(Of course we are not using "complexity" in Rosen's sense here.)
1. I liked the paper. It represents a considerable advance in clarity
over this issue, as compared with many of the other contributions to
this debate.
2. There is a slight conflation between issues of what complexity _is_
and what _causes_ complexity. Not being perfectly ordered (constrained)
and not bieng perfectly disordered (distinct) may be necessary
conditions for complexity but they are not, in general, sufficient.
3. Dictionary defintions of word (especially the etomology) are not a
very reliable guide to useful modern academic usage. Although a very
good starting point.
4. The dependency of complexity on the framework it is being considered
within does not prevent objectivity, if the participants *agree* on a
uniform such framework/language/goal (this may be as a result of a
claimed privileged framework - natural kinds, etc.).
5. Having said that, the disussed connection between complexity and
distinction/connection is a very interesting and illuminating one. Let
me express some of this in my own terms (as I did in the concluding
session of the workshop).
The complexity of an organism from the point of view of a
preditor/competitor (and even cooperator) will be the difficulty in
predicting the organisms actions given its knowledge of that organism.
Thus being complex (in this sense) is one mechanism to achieve
evolutionary fitness (there are others, of course, like simple avoidance
or strength).
The complexity of an organism from its own point of view, is the
difficulty in predicting, modelling and controlling its _own_ state.
Part of this will be the ease with which it can come to decisions about
action (one would could only survive for any length of time as the
donkey that is exactly mid-way between two piles of hay and can not
decide which to go for in a protected environment). This will only be
one corrolory of its own complexity. Others include whether it can
predict the outcome of ites own actions on its own state, and whether it
can preform an intended action. The effect of these difficulties depends
on the survival and reproductive strategies that the organism has
evolved.
(It is interesting to note how this reflects the duality of
representation (internal/external) that Catherine Kennedy has
described.)
The complexity of this situation from our view point is the difficulty
in understanding and modelling the interaction of these (and other
evolutionary processed) given our description and understanding of them
to produce the biological complexity of the resulting organisms (i.e.
the difficulty in understanding the functional and structural workings
of an organism given a kownledge of its genetic makeup).
Thus we have at least *four* different kinds of complexity here. Francis
makes the interaction of the first two producing an increase in the
second two plausible, but there are many possible such interactions
(hence why a local increase in complexity is not inevitable as Francis
points out).
5. Francis's argument that just becuase there is no necessary intrinsic
local increase in (biological) complexity that this would not prevent a
general global increase (at least in the maximum biological complexity)
is quite correct. An interesting comparison is with the physical arrow
of time: most physical processes are locally reversible (Some sub-atomic
processes are only time-charge-spin reversible as I understand it), but
global processes (e.g. entropy are not.
---------------------------------------------------
Bruce Edmonds,
Centre for Policy Modelling,
Manchester Metropolitan University, Aytoun Bldg.,
Aytoun St., Manchester, M1 3GH. UK.
Tel: +44 161 247 6479 Fax: +44 161 247 6802
http://bruce.edmonds.name/bme_home.html