RE: Precision of replication

From: Keith Henson (
Date: Fri 20 Jun 2003 - 01:08:45 GMT

  • Next message: Keith Henson: "Typical progress"

    At 09:30 AM 19/06/03 -0700, Richard wrote:
    >Lawry wrote:


    ><<Richard, is it your thought that memeplexes and memes behave differently
    >when it comes to transmission and replication? I have looked at the case in
    >which only a subset of the memes in a memeplex are transmitted, and what
    >effect that has, but short of this I have been proceeding on the assumption
    >that a memeplex is essentially just a 'big meme' when it comes to
    >I think you've put your finger on it. It makes no sense to talk about a meme
    >being transmitted with less than 100% fidelity unless you are talking about
    >mutation. With a memeplex, you might be interested in less than
    >100%-fidelity transmission as long as the receiving mind exhibits similar
    >behavior as a result of sharing the memeplex.

    A bit of borrowed model and some examples might help here.

    Meme transmission between people is through a classic Shannon information transmission channel. That channel is severely limited as to bandwidth (a few bits per second) and noisy. As was mentioned on this thread, errors creep in at the transmit end, the channel, and the receive end.

    The way errors are corrected in electronic communications is through check sums and retransmission. I.e., you test that the information got through without error and send it again if it had errors.

    Something very much like this is done in teaching children where there is a cycle of feeding information to the children, testing and repeating until you verify that the information has made the jump to the child's mind. I suppose the way a word is spelled could be considered a minimal kind of meme--in which case children learn many thousands of them to a high degree of accuracy and are tested and corrected on spelling constantly (perhaps less today with the advent of spell checkers). [Spelling tends to be highly constrained by dictionaries, but even there you can see minor spelling drifts and divergences, night -> nite, plough -> plow, and color
    -> colour.]

    Now the fewer bits being transmitted the better the information fidelity. Thus short words are misspelled less frequently than longer ones and you have hundreds of millions of people who know "three strikes and four balls" meme that is part of the "memeplex" of baseball but far fewer of them know the fine points of a catcher dropping the ball on the third strike. [I don't feel the need for the term but I recognize that some feel they need for a name for complexes of memes. If I felt the need, I would favor Hofstedter's term, "a scheme of memes." :-) ]

    Thus meme transmission is *never* free of errors. But with lots of error checking and retransmission, the errors can be reduced to arbitrarily small differences.

    I might add that DNA replication is *also* subjected to error checking and correction by a collection of molecular machines.

    Most of the time memes get replicated "good enough." Long as you understand what is going on it is not something to fuss over.

    Keith Henson

    PS. A few more well known transmission error examples. "Play it Sam" ->
    "Play it again Sam." Neil Armstrong's words when he stepped out on the moon and the mistranslation of "poisoner" to "witch" in the King James bible.

    To go *way* back, does anyone think a child half a million years ago could watch a hand ax being made *once* and be able to chip out a flawless one?

    =============================================================== 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 : Fri 20 Jun 2003 - 01:15:27 GMT