Gil
---A final postscript in my response to Paul C. concerning the notion that exchange "expresses something equal": an additional indication that the equivalence relation established by exchange does not establish a relation of equality in the sense required by Marx is the point, carried forward from my exchange with Fred, that absent special assumptions about market conditions, the relationship is not transitive. That is, in general the fact that A exchanges for B and B exchanges for C does not imply that A is or can be exchanged for C. But transitivity is a necessary condition for equality.
Paul ---- If you assume that all of the exchanges are barters, then of course you would be practically right - there would be a discrepancy in the exchange ratio A/C versus A/B * B/C.
But. 1. Even assuming barter, how big would the discrepancy be? Would it be above the level of random noise that occurs in individual exchanges?
2. If it is not above this level, then it is irrelevant since one only determines the laws governing a phenomenon by dealing with ensembles large enough to eliminate such noise. This is standard experimental practice.
3. If we drop the assumption of barter, then A/C takes the form A/m*m/C and A/B * B/C takes the form A/m*m/B * B/m*m/C which are equivalent unless we are to assume that m/B * B/m is not unity where m is the money price of B, which would be illogical.