Norbert Wiener, in the early 1960s, foresaw a parallel between the process of automation and the nature of magic as it has been depicted in countless fantasies, from Goethe’s tale of the sorcerer’s apprentice to W.W. Jacob’s Monkey’s Paw. The characteristic of magic in these stories is its literalmindedness. It grants to magic-users precisely what they ask, but this, in the end, is never what they intend or desire. The use of magic therefore brings ‘with it a host of unimagined dangers. ‘Automation,’ observed Wiener ‘. . . may be expected to be similarly literal-minded.’footnote1 In the mid-1980s that prediction seems particularly apt. For the past two decades, managers in advanced capitalist countries have regarded computer-based automation as the magical cure for two of their most pressing problems: the compulsion to reduce costs (particularly labour costs), and the need to increase the authority and control of the enterprise over its workforce.footnote2 In their pursuit of higher levels of automation, however, they have inadvertently initiated profound changes in the workings of the capitalist world economy itself:
Automation, and the resultant economic changes, have also posed problems for the Left in the industrialized world. Social crises induced by automation have not resulted in an upsurge of radical political activity, but seem, more often, to have produced apathy, anomie and despair. The participation of workers in trade union movements is declining in many countries, while the most successful political offensives of the 1980s have come from the radical Right rather than from the Left.
One reason for this disarray lies, I believe, in the Left’s failure to apply serious analysis to the economic transformations implicit in the growth of automation and in the movement of large enterprises into information-producing activities. The most popularly influential interpretations have come from neo-conservatives who envisage a utopian information society where judicious use of technology will spontaneously resolve the contradictions of industrial capitalism.footnote3 In response to these fantasies, left-wing writers have commonly taken one of two contrasting positions: either they have denied that the contemporary ‘information revolution’ represents any fundamental change in the nature of capitalism, or they have argued that it spells the death agony of the capitalist system. Neither of these positions seems to me to be tenable. What is needed is, rather, new and critical perspectives on the profound economic changes through which we are living—even if, to quote Wiener again, this involves ‘a real risk of heresy’.footnote4
In an earlier article,footnote5 I tried to take a few steps in this direction by arguing that the growth of the so-called ‘information economy’ could be explained in terms of the labour theory of value. My starting point was Ernest Mandel’s hypothesis that widespread automation represents the ‘inner limit of capitalism’, since the reduced use of living labour power in production will ultimately make it impossible for enterprises to extract the surplus value which they need for survival and growth.footnote6 In contrast to this view, I suggested that, in a highly automated economy, surplus-value creation could be maintained, over a fairly extended period at least, by the channelling of living labour into ‘the incessant generation of new products and new methods of production’. This would explain why the spread of automation in most advanced industrialized countries has been accompanied by a so-called ‘softening
My arguments were subsequently challenged by Ian Steedman,footnote7 who accepted the urgency of discussing the consequences of automation but regarded the labour theory of value as an inappropriate framework. To demonstrate this, Steedman turned his attention to Mandel’s vision of an entirely automated economy. The traditional Marxian picture of total automation is one in which no labour is performed, and therefore no value is produced. Capitalism, and indeed economic activity in any meaningful sense of the word, ceases to exist. Steedman proposed, on the contrary, that by applying a simple, Sraffa-type model to a wholly automated economy, one could demonstrate that profits would continue to be generated. ‘What is revealed by full automation,’ he wrote, ‘is not the “inner limit” of capitalism, but rather the “inner limit” of the labour theory of value and surplus value theorizing.’footnote8
One response to Steedman’s comments might be to embark on a lengthy defence of the labour theory of value, but I do not intend to do so. The debate between neo-Ricardians and adherents of the traditional Marxian labour theory of value has been fought out at great length and on many battlefields.footnote9 From time to time, indeed, it has seemed in danger of sinking into a quagmire of semantic confusions. In this article, I should like instead to use some general comments on Steedman’s position as a basis for taking a few more steps along the path of a critique of the ‘information society’. These steps, as it turns out, point towards a conclusion which is at once intriguing and disturbing: namely, that neither neo-Ricardianism nor the traditional labour theory of value provides a wholly adequate basis for an analysis of contemporary economic change.
There are two fundamentally different approaches to the study of economic systems. The first, drawing on the paradigm of Newtonian physics, tries to disclose unchanging laws governing the workings of the economy in all phases of human development; the second, drawing on the evolutionary paradigm, recognizes the existence of radical discontinuities, in which new systems governed by new sets of laws evolve out of the structures of the old. One of Marx’s greatest contributions to economic thought was the coherent articulation which he gave to the second point of view. Sraffa’s analysis of the economy is also clearly, if less explicitly, grounded upon the specificity of social and economic structures. He assumes the existence of a capitalist system of production, in which a surplus is produced by human labour working within freely competing, privately owned economic units. The ideas which he