I wrote: > > But if the question is, How much SNLT is embodied in, > > say, a Ford Taurus, then the clock-time measure is quite reasonable: > > the car embodies a wide variety of different sorts of labour, and, in > > the absence of information to the contrary, it's reasonable to suppose > > that divergences between actual hours spent and SNLT will roughly > > cancel out. And Rieu replied: > Fist of all, we need to measure the divergence between actual > hours spent and SNLT. How can you know that divergences will > roughly cancel out without measuring the divergences itself? It > looks not so much an explanation as a presupposition to me. One can actually _measure_ the divergences only under restrictive conditions, but that's not a major problem. On a reasonable definition of SNLT, the average divergence of actual hours from SNLT across the entire economy has to be zero. But then the labours going into a particular product can be considered a (large) sample from a population with a mean (divergence) of zero. If the sample can be considered random, from a statistical point of view, then there must be a strong expectation that the divergences roughly cancel. My contention is that the many and varied labours that go into producing any given product can reasonably be thought of as a random sample of social labours, absent any specific reason to believe otherwise. Allin Cottrell.
This archive was generated by hypermail 2b30 : Mon Apr 02 2001 - 09:57:30 EDT