Re: [OPE-L] Hume

From: Gerald_A_Levy@MSN.COM
Date: Sun Jan 23 2005 - 05:44:48 EST

----- Original Message -----
From: "Jurriaan Bendien" <>
Sent: Saturday, January 22, 2005 7:40 PM
Subject: Hume

Interesting topic - quick side-comment while I think of it: a good
discussion of the role of causal theory in scientific research is: Mario
Bunge, Causality in Modern Science (Dover Press,1979). His website is here:

Cognitively, I would say most stochastic or probabilistic investigations
ultimately assume, tacitly or explicitly, judgements or theories about
causal relationships, in the same way as analyses of price aggregates refer
to judgements or theories about value or value relationships (even although
the validity of value theory may be denied). Causal understandings are in
fact inherent in human praxis, and likewise value attributions will exist
regardless of exchange, or even regardless of individual volitions.

This view of things would be rejected in a naive empiricist-behaviourist
universe perhaps, where learning is equated with observable behavioural
change resulting from a perception of an association between a stimulus, a
relevant condition, and a response, at certain rates of frequency.
Likewise,  in a Hayekian universe, any objective concept of value is
strictly peaking  meaningless, and therefore, strictly speaking, an
aggregate concept such as  "value added" used in national accounts is
also a fiction. The problem  though is, why stop there? Are not the
aggregates in business accounts in  that case also a fiction?

This kind of consideration leads me to think that a big problem in Sraffian
economics is the utilisation of price theory itself. Just as the
neoliberals  talked about "the market" as if it was one homogenous entity,
Sraffians assume that prices are things which all fall in one category, and
are all ontologically the same. But there is a big difference between actual
prices, anticipated prices, theoretical prices, ideal prices, potential
prices, accounting prices, imputed prices and hypothetical prices. These
are not ontologically all the same, and behind their postulation often
lurks a concept or theory of value.

I think in fact it could be shown logically (although I haven't done it),
that ultimately Sraffian economics still assumes, or makes reference to, a
concept of value, and that this is completely inescapable, just as
reference to the causal principle is ultimately unavoidable, even in
inquiries which claim to be purely probabilistic.

It is one thing to claim that, because of their nature, some phenomena can
only be probabilistically understood. It is another to claim causality is
an outdated concept. Not infrequently it happens that probabilistic inquiry
is only a pathway to a causal understanding of a phenomenon. Ruling out
the possibility of a causal understanding in advance, may turn out to have
been a mistake at a later date. Correlation may not entail causation, but
that is just to say that theory transcends the facts, contrary to a naive
Hempelian empiricism.

As a digression, a peculiar fact about prices, often ignored, is that, even
if they are not real (observable) prices, they can nevertheless influence
real economic behaviour. Strangely though, the learned economists dismiss
economic "values" as irrelevant, an outdated 19th century conception, even
although they are happy to talk about, and work with, prices which do not
exist, and which nobody has actually observed. They are even prepared to
acknowledge that these "ideal prices" can influence actual economic
behaviour, yet they dismiss "values" in one breath as a fiction. A rather
dogmatic view, to say the least!

One of the great merits of socialist economics has been that it has been
able to theorise prices, pricing processes, and the real use of, or
bargaining about, prices in a far more realistic way than in many
neoclassical parables about the subject. The reason is obvious: in a
planned economy a large number of prices are "administered" prices not
set  spontaneously by the forces of supply and demand. Needless to say,
the very existence of administered prices is largely incomprehensible in
neoclassical economics. How can there exist a price at all, which not set
by market-supply and/or market-demand?

As for Marx, he insisted numerous times that in order to solve the basic
problems of economic theory, one had to examine first of all the "pure"
cases, and then introduce various factors which modified the pure case, and
thus ascend step by step from the general-abstract to the real-specific
observables. For this procedure, however, the vantage point, or point of
departure was crucial, and indeed Marx evidently mulled a great deal over
the problem of his order of presentation, for the purpose of building up a
theoretical structure that could integrate more and more socio-economic
phenomena, settling eventually on "the commodity" (the basic principle of
trade combining value-in-exchange with value-in-use) as starting point, and
a "Hegelian" structure for his presentation of economic categories,
mimicking the master's Wissenschaftslogik.

Marx explicitly admitted in Cap. Vol. 3, contrary to what many Marxists
assert, that aggregate output prices did not equal aggregate output values,
that this was a theoretical assumption, but that the assumption was
warranted, because the deviation between the two aggregates was not so
great, in the type of economy he studied.

Marx thought the problem was, for example, to explain what relative price
levels were, if supply balanced demand. That idea opens up the possibility
of a causal understanding of prices and equilibrium, which goes beyond
trite verities according to which supply is explained by demand, and
demand is explained by supply. It is of course possible to explain prices in
terms of other prices, but as soon as we ask hard questions about why
particular prices are deemed relevant to the explanation, it usually turns
out a value-theoretic framework lurks in the background, unacknowledged.

Pursuing this line of inquiry further, it turns out that pricing (attaching
prices) IS valuing, because in relating many different (ontologically
heterogenous) prices, a concept of value is being applied anyway, tacitly
or explicitly. Value relations turn out to exist while prices are being
reconciled and computed, and they may exist, regardless of individual
volitions or expectations.

The latter is of course what Marx argues: where neoliberals talk about "the
market" as a homogenous entity, Marx talks about value which exists
regardless of whether exchange occurs or not, because it so happens that
goods and services take specific quanta of labour to produce, and do not
appear in markets out of nowhere, i.e. ultimately what is being valued in
markets is human labour-time, even although markets can function without
any awareness of this reality (for the purpose of price-negotiation, neither
party to the exchange necessarily has to know what a good is objectively

I think Marx did not just seek of offer an explanation of social relations
under capitalism, but an explanation of the totality of capitalism, and how
particular facets of the system related to the social order as a whole.
Value theory, elaborated on the theoretical assumption of market balance,
thus aims to elucidate the dynamics behind relative price movements and
economic behaviour, under conditions where masses of prices are related.

But that is very much an unfinished project, since Marx not only did not
discuss the spheres of distribution and consumption in any detail, but also
discusses production and circulation only in their pure forms (as
emphasised by Kozo Uno for example). (NB - thinking that Marx's analysis
is complete,  has I think led to numerous errors, such as thinking that
exploitation can only occur at the point of production, and not in the
sphere of distribution or consumption).

The questions for economists are really:

(1) why an (even incomplete) value theory such as Marx's enabled a far
better prognosis of the historical trajectory of the capitalist system than
price theories, and why, even although value-assumptions are constantly
being made in judging prices, it is impermissible to elaborate a consistent

(2) why Marx's theory enables a far more realistic theory of economic
behaviour than an economics which only relies on the price system, and
which typically resorts to "extra-economic factors" outside the aegis of the
discipline when this reliance causes anomalies and paradoxes.

(3) what explains economic behaviour in regard to the majority of assets
owned in society, which are not being traded at any given time and have no
determinate or actual price, although most people would agree they do have

Wassily Leontief, who acknowledged the predictive and explanatory power of
Marx's analysis, called Marx "the great character reader" of the capitalist
system, rather than a true economist (concerned with the optimal allocation
of scarce resources). But a moment's reflection about Leontief's own
theories of economic aggregates will reveal that they still refer to
valuation principles, and hence to a theory of value, even if these are not
made explicit. The question then arises of what warrants this "embedded"
theory of value? Perhaps the "grundlichkeit" of the old lion of the British
Museum has something to recommend after all.

I wouldn't deny that it is possible to theorise the economic phenomena of
capitalism without resorting to values, and only by referring to prices,
but if it is true that concepts of value, value referents and valuations are
being assumed anyhow in theorising the price system or the activity of
pricing, it seems logical to make those assumptions explicit, and theorise
them coherently. So if Marx's theory is rejected, the problem of value
still remains.


This archive was generated by hypermail 2.1.5 : Mon Jan 24 2005 - 00:00:02 EST