Re: RS

From: Chris Taylor (
Date: Mon 15 May 2006 - 10:12:32 GMT

  • Next message: Dace: "Re: RS"

    There is an interesting (if not informative) parallel perhaps in RNA viruses versus, say, us when it comes to fidelity of
    (genetic) information transmission:

    For RNA viruses (AIDS being the most famous) replication is very much a slap-dash affair, resulting in a 'quasispecies'; i.e. in the highly-dimensional space of all possible sequences, where each gene locus is a dimension, there is effectively a diffuse cloud of points, representing all the members of a population
    (and the 'average' sequence may not even be present).

    This is almost always true to some extent excepting true clonal pop.s, but the cloud is _much_ more compact in a species with a much more conservative (i.e. lots of error checking and fallback mechanisms) replication system (our error rates are vanishingly small compared to RNA viruses, even DNA viruses are pretty faithful [like poxes etc.]).

    So the mechanism of transmission of music offers analogues of both -- through the sheet music, much more like a DNA-based system, through 'hearing' (unless you are Mozart copying Allegri's Miserere -- much more like an RNA virus. The success of the latter very much depends on the
    'fitness' of the core idea -- i.e. getting it wrong is unimportant because the fundamentals are so strong that core idea is maintained (freeing the system to explore sequence space much more effectively to look for novel gains).

    Incidentally, as an aside, unless one has perfect pitch, one only ever learns the _intervals_, not the literal notes; also, given the sheet music, some people read it and 'hear' the tune, others need it to be played -- do you have the translation machinery..?

    But anyway on the overall point (baseball, tunes etc.) of whether there are _memes_: This is a fallacious argument through and through.

    To go via an analogy; what is a species!? Well it's easy(ish) if we're looking at animals (and general biology is terrible for really being mostly animal biology) but with plants there is often a continuous (effectively) spectrum of types that gradate from one species to another (consider also 'ring species'); and with bacteria for example, just forget it! The point being that even where we have actual _physical_ genes (or cistrons or whatever) we still only uses the species concept as a convenient label with which to work, rather than it being some sort of objective truth.

    Cut to memetics: _Any_ argument that has a core component that is contingent on a particular conception of a meme is _doomed to fail_ as classic pointless black-box-ery. We need to abandon this misguided attempt to 'discover' the true nature of a meme
    -- it is a convenient label to use and it is _not_ always applicable. All we really have are (related but not equatable) patterns in matter and individual minds; patterns expressed in matter (artefacts, sounds, faces, etc.) in the 'extra-mental' world can impact patterns in a mental world (a single mind) and vice versa. That is it. Any attempt to classify further than that, other than as a summary or thumbnail of reality for the purpose of reductionist modelling or general theorising is misguided at best and fundamentally misleading at worst.

    Memes are not real. Species are not real. Even genes are not really real tbh. If there are exceptions to a concept, then it is flawed unless it is understood from the outset that it is a convenient 'model' or approximation, for the purpose of analysis.

    As a sort of an aside in closing, consider the fact that memes
    (using the thumbnail conception) and genes (another thumbnail tbh) are very likely to be interchangeable in a sense -- if a species whose behaviour is genetically controlled ('hard-wired') evolves a new response, how is that different (except in timescale and the 'choice' of method of transmission) from a meme evolving to produce that behaviour and being passed around? Do we really need to restrict the ideas of memetics (that behaviour, as controlled by 'something', can be heritably
    [factoring in fidelity of transmission, which has an impact for the memetic and the genetic] impacted by features of the world
    [a set that includes all other people]) to the province of memes conceived of as most here might define them? Again a flawed conceptualisation that will unnecessarily restrict scope. Artificially restricted scope will produce unrepresentative theory.

    Cheers, Chris.

    Kate Distin wrote:
    > Dace wrote:
    >>> I agree with your interpretation (below) of what's going on when
    >>> watching a game or listening to a piece of music; but I think that there
    >>> is something fundamentally different going on when we get our cultural
    >>> information via RSs. I think that then there is genuine replication
    >>> going on.
    >> The problem with this view is that the meme for a piece of music comes
    >> across when reading the score but not when listening to it played. It's
    >> just not convincing. Clearly, a theory of memetics should explain how the
    >> meme gets across either way. So too, a theory of memetics that replaces
    >> consciousness with a memeplex is unconvincing. After all, consciousness is
    >> self-evident. A theory of memetics that finds memes in the minds of birds
    >> is also unconvincing, as it attempts to explain phenomena that can already
    >> be handled from a strictly biological standpoint. If memetics is redundant
    >> when it comes to birds, why not humans as well? By emphasizing the RS, you
    >> establish both a role for reflective consciousness, from which RSs emerge,
    >> and a boundary between memetics and biology. For this reason, The Selfish
    >> Meme is essential reading, particularly for those who've been led astray by
    >> Dennett and Blackmore.
    >> I think you've taken two steps forward, but in materializing the RS, you've
    >> taken one step back. Such is life!
    >> ted
    > Well, thank you for the vote of confidence in the book. I think that
    > we're going to continue to disagree about representational systems,
    > because I remain quite comfortable with the idea of their being physical.
    > But I think that you raise a good point about music. You're right, of
    > course: it is totally counterintuitive to say that musica doesn't come
    > across when heard. I transcribe simple pieces for my son all the time,
    > on the basis of my retained memory of hearing them in the past, not of
    > seeing their scores. Had I seen and memorised the written music then
    > I'd be able to replicate it for him. As it is, what he gets is an
    > approximation of the tune he wants, which he can play in C major with
    > his thumbs centred on the piano's middle C. But nonetheless he does get
    > the basics of the tune.
    > So now I am puzzling about this. One option is that music as hummed,
    > sung or played must also be an RS, just as language can be either spoken
    > or written. But I'm not entirely comfortable with this account, because
    > when I hum a note, e.g. middle C, it does not feel intuitively as if I
    > am representing information /about/ middle C. It feels like what I'm
    > humming simply /is/ Middle C. And you have reminded me, quite rightly,
    > that once we start constructing theories which completely undermine/deny
    > the phenomena we set out to explain, we're onto a loser.
    > Another option - which I'm tending towards - is that played/sung music
    > is not representational at all. It is behavioural - by which I mean
    > it's just something that we do, not something that we say. When I copy
    > the tune you have just played or sung, I am imitating your behaviour -
    > it's no different from when I copy your way of rolling out pastry or the
    > way you wear your baseball cap. There are of course different levels of
    > imitation, and on this view the imitation of a tune is on pretty much
    > the same level as the imitation of clapping or waving. Strictly
    > speaking we don't really need memetics to account for it, although it is
    > part of a bigger picture in which there are also more complex elements
    > for which memetics is a useful explanation. I think that memetics does
    > leave room for elements of human culture that are non-memetic (in the
    > sense of being just low-level behavioural imitation).
    > Non-representational things in which there is no distinction between
    > information and its effects because they're not 'about' anything, they
    > just /are/.
    > And maybe there are elements of baseball games that are like this too.
    > I mean that even I could probably try to copy how a bat is swung or a
    > ball is caught/thrown. But I wouldn't do it with any level of
    > understanding. I'd be pretty much a bluetit pecking at the milk. But
    > in order to start to understand what was going on, information has to
    > come in from somewhere. And at this point I still want to maintain that
    > observers bring it into the game rather than getting it from the game.
    > For one thing, I don't think that the reproduction is as predictable as
    > Richard suggests. I personally could watch almost any sport for as long
    > as you'd like me to, without gaining any inkling of the rules (or the
    > point). Hence Richard's comment about how most people have never read
    > the rule book - people's interpretation of what's going on will vary,
    > even when they play the game. Any similarity between their
    > understanding of the rules, and the rules as written in the rule book,
    > comes from what they bring to the game and reconstruct for themselves,
    > rather than from the game itself.
    > Now, we humans are so good at doing this - at reconstructing and
    > inferring - that in a sense it doesn't matter whether we call it 'real'
    > replication or not. Much of the same information gets across, whatever
    > the mechanism. Where I maintain that it /does/ matter, is at the level
    > of a theoretical analysis of what's going on here. Because unless you
    > keep the replicator-phenotype distinction sharp then memetics will flounder.
    > I think.
    > Kate

    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)

    This archive was generated by hypermail 2.1.5 : Mon 15 May 2006 - 10:38:10 GMT