A theory embodies our best knowledge regarding the nature and workings
of a particular class of phenomena. A theory-building tool would serve a
theory, but not contain it. If I'm looking for knowledge about a given
topic, I'd rather study the theory than the tools used to build it.
The distinction also is important because it would affect the way we
treat simulations. If they are theories, then ultimately they must
conform with evidence. If they are only tools, then they need not.
As I wrote previously, I don't subscribe to the idea that, if the output
of a rigorous simulation conforms to hypotheses obtained by interpreting
a discursive theory ("theory" for short), then the simulation can be
said to test and corroborate the theory. The simulation may *feel* like
it is a tool contributing to building the theory, but unless there is
complete logical consistency between simulation and theory, any
"building" that one does to the discursive theory is arbitrary. Instead,
assuming that one has good definitions for the terms of the simulation,
I think that it makes more sense to think of the simulation as a more
rigorous version of the theory, perhaps inspired by the discursive
version, but nonetheless a theory in its own right. The programming
language and associated utilities are the "tools."
> -----Original Message-----
> From: News and discussion about computer simulation in the
> social sciences [mailto:[log in to unmask]] On Behalf Of Scott Moss
> Sent: Friday, November 21, 2003 6:42 AM
> To: [log in to unmask]
> Subject: Re: Theory and Simulation
> If I understand the issues between Kathleen and Rosaria, the
> main question is whether the simulation model as implemented
> is a theory or whether the implementation is a tool of theory
> building. I think I also understand how this difference
> reflects their respective research agenda. What I haven't
> understood is why it matters whether we say that the
> implementation is a theory or a theory-building tool.
> I would be grateful for enlightenment on this.
> Kathleen Carley wrote:
> > Rosaria Conte wrote:
> >> The alternative point of view has been put forward out
> of the AI
> >> and DAI tradition.
> >> The argument there is that the simulation is the theory. If one
> >> believes the Turing test, or the social analog, The
> Social Turing
> >> test then the fundamental mechanisms will be discovered through
> >> the development of a valid computational model.
> >> right: as I said, computer modelling is a fundamental tool
> for theory
> >> building. But still, the theory is not (what is generated by) the
> >> program! This is quite clear to AI-people, often involved
> in logic as
> >> well as cmputational modelling. The agent systems, which is one of
> >> the most promising AI subfields, is based upon the interplay among
> >> three different components, the theory, the computational
> >> architecture of the system, and formal language to describe it.
> > The issue here is not "what is generate by the program" but "the
> > program itself" Again, for many AI people and logicians, current
> > logical formalisms are incapable of representing core ideas in some
> > agent models - such as knowing not requiring infinite regress. For
> > these models - new "logics" are needed. Until then, the
> program itself
> > is the logic and is the theory. The results generated by the
> > simulation are the hypotheses or predictions.
> >> In addition, no verbal theory completely specifies the
> >> ley alone the dynamic unfolding process, particularly
> for complex
> >> adaptive systems. The simulation is needed to uncover and
> >> describe these.
> >> since it generates them, it is needed to uncover or
> observe them. But
> >> it cannot iteslf /describe/ them explicitly. The simulation allows
> >> you to observe the process that ought then to be explicitly
> >> described.
> > Here I would say that the simulation results are used to uncover or
> > observe the processes - but the simulation model is a formal
> > description of the process. The language in the model is just a
> > distinct symbol system for explicit;y formulatingof the theory.
> > I would suggest that this is a philosphy of science question
> > involving, among other issues, as to what are the features
> that make a
> > symbol system adeqaute for formulating theory.