RE: Papers critical of memetics

Nick Rose (
Tue, 9 Feb 1999 15:26:40 -0500 (EST)

From: Nick Rose <>
Subject: RE: Papers critical of memetics
Date: Tue, 9 Feb 1999 15:26:40 -0500 (EST)

Nick wrote;

>>1a. Why do so many people interested in memetics have
>>different definitions of the meme and what is the real
>>A real definition would require some evidence to support
>>it. The problem with 'neural' memes is that we have no
>>way (perhaps only currently) of generating that evidence.
>>If memes are defined as behaviours (ala Gatherer) rather
>>than ideas at least we have something we can measure. A
>>good definition should be like a good theory - testable!

Aaron wrote;

>Good to hear from you, Nick.

Nice to be heard ;)

>I agree that a good *theory* must be testable, but
>*definitions* used within such a theory need not be
>testable in their own right at all. A meme is not some new
>kind of entity; it is merely a subclass of a class
>of entities already recognized and researched, namely,
>memory items.

I have a hard time thinking of a single psychologist or
neurophysiologist who can identify a unit of memory... As
for the definition - well, if it's not testable how are you
to convince others that they exist? This I think is the
weakness in the memes as units of memory definition - not
to mention the fact that there's plenty going on in
our memory which is not at all memetic!

>Specifically, it is the subclass of memory
>items whose causation depended upon previous instances of
>the same memory item. If one already agrees that such a
>subclass of memory items exists, then one is not claiming
>anything new simply by giving a name such as "meme" to it.

Yes, but I'd say (and the majority of psychologists would
probably agree) that not all such memory items are memes.

>Merely giving a new name to an existing class of entities
>is a scientifically conservative move that does not
>require new empirical work. Dawkins was not claiming to
>discover any fundamentally new kind of entity that
>required empirical verification, but was merely giving a
>new name to a subclass of long recognized entities (items
>of brain-stored information).

Dawkins was very unclear as to what a meme was. I suspect
he'd accept that a large amount of 'brain-stored
information' was not memetic.

>He did so only to encourage
>a certain line of evolutionary replicator theory, and it
>is that theory that requires testing, not the definition.
>Of course, if the theory were completely falsified then
>the definition would likely be discarded as useless.

Evolutionary replicator theory is certainly something that
needs testing. How we can test this when our definition of
the replicator which underlies the process is itself
untestable! i.e. how can you justify scientifically the
way you've defined a meme?

>Even meme critics such as James W. Polichak agree that
>such things as "memories" exist, and that "memories" are
>valid subjects of scientific study despite not being
>*directly* observable. When Polichak refers to "false
>memories," for instance, he is referring to something that
>we can measure, even though it is something we cannot
>measure *directly*. We measure it by its effects, much as
>we measure electrons, body temperatures, and many other
>things by their effects. Measuring memes by their effects
>on behaviors therefore does not challenge their status as
>neurally stored.

Sounds like cognitive science to me. This goes back to the
're-inventing the wheel' criticism. Can meme theory say
anything more about how memories are stored or affect
behaviour than all the work in cognition.

>Those who imply or seem to imply that "the meme" is a
>newly discovered kind of entity tend to provoke the
>greatest scientific skepticism. Claiming to find a new
>kind of entity (rather than just naming a subclass of
>already recognized entities) really does require that one
>produce a new body of evidence from the start, and lacking
>such new evidence leads to criticisms such as Polichak's.

Hmm... But I think you'll have a hard job convincing
sceptics that a memory item is a meme and requires no more
refinement than this! And how are you going to refine your
definition unless it is empirically testable?

>My 1991 Journal of Ideas paper (Thought Contagion
>as Abstract Evolution), was clear that the word "meme"
>only signifies a way of partitioning the realm of memeory
>items. However, the reaction of Polichak and others has me
>going to greater lengths to clarify in non-technical works
>that using the word "meme" does not imply a newly
>discovered class of entities.

And I wish you the best of luck with that! However, I
suspect that you'll get constantly asked to defend your
'way of partitioning the realm of memory' and probably in
psychological terms. We'll see I guess...

Nick Rose
"University of the West of England"

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)