Subject: Re: Measuring Memes
Date: Sun, 13 Jun 99 18:48:22 -0000
From: Mark Mills <mmills@fastlane.net>
To: "Memetics List" <memetics@mmu.ac.uk>, <williamc@roman.net>
Bill,
>I must have some notion of the definition or lack of
>definition of memes before I can begin an analysis. Thus far it appears
>there are a number of definitions.
While unable to find any one definition 'true,' I have a system for
categorizing memetic definitions. Perhaps you will find it interesting.
I start with the assertion that 'memetics' must be defined in
relationship to genetics. This doesn't seem far fetched. 'Meme' is a
word play on 'gene.' Dawkins promoted the term in a book on genetics.
Established magazines seeking memetics book reviews turn to PhD's in
evolutionary theory (genetics).
As far as I know, all published memeticists differentiate between 'memes'
and 'genes.' In brief, 'memes' provide replicator function to
evolutionary phenomena devoid of 'genes.' Memes are 'not-genes'. Thus,
memetics is something of an 'anti-theory', it makes sense only in terms
of a reaction to established theory.
I could go further, but you should already have a feel for this. I take
your comments on correlation to imply an understanding of genetic theory
and thus an implicit agreement that memetics is somehow related to
genetics.
Calling memetics an 'anti-theory' forces me to take any system of memetic
understanding back to the source term, 'gene.' The term was first used
in its modern form by Mendel. Mendel grew some many plants and recorded
genealogy through generations and discovered discrete mathematical
correlations. Mendel suggested the source of these correlations was a
logical construct called 'genes.'
At this point, he made no assertion regarding the nature of a gene. The
linkage of 'gene' and DNA only occurred 60 years later when
microbiologists began to prove some Mendelian trait correlations could be
mapped to DNA.
At this point, we have the bifurcation of genetic theory which produces
the 'not-gene' theory we call memetics. The issue regards the nature of
a gene. Is a gene a 'process' or a 'thing?'
To make a long and fascinating story short, the scientific establishment
decided a gene was substrate dependent. To be more precise, the
establishment agreed that genes were based on a DNA substrate. As it
currently stands, most scientists define a gene as a DNA
open-reading-frame (ORF) or ORF assembly. Open-Reading-Frames can be
further defined in terms of specific codon assemblies, but there is no
need for such detail here.
The substrate model is not universally accepted. A variety of scientists
and philosophers find the ORF definition unsatisfactory. I like to
point to human language as an example of this. It is clear that humans
inherit a capacity for language. No other animal inherits the same
capability. Language capability is clearly an evolutionary product, but
no DNA linkage has yet to be found. Some can say the mechanical link
between ORF and language will be found, but others doubt this
eventuality. This discomfort is the source of memetics, a reaction to
slavish ORF genetic dogma. Memetics does what ORF genetics cannot do,
provide an evolutionary path to language and culture.
As a reactionary concept, memetics has no central thesis, only a variety
of differences between itself and genetics. The differences generally
reflect the interests and assumptions of their author.
To make sense of the seemingly infinite variations on memetic
definitions, I must return to the term 'gene.' There are two accepted
definitions of 'gene' among the scientific community: the original
Mendelian gene (substrate free) and the ORF definition (substrate
defined). The Mendelian definition leaves open the possibility that a
gene is process without specific form. The ORF does not leave this
flexibility to the researcher. A gene is defined by specific DNA codon
sequences.
For lack of better terms, I'll say the Mendelian gene is a 'substrate
free' definition and the ORF is a 'substrate dependent definition.'
These two terms will become central to our confusion about memes. (yes,
I will get back to memes.)
Microbiologists generally ignore the older, substrate-free definition.
The press and educational establishment have come to use the ORF
definition with dogmatic regularity. It seems that only philosophers and
students of linguistic/cultural evolution are troubled by the problems a
substrate dependent definition of 'gene' produces.
Based on my 'substrate-dependent versus substrate-free' categorization of
genetic theory, I assert one can make sense of the variety of memetic
theories using the following classification scheme. All memeticists use
one of the following four assumptions. Some are inconsistent and jump
from assumption to assumption, but the best articles are pretty
consistent about their fundamental definitions:
a) genes are substrate dependent, memes are substrate free
b) genes are substrate dependent, memes are substrate dependent
c) genes are substrate free, memes are substrate free:
d) genes are substrate free, memes are substrate dependent
In terms of the lingo here, Dawkins A is substrate dependent (memes live
in the brain), Dawkins B is substrate free (memes are words, songs,
coordinated human activities,...).
In terms of my above logic, a Mendelian gene is substrate free, an ORF
gene is substrate dependent.
One can take a popularity poll for each of the above positions. It
wouldn't prove anything. Few are willing to be limited by such simplistic
categorization, and I've received several complaints about my
reductionism. I feel the opportunity to understand memetic developments
worth the troubles produced.
Depending on one's chosen premise, each of the above 4 classes can be
defended strongly. Class a) is where most people start. Most come from
the dogmatic assumption that genes are ORFs (or the less specific 'chunk
of DNA') and memes 'ideas' (by implication, substrate free). After more
work in memetics, some find they dislike the lack of parallel between
gene and meme. These folks end up suggesting that an ORF definition of
gene requires a substrate dependent definition of meme. A few come to
memetics with a philosophical bent and end up with the notion that both
genes and memes are substrate free. I've never encountered anyone
suggesting a gene is substrate free while claiming a meme is substrate
dependent. There seems to be no cognitive development path to this
perspective, but I'm sure it exists somewhere.
I don't know if the above helps any. For me, it clarifies the term
'meme' in a variety of settings. For example, it allows me to talk with
geneticists about memetics. All the microbiologists I've met use the ORF
definition for gene. I point out the opportunity for neural tissue to
provide an analogous substrate to DNA. Upon this substrate could be a
code and units of this code would be memes. I always get a positive
reception to the idea of substrate dependent 'meme.' It doesn't threaten
anything they are doing, nor does it undercut their assumptions about
genetics. They may think I'm wasting my time, but there are no logical
problems.
This logic doesn't work as well for my conversations with those in the
social sciences. Frankly, these folks are uncertain about their
definition of 'gene.' They pay lip service to the rigors of an ORF
definition of 'gene,' but privately rely on the substrate-free Mendelian
gene. I basically agree with them, since I see no way to reconcile
linguistic/cultural evolution with ORF mechanics. Only the Mendelian
gene offers a plausible evolutionary logic to help understand the
evolution of human linguistic abilities. In brief, too many higher level
organizational structures above the ORF are required (neural development
is recursive, requiring higher order developmental bifurcations than ORFs
can provide. I doubt an ORF can define an neural network.).
The microbiologist senses the social scientist's lip service to ORF
orthodoxy and laughs at their lack of 'scientific rigor.' The social
scientist laughs at the reductionism of the microbiologist.
Regardless one's views about genes, memes and substrate dependence, I
think it useful to make meme and gene logically consistent. A meme is a
'not-gene,' and the only way to avoid overlap is logical consistency.
Many memeticists will disagree with me and push a disconnection of
memetic and genetic logic. I wish them well, but doubt any useful work
will emerge from this foundation. IMHO, a meme is a 'not-gene' and makes
no sense without a firm connection to genetics.
IMHO, the march of science seeks a unified evolutionary theory, one that
renders all evolutionary phenomena understandable. Our popular substrate
dependent gene fails to accomplish this.
One can pick their battlefields and conserve energy. I suggest memetics
focus on the huge body of data regarding inherited linguistic ability.
It is waiting to be explained via evolutionary logic. I'm sure there are
a million other places to start. The main thing is to push the genetic
model (both gene and 'not-gene' meme). There are a host of cultural and
cognitive phenomena with plausible inherited dependencies which ORF logic
cannot address.
I liken the situation to that of Ptolemy and Copernicus. Ptolemy did a
great job on 90% of known observations. 90% is not good enough. The
ideas espoused by Copernicus had selective advantages and eventually
displaced Ptolemy.
Assuming I've made an adequate case against inconsistent meme/gene
definitions, I can turn to the problems facing anyone adopting the 2
logical options remaining: 1) both gene and meme are substrate free, 2)
both gene and meme are substrate dependent.
the substrate dependent model. After all, microbiologists are in love
with the ORF (substrate dependent) definition of 'gene.' Microbiologist
seem to 'own' the term.
I favor adoption of Mendelian genetic logic (substrate free). Instead
of giving lip service to the ORF crowd, memeticists can adopt the
philosophic work defending substrate free genetic logic. ORFs still hod
a place, but get displaced from their position as 'gene.' Instead, they
become 'gene' support systems.
I foresee two problems with this. First, it would require a
confrontation over the definition of gene (I'll skip the controversy over
meme). Few in the social sciences gaining utility from a substrate-free
definition have the background to promote it against hordes of PhD's
based on ORF orthodoxy.
The second problem is more subtle, why differentiate meme from gene if
substrate dependencies vanish?
I'll spare you from my speculations on these last questions.
So, there it is. A logical framework for all memetic definitions and a
projection of possible developmental paths for the memetic field. My
hope is to clarify directions memeticists can successful push the
science. My central assertion is this: memetics is an unconscious
reaction to inadequacies of the popular ORF definition of 'gene.' These
inadequacies are most evident in the evolution of human linguistic
capacity and cultural behavior/artifacts emerging coincident with
linguistic ability.
>So far it is unclear to me whether this
>is because the discipline can not get beyond mere unilatereral declarations
>or if the crucial issues have not yet crystalized or if there really is no
>such thing as memes. Perhaps it is simply a specialization in search of a
>definition? This can be a formula for hollow expertise.
I hope the above address your concerns.
Mark
===============================================================
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