I think Gordon and Steedman are confused about the very point Althusser is making. The Althusser quote Angela posted simply says that Marx took the classical labor theory of value and brought it to life, made it useful for understanding capitalism, by adding the distinction between labor and labor power. That difference is a measure of capitalist exploitation, and it is the starting point for all of Marx's analysis of capital and capitalism. Gordon/Steedman do not seem to understand this distinction well enough to make use of it in their analysis you present.
Let me go comment on specific parts of the text.
> Whatever light Althusser here throws on Marx's critique of classical
> economics (and honestly I have trouble following him), Marx remains open
> here to one of Ian Steedman's major criticisms (the other being the
> problem of joint production), nicely summarized in his own way by David
> Gordon in Resurrecting Marx (transaction, 1990):
>
> "According to Marx's account, wages (and hence surplus value) depend on the
> commodities required to sustain the laborer. To simplify, imagine that
> laborers need only bread to survive. Then, according to Marx, wages depend
> on howmuch bread the workers require.
Actually, it is the value of variable capital (the value of the reproduction cost of productive labor (only, excluding unproductive labor)), not wages, that is determined by the value of the consumption basket (bread in the simplification) you describe. And, for clarity, it is the cost of social, not physical, subsistence that matters. Thus, some wages are paid out of surplus value--all of the wages of unproductive labor and those of productive labor that exceed the cost of their social subsistence. Likewise, some workers are paid wages that less than the cost of their social subsistence. With a given reproduction cost of constant capital, it is v that determines surplus value, and wages that determine profits. Two very different sets of numbers.
But so far, you may argue that these mistakes aren't very important; they don't seem to get at the problem of circularity that so troubles you and them.
> "An ambiguity lies concealed in the phrase 'how much.' Does this mean how
> much in value terms, or in physical terms? That is to say, is the question
> at issue what quantity of bread the worker needs--or is it, rather, what
> the value of this quantity is? Marx took the question in the second way.
> The value of wages, in his view, depends on the value of the bread the
> worker needs.
Yes, if you substite v for wages.
> "Steedman's objection is that once the physical quantity of bread at issue
> is given, this suffices to determine wages (and hence profits). The value
> of bread 'drops out' of the picture altogether: no recourse to it is
> necessary in order to explain wages.
>
> "Steedman proves this through the presentation of a simple model, which he
> later complicates. At first, for example, he assumes that only one
> technique of production is necessary; later, he allows choice of technique.
> Although Steedman's argument does not require very difficult mathematics, I
> do not think I have a sufficient grip on it to try to explain his reasoning
> in detail. Rather I shall attempt to present the basic point of his
> argument in my own way. "If wages depend on teh value of bread, then we
> obviously need to know that value. But bread, like all other commodities,
> is subject to the labor theory of vlaue, in Marx's view. To determine the
> value of bread, one needs to determine the value of the labor that produces
> bread. But the wages of bread produces, like all other wages, depend on the
> value on the value of labor power. This in turn depends on the value of
> bread, in the way earlier explained.
>
> "The circularity here is apparent. Before one can determine the value of
> labor power, one needs to know the value of bread, a value that in turn
> depends on the value of labor power. Unless one can break out of the
> circle, the labor theory's explanation of wages and profit fails.
There is no circularity in the way Gordon describes. The value of the bread depends on the value of the *labor* expended, not the *labor power*. The value of the bread, the total value to be converted at exchange, is determined by the value of the total labor, living and dead, c + v + s, not just v. The value of v matters only to questions of *distribution* of product (e.g., exploitation), and thus it is irrelvant to the question of the value of the bread. In quantitative terms, the value of a product equals the total labor expended to produce it. The value of labor power then determines the distribution of product between capital and labor. In exchange, then, capitalists sell the bread for a price and bargain with workers over how much they (the workers) will get as a wage. This bargain may affect the price, but not the value of the bread.
Btw, capital does have the ability, however, to affect the value of workers' subsistence basket. One of the ways, as Jim O'Connor pointed out here recently, is through increases in productivity (e.g., through subsitution of capital for labor in the production of subsistence goods).
Roger