On the issue of reading foreign languages, scholarship, and    
opportunity cost (leaving manners aside).  
  
While I would agree that one ought to consider one's own opportunity    
costs in deciding whether to read a text in its original language    
rather than in translation, I would point out that costs, in terms of    
scholarship, are too easily underestimated.  
  
First, because some authors are very difficult, if not impossible, to    
translate: Marx's dialectical terminology and language, for instance,    
is absolutely not caught in the English version (I cite some examples    
in footnotes to an article on Tugan Baranowsky's marxism, RPE April    
2006), so that studying Marx in English often gives a misleading    
impression of his ideas. Conversely (I am told), the German    
translation of Keynes is pretty awful; or, again, the Italian    
translation of Sismondi at places makes him tell the opposite of what    
he wrote in French. Generally speaking, if one wants to carefully    
study an author, one ought to be able to read that author's language    
and to understand the subtleties hidden behind the expressions that    
author uses (as pointed out by Jesse Vorst); moreover, not everything    
has been translated. This may sound obvious, but Marx's example    
indicates that it isn't that obvious.  
  
Second, because ignoring a language may lead to ignore what has not    
been translated from that language, and this could mean entire chunks    
of history. For instance, if one only looks at the English language    
literature, one could (almost) claim that after the gluts debate and    
until 1936, Say's law was generally accepted in the academia, while    
in reality it was widely discussed and rejected, or at least    
explicitly set aside, in the non-English literature, out of which    
grew some of the most interesting business cycle theories in the two    
decades before WW1 (in the German-speaking area) or where the    
premises of economic dynamics were laid (Italy).  
  
This, of course, goes on top of Anthony Waterman's correct    
observation on the dominance of French in the XVIII century, to which    
one also ought to add that the dominant scientific language at the    
end of the XIX century was German --in economics to a lesser extent    
than in natural sciences, but some of the leading journals were    
printed in German. And academic journals specialised in economics    
were printed in French, German and Italian before the QJE, EJ and JPE    
were funded (see http://cepa.newschool.edu/het/essays/journal.htm).  
  
Daniele Besomi