On Wed, 19 Nov 1997, Bill Schell wrote:
> Andy wrote that marxism is heuristic but that it is a testable, useful way
> to look at history. Later he wrote that Weber's models "ideal types" are
> heuristic and hence untestable. Which is it Andy? You can't have it both ways.
This is a misrepresentation of my position (but in part understandable
since my position has not been explicated in any great detail). Things are
never so simple, Bill. Allow me to clarify.
First, an ideal type is a type of heuristic, true; but not all heuristics
are ideal types. Marx's heuristic models are not generally ideal types.
For example, his most broad statements, such as that found in the
"Preface" to A Contribution to a Critique of Political Economy is a very
general heuristic model; but unlike Weber's bureaucracy, Marx's heuristic
is processual and analytical, not typological. Huge difference. Marx's
idealized capitalist system in Capital, buttressed by massive amounts of
empirical data, constitutes another type of heuristic, here somewhat
closer to Weber's notion of ideal types, perhaps, but still analytical and
theory generating. For another example, take Weber's class model, which
was an arbitrary static stratificatory model (based, as it was, in Weber's
liberal pluralist conception of political economics), which one may use to
describe class structure (I don't really find it useful even for this);
whereas Marx's class model is an analytical tool from which to theorize
about social class (which is why Marx's works present such different class
structures, because different times and different places have different
class structures due to the particular dynamic in force). From these
bases, Marx generates very specific theories (many of which carry
impressive predictive validity). Weber defines, conceptualizes, and
categorizes, but does not generally theorize.
So, to clarify--and this is essential--historical materialism is at one
level a heuristic model (actually several models), but it is a heuristic
model from which special theories are generated. These theories are in
turn testable. Weber developed his ideal types and basically left it
there. Beyond this there really were no theoretical statements in Weber.
Moreover, at the meta-theoretical level, a scientific research program,
such as described by Lakatos (a heuristic), is generally not directly
testable, rather it is a set of core theories that comprise a negative
heuristic. One must go beyond the heuristic and generate theory(ies),
either deductive conjectures tested against empirical data, or inductively
generated theoretical generalizations. Through a retroductive process the
heuristic is adapted and expanded, increasing its explanatory power (the
positive heuristic). Clearly significant parts of both Marx and Weber's
systems are expounded in the form of models. But only one model has
produced genuine theories. And, in my view, only model (between these two)
can.
So, again, it all wasn't as easy as Bill Schell would wish to make it. And
I didn't try to have it both ways. A heuristic is like a hammer--it is all
about picking it up and building something with it (rather than, not
exactly understanding what a hammer is used for, attempting to beat
somebody over the head with it).
Andy