How Models Fail
A Critical Look at the History of Computer Simulations of the Evolution of Cooperation
4.4 “There are no alternatives to modeling”
Argument: There is no real alternative to modeling, anyway.
If you try to do without models, merely relying on verbal explanations,
you are just making use of implicit models that are never fully articulated.
Surely, explicit modeling is better than relying on implicit models.
Without models nothing could be explained. (See also Epstein (2008),
who employs a variant of this argument.)
Response: It is at least for the time being (the distant future
of science may of course prove me wrong) practically impossible to
express everything that can be expressed verbally in mathematical terms
or with formal logic. This includes many of the causal connections
that we are interested in when doing social sciences. Otherwise, how
come that among the many books published about the causes, course and
consequences of the First World War these days, there is no game theoretical
or otherwise model-based study that could rival the conventional historical
treatments? Otherwise, how come that lawyers, attorneys and judges
- their job being to a large part one of logical reasoning, as one
should think - do not use formal logic to express the legal connections
they ponder over?