This increase in the aggregate of use-values, moreover, indicates also a change in the value relationships. As Marx argues so convincingly, basing his whole theory on this axiom, the progressive development of labour productivity reacts on both the composition of acc.u.mulating capital and the rate of surplus value so that they cannot remain constant under conditions of increasing acc.u.mulation of capital, as was a.s.sumed by the diagram. Rather, if acc.u.mulation continues, _c_, the constant capital of both departments, must increase not only absolutely but also relatively to _v + c_ or the total new value (the social aspect of labour productivity); at the same time, constant capital and similarly the surplus value must increase relatively to the variable capital--in short, the rate of surplus value, i.e. the ratio between surplus value and variable capital, must similarly increase (the capitalist aspect of labour productivity). These changes need not, of course, occur annually, just as the terms of first, second and third year in Marx's diagram do not necessarily refer to calendar years but may stand for any given period. Finally, we may choose to a.s.sume that these alterations, both in the composition of capital and in the rate of surplus value, take place either in the first, third, fifth, seventh year, etc., or in the second, sixth and ninth year, etc. The important thing is only that they are allowed for somewhere and taken into account as periodical phenomena. If the diagram is amended accordingly, the result of this method of acc.u.mulation will be an increasing annual surplus in the consumer at the expense of producer goods. It is true that Tugan Baranovski conquers all difficulties on paper: he simply constructs a diagram with different proportions where year by year the variable capital decreases by 25 per cent. And since this arithmetical exercise is successful enough on paper, Tugan triumphantly claims to have 'proved' that acc.u.mulation runs smoothly like clockwork, even if the absolute volume of consumption decreases. Even he must admit in the end, however, that his a.s.sumption of such an absolute decrease of the variable capital is in striking contrast to reality. Variable capital is in point of fact a growing quant.i.ty in all capitalist countries; only in relation to the even more rapid growth of constant capital can it be said to decrease. On the basis of what is actually happening, namely a greater yearly increase of constant capital as against that of variable capital, as well as a growing rate of surplus value, discrepancies must arise between the material composition of the social product and the composition of capital in terms of value. If, instead of the unchanging proportion of 5 to 1 between constant and variable capital, proposed by Marx's diagram, we a.s.sume for instance that this increase of capital is accompanied by a progressive readjustment of its composition, the proportion between constant and variable in the second year being 6 to 1, in the third year 7 to 1, and in the fourth year 8 to 1--if we further a.s.sume that the rate of surplus value also increases progressively in accordance with the higher productivity of labour so that, in each case, we have the same amounts as those of the diagram, although, because of the relatively decreasing variable capital, the rate of surplus value does not remain constant at the original 100 per cent--and if finally we a.s.sume that one-half of the appropriated surplus value is capitalised in each case (excepting Department II where capitalisation exceeds 50 per cent, 184 out of 285 being capitalised during the first year), the result will be as follows:

1st year: I. _5,000c + 1,000v + 1,000s = 7,000_ means of production II. _1,430c + 285v + 285s = 2,000_ means of subsistence

2nd year: I. _(5,428 4/7)c + (1,071 3/7)v + 1,083s = 7,583_ means of production II. _(1,587 5/7)c + (311 2/7)v + 316s = 2,215_ means of subsistence

3rd year: I. _5,903c + 1,139v + 1,173s = 8,215_ means of production II. _1,726c + 331v + 342s = 2,399_ means of subsistence

4th year: I. _6,424c + 1,205v + 1,271s = 8,900_ means of production II. _1,879c + 350v + 371s = 2,600_ means of subsistence

If this were a true picture of the acc.u.mulative process, the means of production (constant capital) would show a deficit of 16 in the second year, of 45 in the third year and of 88 in the fourth year; similarly, the means of subsistence would show a surplus of 16 in the second year, of 45 in the third year and of 88 in the fourth year.

This negative balance for the means of production may be only imaginary in part. The increasing productivity of labour ensures that the means of production grow faster in bulk than in value, in other words: means of production become cheaper. As it is use value, i.e. the material elements of capital, which is relevant for technical improvements of production, we may a.s.sume that the quant.i.ty of means of production, in spite of their lower value, will suffice for progressive acc.u.mulation up to a certain point. This phenomenon amongst others also checks the actual decline of the rate of profit and modifies it to a mere tendency, though our example shows that the decline of the profit rate would not only be r.e.t.a.r.ded but rather completely arrested. On the other hand, the same fact indicates a much larger surplus of unsaleable means of subsistence than is suggested by the amount of this surplus in terms of value. In that case, we should have to compel the capitalists of Department II to consume this surplus themselves, which Marx makes them do on other occasions; in which case, and in so far as those capitalists are concerned, there would again be no acc.u.mulation but rather simple reproduction. Alternatively, we should have to p.r.o.nounce this whole surplus unsaleable.

Yet would it not be very easy to make good this loss in means of production which results from our example? We need only a.s.sume that the capitalists of Department I capitalise their surplus value to a greater extent. Indeed, there is no valid reason to suppose, as Marx did, that the capitalists in each case add only half their surplus value to their capital. Advances in labour productivity may well lead to progressively increasing capitalisation of surplus value. This a.s.sumption is the more permissible in that the cheapening of consumer goods for the capitalist cla.s.s, too, is one of the consequences of technological progress. The relative decrease in the value of consumable income (as compared with the capitalised part) may then permit of the same or even a higher standard of living for this cla.s.s. We might for instance make good the deficit in producer goods by transferring a corresponding part of surplus value I to the constant capital of this department, a part which would otherwise be consumed, since this surplus value, like all other products of the department, originally takes the form of producer goods; 11 4/7 would then be transferred in the second year, 34 in the third year and 66 in the fourth year.[340] The solution of one difficulty, however, only adds to another. It goes without saying that if the capitalists of Department I relatively restrict their consumption for purposes of acc.u.mulation, there will be a proportionately greater unsaleable residue of consumer goods in Department II; and thus it becomes more and more impossible to enlarge the constant capital even on its previous technological basis. If the capitalists in Department I relatively restrict their consumption, the capitalists of Department II must relatively expand their personal consumption in proportion. The a.s.sumption of accelerated acc.u.mulation in Department I would then have to be supplemented by that of r.e.t.a.r.ded acc.u.mulation in Department II, technical progress in one department by regression in the other.

These results are not due to mere chance. The adjustments we have tried out on Marx's diagram are merely meant to ill.u.s.trate that technical progress, as he himself admits, must be accompanied by a relative growth of constant as against variable capital. Hence the necessity for a continuous revision of the ratio in which capitalised surplus value should be allotted to _c_ and _v_ respectively. In Marx's diagram, however, the capitalists are in no position to make these allocations at will, since the material form of their surplus value predetermines the forms of capitalisation. Since, according to Marx's a.s.sumption, all expansion of production proceeds exclusively by means of its own, capitalistically produced means of production and subsistence,--since there are here no other places and forms of production and equally no other consumers than the two departments with their capitalists and workers,--and since, on the other hand, the smooth working of the acc.u.mulative process depends on that circulation should wholly absorb the aggregate product of both departments, the technological shape of enlarged reproduction is in consequence strictly prescribed by the material form of the surplus product. In other words: according to Marx's diagram, the technical organisation of expanded production can and must be such as to make use of the aggregate surplus value produced in Departments I and II. In this connection we must bear in mind also that both departments can obtain their respective elements of production only by means of mutual exchange. Thus the allocation to constant or variable capital of the surplus value earmarked for capitalisation, as well as the allotment of the additional means of production and subsistence (for the workers) to Departments I and II is given in advance and determined by the relations between the two departments of the diagram--both in material and in terms of value. These relations themselves, however, reflect a quite determinate technical organisation of production. This implies that, on the a.s.sumptions of Marx's diagram, the techniques of production given in each case predetermine the techniques of the subsequent periods of enlarged reproduction, if acc.u.mulation continues. a.s.suming, that is to say, in accordance with Marx's diagram, that the expansion of capitalist production is always performed by means of the surplus value originally produced in form of capital, and further--or rather, conversely--that acc.u.mulation in one department is strictly dependent on acc.u.mulation in the other, then no change in the technical organisation of production can be possible in so far as the relation of _c_ to _v_ is concerned.

We may put our point in yet another way: it is clear that a quicker growth of constant as compared with variable capital, i.e. the progressive metamorphosis of the organic composition of capital, must take the material form of faster expansion of production in Department I as against production in Department II. Yet Marx's diagram, where strict conformity of the two departments is axiomatic, precludes any such fluctuations in the rate of acc.u.mulation in either department. It is quite legitimate to suppose that under the technical conditions of progressive acc.u.mulation, society would invest ever increasing portions of the surplus value earmarked for acc.u.mulation in Department I rather than in Department II. Both departments being only branches of the same social production--supplementary enterprises, if you like, of the 'aggregate capitalist',--such a progressive transfer, for technical reasons, from one department to the other of a part of the acc.u.mulated surplus value would be wholly feasible, especially as it corresponds to the actual practice of capital. Yet this a.s.sumption is possible only so long as we envisage the surplus value earmarked for capitalisation purely in terms of value. The diagram, however, implies that this part of the surplus value appears in a definite material form which prescribes its capitalisation. Thus the surplus value of Department II exists as means of subsistence, and since it is as such to be only realised by Department I, this intended transfer of part of the capitalised surplus value from Department II to Department I is ruled out, first because the material form of this surplus value is obviously useless to Department I, and secondly because of the relations of exchange between the two departments which would in turn necessitate an equivalent transfer of the products of Department I into Department II.

It is therefore downright impossible to achieve a faster expansion of Department I as against Department II within the limits of Marx's diagram.

However we may regard the technological alterations of the mode of production in the course of acc.u.mulation, they cannot be accomplished without upsetting the fundamental relations of Marx's diagram.

And further: according to Marx's diagram, the capitalised surplus value is in each case immediately and completely absorbed by the productive process of the following period, for, apart from the portion earmarked for consumption, it has a natural form which allows of only one particular kind of employment. The diagram precludes the cashing and h.o.a.rding of surplus value in monetary form, as capital waiting to be invested. The free monetary forms of private capital, in Marx's view, are first the money deposited gradually against the wear and tear of the fixed capital, for its eventual renewal; and secondly those amounts of money which represent realised surplus value but are still too small for investment. From the point of view of the aggregate capital, both these sources of free money capital are negligible. For if we a.s.sume that even a portion of the social surplus value is realised in monetary form for purposes of future investment, then at once the question arises: who has bought the material items of this surplus value, and who has provided the money? If the answer is: other capitalists, of course,--then, seeing that the capitalist cla.s.s is represented in the diagram by the two departments, this portion of the surplus value must also be regarded as invested _de facto_, as employed in the productive process. And so we are back at immediate and complete investment of the surplus value.

Or does the freezing of one part of the surplus value in monetary form in the hands of certain capitalists mean that other capitalists will be left with a corresponding part of that surplus product in its material form? does the h.o.a.rding of realised surplus value by some imply that others are no longer able to realise their surplus value, since the capitalists are the only buyers of surplus value? This would mean, however, that the smooth course of reproduction and similarly of acc.u.mulation as described in the diagram would be interrupted. The result would be a crisis, due not to over-production but to a mere intention to acc.u.mulate, the kind of crisis envisaged by Sismondi.

In one pa.s.sage of his _Theories_,[341] Marx explains in so many words that he 'is not at all concerned in this connection with an acc.u.mulation of capital greater than can be used in the productive process and might lie idle in the banks in monetary form, with the consequence of lending abroad'. Marx refers these phenomena to the section on compet.i.tion. Yet it is important to establish that his diagram veritably precludes the formation of such additional capital. Compet.i.tion, however wide we may make the concept, obviously cannot create values, nor can it create capitals which are not themselves the result of the reproductive process.

The diagram thus precludes the expansion of production by leaps and bounds. It only allows of a gradual expansion which keeps strictly in step with the formation of the surplus value and is based upon the ident.i.ty between realisation and capitalisation of the surplus value.

For the same reason, the diagram presumes an acc.u.mulation which affects both departments equally and therefore all branches of capitalist production. It precludes expansion of the demand by leaps and bounds just as much as it prevents a one-sided or precocious development of individual branches of capitalist production.

Thus the diagram a.s.sumes a movement of the aggregate capital which flies in the face of the actual course of capitalist development. At first sight, two facts are typical for the history of the capitalist mode of production: on the one hand the periodical expansion of the whole field of production by leaps and bounds, and on the other an extremely unequal development of the different branches of production. The history of the English cotton industry from the first quarter of the eighteenth to the seventies of the nineteenth century, the most characteristic chapter in the history of the capitalist mode of production, appears quite inexplicable from the point of view of Marx's diagram.

Finally, the diagram contradicts the conception of the capitalist total process and its course as laid down by Marx in _Capital_, volume iii.

This conception is based on the inherent contradiction between the unlimited expansive capacity of the productive forces and the limited expansive capacity of social consumption under conditions of capitalist distribution. Let us see how Marx describes this contradiction in detail in chapter 15 on 'Unravelling the Internal Contradictions of the Law'

(of the declining profit rate):

'The creation of surplus-value, a.s.suming the necessary means of production, or sufficient acc.u.mulation of capital, to be existing, finds no other limit but the labouring population, when the rate of surplus-value, that is, the intensity of exploitation, is given; and no other limit but the intensity of exploitation, when the labouring population is given. And the capitalist process of production consists essentially of the production of surplus-value, materialised in the surplus-product, which is that aliquot portion of the produced commodities, in which unpaid labour is materialised. It must never be forgotten, that the production of this surplus-value--and the re-conversion of a portion of it into capital, or acc.u.mulation, forms an indispensable part of this production of surplus-value--is the immediate purpose and the compelling motive of capitalist production. It will not do to represent capitalist production as something which it is not, that is to say, as a production having for its immediate purpose the consumption of goods, or the production of means of enjoyment for the capitalists. (And, of course, even less for the worker. R. L.) This would be overlooking the specific character of capitalist production, which reveals itself in its innermost essence. The creation of this surplus-value is the object of the direct process of production, and this process has no other limits than those mentioned above. As soon as the available quant.i.ty of surplus-value has been materialised in commodities, surplus-value has been produced. But this production of surplus-value is but the first act of the capitalist process of production, it merely terminates the act of direct production. Capital has absorbed so much unpaid labour. With the development of the process, which expresses itself through a falling tendency of the rate of profit, the ma.s.s of surplus-value thus produced is swelled to immense dimensions. Now comes the second act of the process. The entire ma.s.s of commodities, the total product, which contains a portion which is to reproduce the constant and variable capital as well as a portion representing surplus-value, must be sold. If this is not done, or only partly accomplished, or only at prices which are below the prices of production, the labourer has been none the less exploited, but his exploitation does not realise as much for the capitalist. It may yield no surplus-value at all for him, or only realise a portion of the produced surplus-value, or it may even mean a partial or complete loss of his capital. The conditions of direct exploitation and those of the realisation of surplus-value are not identical. They are separated logically as well as by time and s.p.a.ce. The first are only limited by the productive power of society, the last by the proportional relations of the various lines of production and by the consuming power of society. This last-named power is not determined either by the absolute productive power or by the absolute consuming power, but by the consuming power based on antagonistic conditions of distribution, which reduces the consumption of the great ma.s.s of the population to a variable minimum within more or less narrow limits. The consuming power is furthermore restricted by the tendency to acc.u.mulate, the greed for an expansion of capital and a production of surplus-value on an enlarged scale. This is a law of capitalist production imposed by incessant revolutions in the methods of production themselves, the resulting depreciation of existing capital, the general compet.i.tive struggle and the necessity of improving the product and expanding the scale of production, for the sake of self-preservation and on penalty of failure.

The market must, therefore, be continually extended, so that its interrelations and the conditions regulating them a.s.sume more and more the form of a natural law independent of the producers and become ever more uncontrollable. This eternal contradiction seeks to balance itself by an expansion of the outlying fields of production. But to the extent that the productive power develops, it finds itself at variance with the narrow basis on which the conditions of consumption rest. On this self-contradictory basis it is no contradiction at all that there should be an excess of capital simultaneously with an excess of population. For while a combination of these two would indeed increase the ma.s.s of the produced surplus-value, it would at the same time intensify the contradiction between the conditions under which this surplus-value is produced and those under which it is realised.'[342]

If we compare this description with the diagram of enlarged reproduction, the two are by no means in conformity. According to the diagram, there is no inherent contradiction between the production of the surplus value and its realisation, rather, the two are identical.

The surplus value here from the very beginning comes into being in a natural form exclusively designed for the requirements of acc.u.mulation.

In fact it leaves the place of production in the very form of additional capital, that is to say it is capable of realisation in the capitalist process of acc.u.mulation. The capitalists, as a cla.s.s, see to it in advance that the surplus value they appropriate is produced entirely in that material form which will permit and ensure its employment for purposes of further acc.u.mulation. Realisation and acc.u.mulation of the surplus value here are both aspects of the same process, they are logically identical. Therefore according to the presentation of the reproductive process in the diagram, society's capacity to consume does not put a limit to production. Here production automatically expands year by year, although the capacity of society for consumption has not gone beyond its 'antagonistic conditions of distribution'. This automatic continuation of expansion, of acc.u.mulation, truly is the 'law of capitalist production ... on penalty of failure'. Yet according to the a.n.a.lysis in volume iii, 'the market must, therefore, be continually extended', 'the market' obviously transcending the consumption of capitalists and workers. And if Tugan Baranovski interprets the following pa.s.sage 'this eternal contradiction seeks to balance itself by an expansion of the outlying fields of production' as if Marx had meant production itself by 'outlying fields of production', he violates not only the spirit of the language but also Marx's clear train of thought.

The 'outlying fields of production' are clearly and unequivocally not production itself but consumption which 'must be continually extended'.

The following pa.s.sage in _Theorien uber den Mehrwert_, amongst others, sufficiently shows that Marx had this in mind and nothing else: 'Ricardo therefore consistently denies the necessity for an _expansion of the market_ to accompany the expansion of production and the growth of capital. The entire capital existing within a country can also be profitably used in that country. He therefore argues against Adam Smith who had set up his (Ricardo's) opinion on the one hand but also contradicted it with his usual sure instinct.'[343]

In yet another pa.s.sage, Marx clearly shows that Tugan Baranovski's notion of production for production's sake is wholly alien to him: 'Besides, we have seen in volume ii part iii that a continuous circulation takes place between constant capital and constant capital (even without considering any accelerated acc.u.mulation), which is in so far independent of individual consumption, as it never enters into such consumption, but which is nevertheless definitely limited by it, because the production of constant capital never takes place for its own sake, but solely because more of this capital is needed in those spheres of production whose products pa.s.s into individual consumption.'[344]

Admittedly, in the diagram in volume ii, Tugan Baranovski's sole support, market and production coincide--they are one and the same.

Expansion of the market here means extended production, since production is said to be its own exclusive market--the consumption of the workers being an element of production, i.e. the reproduction of variable capital. Therefore the limit for both the expansion of production and the extension of the market is one and the same: it is given by the volume of the social capital, or the stage of acc.u.mulation already attained. The greater the quant.i.ty of surplus value that has been extracted in the natural form of capital, the more can be acc.u.mulated; and the greater the volume of acc.u.mulation, the more surplus value can be invested in its material form of capital, i.e. the more can be realised. Thus the diagram does not admit the contradiction outlined in the a.n.a.lysis of volume iii. In the process described by the diagram there is no need for a continual extension of the market beyond the consumption of capitalists and workers, nor is the limited social capacity for consumption an obstacle to the smooth course of production and its unlimited capacity for expansion. The diagram does indeed permit of crises but only because of a lack of proportion within production, because of a defective social control over the productive process. It precludes, however, the deep and fundamental antagonism between the capacity to consume and the capacity to produce in a capitalist society, a conflict resulting from the very acc.u.mulation of capital which periodically bursts out in crises and spurs capital on to a continual extension of the market.

FOOTNOTES:

[332] _Capital_, vol. i, pp. 593-4.

[333] Ibid., p. 594, note 1.

[334] Op. cit., vol. ii, p. 384.

[335] Ibid., pp. 400-1.

[336] Ibid., p. 488.

[337] _Capital_, vol. iii, p. 568.

[338] _Theorien_ ..., vol. ii, part 2, 'The Acc.u.mulation of Capital and Crises', p. 263.

[339] 'It is never the original thinkers who draw the absurd conclusions. They leave that to the Says and MacCullochs' (_Capital_, vol. ii, p. 451).--And--we might add--to the Tugan Baranovskis.

[340] The figures result from the difference between the amounts of constant capital in Department I under conditions of technical progress, and under Marx's stable conditions.

[341] _Theorien uber den Mehrwert_, vol. ii, part 2, p. 252.

[342] _Capital_, vol. iii, p. 285 ff.

[343] _Theorien_ ..., vol. ii, part 2, p. 305.

[344] _Capital_, vol. iii, p. 359.

_CHAPTER XXVI_

THE REPRODUCTION OF CAPITAL AND ITS SOCIAL SETTING

Marx's diagram of enlarged reproduction cannot explain the actual and historical process of acc.u.mulation. And why? Because of the very premises of the diagram. The diagram sets out to describe the acc.u.mulative process on the a.s.sumption that the capitalists and workers are the sole agents of capitalist consumption. We have seen that Marx consistently and deliberately a.s.sumes the universal and exclusive domination of the capitalist mode of production as a theoretical premise of his a.n.a.lysis in all three volumes of _Capital_. Under these conditions, there can admittedly be no other cla.s.ses of society than capitalists and workers; as the diagram has it, all 'third persons' of capitalist society--civil servants, the liberal professions, the clergy, etc.--must, as consumers, be counted in with these two cla.s.ses, and preferably with the capitalist cla.s.s. This axiom, however, is a theoretical contrivance--real life has never known a self-sufficient capitalist society under the exclusive domination of the capitalist mode of production. This theoretical device is perfectly admissible so long as it merely helps to demonstrate the problem in its integrity and does not interfere with its very conditions. A case in point is the a.n.a.lysis of simple reproduction of the aggregate social capital, where the problem itself rests upon a fiction: in a society producing by capitalist methods, i.e. a society which creates surplus value, the whole of the latter is taken to be consumed by the capitalists who appropriate it. The object is to present the forms of social production and reproduction under these given conditions. Here the very formulation of the problem implies that production knows no other consumers than capitalists and workers and thus strictly conforms to Marx's premise: universal and exclusive domination of the capitalist mode of production.

The implications of both fictions are the same. Similarly, it is quite legitimate to postulate absolute dominance of capital in an a.n.a.lysis of the acc.u.mulation of individual capitals, such as is given in _Capital_, volume i. The reproduction of individual capitals is an element in total social reproduction but one which follows an independent course, contrary to the movements of the other elements. In consequence it will not do simply to take together the individual movements of the respective capitals in order to arrive at the total movement of social capital, since the latter is essentially different. The natural conditions of reproducing individual capitals therefore neither conform with one another, nor do they conform to the relations of the total capital. Under normal conditions of circulation, every individual capital engages in the process of circulation and of acc.u.mulation entirely on its own account, depending upon others only in so far, of course, as it is compelled to find a market for its product and must find available the means of production it requires for its specific activities. Whether the strata who afford this market and provide the necessary means of production are themselves capitalist producers or not is completely immaterial for the individual capital, although, in theory, the most favourable premise for a.n.a.lysing the acc.u.mulation of individual capital is the a.s.sumption that capitalist production has attained universal and exclusive domination and is the sole setting of this process.[345]

Now, however, the question arises whether the a.s.sumptions which were decisive in the case of individual capital, are also legitimate for the consideration of aggregate capital.

'We must now put the problem in this form: _given universal acc.u.mulation_, that is to say provided that in all branches of production there is greater or less acc.u.mulation of capital--which in fact is a condition of capitalist production, and which is just as natural to the capitalist _qua_ capitalist as it is natural to the miser to ama.s.s money (but which is also necessary for the perpetuation of capitalist production)--what are the _conditions_ of this universal acc.u.mulation, to what elements can it be reduced?'

And the answer: '_The conditions for the acc.u.mulation of capital are precisely those which rule its original production and reproduction in general_: these conditions being that one part of the money buys labour and the other commodities (raw materials, machinery, etc.) ...

Acc.u.mulation of new capital can only proceed therefore under the same conditions under which already existing capital is reproduced.'[346]

In real life the actual conditions for the acc.u.mulation of the aggregate capital are quite different from those prevailing for individual capitals and for simple reproduction. The problem amounts to this: If an increasing part of the surplus value is not consumed by the capitalists but employed in the expansion of production, what, then, are the forms of social reproduction? What is left of the social product after deductions for the replacement of the constant capital cannot, _ex hypothesi_, be absorbed by the consumption of the workers and capitalists--this being the main aspect of the problem--nor can the workers and capitalists themselves realise the aggregate product. They can always only realise the variable capital, that part of the constant capital which will be used up, and the part of the surplus value which will be consumed, but in this way they merely ensure that production can be renewed on its previous scale. The workers and capitalists themselves cannot possibly realise that part of the surplus value which is to be capitalised. Therefore, the realisation of the surplus value for the purposes of acc.u.mulation is an impossible task for a society which consists solely of workers and capitalists. Strangely enough, all theorists who a.n.a.lysed the problem of acc.u.mulation, from Ricardo and Sismondi to Marx, started with the very a.s.sumption which makes their problem insoluble. A sure instinct that realisation of the surplus value requires 'third persons', that is to say consumers other than the immediate agents of capitalist production (i.e. workers and capitalists) led to all kinds of subterfuges: 'unproductive consumption' as presented by Malthus in the person of the feudal landowner, by Vorontsov in militarism, by Struve in the 'liberal professions' and other hangers-on of the capitalist cla.s.s; or else foreign trade is brought into play which proved a useful safety valve to all those who regarded acc.u.mulation with scepticism, from Sismondi to Nicolayon. Because of these insoluble difficulties, others like v. Kirchmann and Rodbertus tried to do without acc.u.mulation altogether, or, like Sismondi and his Russian 'populist' followers, stressed the need for at least putting the dampers on acc.u.mulation as much as possible.