Re: On Gatherer's behaviourist stance

Aaron Lynch (
Sat, 05 Sep 1998 12:38:42 -0500

Message-Id: <>
Date: Sat, 05 Sep 1998 12:38:42 -0500
From: Aaron Lynch <>
Subject: Re: On Gatherer's behaviourist stance
In-Reply-To: <000801bdd8b5$74e09f40$da4e95c1@pc>

At 11:09 AM 9/5/98 +0100, Paul Marsden wrote:
>A Brief Reply to Derek Gatherer's Behaviourist Stance.
>Ok Derek, I hope to have the chance to answer more fully in a commentary on
>your stimulating and provocative paper but I'd like to make a couple of
>points partly because both Richard and Aaron (in an aside on software
>authorship) have weighed in on this one, but mostly because I think you have
>touched on THE critical issue currently facing memetics - how to
>operationalise the paradigm - i.e. stop talking about it, and start doing
>First of all, it seems to me that you are objecting to only one part of the
>thought contagion metaphor - the THOUGHT part. The contagion metaphor
>remains central to your (and most other memeticists, I believe)
>conceptualisation of the paradigm. Let's not throw out the baby with the
>bath water. The contagion metaphor is good.
>Your key point is to restrict memetics to behaviour - which renders memetics
>SYNONYMOUS with behavioural contagion research is social psychology. (This
>is not necessarily a bad move, there's a wealth of empirical data to exploit
>out there - but a critical one.) This is implicit in your paper but not
>underlined as THE "take home message".
>Your reason for restricting memetics to behaviour appears to be due to a
>healthy scepticism of homuncular cognitive psychology which attempts to
>explain behaviour in terms of putative internal states. Whilst
>deconstructing REAL homuncular intentionality is entirely laudable (I'd hope
>nobody on this list believes in real irreducible individual intentionality),
>it doesn't stop us from using intentionality as a heuristic device, or a
>Dennett puts it taking the Intentional stance (because the evolutionary loop
>of the Darwinian algorithm will give behaviour an "as if" intentionality.)
>The best way for me to accurately predict your behaviour, (or that of a
>chess computer) is to posit intentions, either by asking you what they are,
>or by simply assuming the chess computer "wants" to win - but of course
>there is no such thing as REAL, OBSERVABLE INTENTION THINGS in your brain -
>it is just a STANCE that helps me understand, explain and predict your
>behaviour. Now my point is this: Rather than adopting the intentional
>stance for individuals (X did Y because X wanted to do Y), memetics adopts
>the intentional stance for the intention itself (X did Y because Y wanted X
>to do Y). Natural selection operates on these heuristic devices "as if"
>they want to spread. Now, granted this is taking a representational theory
>of mind to a second order - but we should not confuse *real* intentionality,
>with as if *intentionality*. Representational theory of mind is a heuristic
>device for explaining behaviour and nothing more, your intentionality is my
>construct, not the other way around.
>The proof of the memetic pudding will be in its eating, if memetics can
>explain and predict certain aspects of, and yes you are right, HUMAN
>BEHAVIOUR better then homuncular psychology. The whole point of positing
>internal states in psychology is to predict and explain behaviour - nothing
>more. But these internal states are heuristics and NOT REAL in any naive
>sense. We just posit intentions with intention. To reiterate - for me, the
>beauty of the memetic stance - as I call it, is to explain human behaviour
>not be positing intentionality (historically and forecast) to individuals
>but to intentions themselves. Of course there is no such thing as a thought
>contagion, but if it helps me predict your behaviour then all well and good.
>Thus a meme is not a REAL object itself, it is a heuristic device, that we
>use to explain behaviour. In other words it is a functional object, not a
>real object.
>One way to keep every body happy in this debate would be to use not the
>behavioural contagion metaphor, nor the thought contagion metaphor, but the
>SOCIAL CONTAGION metaphor - which implies neither bare-foot behaviourism or
>universally rejected 19th century introspectionism. The SOCIAL CONTAGION
>metaphor would include the spread of both internal states (beliefs,
>thoughts, attitudes and intentions, and, critically, their resultant
>behaviour between individuals i.e. it is a social process. I know this
>seems a wishy washy middle point between you, and that happy couple Aaron
>and Richard - but I think by looking at memetics as an explanatory framework
>with an innovative heuristic device might prove to more useful, rather than
>becoming to pre-occupied over the unanswerable question over the ontological
>status of a meme.
>I know I have not addressed all the important points you raised - but I
>wanted to add my thoughts for what I think is THE central debate in memetics
>at the moment.
>Paul Marsden
>Graduate Research Centre in the Social Sciences
>University of Sussex
>tel/fax (44) (0) 117 974 1279
>Journal of Memetics - Evolutionary Models of Information Transmission:


Thank you for this thoughtful analysis.

I would like to add that many other theoretical constructs in science could
be labeled as "heuristics" and "stances" as well. Quarks, for instance,
have not only never been observed directly, but the "standard model" of
quantum chromodynamics (quark and gluon theory) implies that free quarks
don't exist. One could thus argue that quarks are "heuristics" and quantum
chromodynamics is a "stance." It is, of course, quite good at explaining
and predicting vast classes of observations. But if there were a
"behavioral school" of particle physics, we might hear members saying that
there are no quarks, we should not study quarks, and we should only study
particle behavior. The dispute might be rather calm, though, as there is
less emotion connected to subatomic particles than there is to human social

Physicists use indirect methods of detecting quarks, electrons, black
holes, etc. Likewise, universities use indirect methods, such as "exams" to
detect the "knowledge" of students, including, for instance, the death date
of Napoleon. They are specifically interested in drawing conclusions about
lasting internal beliefs and ideas, not about mere behavior on a particular
day. If you can show that your "poor" behavior in a particular exam was due
to epilepsy, you might get a second chance exam to demonstrate "knowledge"
as distinct from mere behavior. As imperfect as their methods may be, the
instructors become quite upset if they catch people deliberately
circumventing the correlation between test results and enduring
"knowledge," as by bringing in an answer sheet that says "Question 1: A,
Question 2: D, etc." They are interested in making inferences about when
you *thought* Napoleon died, and only resort to "exams" as an instument to
assist in making such inferences.

Pollsters and cognitive psychologists also have their imperfect indirect
methods of assessing "knowledge," "belief," etc. Indirect methods are used
to detect "computer viruses" and other "software" on computers, too--though
the brain and the PC have drastically different workings.

--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)