First, a disclaimer--I'm not a historian of Economic Thought. Having owned  
   up, I'd like to suggest that the very interesting contributions to "why  
   teach..." ought to consider another aspect of this issue. I am a member of  
   an economics department that at one time had a history of thought course  
   required for Ph.D. students, but no longer has that requirement. There are  
   some particular "local" reasons for that (the retirement of a senior faculty  
   person,etc etc), but there is a broader issue associated with the demise of  
   the requirement. That broader issue strikes me as extremely relevant to "Why  
   Harvard  [or  wherever] does not teach History of Thought," and is not  
   well-captured by the specific wording of Womack's original query, or by many  
   of the responses. So this is a plea for consideration of this issue.  
  
   So here it is. The way Womack posed the grumpy Harvard claim "Why teach  
   error?"  suggests  that an answer be of the form "because we learn the  
   following useful things by studying error."(I'm ignoring Quinn's well-taken  
   point that HET is not just a study of error...)But that style of answer is  
   necessary to but insufficent for making the case for requiring a history of  
   thought course. Given the way economists view the world, making a case for a  
   required HET course for Ph.D. students requires NOT JUST that the course  
   have a positive yield ("teaches us useful things"), but that the yield is  
   higher than the most attractive foregone alternative use of course time.  My  
   memory  (which  is  sometimes flimsy) suggests that the loss of an HET  
   requirement at my university was based in large part on the argument that,  
   while HET was valuable, an additional semester requirement in Econometrics  
   had a distinctly higher yield for our current students.  
  
   By the way, whatever the true unstylized facts are about what med schools do  
   or do not require, it surely must be the case that a telling argument about  
   whether or not history of medicine should be required would depend on the  
   yield from studying that versus the yield from the course being "displaced."  
   "Just" showing that there is a positive yield (not counting opportunity  
   cost) of studying history of medicine is simply insufficient to making a  
   really strong case for teaching it. Even humanities-oriented medical types,  
   for example, might want to ponder the relative payoffs from "forcing" a  
   history of medicine requirement to that of "forcing" a medical ethics course  
   requirement....  
  
   Robert Goldfarb