Since Paul has brought up the issue of money, I'd like to make a couple
of points to try share some hard-won understanding.
Marx's theory of money is first of all a theory of the money function, as
store of value, medium of exchange, and means of payment. In different
societies different valuable things can function as money (cattle,
cigarettes, gold, silver, the debt of a well-regarded private agent, or
the debt of the state are examples). Marx usually analyzes the situation
where gold is money, and where, in particular, the debt of the state is
convertible into gold at a fixed rate of exchange (the mint price)
through a credible legal commitment. (For example, the first U.S.
Congress fixed the value of the dollar at 1/20th of an ounce of gold of a
particular fineness.) Marx's theory explains quite systematically how, in
such a system, the money price of commodities is ultimately regulated by
the production costs of gold. To make this story complete I think you
need to consider the fact that durable assets like gold are actually
valued through speculation on their future relative production costs in
the presence of uncertain technical change.
In the 20th century the convenience of using the debt of the State
as a circulating medium led to a situation where the State could issue
small denomination certificates of debt without paying interest on them
and the public would hold them as a medium of circulation. In many
countries the State outlawed private circulation of banknotes, or, as in
the U.S. taxed them out of existence. It is also clear that small
denomination non-interest bearing notes would not circulate if the State
also issued small denomination interest-bearing debt (as long as it had
the same advantages of anonymity and transferrability, like a "bearer
bond".) The value of the State currency rests on the same principles as
the value of the other components of the State debt, of which it forms a
part, as the accounting conventions of central banks show.
Thus it is not true, as the current neoclassical theory of money seems to
assume, that money is a "bubble", that is, an asset whose value is
maintained purely by speculation and its own scarcity. The State has
substantial assets, including its own future taxing power, to back its
debts, including currency. There is a question as to how the State debt,
denominated in abstract units of account such as the dollar or the pound
or the yen or the franc, is valued in terms of produced commodities. When
the State has pegged the dollar price of a commodity (such as gold, in a
convertible gold standard system) it isn't hard to see how this works; it
is much more puzzling when there is no such convertibility, as in most
countries at the present time.
The evidence of ownership of a claim on the state for $10, or L50, is of
secondary theoretical significance, as far as I can see. This evidence
may be an engraved banknote in one's possession, or an entry in a book at
a bank, or on the equivalent file in a bank's computer, or whatever. The
important economic point is that the monetary asset represents a claim on
the State. You could even think of state currency as a system of tax
prepayment certificates.
Fortunately a great part of Marx's theory of money is completely
independent of the exact valuable asset that forms the social equivalent,
and can be applied to modern systems of State debt money. For example,
the dollar still can represent a certain quantity of social labor time
(whatever resolution one accepts to the problems of defining this
relation), and thus the whole of Marx's analysis of the social relations
of capitalism remains perfectly coherent. The problems that are not so
easily resolved are those connected with the valuation of the state debt
in general, its role in facilitating production and consumption
transactions, and thus altering the turnover times of capital, and its
role in situations of crisis, where the State has the power (and
frequently uses it) to socialize private debts in order to avoid a
complete collapse of the credit system.