RE: Meme definition (was: birthdays)

From: Keith Henson (
Date: Wed 18 Jun 2003 - 22:17:05 GMT

  • Next message: Richard Brodie: "RE: Meme definition (was: birthdays)"

      "Richard Brodie" <> wrote:
    >To: <>
    >Subject: Meme definition (was: birthdays)
    >Date: Sun, 15 Jun 2003 15:56:58 -0700
    >Hey, a memetics discussion!
    > >"Meme" was proposed by Dawkins as a name for a cultural replicator
    > >to the gene, and later refined by him and Dennett to be a mental
    > >As I pointed out in my 1995 book, meme is not the only cultural replicator
    > >nor THE unit of cultural evolution. Artifacts and subcultures can also
    > >fruitfully be looked on as replicators.
    >Keith wrote:
    ><<I would like to suggest you consider a simplification consistent with
    >Darkins and Dennett, namely memes as pure information, independent of
    >So the meme of chipped arrowheads would amount to the information about how
    >to make and use them. Taking only the making part, the information could
    >exist in a human mind, on paper (though for sure it would be tough to
    >successfully describe how to chip rocks in text alone), in a video of
    >someone chipping out an arrowhead, or to some extent in the object of an
    >arrowhead itself if a person who knew rock chipping but not arrowheads
    >could duplicate one from a sample.>>
    >The crux of the matter is that the information must be in a mind in order
    >for replication to occur.

    For the information to have any *real world consequences*, I completely agree with you it must be in a mind, just like a gene has to be in a cell for it to have real world consequences.

    But a person can write down a meme on paper, and someone who can't even read can make copies and distribute those copies to people who never meet the person who originally wrote down the meme. Is a meme replicated by the copying and distribution process? In my opinion yes because the end result
    (more copies of the meme in minds) after people read the copies is the same as if they were told the meme by the person who wrote it down. The metameme of spreading memes via print contributed mightily to the rise and current preeminance of Western culture. (This metameme spread very slowly into Islamic culture according to Bernard Lewis.)

    [Metameme: A meme that has major effect on the propagation and/or survival of other memes. The scientific method, logic, and public education are examples.]

    >Sometimes the meme goes out in the world encoded
    >in a straightforward, one-to-one way into a vehicle, like the method for
    >making arrowheads being made into an instructional video. Other times a meme
    >can replicate through a much more complex, probabilistic series of events in
    >which encoding may not be the best model to represent the replication. For
    >instance, suppose I am learning to play poker tournaments without the use of
    >a book. I start off really bad, but through practice I start to come up with
    >some strategies in my mind, ones I think the better players use to gain an
    >advantage. Sometimes I will be mostly correct and their strategy-memes get
    >replicated in my mind. Is it useful to say those memes are encoded in the
    >poker game? Possibly, but it's a stretch compared with the book on poker or
    >the instructional video. I may misconstrue the play and invent my own
    >strategy, a different one from what I think my opponent is using. I may then
    >use that strategy and a third player may infer the original strategy-meme my
    >opponent was using in the first place, misconstruing mine. I'm not sure
    >encoding is the best metaphor here.

    I would say memes are not involved in such de novo learning. There certainly are memes such as the rules for card games in general and the specific rules of poker that a person would have to learn in order to play but idiosyncratic playing strategy I don't think would count as memes unless it was obtained from observation of other players and imitated.
    (I.e., don't draw to a small pair.)

    >How about a child whose mother is a strict disciplinarian? The child blames
    >the strictness for her woes and makes a decision to be lenient with her own
    >child, who comes to the opposite conclusion and once again adopts the
    >strictness meme. Where is that generation-skipping meme encoded?

    In other minds. Assume child raising strategies 1 and 2 that are both common memes in the culture. In the case you described, a meme that could take the path from parent to child failed to make the jump and the child picked up the alternative from others. Same thing with the second generation. Incidentally, this would be more likely if both of the children were later born and thus more likely to rebel. (See the excellent _Born to Rebel: Birth Order, Family Dynamics, and Creative Lives_ by Frank J. Sulloway)

    >So as you see, I don't consider it a simplification to say that memes are
    >pure information regardless of medium. Some information is not fruitfully
    >viewed as a meme,

    Certainly agree with you here. Recent discussed examples were your current location and the time. Both are non-persistent information. Other examples are concepts like "unsupported objects fall" which are learned anew by every child from the environment and don't have to be transmitted as memes. But replicating *cultural elements* like how to chip rock into arrowheads or how to multiply single digit numbers by 9 or "look both ways before you cross" qualify.

    >such as molecular structures and DNA strings (when people
    >aren't studying them of course).

    Computer viruses, DNA strings (genes) and memes should all be considered pure information because they can be freely interconverted from one media to another without loss. (It would be silly to do so, but a computer virus could be encoded into DNA, put in a cell, duplicated, read out by a gene sequencer and converted back into an identical computer virus.) They differ mainly in the place in which the information has real world effects. A computer virus has to be executing in the proper computer, a DNA string must be in a cell, and a meme must be "running" in a human brain to have effects. They all belong to the more general class "replicators."

    >Some memes, such as poker strategies and
    >parenting styles, are sometimes replicated probabilistically, without an
    >obvious paper trail. Other times they are encoded in instruction books.

    See above.

    ><<A subculture is just a collection of memes. The fact that a whole bunch of
    >them are bound up in a mutually supporting package is not unlike Dawkins
    >pointing out that genes are bound together in much the same way in a
    >It's a collection of memes, artifacts, and people, less like a genome and
    >more like an organism. However, a cultural organism is not subject to the
    >biological constraint that modifications can only come through changes in
    >the seed, so it has some characteristics of a genome and some of an
    >organism. It also doesn't really have defined generations. But it can, I
    >think, fruitfully be studied as a replicator in its own right: its
    >existence, given the right environment, causes more of it to exist in the

    Or fewer if a culture fails.

    Let me suggest two analogies/examples for you.

    First, there are some vines and other plants that only propagate vegetatively, i.e., they have no seeds at all--somewhat like your description of a people and culture.

    Second, I recently finished Steven LeBlanc's Prehistoric Warfare in the American Southwest. After there was very widespread movement into large defensive pueblos, 24 of 27 of them failed and were abandoned. The extent this was due to an unsuited environment is not known, but some of them may have failed due to cultural factors. I.e., their culture which had worked ok where they were dispersed corn farmers was not suited for or did not evolve fast enough into a form where the numbers required for defense could live in large dwelling places. Or the failure of these groups to survive could have been genetic, or some combination of genetic/memetic or perhaps just bad luck.

    In both these you *can* look at the "replication" as replicating a higher level, plants with genes or tribes with a particular cultures. You can also consider it from a lower level, genes of the plant and memes and genes for the tribes.

    Dawkins and others have spent a lot of time considering these cases. I am not so sure that these cases contribute that much to the understanding of replication processes other than to make it clear that other levels can be considered. I find the co-evolution of genes and memes to a much more interesting area to think about.

    >As I agree with almost everything you've ever written on memetics, I view
    >this particular quibble as very minor. But if you keep criticizing me I may
    >threaten to go over to the "performance model." ;)

    <grin> Aaron Lynch and I went many rounds on this topic years ago. It is archived by Google in alt.memetics, Dec. 1996. We ended in minor disagreement but mutual respect. There are some good posts from that era archived. For example:

    (Robin Faichney)

    In article <B3C72D85966853406@>, Ton Maas <> writes
    >As I've said before, "information" is a real tough one to tackle,
    >especially if we want to decide on what it "is". One could argue that there
    >is no information in the DNA at all; that whatever is "in" the DNA becomes
    >"information" only when it is being processed in situ - in context. This is
    >where "meaning" or "relevance" emerges, depending on various factors
    >including the content of the DNA.

    Information may be tough, but it's not impossible. (In fact, I don't even agree it's tough.) The concept of information appears in physics, especially thermodynamics, where it is basically just the *form* or structure of physical reality
    ("information" and "form" are very closely related). Information in communications theory is physical information encoded, where encoding is any reversible transformation, and it is subjective, by which is meant, it resolves uncertainty, and therefore is dependent on the prior state of the receiver, [and] the nature of the uncertainty. Context determines that, and (what is much the same thing here) decoding. Meaning is subjective information.

    Keith Henson

    =============================================================== This was distributed via the memetics list associated with the Journal of Memetics - Evolutionary Models of Information Transmission For information about the journal and the list (e.g. unsubscribing) see:

    This archive was generated by hypermail 2.1.5 : Wed 18 Jun 2003 - 22:23:50 GMT