Re: Analog Feedback and Ringing

Dave Pape (
Wed, 11 Jun 1997 01:04:26 +0100 (BST)

Date: Wed, 11 Jun 1997 01:04:26 +0100 (BST)
Message-Id: <>
From: Dave Pape <>
Subject: Re: Analog Feedback and Ringing

At 18:43 10/06/97 -0500, Timothy Perper/Martha Cornog wrote:

>>... what's the justification for saying that /cells/ are
>>self-replicating? I was wondering whether they were themselves maybe
>>/apparent/ self-replicators, emergents of the self-referential system of the
>>Earth's organic chemistry. The reason we see self-replication rather than
>>wider self-reference, might be because the self-reference in those systems
>>is very intimate.
>>Sorry if this doesn't look like it relates to memetics... I think it does,
>>but want to float this question before digging myself any deeper into my
>>armchair memeticist's theories.

>Might as well take a break from the heavy duty stuff about ethics to
>comment about other heavy duty stuff.

Thanks for your neural resource!

>1.) The idea that cells are the true self-replicators is empirical. If
>you put cells in the proper medium, you get more of them.


>Cells turn out to be the smallest units that self-replicate.

>Cells, however, do appear to be some kind of complicated emergent from the
>Earth's organic chemistry, at least at one period long ago.

Okay... I think there's theories out there [oh good /citations/, Dave!]
about DNA precursors being things like... nucleotide chains linking on
crystalline clay molecules, and other complicated but non-biological
chemical reactions, all of which have this property of autocatalysis.
Meaning that >= 1 of the products of the reaction set catalyses... its own
I'd say that such systems are like the audio feedback system I described.
There's a self-referential system, from which an apparently self-replicating
emergent... erm... emerges [sorry- not enough synonyms. Was trying to avoid
saying "skanks forth" :)]. It's the feedback thing that's the apparent
self-replicator... only it's actually a very tight self-reference in the SYSTEM.

And I'm thinking that if cells require a medium, that's like saying they're
a thing, more of which is produced by self-reference in an autocatalytic
system composed of /the medium plus the cells/.

Thus, current chemistry, including biological chemistry, would /still/ be
the working of the autocatalytic chemical system that was bootstrapping
before DNA, and cells, were on the scene. Cells are STILL an emergent of
autocatalytic (read self-referential) chemical processes.

>2.) Amplification and feedback. ALL RIGHT! GOOD STUFF!!

Wahey! Joy dancing! :)

>Jumping in at
>the deep end, Dave, we can analogize memes not to these Platonic thingies
>but to information-bearing packets circulating in a social system. It now
>follows that the behavior of the system (which we assume has an energy
>source that lets it make more memes, or people for that matter), depends on
>how its feedback loops are built. It can be hooked up so that it rings and
>will eventually oscillate.

Is a group, or a team, an emergent of self-referential cultural interactions
between its members? A interacts with B, causing B to interact with C,
causing C to interact more with A... bootstrapping a group?

>Indeed, that is how one designs an analog

Wahey again! Analog oscillators are cool. I've got a couple of old analog
keyboards, and I used to piss about with resonance (=feedback) a lot. Little
did I know it'd turn out to be a universal principle ha ha ha ha... The cool
thing about resonance is, it's like, something from nothing.

>We need several kinds of energy storage components (inductors
>and capacitors) in the electronics, which suggests that maybe we have
>different kinds of meme storage components also. But we can also break the
>loop in various ways to stop oscillations.

Now then... the place I see the link is... the formation of meta-memes
(meta-concepts, category concepts...). A while ago I had loads of cognitive
dissonance going, because I knew that ideas were defined in terms of other
ideas, but that I treated them as entities in their own right. I started
thinking about how you teach people a new idea. What you do is activate in
their brains memories, neural traces, which correspond to already-known ideas.

As an example... I learnt my concept of "memes" by thinking about evolution,
psychology (my degree subject), ecology, emergent properties, etc. All those
ideas were active in my mind. Meaning that neurally, patterns corresponding
to those ideas were active in my cortex. Now, what happens when neurons are
active at the same time is, they become more closely synaptically joined
[neurophys doesn't come much more brutal than this, sorry, it's all a bit
preliminary]. Meaning that the neural groups which, when activated, give the
impression of me feeling [being? :)] those ideas, become linked... and it
feels like those ideas are associated together. So, activating one of the
concepts involved will tend to activate others... which may re-activate the
first-activated concept. ...Meaning that initial activation will produce a
more lasting "output" than before, and one which has different overtones.

Now, linking this to the audio feedback example... that's a bit like the
feedback loop getting tighter, and the amplified singer's voice "ringing"
with feedback overtones.

What I was wondering was.... is the point at which you say a new concept has
formed, the point at which this sort of cognitive self-reference produces an
apparently self-replicating (within cortex) "feedback tone"? The tone would,
I'm assuming, be pretty complex, time-based, and rich... and it'd code for A
NEW IDEA. So... my concept of "memes" [massive over-simplification] is the
feedback emergent from interaction between my concepts of "ideas",
"ecologies", "feedback", etc.

I dunno. I'd love to know what anyone thinks about this. I'd especially like
to work out whether it fits in with models like William Calvin's hexacode
for Darwinian selection of ideas in cortex... I'm a bit worried, I'm
wondering whether I'm talking more about long-term memory, rather than
online workspace seething... hey, maybe I'm talking about complete bollocks.

>This approach I find infinitely more appealing than all the digital models
>of how the brain works. The trouble with analog electronics, however, is
>that the math is a lot harder (Laplace transforms, for example). But I
>think the most fruitful avenue for memetic research will prove to be
>analog, and not digital.

I suppose EVERYTHING is digital if you accept the validity of quanta... I
guess we use analog maths because, although it's harder than digital maths,
you'd have to do a LOT of digital maths on quanta to model anything you're
likely to get beaten up by, so doing analog math becomes... erm... an
easier... option?

I don't have a problem with (kind of) digital neurons working in huge groups
to cause (kind of) analog effects...

Dave Pape
I am ready.

Phonecalls: 0118 9583727 Phights: 20 Armadale Court
Westcote Road
Reading RG30 2DF

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)