitted verbally or by repeated action from one mind to another.

[Shortening (modeled on GENE) of mimeme, from Greek mimēma, something imitated, from mimeisthai, to imitate. See mimesis.]

** When I started, promising only to write one paragraph, Beth said, “You’re writing a one-paragraph entry on memes? That’s not possible!” And she was right.


  1. Sure you’ve seen this movie before… For instance:

    If we were doing this Linnaean style I would propose the following:

    Kingdom: Movie

    Phylum: Romantic comedy (A and B are not in love at beginning –> they meet and are drawn to each other –> they overcome an obstacle –> they finish in love)

    Genus: Obstacle is a misunderstanding of A’s intentions toward B or vice versa (rather than, say, a domineering father, or the Nazis, or A’s lack of self-confidence owing to his absurdly long nose)

    Species: Misunderstanding stems from A’s initially cynical feelings about B, which become genuine over time

    Can anyone propose a better “meme tree”?

    Posted by Adam on |
  2. I don’t think Kingdom: Movie is a good way to begin. Whether it’s in a movie, a book, a play, whatever, the essentials are the same. The “movie” aspect is actually the most superficial rather than the most fundamental.

    Also – this is partially devil’s advocate because your approach actually seems reasonable to me at first blush – does your Phylum: Romantic Comedy designation really matter? Couldn’t the mechanics of this misunderstanding also occur in another kind of plot entirely and still be identifiable as “one of those”?

    The way we recognize these things – the way they get shuffled during improv comedy – they’re highly interchangable. Like Legos. The problem with a tree-structure taxonomy is that it locks things into very specific positions within an overall scheme, whereas these … mimetic memes … need to be catalogued in a way that leaves them free for all sorts of reconfiguration. The Linnaean system is appropriate for lifeforms because of the way that lifeforms develop from one another – it reflects historical linkages. But a dictionary organized on a semantic tree structure would be irritatingly limiting. Even a thesaurus, which is meant to show linkages, doesn’t try to categorize further – that would go counter to the way we actually use words.

    I really think that if this were to be done for memes, it would have to be pure dictionary form – but then again, you’d want to show variant forms and related groups, etc. Maybe the project should be compared to the folk tale classification projects – none of which are particularly satisfying. Or think of it as like this. I think the answer will be some thousands of discrete items, most of which then sprout tree-like schemes for differentiating variants.

    My gut feeling is that this particular item’s essential nature is the “tragic misinformation,” which encompasses Romeo and Juliet as well as How to Lose a Guy in Whatever. Which is to say that I can’t think of anything that would make me say, “yeah, in some ways that’s kinda the same thing” that wouldn’t qualify as “tragic misinformation.” Then within “tragic misinformation” this is a particularly overused variety – “revelation of longstanding deception initiated prior to change of heart.” But then we need a way of indicating the satellite items associated with this one – e.g. the sincere but thwarted attempt of the deceiver to reveal the deception, earlier.

    Then again, it’s all much more complicated. “Change of heart” is its own little world, and might not this be a subset of that world? Or do they only exist in parallel? Anything tree-like begins to be trouble, no matter how limited the scope.

    Perhaps the only solution is to reduce all concepts to really atomic units – “change of heart,” “deception,” “reportage,” and then consider each combination as a sort of molecular formula. Certain molecules are more common or more likely than others, but that information is not part of the formulae with which they are notated.

    The problem here is that the system of codification rapidly approaches, you know, just everyday language. And what can that possibly teach us?

    Gotta go, more on this later.

    Posted by broomlet on |
  3. Plus ca change, plus c’est la meme chose. H.

    Posted by Anonymous on |
  4. Given the choice between a vertical, Linnaean classification scheme and a horizontal, dictionary one, I think the Linnaean system is more useful and not unfaithful to how memes operate in cultural works. For example, “the sincere but thwarted attempt of the deceiver to reveal the deception” (let’s call this “thwarted pre-revelation” for short) is not a separate meme, it’s a subelement of this particular species of plot, the way black and white stripes are a subelement of zebras. You couldn’t have “thwarted pre-revelation” in, say, a detective story, or even a romantic comedy of another species, like “downtown boy/girl meets uptown girl/boy” (see Lady and the Tramp, or The Way We Were), or “I hate you I hate you I love you!” (The Taming of the Shrew, When Harry Met Sally), or “My tragic fatal disease will help you appreciate life anew” (parts of Love! Valour! Compassion!, or that dreadful Richard Gere-Winona Ryder movie).

    In other words, I don’t think memes get combined horizontally, except possibly and only partially in the weirdo world of improv. You can’t just do “stranded on a desert island” + “we’ve only got twelve hours before it detonates!” + “she’s secretly a witch” and have it be pleasurable except on a meta level. And of course improv is all about meta.

    When these sorts of memes do get mixed and matched they tend to be from related familial groups — think of Love Story, which is “uptown-downtown” in the first half and “fatal disease” in the second half. To me it’s no surprise that Love Story feels icky and contrived — they don’t map onto each other very easily.

    So while I am attracted intellectually to the “Lego” model, I think the “tree” model works better.

    Oh, and I did Kingdom: Movie deliberately. Some narrative memes, like “I grew up and became a writer,” belong only to novels; some, like “a mysterious foreign man is initially enchanting, and the main character is enamored of him, but he turns out to be a cad,” belong to TV shows. This is not just a matter of length but of resolution: a TV show has to end up where it started, while it would be fatal for a movie to do so. I don’t think the “tragic misconception / initially cynical attraction turns sincere” meme would work in either a TV show or a novel. Maybe a play — I could call it Kingdom: Movie/Play, if you’d rather. Or perhaps it should be Kingdom: Medium-Length Non-Recurring Narrative. But that’s unwieldy.

    Posted by Adam on |
  5. I think you are confusing a dictionary with a style guide. The reason you can’t just do “stranded on a desert island” + “etc.” + “etc.” isn’t that it’s impossible or even “ill-formed” in any strict dramato-syntactic sense – it’s just that it would be stupid. So too would it be stupid for this “revelation of a longstanding deception” event to happen seven times in the same plot. This isn’t a taxonomically relevant feature of the event-meme itself; it’s just a question of sensitive usage.

    You’re simply wrong to say that you couldn’t have “thwarted pre-revelation” in other genres e.g. a detective story. The mysterious dame and the shamus are falling for each other; in a moment of weakness she almost comes out and tells him that she did it, but then thinks better of it, and then in the end, he figures it out for himself and has to kill her. Then in retrospect he sees that she must have really loved him because she tried to spill the beans.

    A real example would have been better but I can’t think of one right now, and while you can argue about whether the above is any good, you can’t argue that it’s perfectly legit as a recognizable specimen of recent western culture. My point being that while the meme might have a relatively strong connotative association with romantic comedy, it is not defining.

    And your defense of narrative “types” as the most basic categorization does not persuade me. I mean, “I grew up and became a writer” happens in movies all the damn time. Too often. Just the fact that you say “I don’t think X would work in a Y format” tells me you’re on the wrong track. Why should this have anything to do with taxonomy? That’s like arguing that the words in a dictionary should only be organized alphabetically after first being separated into the categories “can be used to talk about food” and “can’t be used to talk about food” because “I don’t think the word ‘oracular’ could ever describe food, so the category system is valid.” Well, that may or may not be true, but beyond validity we’re talking about relevance, and this all seems irrelevant to me. “Neither here nor there” is the expression.

    Maybe this problem arises because I haven’t worked out (or at least expressed) what constitutes relevance in this little thought experiment. I need more rigor to approach this issue! Still thinking about it.

    Posted by broomlet on |
  6. I think you need to be less resistant to Adam here. Sure, you can have “thwarted pre-revelation” in something other than a film. But in those other genres, it becomes a different meme. The revelation may be labeled similarly in a detective novel as in a comedy film, but the feel is entirely different. I don’t know if that’s because the actions are actually different, or if it’s because the viewer/reader attaches different connotations to it, but it’s certainly different. We prepare ourselves differently when we watch a movie (especially when we know the genre of the film), and the events take on different meanings. But I suppose that’s the tricky realm we’re getting to here. The “but she already knew he was her father! (or whatever)” in a Christie novel is entirely different from the “she already knew it!” in a How to Lose a Guy-typey movie. However you spin it, those memes are different.

    Okay, I don’t really have much else to add at this point, but I do want to thank you guys for keeping this blog going. I know this seems pathetic, but I really appreciate being able to occasionally escape into this kind of intellectual-nerd-with-self-awareness world that I really miss from college. I’m sorry that I don’t have better Internet access and can’t be a more active participant. But I do enjoy it when I can. Also, where’s my copy of Moby Dick? Joan already sent me Indian food and pens!

    Posted by Mary on |
  7. I have worked out some further thoughts on this topic that I mean to post, soon, as a new entry. But I just want to respond to Mary’s comment.

    The question here is what ought to constitute an individual entry in the meme-lexicon. If the idea of the lexicon is going to have any legitimacy, we have to believe that the listed memes are, in some essential way, always identical to themselves, even when they appear in different works. But in deciding what kinds of things deserve their own entries, we face a spectrum. We could say, at one extreme, that since the feel of a given meme is subtly different depending on its context, each work consists of elements unique to that work, which can be compared but not truly identified with one another from work to work. This pretty much makes the lexicon useless and the term “meme” inappropriate.

    The other direction on the spectrum tends toward very broad groupings. There you’d find more traditional categorizations like “romantic complication,” where the variations among individual cases of “romantic complication” are viewed as basically inessential. This also makes a lexicon unecessary, and again the word “meme” seems inappropriate.

    This notion of a lexicon requires us to choose a point somewhere in between. You and Adam want to place that point further toward the contextual-distinction end of the spectrum than I do. But you’re both defending your position by saying that the sort of distinctions you favor are undeniably relevant to one’s experience of a meme. I don’t deny that; I’m just saying that they’re not necessarily a good place to start a taxonomy. I think that if you follow the path that starts with dividing into genres, etc., you’re going to end up with a fractured and inflexible lexicon.

    Put it this way: if I look up “suit of armor in hallway,” I want to find citations for both Scooby-Doo and Beauty and the Beast. Somewhere within the entry, I want it to explain how those uses differ. Maybe even in an outline/tree structure – Scooby-Doo is I.a while Beauty in the Beast is I.c, or something. But I definitely want them to be under the same entry.

    Then, as some kind of appendage to “suit of armor in hallway” there would be dependent sub-memes. “Comes to life,” “drops blade,” “someone hiding inside,” “knocked to pieces,” etc. Under “comes to life” you might then have “eyes appear” and “does battle.” Etc. How to organize all these is a complicated matter. But it seems clear to me that there need be no higher tier of categorization than “suit of armor in hallway.” This item might be identified as type “decor” and have the genre tags “mystery / historical / comic book / children’s /…” etc., but it should not be held in a tree structure by these tags – just as a word in the dictionary might have the designation “noun” and the derivation “from the Greek mimēma,” even though the dictionary is not organized by part of speech or language of origin.

    Also: Sorry about Moby Dick! I stopped being able to access the OED online and that sort of killed my reading/vocab project, so the whole thing slipped my mind. But life finds a way and I will be returning with more rambling glossaries in a few weeks. At which time you’ll very much be wanting your copy to have arrived already. I’ll get on that.

    Posted by broomlet on |

Comments are closed.