In reply to Barkley Rosser: Of course, much depends on the particular questions particular
models try to address, but for all I know, in general, conventional economic theory does
not ignore the constraints imposed on human agency, nor does it ignore how the constraints
are reproduced.  And this is not a "micro" statement.  It applies pretty generally to
conventional theory in its modern version.
   
My impression is that people on the heterodox side may be clinging a bit too much to the
old Cambridge Capital Controversy, perhaps hoping for a Sraffian/Post-Keynesian revival.
IMHO, that is extremely unlikely. Not so much because of interests vested in blocking such
development or because of perverse academic/ideological inertia, but mainly because of
valid analytical reasons.  As important as the issues raised in the CCC were in the
Marshallian pre-history, they've been made largely irrelevant by the neo-Walrasian shift
in modern theory.  Nonconvexities (e.g., those leading to re-switching and ill-behaved
measures of aggregate capital), like Knightian uncertainty and n other features of the
real world, ought to humble economic theoreticians, but there's no reason to draw from
them nihilistic conclusions.
   
We may not like the form of Frank Hahn's 1982 piece in the CJE, but dynamic general
equilibrium has effectively integrated the old issues of growth and distribution.  And it
has done so in a fairly coherent framework that bypasses the capital critique altogether.
That said, as far as I know, competent economists do not view general equilibrium, even
its more elaborated versions, as a description of the world as is, but simply as a
departing limit case, a platform from which to launch theoretical expeditions under
alternative (and more interest/immediately relevant) assumptions.  In fact, in the last 25
years or so, much of the theoretical interest has shifted *from* merely reinforcing the
Arrow-Debreu edifice via conceptual re-definitions and extensions *to* sorting out the
implications of alternative assumptions.  And the latter -- by necessity -- highlight the
limits in the robustness of the Arrow-Debreu edifice.
   
John Womack wrote here recently that economists in Harvard deem the history of ideas
useless.  While we find everywhere people who mistake abstractions for real life, I
frankly doubt that many high-caliber theoretical economists share that view.
   
And while I'm at it, I'll say (in reply to Nicholas J. Theocarakis and Sumitra Shah) that
we should not believe that, as a general rule, theoretical economists of the conventional
type are less sophisticated in their appreciation of the complexity of social life than,
say, historians or HES members.  Nicholas (and Sumitra?) appear to view mathematical
formalization as part of the problem, if not *the* problem.  Nicholas says economists
"trivialize" the study of constrained human agency when they use math models.  He views it
as "forcing" real life into abstract mathematical models.  Obviously I disagree strongly
with this view of the use of math in economic theory.  To my mind, all math does here is
provide a kind of transparent, public-domain logical discipline to the analysis of "IF X,
THEN Y" statements.  People who distrust the results of economic theory are thus in a
position to pinpoint by themselves the flaws in premises and/or derivation.
   
Back to Barkley, the conventional approach to growth and distribution is not necessarily
tied to representative-agent type models.  Better-informed people should correct me if I'm
wrong, but my impression is that in macro policy applications, the representative agent
(but not general equilibrium) has been virtually abandoned in the last fifteen years
(again, much depends on the specific purpose of each specific model).  In macro, growth,
development, etc., there's an explosion of alternative approaches that owe much to the
standards of analytical rigor set by general equilibrium.  The Debreu-Mantel-Sonnenschein
theorem of the 1970s pulled the rug off the feet of the dogmatic, ideological approach to
micro-foundations (namely the New Classical approach) and we're just starting to notice
the effects...
   
Julio Huato