Re: Replicators, was Non Homuncular Memetics

Aaron Lynch (
Wed, 08 Oct 1997 14:11:32 -0500

Message-Id: <>
Date: Wed, 08 Oct 1997 14:11:32 -0500
From: Aaron Lynch <>
Subject: Re: Replicators, was Non Homuncular Memetics
In-Reply-To: < >

Aaron Lynch responding to Valla Pishva:

>> I agree with your points about systems built upon systems. Infants, for
>> instance, generally learn the basic phonemes from their parents in order to
>> proceed to further levels of language learning. Some of babbling phonemes
>> they make are reinforced while other babblings are not. Language learning,
>> in turn, opens the door for all sorts of other learning.

> A good point; this brings up the issue of a "backlash effect"
>between systems as a result of environmental issues. So while parents are
>passing on their specific language meme (or two) to their child, they are
>actively excluding, or limiting the future scope of, other language memes,
>since, once the child is past a certain age, it is much more difficult to
>pass on other language memes. This is sort of an innoculation against
>other language memes based directly on the "lower" genetic system. The
>line blurs, though, when we look items such as religion; is there a
>certain age after which it would be highly unlikely, or destructive, to
>make someone disavow religious conviction?

I don't know of any studies regarding difficulties of changed or disavowed
religious convictions versus age.

>If it's not based on a
>physical barrier, then the relative activeness of a meme should be based
>on a conceptual system when promoting (1) the meme and (2) innoculation
>against other memes. Genes might be seen as active/passive in the same
>way if we take it that memes "know" the "information constant" underlying
>physical development and use that to their advantage occasionally, while
>genes "know" the "information constant" underlying the environment that
>facilitates them, and use that info to their advantage. Suppression
>genes and green beard effects (something which I independently theorized
>before learning of, Aaron :) ) might be examples, but I guess this might
>be stretching it.

Looks like we have something in common in theorizing first, reading later!
I'd say it raises the likelihood that you will theorize something that no
one else has theorized yet.

>> The idea that replication happens only with respect to an abstraction is
>> actually a correlary to a more general principle that the "sameness" of two
>> entities exists only with respect to an abstraction. If its any comfort,
>> recall that physicists have long ago adjusted to the idea that the
>> "simultaneity" of events exists only with respect to a coordinate system, a
>> framework of time/location abstractions. (I normally cringe at using the
>> word "relativity" in connection to culture, but it might be appropriate in
>> this context.)
> I like this alot. The epistemeological issues it brings up seem very
>relevant to consciousness issues, computer or otherwise.

Relevant to a lot of topics in science, really. Still, we have a lot of
people who seem to insist that memetics cannot be science unless some kind
of "fundamental units" of (interpersonal) cultural transmission are
discovered. The question sounds a bit different when you rephrase "what are
the units?" as "what are the abstractions?" Arguments over how to define
"the gene" are also best understood as arguments over which abstraction
systems are most useful or most "powerful."

Physics uses arbitrary units such as the meters and seconds, and has even
concluded that measurements made in these arbitrary units depend on frames
of reference in a universe that has no absolute coordinate system. That's
not to say that all coordinate systems or abstraction systems are equally
useful in understanding phenomena. Much as physicists choose their
coordinate systems to fit the subject of investigation, so must memeticists
(and geneticists) choose their abstractions to fit subjects of investigation.

As for computers, the relevance begins at the lowest units of information.
For example, engineers have devised somewhat arbitrary conventions for
calling all "signals" between 3.4 and 5.0 volts "the same" with respect to
"logic assertion," and all signals between 0 and 1.7 volts "the same" with
respect to "logic negation." And the specific voltage intervals range
widely from vacuum tubes to the latest microprocessor CMOS. In absolute
terms, all "1s" are not the same, and all "0s" are not the same, even
within a given "logic family."

It is often forgotten, too, that "digital" circuits are not fundamentally
digital--their transistors, vacuum tubes, etc have merely been wired and
powered to respond in a mostly "bimodal" fashion. The two modes are then
made by cultural convention to represent abstractions of "1" or "0," or
"true" or "false," etc. "NOR gates," "bistable multivibrators,"
(flip-flops) etc. are composed of macrosopically analog transistors or
tubes wired to produce mainly "bimodal" outputs from "bimodal" inputs.
Something to bear in mind if someone claims that memetic evolution cannot
happen because culture is not "digital."

>>From another posting:
>> Different classes of replicators are generally not isomorphic to
>>each other. The lack of isomorphism between genes and memes, for
>>instance, is a good reason to avoid too much dependence on analogy and
>Isnt the fact that they lack isomorphisms between them almost a
>definitional in differentiating different classes of replicators? Or can
> different classes be represented as subsets of other classes, but
>different in that they create contextually easier concepts to work with in
>understanding specific domains?

Classes of replicators can be defined in terms of their material
(information storage) substrates, but this does not rule out isomorphisms
between the systems of abstractions used to describe the replicators.
Consider, for instance, changing the substrates of "digital logic."
Substitute N-type silicon for P-type silicon, and vice versa and reverse
the power supply voltages. You can then build a whole network of these
deviant computers, and discover that all sorts of things such as computer
viruses exhibit abstract isomorphisms to the original network. Still, the
lack of isomorphism between DNA replicating in paired chromosomes and memes
replicating in brains forces us to avoid any attempt to build memetics on
forced metaphors and analogies. The metaphors and analogies make nice
inspirations and starting points, as well as pedagogic tools. But the
science itself should be expressable without them. Forced metaphors in the
face of non-isomorphism have also caused serious scientists to have an
immune reaction to memetics.

--Aaron Lynch

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)