Re: Sticky Memes

From: Chris Taylor (
Date: Wed 25 Jan 2006 - 01:54:23 GMT

  • Next message: Chris Taylor: "Re: Sticky Memes"

    Okay this is where more is needed to flesh out the (~)memes-only view, but it is there for the having, as exemplified by the Karl Sims thing I mentioned. He had evolving rewiring in fish between motor centres and muscle blocks where if a given change meant that the fish moved in a more coordinated manner (which meant swimming better) those links were reinforced and preserved. Basically the resultant behaviour was only one of the very many possible behaviours that network could encode, but by a trick that essentially exploits the fish's genetically-heritable hydrodynamic shape, it learns to swim. Nowhere did that swimming motor program exist in the fish's inherited legacy of genes and a bit of cytoplasm; it was truly emergent. A very compact way to encode a very complex thing (sequential firing of muscle blocks from front to back in two sets acting half a wavelength apart); essentially seed the system (the rule that motion is good) and let the fish rediscover its abilities.

    Now how might that work for the most basic behaviours in humans, like coordinating eyes at the same thing; perhaps as simple as a rule that when the same stimuli create less complexity in part of the visual cortex (two images compress to one stereo image) remember the muscle settings? (Bear in mind that the muscles controlling your eye movement are far from being three servo motors for the three axes, its a bit of a mess frankly and a real challenge to use as well as we do).

    Now how about learning to communicate? Is it possible that at a higher level the same rule applies? Getting a reaction to a prompt is one of the first really complex things babies do for fun; then there's the copying and being copied game. Again I can see generic rules that could control that without complex behaviours to be social being programmed in.

    I could go on but I think you see the principle; get for free what would otherwise be a bugger to encode (remember, more genes for your liver than your brain, by several thousand). And let us not forget that evolution is above all lazy. Why are our noses small and useless? Possibly because as soon as dogs started doing the sniffing for us, selection to maintain them relaxed. Now how about any genetically coded behaviour that could potentially be maintained via the meme pool, or by designing for the accidental discovery of the behaviour, wouldn't evolution jump at it and let those genes drift away?

    To my mind there are two important meme sources for this, which are (1) memes modelled from experience of the world ('proper' memes) and (2) memes that pop into existence as a result of experience alone, whether of the world, or of your sensory experience of yourself. I imagine low-level (pre)motor programs that are in essence (i.e. in 'substance') the equivalent of the memes in your forebrain, but they live elsewhere and differ somehow; memes that involve imagined action are somehow picked up by premotor resident patterns (some sort of resonance is the best I can offer) and passed on to become motor instructions. Why couldn't the way to touch thumb to forefinger be a 'motor meme' (abusing the meme name cos these things wold never be copied -- like comparing email ('proper' memes) to machine code
    (or maybe assembler actually)).

    I'm a fanatic...

    Kate Distin wrote:
    > Chris Taylor wrote:
    >>> As a slight aside, this is one of the key reasons why I don't go for
    >>> the mind-as-a-memeplex view. Philosophers talk about "propositional
    >>> attitudes" - i.e. the fact that we can take a variety of different
    >>> attitudes to any given proposition: we can
    >>> wish/fear/believe/disbelieve that "it's going to rain today" or "the
    >>> earth is flat" or whatever. We respond to incoming memes - we are
    >>> not identical with them.
    >> This is not incompatible, in fact the memes-only thing provides the
    >> most compact explanation. The diversity of responses is purely a
    >> result of the diversity of memes in there, and the difference between
    >> awareness and acceptance relates to previous discussions about having
    >> the 'same' idea.
    > I don't think it can be *purely* a result of the diversity of existing
    > memes, or how could this account explain what goes on in newborn (or,
    > going back in time, newly evolving) brains which have not yet got any
    > memes? And where would memes have come from in the first place?
    > A more likely explanation, for my money, is that the human brain has a
    > certain innate potential, which can vary like any other innate potential
    > (e.g. height) and develops as a result of interaction with incoming
    > memes. Kind of like a muscle, which develops with exercise. Obviously
    > I'd agree with you that as we acquire more and more memes, our existing
    > stock must influence our reaction to incomers; but I think it's also
    > important to acknowledge the influence of our innate mental capacities -
    > which will vary from person to person. And once you do this, there's a
    > gap between people and their memes.
    > To put this in non-memetic terms, would you say that we are purely
    > cultural creatures? Blank slate at birth? That *seems* to be implied
    > by what you're saying (possibly only to me!). What, though, of
    > genetically-related people who meet late in life and discover many
    > freakish similarities in personality and habit?
    > 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)
    > see:

    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 : Wed 25 Jan 2006 - 02:15:05 GMT