Chapter 14 The Dual Ontology of Capital

In: The Spectre of Capital: Idea and Reality
Christopher J. Arthur
Search for other papers by Christopher J. Arthur in
Current site
Google Scholar
Open Access

This chapter is not part of the systematic presentation itself (hence its lack of section numbers denoting the introduction of a new form). Rather, it continues the reflections of Part 1, in that I stand back, as it were, to situate what was presented in the previous chapter in the context of a ‘bigger picture’ of my diagnosis of the inverted world of capital.

The transformation problem (of supposed ‘values’ into supposed ‘prices’) is usually thought to require a mathematical fix; if such a fix is ruled invalid then the Marxian theory is to be abandoned. In truth, the supposed problem is due to a fundamental confusion between the realm of material processes and the ideal movement of capital; these combine only uncertainly. Right at the start value is to be taken here as distinct from, indeed opposed to, use-value. Later, however, once the value system is shown to take possession of the material process of production, use-value must be granted as having its economic determinacy. But – and this is crucial – to collapse the value system into material determinants (for example, concrete labour times) is to reduce value theory to a naturalism. This neglects the specific determinacy of capitalist social relations. Such a neglect is encouraged by the way capital occludes social relations and makes it look as if material production is the be-all and end-all.

In truth the transformation problem has an ontological basis. Once one locates the transformation procedure in the context of the interpenetration of two opposed systems of determination, two things follow. First, one understands why there is a problem for the naturalistic approach; second, one understands there is really no problem at all. The problem arises only if it is assumed that c and v are measured first in simple prices (SP); but there is no reason to make such an absurd assumption and every reason not to do so, for the magnitude of these prices is systemically achieved only through ‘reproduction price’ (RP).

In the reproduction of capital as a system, the relation of competition between capitals distorts, or even subverts, the capital relation in the narrow sense. However, this is of no moment since simple prices derived from the latter are in truth merely virtual, just numerical illustrations of the abstract shape of the capitalist relations.1

This is because the life of capital is concretely determined only as competition between many capitals. Once competition is taken seriously the difference between capitals makes a lot of difference to the constitution of the whole. Because capital reproduces itself only through interchanges between departments at prices of reproduction, these categories are central to the capital concept. I argued above that prices of reproduction are the finished form of value because they are intrinsic to capital’s reflection on itself in its total interconnectedness. What causes the emergence of a problem is the requirement that departments balance which means that so-called input prices and so-called output prices must be the same. However, the reproduction prices ensure there is a uniform rate of profit and departments are integrated.

While I have questioned the meaningfulness of simple prices, I accept the great importance of what underlies them, namely the labour times required to produce the wage bundle and the surplus product. The rate of ‘exploitation’ given in the ratio of the labour ‘embodied’ in surplus goods (whether luxuries or capital goods) to the labour ‘embodied’ in wage goods is prior to, and independent of, the determination of prices. Moreover, it gives something of a parallel to exploitation in pre-capitalist systems such as the direct appropriation of slave labour.

Even if it were allowable to give sense to simple prices of wage goods and surplus product, it still does not follow that any conservation of their aggregates should obtain. We are dealing with quite different states of the same thing; the initial state is determined by the one-dimensional structure of exploitation; the second, two-dimensional state, re-determines the first by the impact of relations between capitals.

If v and c are calculated at SP their magnitudes will normally differ from those calculated in terms of transformed values, of RP. If it is stipulated that total SP is unchanged when transformed to total RP, the balance of v and c within all cost prices, and their aggregates, change. This means that the total profit (added value minus v) in SP terms will differ from that in RP terms, and the relevant rates of profit also.

This can of worms is avoided if v and c are given as RP in the first place, hence not in need of any transformation. What remains true at the level of the aggregate is that, with all measures in RP, the total M invested yields the total M′, with the added value, total m, given, and the total sv easily derivable, together with the GRP. The MELT links the total added value with the total socially necessary labour time, as we have seen. This fundamental theorem, so to speak, lies at the basis of the determination, in the sense of pre-condition, of prices and profit, albeit that it leaves RP s yet to be explained in full.

Incommensurability of different states of the same thing is well known in physics: ice is bigger than the same amount of liquid water, indeed here even the normal unit of measure differs, litres and cubic centimetres. In our case, we are dealing with socially determined ideal magnitudes. How much less, then, should we expect conservation across situations in which additional social relations come into play? It helps when dealing with the transformation of value into more finished forms to recall that value is an ideally constituted substance. Thus it is ‘soft’, so to speak, and re-formed according to the forces to which it is subjected.

The general rate of profit in money terms, and reproduction prices, are mathematically co-determined (whether simultaneous equations, or an iterative procedure, are utilised). But ontologically prior is the physical mass of waged goods and surplus product. This mass cannot be changed by how it is priced, and how value is distributed in such terms. If the distribution of the product cannot be changed in physical terms, the social measure of its fractions is not theoretically determinate until the level of reproduction prices. However, one can always retrospectively impute such ‘finished’ values to the inputs of immediate production, and then the ‘conservations’ hold tautologically, because the inputs are now given in ex post prices, ‘already transformed’ so to speak.

The material process, namely the regulation and exploitation of labour, other things being equal, would express itself at a level of phenomenal prices in linear fashion. However, capital, as a peculiar form of production, exerts formal determination on the supposed content. There is a ‘problem’ if we recall (§52) that materially living labour is subsumed, appropriated, and exploited, by capital, but it is in itself other than capital. So we need now to show how this plays out within the ideal system of pricing. It is capital, according to its self-determining movement, that locates living labour as the other that must be appropriated. It is capital, with its own measure of itself as a rate of accumulation, that determines labour is to be measured in time, not in intensity or whatever, and that a relevant issue is not merely the rate at which new value is added but the rate at which the constant capital is resurrected.

The dual ontology of capital refers to the original separation between material and ideal levels of reality. The latter is the value system and, as such, is itself a single system; there is no question of counterposing value and price because value is actual only as price, and the transformation procedure moves from an overly abstract account of value to a more concrete, because systemically determined, one.

The importance of the material level is underlined by Riccardo Bellofiore.2 He stresses the salience of the category ‘real wage’ (which is taken as given by Marx through most of his argument). He draws attention to the crucial role played by the division of the product between the real wage and the surplus product. Indeed, if the labour time to produce the real wage took the whole day there would be no surplus for capital to appropriate. This ratio is a real measure of exploitation and our theory must explain how it is reproduced. In order to do this, it is convenient to keep the real wage constant through the analysis (until the tendency to the falling rate of profit is addressed).

It is perfectly consistent with this presupposition to assume at the value level that so-called variable capital is likewise fixed. There is no need to favour one constant over the other. The value of the real wage bundle must be ultimately explained; but at the start it may be taken as given, and not requiring any transformation. (See §83.3.)

It follows that the duality gives rise to two different definitions of ‘necessary’ (and hence ‘surplus’) labour: one refers to the labour required to produce the real wage bundle; the other refers to the labour yielded to get the money wage to purchase it, at prevailing prices. These are two different magnitudes if, for instance, we take the real wage bundle at production prices.

One problem with capital is the way the underlying material reality is obscured by the value system through which capital cognises it. If the source of surplus value is surplus labour this qualitative point is obscured by the highly mediated determination of its magnitude, and the apparent ratio of exploitation in value terms. That is why it is essential to begin with the elementary capital relation in order to show how the surplus is produced in the struggle of class against class before addressing the more complicated relationships consequent on capital competition. Nonetheless the systemic determination of value is also crucial.

The ontological duality reaches right back to the absenting of use-value at the start of the value form dialectic, because ideality, although a fold in materiality, has its own effectivity; the two realms interpenetrate, such that each movement conditions the other. There is no ‘third’ to these two; so the capital system is intrinsically unstable, albeit it may find temporary ‘fixes’ to its problems.

When we accept some such equilibrium derivation of RP and GRP as Sraffian theory proposes, this does not dispose of the deeper ontological reality intrinsic to the capital relation. Moreover, even within the value system, the magnitudes of c and v theoretically discovered by simultaneous equations cannot displace the explanation of their actual production through the exploitation of living labour in the capital circuit.

The TP derives PP timelessly in making a mathematical calculation of how surplus value is distributed to each capital. However, the origin of total surplus value depends upon class struggle at the point of production over capital’s effort to pump out labour from the recalcitrant labour force. It is not necessary to know what determines c and v in order to explain the production of surplus value as the outcome of exploitation.

Albeit the traditional transformation ‘problem’ is a false one, the material organisation of living labour must be taken as the reality presupposed to the price and profit questions. Certainly, I think it is a problem that the price determinations occlude the fundamental material relation between capital and labour, in which the real wage allocated to labour is a key parameter reproduced by the system. (Notice that any transformation from simple prices to production prices will change the general rate of surplus value, because at production prices aggregate surplus-value and aggregate wages, sufficient to cover the real wage, will normally differ from the same magnitudes as they were expressed in simple immediate prices.)

Capital’s regime of truth confuses the ideal and material realms; so we must deconstruct this double-sidedness of capital. In the capital system there is an identity-in-difference; for the same ‘stuff’ is differently form-determined when considered in the context of a class relation, and when considered as ‘valued’ in its ideal actualisation through capitalist competition. Explaining the division of net material output is as important as explaining the monetary increment appropriated by capital.

Certainly, because this system is exclusively in monetary terms, it makes no sense to refer prices back to some sort of ‘labour values’ measured in time.3 However, such monetary realisation of value, and surplus value, does not mean that the physical rate of exploitation, expressed in virtual labour hours, is irrelevant. The physical rate compares the labour required to produce the real wage with the expropriated labour required to produce surplus goods. Moreover, this rate is of great interest to both classes, and determined in the struggle at the point of production. Whatever the price rule actually in operation this physical rate remains a crucial underlying parameter.

It is not just a question of abstracting out an important variable for study, for this abstraction is rooted in the ontological duality characteristic of the capital system. On the one hand, we have the material side of the economic metabolism, of prime interest to workers; on the other we have the ideal side, predicated on the self-movement of capital, wherein it re-conceptualises the material variables under its alien measures. Thus the working of the system must be appreciated from two class standpoints.

If the concretisation of the abstract notions, through which capital comprehends itself, fails to preserve the material register of exploitation this is no reason to neglect this fundamental material form. There is here a deeper truth, obscured by the measures capital takes to be real, namely the truth of exploitation in all its rawness.

In all this the real wage is of basic importance to the working class. C–M–C′ is the workers’ circuit which allows labour power to be made available for purchase by capital, but which is reproduced domestically insofar as money wages purchase the real wage bundle. The workers are interested in how long it is necessary to labour in order to fund the real wage; anything above this is surplus labour for capital. This is a key ratio regardless of how the real wage is priced. It is the same as the rate of exploitation in money terms. In value analysis it is this time that counts as necessary labour; surplus labour time above that is appropriated by capital.

This ratio of necessary and surplus labour is not necessarily identical with that between the labour required to produce the real wage bundle itself and the labour required to produce the surplus product. As we see in the table of TP, if capital a is a representative fraction of the means of production department, and capital b is a representative fraction of the real wage department, it follows that the rate of exploitation registered in the simple price of each output is changed if the rate is recalculated in production prices. But the final determination of the rate of surplus value in money terms is of no importance for the workers, and their own competition. It is of concern to capital as the outcome of capital’s competition.

Which system do the workers care about? Trapped as they are in the duality, they are doubly determined; immediately they want a money wage to cover the cost of the real wage, however priced, so that leads to one definition of ‘necessary labour time’; but the underlying reality of the labour time required to produce the real wage is another way of looking at their exploitation. The second has the (dis)advantage of continuity with other historical forms of expropriating labour that were less disguised than here; for here the intuitively obvious material measures are overturned by money measures. Given capitalist mystification, uncovering this layer of reality is an achievement quite as much as insight into how the ideality of profit works.

In considering this ontological duality, there is felt a tension between two methodological principles. We are surely concerned to identify, and analyse, the basic law of motion of a system, disregarding superficial perturbations. This might lead us to prioritise ‘labour values’ measured in simple prices. Yet, at the same time, since it is a basic principle of dialectic that truth pertains only to system, then such labour values, if wrongly taken as substantively valid, are nothing but a false concretisation, because the true concrete is the totality, of which every part is subject to systemic determination. From this second point of view, simple prices are, at best, purely virtual, illustrating a fundamental relation to be sure, but a relation (the capital relation) that is co-determinant with others (the relation between capitals, for instance) that constitute the totality. To insist that physical exploitation (the ratio of labour time embodied in the surplus product to that embodied in the real wage) must be conserved in its socially measured shape is an abstract materialism that does not grasp the ideality of social form.

This last consideration is very powerful from the point of view of systematic dialectic. Yet it surely puts in question the fundamental Marxian theorem: that all value arises from the exploitation of labour. In my opinion this last can be supported only on ontological grounds, not on the claim that labour inputs are the best predictor of empirical prices, for instance. The exploitation of living labour is the explanation of the genesis of new value. This ontological truth remains fundamental, whatever the prices socially assigned.

So far from value being an ahistorical form distorted by capitalist relations it is only within capitalism that value becomes a truth. The peculiarity of capital is that the source of value in labour is so hidden, that it looks as if value is the reward for capital’s toil and trouble. (How could it be otherwise if there is a pro rata reward for the whole of capital?) Nevertheless, once it is recognised that capital is a social form with ontological depth, categories holding at one level may be redefined, or even inverted, at another. Because there is interpenetration of opposites, ideality and materiality, each side is reproduced by its other; but neither system reduces otherness to its own other.

First capital encounters labour power in exteriority, then internalises it through the wage-form, and then subsumes living labour under the capital relation that reproduces it in its position as capital’s own other. The determination of magnitude looks as if it flows from labour time, but this is a determined determinant for it is capital that pumps out living labour in the most efficient way available to it. In this way, the linear determination of value by labour is at the same time intrinsic to the spiral movement of capital accumulation.

Value is real only as a highly mediated result, because it is embodied in a totality of relations, not a single commodity. I myself simplify value when I identify it with reproduction prices in abstraction from market contingencies such as supply and demand. Many would say this is wrong because socially necessary labour time is a function of demand, e.g. an oversupply and a fall in value is consequent on ‘wasted labour’. For me it is legitimate to distinguish value from market prices if the former is a function of the minimum set of determinations that yield a concept of self-mediated reproduction no matter how abstractly this is taken. Hence the finished form of value is the reproduction price and then more concrete determinations further modify prices on the market.

However, if one gives magnitudes at too abstract a level, these merely illustrate the simple determinations concerned, but must not be considered as being real values, merely taken apart from secondary influences. They are the result of a violent abstraction from the real process of value determination. But they reflect something that is real, namely the material process of exploitation which in one sense is the very nature of the economic metabolism, despite the impossibility of going smoothly from so-called ‘labour values’ to prices of production. This impossibility is founded on the ontological inversion that transforms in a very deep sense the magnitudes concerned. Contrary to neo-Sraffians, the labour ratios are not transformed out of existence; they remain as a discrepant surd in the ideal value system, reflecting the material reality of capitalism.

Although distribution cannot change the total mass produced, it is necessary to consider carefully what this mass is. It is the physical mass of workers’ real wages and the corresponding surplus goods. But how these are socially expressed as value can be changed. Under the (virtual) price rule of ‘simple prices’ we get one aggregate value, under the price rule of ‘reproduction prices’ we get another. The former makes explicit the underlying class relation, the latter is the form under which capitals fight for their shares. Because these measures are rooted in different ways of conceptualising production, there is no possibility of ‘deriving’ one from the other. Yet both measures show us something real about the economy; it is not a matter of choosing one as more adequate. This becomes clear if we attend to the basic fact of expropriated labour. At the material level, there obtains a ratio between the labour times required to produce the real wage and that required to produce surplus product. At the ideal level, there is a different measure: the labour required to gain the money wage and the surplus labour that underpins surplus value. Both material and ideal measures co-exist.


This chapter looks at the ontological basis of the transformation ‘problem’ in the dual ontology of capital, as it contains within itself ideal and material relations. Over-prioritising the material side leads to a naturalism of the value form, neglecting its socially specific character. Over-prioritising the ideal side leads to the occlusion of the economic effectivity of material determinants. The real wage and the surplus product are certainly material presuppositions of the TP. What is problematic is their reading as conditions of social forms, and especially of the systemically determined value measures. In this broad context are situated the forms addressed in the previous chapter, namely simple rice, production price, and reproduction price.


This has not been generally recognised; for F. Engels (in his Supplement to Marx’s Capital III) persuasively advanced the view that a regime of ‘simple commodity production’ would yield determinate values, and that such determinacy would carry over to what he called the ‘modified’ prices of capitalist production. I have elsewhere demolished Engels’s view; see Arthur 2005b.


Bellofiore 2004, p. 208.


‘Labour value’ is not a term used by Marx. Both Fred Moseley and Geert Reuten espouse a ‘single system’ of value relations, but each for different reasons. See Moseley 2016 and Reuten 2019, pp. 69–71 and 74–6. Moseley 2016 claims to be an interpretation of Marx, but Reuten 2017 has a different take on Marx’s exposition.

  • Collapse
  • Expand