[OPE-L:7253] [OPE-L:778] Re: Re: Re: Commodity

Gil Skillman (gskillman@mail.wesleyan.edu)
Fri, 26 Mar 1999 19:26:49 -0500

Thanks to Michael for his thoughtful comments. He writes:

>In the thread 'Sequiturs?', Gil writes:
>> unless " generalized commodity production and exchange" is
>>taken to mean "exchange without transaction-specific bargaining,
>>discriminatory pricing, incomplete information, or transaction and mobility
>>costs", then this cannot imply LOOP.
>I realise this is the nub of the problem, and have been following this
>fascinating thread with interest. And I'm not sure I have come even to a
>temporary equilibrium in thinking about it. But, taking my credibility in my
>hands, I will try thinking out loud a bit.
>First we are indeed talking about the abstract conceptual model. I do not
>deny the empirical existence of transactions-specific bargaining, etc.. The
>question is whether these practices are to be seen as contingent (with
>respect to the reproduction of the capitalist (for short) system), necessary
>to it but subordinate to the dynamics of the value-form, or systemically
>necessary to the reproduction of the capitalism? (Note that none of these
>classifications invites us to ignore the existence and effects of these
>practices in investigating capitalism, or says a priori that they are
>'unimportant' in any sense.)
>Since I do not think that abstraction and the development of conceptual
>models should be contingent, an attempt to answer this question with due
>care would require a rerun through the systematic process of unfolding and
>transcending contradiction that imo drives the development of adequate
>conceptual models. But as a tentative beginnning, I would like to suggest
>that the practices at issue (with the probable exception of 'transactions
>and mobility costs') are indeed contingent wrt the reproduction of the
>system. Any residual uneasiness about this notion derives, I think, from the
>fact that they seem to typify what we might think of as normal behaviour of
>capitalist economic agents, and that they appear to be empirically
>ubiquitous. However, whilst these thoughts certainly invite us to have a
>conceptual model that can deal with these practices, it doesn't preempt the
>conceptual question of *how* we should do so.
>In a more pragmatic methodological vein, if Marx is indeed enaging in an
>'even if' type argument in Vol I - 'even if all exchanges occur at value,
>there is demonstrable exploitation in a capitlaist system' - then that might
>include the argument 'even if there is no assymetric power at work
>associated with 'transaction-specific bargaining, discriminatory pricing or
>incomplete information', there is exploitation. One might care to think of
>an analogy with a frequent defence of the 'instrumental rationality'
>assumption in orthodox microeconomics: that without it there will be a
>tendency to offer trivial proximate causes for phenomena in terms of
>unexplained changes in subjectively perceived preferences or constraints.

I'm willing to accept for the sake of argument that Marx abstracts from
such conditions as transaction-specific bargaining in his Chapter 1
analysis of commodity exchange. To anticipate a bit, though, even with
this stipulation the remainder of his Chapter 1 argument does not follow.
This turns out to be costless to his theory of exploitation, since it's
possible to support the latter without asserting that abstract socially
necessary labor is the basis of something called the value of a commodity,
or even that commodities have a thing called "value" understood as distinct
from exchange-value and use-value.

>Gil again:
>> Remember, the question is whether "commodity exchange" implies LOOP as a
>matter of *definition*, not simply contingent *abstraction.*
>I am not sure what is meant here, although I may have covered Gil's concerns
>above. Imo it is precisely 'definition' that tends to be 'simply
>contingent', unless its is developed as part of an internally coherent,
>systemic conceptual model that articulates (and rectifies) our abstractions.
>This involves, amongst other things, purging as much contingency from the
>system as we can (and then keeping that which we cannot in mind in the
>further interpreation of our conceptual models).

This issue is not central to the thread with Michael and me. Alan has
asserted that LOOP is part of the *definition* of a commodity, and
furthermore that this is a standard definition. I have reason to doubt
this, but that has no impact on the substance of the present discussion.

>Gil goes on:
>>No, contradiction. Michael has exactly missed my point. The relation of
>>"equality" that Marx holds to flow from systematic exchange is *not* shown
>>to depend on the exchange of *commodities* in particular. [Note in this
>>connection that Alan's representation of an "equality relation" *in no way*
>>depends on the elements that enter into the relation, but rather *only* on
>>the properties of the relation R itself.] Therefore if it is true that
>>systematic exchange (however qualified) establishes relations of equality,
>>it must hold for non-commodities as well as commodities. But that's where
>>the contradiction necessarily arises, since unimproved land and boot-polish
>>can be exchanged, and thus "equalized" in Marx's sense, and yet obviously
>>aren't both products of labor, contrary to his conclusion.
>I can see that this is crucial to the thread argument so far, but I am not
>sure that I grasp the point.

Let me add something here in anticipation of subsequent discussion. Marx's
assertion that "a common element of identical magnitude exists" in
exchangeable commodity bundles is based on his claim that "the valid
exchange-values of a particular commodity express something equal", and
thus that an exchange relation "can always be represented by an equation."

But this claim is *not* shown to depend on the fact that *commodities* in
particular are being traded; as argued, it follows only from the fact of
systematic exchange (possibly abstracting from transaction-specific
bargaining and the like). Consequently, any such "equalizing" properties
of exchange must, so far as we know, also obtain when *non* commodities
like tracts of unimproved land or insurance policies are being traded. But
as argued earlier, this leads to a contradiction with Marx's subsequent
inference that the "common element" among exchanged bundles can only be
abstract labor.

In a separate post to Brendan, I use as an illustration an argument which
asserts that baker's yeast causes rye bread dough to rise. If the
explanation of this claim is based only on the properties of yeast acting
on bread dough, then we've no basis for thinking that the result--bread
dough rising--is unique to *rye* bread. So far as we can tell from the
argument, this restriction is incidental to the properties being asserted.
Just so with the "equalizing" properties Marx associates with exchange.

>So just a few comments that might stimulate more of the excellent
>clarifications that have been so far generated.
>First, *if* Marx's argument in Chapter 1 can only be interpreted as an
>attempt to make the case that there is necessarily some pre-existent
>substance that is equalised by any isolated exchange (and that that
>substance is abstract labour), then I agree his argument is flawed. To this
>since any such exchange can only be understood in terms of its embeddedness
>in a system of capitalist generalised commodity production and exchange. It
>is this system that drives the real abstraction that reproduces abstract
>labour as the substance of value, and that generates the prices and Money
>that are the *only* empirical expression of value,
>Any metaphor of homogenous labour as prexisting all capitalist commodity
>exchange, especially one that envisages exchange value as quantitively
>determined prior to exchange, but even one that has exchange 'chopping up'
>the pre-existing substance seems to me to be fundamentally misleading.
>Of course, given that the whole system is up and running, the agents in it
>will, of course, have a good idea of what the subsequent price of their
>product will be. And indeed, in the absence of a unified monetary unit in
>which to do so (e.g. in international comparisons) actual labour content
>might help the prediction of prices. But that pragmatic apparatus does not
>need the metaphysics of a pre-existing value substance in order to work.
>Later Gil argues:
>>But this "alternative interpretation" begs the question at hand, since it
>>presupposes exactly what must first be proven, that "abstract labor" has
>>anything fundamentally to do with exchange or exchange-value.
>I'll have to think about this. For now, I am not of the view that Marx
>wanted or needed to demonstrate a labour theory of value, since it seemed
>obvious to him that value derived from the sole non-economically
>reproducible ubiquitous element of production.

But this statement assumes what must first be proven, namely that there
*exists* a thing called "value", understood as distinct from use-value or
exchange-value. That said, I agree that Marx didn't need to demonstrate a
labor theory of value. I think Marx's Chapter 1 argument is not only
invalid, it's superfluous.

> He sought, rather, to work
>out the implications of this for a capitalist commodity system. (I think
>Alan recently argued something similar.) If I wanted to try and get this
>across to someone to whom it was not so obvious, I might make grand
>reference to labour as the social form that the human life-force takes
>under capitalism, or more pragmatic reference to some kind of opportunity
>cost type argument. I might also preempt some potentially 'hard' cases by
>arguing that whatever the nature of the 'robotocracy' that someone mentioned
>recently, it would not be capitalist. Indeed, automation and robotisation
>help to generate one of the contradictions of capitalism: that between its
>increasing productive power and the continued dependence of distribution of
>the product upon contribution to its creation.

>>Marx defines "abstract labor" as simply that aspect
>>of labor that remains once one *intellectually* abstracts from the concrete
>>forms of labor.
>No amount of 'intellectual' abstraction can effect the way the world works,
>unless it systematically informs human action. *If* this is indeed the only
>way to interpret Marx;'s argument, then we need some rectification of it.
> >Thus one need not have "generalized" commodity production,
>>whatever that means, in order to define "abstract labor."
>I am not sure what aspect of 'capitalist generalised commodity production
>and exchange' is unclear to Gil?

I doubt if this still matters, since for the sake of argument I'm granting
the abstraction from transaction-specific bargaining power, etc. Again,
this particular complication flowed out of a separate argument with Alan,
and doesn't seem to be central to the issues Michael is pursuing.

>However, I would argue that for a coherent understanding of abstract labour
>that can play its part in reproducing specifically the capitalist system, we
>need something more than some general notion of human muscle and brain power
>that physically underpins not only concrete labour, but also pre-capitalist
>forms of human productive activity in interaction with nature. Once again,
>what we need is not arbitrary, stand-alone definition, but a coherent
>abstract conceptual model.
>>And by the way, what does "pure" mean in this context?
>See above on abstract conceptual models and their interpretation. It refers
>to what is currently (inevitably pro temps) a necessary moment of our
>abstract conceptual model of the system - ie is necessary to the
>reproduction of that system as what it is. Perhaps your discussion with Alan
>et al could be couched in terms of whether indeed LOOP is such a necessary
>As I said previously:
>>> it remains to be seen whether any of the counter
>>>examples to LOOP can be shown to be ubiquitous and systemically important
>>>enough to threaten LOOP as a moment of the pure conceptual model.
>With reference to what Gil sees as a diversion into 'even-if' arguments, he
>>I have a separate critique with
>>respect to Marx's Chapter 5 argument on this point. Briefly, Marx asserts
>>that it is in some sense necessary to explain exploitation of labor under
>>capitalism on the hypothesis that all commodities exchange at their
>>respective values. This claim, as I argue elsewhere, is invalid.
>I am not sure that the only, let alone the best way to interpret Marx here
>is that it is 'necessary' in any strong sense to explain exploitation on the
>basis of exchange at value. Rather (see above) it may be a pragmatic
>methodological imperative as part of a larger argument that there is a
>specifically capitalist form of exploitation, appearances notwithstanding,
>even if we assume away all assymetries of power that, as it were,
>improvements to the functioning of the capitalist economy may, in principle,
>be expected to 'internalise'.

I'm not sure if we want or need to get into the details of my Chapter 5
critique, but it makes a stronger claim than simply contradicting Marx's
concluding assertion that "The transformation of money into capital has to
be developed on the basis of the immanent laws of the exchange of
commodities, in such a way that the starting-point is the exchange of
equivalents." [pp 268-9, Penguin ed.]

It's true this claim is invalid, and doesn't follow from the arguments Marx
advances in Chapter 5, but I also make the point that this representation
of the problem is *essentially misleading*, and obscures what is *really*
central to the explanation of capitalist exploitation in historical
materialist terms. I lay out this critique in detail elsewhere, and it's
not obvious that we should or need to get into it now.

>Finally from Gil:
>>OK, but I don't see this caveat as relevant to the exception Alan takes to
>>my use of the term "tautology." Granting that "no real world object is
>>identical to any other real world object", it is still the case that a
>>theorem, as typically understood, is not simply a case of circular
>Fine. My purpose in raising this point was not specifically to take Alan's
>side of the argument, but rather to help to clarify the 'abstract conceptual
>model'/interpretation thereof/the real world in thought, distinction upon
>which my arguments rest (and which, by good fortune, may also help you to
>grasp what a I mean by 'pure' in this context!)

Bottom line: hindsight is 20/20, and one can never tell going into a
discussion what issues are going to emerge as central. Given what seems
now to be the lasting source of disagreement, I'm happy to grant for the
sake of argument the postulate that Marx abstracted from factors which
preclude the operation of the "law of one price" in commodity exchange.
[I'm still not clear where this stipulation is made in _Capital_, and I
continue to have my doubts about Alan's definition of "commodities", but
these concerns are not central here.] This stipulation allows a focus on
this central issue: I see no valid basis in Marx's Ch. 1 argument for the
claim that commodities have a "value", understood as distinct from
exchange-value or use-value, let alone for the subsequent claim that
abstract labor is the substance of this value, or socially necessary labor
time its measure.