Kevin wrote:  
  
>I agree with Matt Forstater that human monetary systems typically  
involve   
>infinite regress valuation.    
  
I am apparently not expressing myself very clearly.  First, James  
thought I was saying "fiat currency" is "undesirable," and now Kevin  
thinks I was saying that I support the infinite regress argument.   
  
I meant to characterize Kevin's view as the "typical infinite regress  
argument" and to indicate that the tax liability payable in government  
currency is the way out of the circular argument. As Abba Lerner put it,  
in Encyclopaedia Britannica:  
  
Any particular seller will accept as money what he can use for buying  
things himself or for settling his own obligations.  This seems to say  
that a means of payment will be generally acceptable if it is already  
generally acceptable, and it looks like a circular argument.  But it  
only means that general acceptability is not easily established.  
General acceptability may come about gradually.  If a growing number of  
people are willing to accept payment in a particular form, this makes  
others willing to accept that kind of payment.  General acceptability  
may be established rapidly if very important sellers or creditors are  
willing to accept payment in a particular form of money.  For example if  
the government announces its readiness to accept a certain means of  
payment in settlement of taxes, taxpayers will be willing to accept this  
means of payment because they can use it to pay taxes.  Everyone else  
will then be willing to accept it because they can use it to buy things  
from the taxpayers, or to pay debts to them, or to make payments to  
others who have to make payments to the taxpayers, and so on.  (Lerner,  
1946, p. 693)  
  
Beyond that, I leave it to Eric Tymoigne's excellent post recently  
forwarded by John Henry to sort out the rest.  
  
Best,  
  
Mathew Forstater