A theory of gravity is supposed to tell a story about how   
physical objects are attracted to each other, by some agreed   
measure of attraction, and to explain and predict observable   
phenomena based on this story.  So a theory of choice,   
I take it, should tell a story about how choices are made,   
by some agree measure of choice, and explain and predict   
observable phenomena based on this story.  The agreed   
measure of choice for most economists appears to be observed   
behavior.   
  
Gravitational physics is a field where theories (pl.) of   
gravity are pitted against each other to determine their   
relative usefulness.  We would not say "gravitataional   
physics is the theory of gravity" for a few reasons.  I'll   
mention two.  First, it is a field of study, not a theory.   
Second, theories (plural) are studied in this field of   
study, not "the" theory.   
  
A given theory distinguishes itself from other theories by   
ruling in and out various factual claims.  A theory of   
gravity that failed to rule out some imaginable, observable   
(though possibly in a frequentist sense) behaviors of   
physical bodies would be rightfully dismissed.  It is not   
a scientific theory.  A theory of choice should similarly be   
dismissed if it cannot rule out some imaginable, observable   
human behaviors.   
  
I understand Pat Gunning to insist on two claims.  The   
first,  which he seems inclined to focus on, is a very weak   
claim: that we often find it useful to interpret human   
behavior as purposive.  Nobody has disputed this, I think.   
As I said earlier, this kind of folk psychology is pervasive   
and does not distinguish economists from others except, as   
Pat suggests, many economists are inclined to spend a lot of   
time thinking about what might follow from this.   
  
Of course we also find it useful to interpret animal   
behavior as purposive, and even evolutionary biologists have   
found teleological thinking to be provisionally useful.  if   
we find such teleological thinking useful, that is   
justification enough for using it.  Provisionally.   
  
But as our profession has proved (and as the recent HES list   
discussion highlights), NO behavior is ruled out by the   
brute admonition to interpret behavior as purposive.  For   
example, in violation of the Stigler and Becker (1977)   
admonition to economists, James will "explain" suicide   
bombers by giving them special utility functions (real ones,   
apparently!) to maximize.  Thus even folk psychology, which   
Pat proposed as the root of economics, gains usefulness only   
when supplemented substantially with concrete content.   
A "theory of choice" cannot simply be the assertion that   
observed behavior is purposive.  This brute assertion has no   
empirical content and by itself does not even suggest any   
models of behavior.  That is why I and others on this list   
have insisted that even that part of economics that focuses   
on purposive behavior---let's call it "choice-theoretic"   
economics---gets its content elsewhere.   
  
The second, strong claim Pat (and others) have made is the   
economics *is* the theory of choice.  I previously raised   
questions about what kind of claim this is, and others on   
the list have done so as well.  I suppose it is intended as   
a demarcation criterion for the profession---a kind of "do   
things my way or you're not an economist." I am not really   
interested in demarcation criteria, but this one has serious   
difficulties, as it is evident that many professional   
economists are engaged in activities not meeting this   
criterion for belonging to the profession.  Perhaps an even   
more powerful objection to any proposal that "economics is   
the theory of choice" is that *identifying* a field of study   
with a single theory is a suspect move at many levels.   
  
Cheers,   
  
Alan Isaac