[OPE-L:1097] Re: Definitions and Tautologies

Gilbert Skillman (gskillman@mail.wesleyan.edu)
Fri, 16 Feb 1996 19:33:08 -0800

[ show plain text ]

Andrew writes:

> Gil's response has changed the subject. I had said that unless Marx is
> shown to have been logically inconsistent, e.g., in the value/production
> price transformation, then interpretations must be judged by how well
> they make sense of the whole--my jigsaw puzzle parable. Gil responded
> that the TSS interpretation "necessarily" leaves out pieces of the puzzle;
> i.e., it contradicts Marx's definition of a commodity's value as the
> labor-time required for its production (or the labor it contains), AND
> he (Gil) challenged me to show that this wasn't the case. The
> implication of this challenge was therefore that, if I did show the
> TSS interpretation of value does not contradict Marx on this score, the
> pieces of the puzzle are not hiding under the rug, the TSS interpretation
> is thus superior interpretation of Marx's value theory precisely because
> it does make sense of the whole, replicating Marx's theoretical conclusions
> again and again, whereas other interpretations cannot.

But I have not changed the subject at all, if we reject demonstrations of
"consistency" which *require* us to fundamentally alter the basic meaning
of words, especially if this alteration is in no way suggested by the context of
passages in question. And I think we *must* reject such
demonstrations, because otherwise any statement can be shown to be
consistent with any other statement by the appropriate
"interpretation". One thinks of Orwell: war is peace, etc.

To anticipate, I don't consider as in any meaningful sense "superior"
an interpretation that requires us to do fundamental violence to the
contextually indicated meaning of the writer, even if this practice
"makes sense of the whole." [In what must surely be a very special
notion of "making sense of".]

Recall Marx's stipulations: "Commodities...which can be produced in
the same time, have therefore the same value." "The value of a
commodity would therefore remain constant, if the labour-time
required for its production also remained constant."

The point of my argument is that the TSS expression for value "does
not contradict Marx" *only* if one accepts that "the labour time
required for the production" of a commodity is identical to p(t)*A(t) +
L(t). The mere fact that the latter term can be transformed into
an expression measured in hours does not do the trick.

Now, it is clear from context that Marx does not *define* the labor
time required for the production of a commodity as p(t)*A(t) + L(t);
indeed, prices are nowhere mentioned when he discusses the notion of
socially necessary labor time on p. 129 of V. I (Penguin edition).
Therefore the identity must be established somewhere else in Marx's
writing, invoking some other mechanism than definition.

Now I grant Andrew this: it could be that Marx can be construed to
assert this somewhere [though that's not obvious to me--which passage,
Andrew?], but since we know that any such assertion would in general be
nonsensical--changes in *product* demand can change p(t)*A(t), even if the
technique of production is fixed, and by any rational usage of language demand
conditions aren't production conditions--there is a deep question as
to the sense in which this approach defends or validates Marx. But I
don't believe Marx says any such thing.

> Notice that I'm not just claiming this. The discussion between Gil and I
> reached the point where he himself suggested that this conclusion would
> be correct--if only I could show that the TSS interpretation of commodity
> value is not "necessarily" inconsistent with Marx.

> I DID SO in ope-l 1044.

Only if we grants a reading of the sense of Marx's words which is
violates the standard meaning of words and which is not indicated by
Marx's context, at least as far as I can tell.

> Thus we are back to the conclusion of my jigsaw
> puzzle parable: the results speak for themselves. If those whose
> interpretation of the instructions makes them unable to solve the
> puzzle complain that they find a contrary interpretation, on the basis
> of which others have solved the puzzle, to be implausible--and now I'll
> add sophistical and unnatural--it doesn't matter in the least. I don't
> care that the instructions to Ricardo's or Morishima's or Roemer's
> puzzle stipulate that pieces must be interlocked. Using ALL the pieces
> of Marx's puzzle, the TSS interpretation has been able to replicate the
> picture on the cover of the box by interpreting p. 265 of Vol. III
> (Vintage)

[which passage on p. 265?--GS]

> and several similar passages differently; along with which
> goes a different understanding of labor-time needed for production.
> Now note well: Gil DOES NOT, DOES NOT, dispute the fact that I have
> indeed shown what he asked me to show.

But yes I do. See above.

> What he says is that I show the consistency of the TSS interpretation of
> commodity value to Marx's definition by assuming what must be proved.
> Excuse me. I didn't know I had been challenged to "prove" anything
> other than the logical possibility that the TSS interpretation is
> compatible with statements that value is determined by labor-time
> required for production or labor-time contained.

Yes, that's exactly it: you have to prove on contextual grounds that
the TSS interpretation is compatible with what Marx says, according
to his probable meaning. It is not a proof if one begins with the
stipulation, "assume Marx states that the labor time required for
production is a function of commodity prices". That is exactly the
point at issue: I do not see grounds for the claim that Marx
believes the LTRFP can be understood in these terms.

> Of course, Gil is right about another point: if I must begin from HIS
> interpretation of these statements, according to which prices cannot
> affect the labor-time needed for the commodities' production or the
> labor-time contained in them, then I cannot arrive at the opposite
> conclusion. Sure. But this is wholly irrelevant, and had nothing
> to do with Gil's original challenge.

No, it is assuredly not just MY interpretation. What dictionary in
the world, including Marxist ones if there are any, would admit an
interpretion of "the labor time required for production" which
depends on commodity prices for its *expression*? If Andrew can show
me one (not written by a proponent of the TSS approach), then we'll discuss
the possibility that it's just my interpretation. But until then, my
arguments are necessarily relevant.

> I had stipulated clear criteria
> according to which my interpretation of Marx would be falsified. Gil
> thought he could indeed falsify my interpretation. He did not.

Under the proviso that one's interpretation should not do fundamental
violence to the meaning of language, I did.

> My challenge to Gil and everyone else on the list is for *you* to now
> specify the conditions under which you'd accept the superiority of the
> TSS interpretation *as an interpretation of Marx's own value theory*.
> Thus far, I haven't seen any such criteria proposed. Gil is claiming
> that he can show Ch. 5 of Vol. I is wrong. I haven't followed all the
> nuances of this debate, and so cannot answer this at the moment, but
> it is a completely DIFFERENT issue. I'm asking for criteria to judge
> the relative adequacy of different interpretations *as interpretations.*

That's entirely fair, especially since Andrew has taken the trouble
to specify the grounds on which the TSS approach to interpreting Marx
can be justified, which Allin has referred to as the Principle of
Charity: other things equal, adopt that interpretation that is
consistent with the greatest part of the writer's account.

But I suggested an alternative principle when I introduced the Ch.5
argument, and it is not a "completely DIFFERENT" issue at all.
Suppose that one interpretation permits invalid and substantively
contradictory conclusions to stand, while another rejects the
objectively invalid arguments and in so doing affirms a valid and
internally coherent portion of the account. Then I think there are
presumptive grounds for the latter approach.

In Andrew's sense, this approach does a certain amount of
violence to Marx's account: in this case, it insists that Marx
maintains two mutually inconsistent hypotheses as to the connection
between prices and values, and that one of the hypotheses [i.e. the
one that considers price-value equivalence to be in some sense the
"pure" case of commodity exchange] is invalid and should be rejected.
[But for what it's worth, the invalid part was the product of at most
a year or so of Marx's work, while the valid part was in evidence in
all of his working notebooks up through the material published as
Parts 4-6 of Volume I.]

But in another sense, I think, it does much less violence to Marx's
account than Andrew's principle, as realized in this case: it does not
compel us to redefine Marx's words in fundamental and contextually
unjustified ways; it does not require us to maintain manifestly invalid
conclusions simply because we are loath to reject pieces of Marx, etc. etc;
and it allows what can be defended as the more relevant and coherent portion
of his argument to stand out.

Do I have to keep going on and on about this? Haven't I made myself
> clear?

Yes, but.....

> Gil says I assume what I need to prove. Ironically, he shows this only
> by assuming what *he* needs to prove. He invokes Morishima to tell us
> that the labor-time needed for a commodity's production is "straight-
> forward and unambigous"--i.e., determined by technological coefficients
> only. But is Morishima Marx?

Marx, not Morishima, defines socially necessary labor time solely in
terms of production conditions (as opposed to market conditions) on
p. 129, right before Marx, not Morishima, issues the stipulations
quoted near the beginning of my post. When Marx distinguishes the
"living" and "dead" components of a commodity's value on p. 294,
Marx, not Morishima, invokes only production conditions and says
nothing about market conditions. The only point behind invoking
Morishima was to verify that Marx's practice of referring to a
commodity's value in terms of direct and indirect labor *or* total
direct labor was legitimate.

>If, but only if, one accepts Morishima's
> view, then input prices cannot have an affect on values.

If, yes, but not only if. Indeed, the single-system approach
seems to offer the only case in which input prices are judged to be
part of a commodity's value (other than the commodity called labor

>Gil's whole demonstration is, to use one of his favorite words, tautological.

Manifestly not so. See above.

> Gil misunderstood some of my subsequent remarks in ope-l 1044. Of course
> he's trying to understand Marx. My point was that one will not succeed
> in doing so if one does the following: translate Marx's statement A
> as X and his statement B as Y, and then claim he contradicted himself
> because X and Y are incompatible. This point was not directed at Gil
> specifically. Almost all the objections to the TSS interpretation on
> this list have employed some variant of this method. This is INVARIABLY
> the method used to "prove" the "internal inconsistency" of Marx's value
> theory. The amazing thing is that anyone buys it. All this method
> shows is that X and Y are incompatible, not that Marx's actual statements
> A and B are. So I'd like Gil to comment on this methodological point,
> if he can, now that he understands I wasn't trying to say that he didn't
> want to understand Marx.

I didn't really think you were accusing me of not trying to
understand Marx, Andrew, but I do have a comment on the
methodological point: something more is being asserted than simply
"if one interprets A as X and B as Y, X and Y are inconsistent, so A
and B are inconsistent". Rather, the argument is that on contextual
and logical grounds A can *only* be reasonably interpreted as X and B
as Y, then etc. Conversely, to the extent the TSS approach requires
us to allow that "the labor time required for production" is
*directly* affected by changes in p(t), this is held to be
problematic on the grounds that it does violence to normal meanings
in language, and to Marx's own meaning, as indicated by the context
of his discussion on pp 129-130 and elsewhere in Vol. I. I think
this certainly constitutes a coherent and valid criticism of the notion that
the TSS approach is "consistent" with Marx.

> So of course I don't think the determination of logical fallacy depends on
> what class one is in. Rather the above method of refutation is flawed,
> obviously flawed, and those like Bortkiewicz who started it had
> an antagonistic intent.

Yes, I believe that's true. But I think we might also agree that an
interpretation which rejects part of marx's argument as invalid does
not necessarily signify an antagonistic intent--especially if the
ultimate objective is to champion the necessary and powerful
relevance of a major portion of Marx's argument.

Bottom line: I believe that Andrew's 'demonstration' of the
consistency of the TSS approach with Marx's stipulations on p. 130
requires us to believe _ab initio_ that Marx allowed p(t) to enter into the
expression of "the labor time required for production." But if this
is not the case, then the TSS approach is manifestly inconsistent
with marx.But this is precisely the question at issue--it can't be
assumed, as Andrew does in his previous post.