Re: Dawkins' Mutation Test for Replicators

James McComb (
Sun, 29 Aug 1999 00:33:26 +1000

From: "James McComb" <>
To: "Memetics Discussion List" <>
Subject: Re: Dawkins' Mutation Test for Replicators
Date: Sun, 29 Aug 1999 00:33:26 +1000


I have scanned Chapter 6 of "The Extended Phenotype". I do not see where
Dawkins applies the mutation test to cultural artifacts and behaviors.

It is true that he regards memes as residing in the brain, and artifacts and
behaviors as phenotypes of memes. However, I did not see that he gave any
reason for doing so. I doubt if it had anything to do with the mutation
test. (Did I overlook something?)


It is true that Dawkins doesn't use the mutation test on artefacts
EXPLICITLY. However, Chapter 6 is ABOUT the mutation test, and Dawkins' use
of it to attack alternatives to his theories. For example, he uses the
mutation test to attack organism- and species-level selection theories, in
order to argue for his gene-level theory. Then he 'clarifies' his
understanding of memes. This passage does not appear in Chapter 6 as an
afterthought, so presumably the mutation test figures in his reasoning for
abandoning the idea that memes could be something besides units of
information residing in brains.

If Dawkins DID use the mutation test to decide that cultural artefacts are
not true replicators, and we decide that many cultural artefacts do in fact
pass the test, then one of Dawkins reasons for switching to an L-meme-style
definition would be undermined.

Robin (?) [as attributed by Jake]:

Surely artefacts and behaviours pass this test with flying colours! The
copying of them as is, is ubiquitous.


I am not as certain. As is? Blackmore addressed this by her copying of the
instructions versus copying of the product. Not all mutations are passed
on. There is necessarily some conceptual work that goes into the copying
(either behaviors or artifacts) to preserve fidelity and prevent degredation
over generations. We have idealized conceptions of the artifacts and
behaviors based on our experiential interaction with them. It is those
idealized conceptions that get replicated in an artifact or behavior, not
the actual artifact or behavior itself.


Good point. In copy-the-product transmission, the 'replication' event is
rather vague. Because it is an idealization, an 'instruction' is more
clearly defined, and hence easier to model.

I have my copy of The Extended Phenotype right here. If I damage it, say, by
pulling out the pages of Chapter 3, no mutated copies of the book will be
spawned. In fact, in order to mutate the book I would have to ask its
publishers to change it, or perhaps modify a printing press which is making

Aren't cultural artefacts usually like this? That is to say, don't they
usually emanate from a few canonical copying sources (like a printing
press), instead of being copied around like whispers in a game of Telephone?
If so, then most cultural artefacts are highly resistant to mutation, and
with good reason. No-one would buy The Extended Phenotype unless people
believed it to be an accurate copy of Dawkins' original manuscript!

James McComb -

Fidelity! Fecundity! Longevity!

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)