Mike and David,
I very much share you views on the yes/no issue of
Sugarscape but what is even more worrying to me is that you encounter this
simulation problem also in 'real' data. I am currently conducting a
correspondence analysis using the UK General Household Survey leisure/sport
(cultural) data and the only thing that has ever been asked ever since 1973 is
'have you done activity x,y,z in the last 4 weeks' - answer: yes,no.
Ironically, Sport England has refused funding to
develop the leisure/sport questions any further on the basis that there has been
no significant change over the last 25 years and hence, national statistics were
perceived too boring.
I am not simulation practitioner but if anyone
has done any simulations that address the issue of 'minor changes' using
'real' data I would like to hear from you.
Thanks,
Iris
----- Original Message -----
Sent: Monday, March 31, 2003 6:03
PM
Subject: Bit-flipping model of
culture
At 02:40 PM 3/30/2003 -0600, you wrote:
> ...[Missing out
on the
> funding is karma coming after me... ironically, I have been
> somewhat critical of Sugarscape myself, specifically the
>
'bit-flipping' model of culture...]
I'd be interested in criticisms of
the 'bit-flipping' model of culture.
Do you have any papers on this, or
care to discuss it here? At a
certain level of abstraction it seems
to me to work pretty well -- what
are the problems you see with
it?
Thanks,
Mike Sellers
Online Alchemy
Mike,
I
will get out of my depth very quickly here, so I'll immediately say that I
think you're right: the model is OK at a certain level of abstraction. I
just think that the level of abstraction happens to be very, very high.
Representing a culture as a collection of discrete yes/no views that
may be individually switched on or off, that are unrelated to one another and
also are more or less unrelated to behavior just seems way
over-simplified. My unease is also due in part to being unconvinced by
the concept of the meme, which after all is undefined in detail, and also
non-empirical in any meaningful sense.
However, as I say, I'm out of my
depth here... I'd be happy to learn more from an informed discussion of
cultural bit-flipping and memes among others on the list!
My criticism
in print is along the lines of that above -- just that it's a rather
impoverished representation of culture, and is made (I'm afraid) rather in
passing. It appeared in this paper
O'Sullivan D and Haklay M 2000
Agent-based models and individualism: is the world agent-based? Environment
and Planning A 32 1409-25
I'm also a little uncomfortable
with the way that 'tribes' are made to emerge in Sugarscape using this
cultural model -- since it presumes that tribes don't pre-exist
individuals. Of course they don't pre-exist individuals in the abstract,
but they do pre-exist specific individuals. This, of course gets us into
rather murky 'emergence' debates about whether various collective phenomena
(language, religion, culture etc) are merely 'aggregate' things, or whether
they have meaningful existence in themselves. For (much) more on that,
see the CMOT paper I cited in a recent post, based on a discussion on this
list back in Summer 2000.
David
--------------------------------------
Penn State Geography
- GeoVISTA Center
University Park, PA 16802, USA