Where does Blackmore's Replicator Power Come From?

Paul Marsden (PaulMarsden@msn.com)
Mon, 29 Mar 1999 16:54:23 +0100

From: "Paul Marsden" <PaulMarsden@msn.com>
To: "memetics" <memetics@mmu.ac.uk>
Subject: Where does Blackmore's Replicator Power Come From?
Date: Mon, 29 Mar 1999 16:54:23 +0100

I would like to run an idea past you in order to evaluate its worth.
Memeticists often speak of self-replication, replicators, and some hold that
memes are interactors as well. That debate has been done - and I do not want
to repeat it here. But if e take memes to be non-genetically transmitted
information (i.e. culture), or indeed a sub-unit of it; socially learned
behaviours, units of imitation, representations, instructions or strategies,
then it is difficult to see how this non-genetic information can properly be
called a replicator, it may be replicated, but it doesn't actively
replicate, unless a miracle ingredient, such as Blackmore's "replicator
power". Now, our brains may be hardwired to appropriate behavioural
strategies (or any other conceptualisation of a meme you many choose) under
certain circumstances, and these might be called memes, but that leaves
memes with an entirely passive role. However, having read Cloak (1986) I
think there might be a sense in which cultural information may act as a
replicator - and this is it: (Comments Please)

One central insight of memetics is that this process of differential
replication or imitation will result in some cultural information being
spread simply because it is good at being spread, rather than for the
benefit of the agents that are spreading the culture. This memetic Úlan, or
replicator power as Blackmore calls it, is a simple consequence of the
casual logic of natural selection. It is perhaps unfortunate that Blackmore
does not make explicit the mechanism by which memes may legitimately be
understood in terms of replicators, because this leaves the reader and
critic in danger of misunderstanding the whole paradigm and conceptualising
culture as some kind of active parasite. It would be more proper to say that
cultural information is replicated rather than replicates itself.
Nevertheless, cultural information does have a self-replicative effect, or
more properly, an effect of self-emplacement, by virtue of the fact it
modifies the environment within which it operates. The causal logic here is
identical to that of natural selection of genetic information. Imagine a
population of information processors in an environment of information. Some
processing is completely transparent, that is, the input is identical to the
output, whilst other processing modifies the informational environment by
outputting information that is different to the input. The usual
evolutionary scenario is to imagine that the processing activity of the
processors results in differential survival chances for the processors, and
given a replicating population of processors and heritable processing,
differential replicative success. Processing that enhances survival and
replication chances, because it allows the processors to exploit an
environment upon they are dependent, tends to increase in prevalence, simply
because those processors have a higher probability of successfully
replication. Now, given a degree blind variation in replication, and an
environment of insufficient resources for the population, relatively
maladaptive processing may become extinct, whilst relatively adaptive
processing will become the norm, until a new blind variation is selectively
retained because it has relatively more evolutionary fitness. In short, the
population will evolve through a mechanism of natural selection, processing
that enhances the chances of that processing's self-emplacement will spread
through the population.

However, by shifting the focus from the processors to the information that
the processors are processing, a second evolutionary mechanism may be
identified, the natural selection or self-emplacement of the information
itself. Because some processors output information that is different to the
input, the informational environment is changed, and because output will
depend on input, outputs will have some causal effect in determining the
probability of a similar output being elicited. In other words, some
information may result in outputs that enhance the likelihood of that output
occurring again, whilst others may modify the environment to make such a
reoccurrence less likely. As more information is processed, information that
increases the chances of its own output will naturally increase in
prevalence whilst information that reduces the probability will decrease.
Over time, information that acts so as to increase the probability of this
self-emplacement will become more prevalent, and if there is more
information than can be processed, only that information that has a force of
self-emplacement will be processed. In other words, Blackmore's "replicator
power" can be seen as the product of the non-miraculous causal logic of
natural selection, that is, self-emplacement.

Well, what do you think (as I dive for cover)?

Paul Marsden
Graduate Research Centre in the Social Sciences
University of Sussex
e-mail PaulMarsden@msn.com
tel/fax (44) (0) 117 973 2051

Journal of Memetics - Evolutionary Models of Information Transmission:
http://www.cpm.mmu.ac.uk/jom-emit/

===============================================================
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: http://www.cpm.mmu.ac.uk/jom-emit