My aim in this discussion is to determine whether and how the distinctions and
issues raised by Keith, Alan, Paul et al. should inform or change social
simulation modelling.  The only direct claim I have understood in this regard is
Keith's

> sociology is the place where you
> go to get your model validated, in my view. Otherwise, you have to redo
> all of the sociology to validate it yourself. Why reinvent the wheel, when
> the sociological knowledge could help you get to your goal faster?

In the normal terminology, you get your model validated by comparing the behaviour
of agents and multi agent systems with the people, groups of people, organisations
or whatever your agents and multi agent systems are intended to represent.  Some
believe that we should have formal constructs to constrain the ways in which we
specify models provided that those constraints guide us to models that will be
validated against data.  At least, this is what I take the foundational end of the
ABSS community to see as the benefit a general social theory.  Whether there will
ever be such a general social theory remains an open research question.

I am not sure whether Keith in the above quote is saying that there already is
such a well validated general theory or that sociological observation is where we
should go to validate our models.

On the issue of methodological individualism vs social realism, I don't feel it an
insult to be told that what I do is the former.  I guess the reason I model
individuals -- attempting to use well validated specifications of cognition -- and
the consequences of their interactions for their own behaviour and freedom of
action is that I am interested in the emergence of behavioural norms and how
social policies can change such behaviour (e.g. in response to the effects of
climate change) or how organisational structures constrain or support effective
behaviour (e.g. in the management of critical incidents).  To represent a social
system as a multi agent system with interaction among agents seems to me to be a
sensible approach  to the analysis of that sort of problem.  Since this is
precisely the sort of problem that is widely addressed by the social simulation
community (at least in Europe), it may be that those of us working on such
problems have found that to be the most effective or natural approach.

Sometimes, social simulation modellers are interested in effects on individuals
arising within existing norms or organisational structures.  In such cases, it
seems plausible to me to take those social norms and structures as given and to
model behaviour within that normative or organisational context.  I thought I did
that in my North West Water model <http://jasss.soc.surrey.ac.uk/1/4/1.html> which
might then have been socially realist.

I just see these as pragmatic modelling decisions.  If these decisions are
usefully to be informed by sociological knowledge, I would be very keen to know
how -- preferably by demonstration or concrete example.



--
Professor Scott Moss
Director
Centre for Policy Modelling
Manchester Metropolitan University
Aytoun Building
Manchester M1 3GH
UNITED KINGDOM

telephone: +44 (0)161 247 3886
fax: +44 (0)161 247 6802

http://www.cpm.mmu.ac.uk/~scott




%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%