top of page

The secret history of automation

David Noble

Here are some notes on an unjustly neglected book by the great Marxist David F. Noble (social historian of technology) called Progress Without People: New Technology, Unemployment, and the Message of Resistance. It was published in 1995, but it was ahead of its time, since its warnings resonate even more profoundly today:

The information highway is barely under construction, the virtual workplace still largely experimental, but their consequences are readily predictable in the light of recent history. In the wake of five decades of information revolution, people are now working longer hours, under worsening conditions, with greater anxiety and stress, less skills, less security, less power, less benefits, and less pay. Information technology has clearly been developed and used during these years to deskill, discipline, and displace human labour in a global speed-up of unprecedented proportions. Those still working are the lucky ones. For the technology has been designed and deployed as well to tighten the corporate stranglehold on the world’s resources, with obvious and intended results: increasing dislocation and marginalization of a large proportion of the world’s population—within as well as without the industrial countries; growing structural (that is, permanent) unemployment and the attendant emergence of a nomadic army of temporary and part-time workers (the human complement of flexible production); a swelling of the ranks of the perpetually impoverished; and a dramatic widening of the gap between rich and poor to nineteenth-century dimensions.

All these tendencies have reached crisis levels by now, and even mainstream economists are warning that in the coming years automation will eliminate millions of jobs. –On the other hand, automation could make possible an unprecedented life of leisure for everyone—for instance, a four-day or three-day workweek—if we could only shift the priorities of government.

One of the uses of the book is as a reminder that it is the government, not the private sector, that’s responsible for the most important technological and scientific innovations. The only reason the capitalist economy works at all is that government (i.e., the public, the working and middle classes) shoulders most of the burden and the cost of innovation. The public sector and what you could call the “public-private sector” (involving public-private partnerships, most notably in high technology and science) is the most dynamic part of the economy. E.g.:

What mechanization was to the first industrial revolution, automation was to the second. The roots of the second industrial revolution lay in the state-sponsored technological developments of World War II. Military technologies—control systems for automatic gunfire, computers for ballistics and A-bomb calculations, microelectronics for proximity fuses, radar, computers, aircraft and missile guidance systems, and a host of sensing and measuring devices—gave rise to not only programmable machinery but also “intelligent” or self-correcting machinery. In the postwar years, the promotion of such technologies was fuelled by Cold War concerns about “national security,” the enthusiasm of technical people, management’s quest for a solution to its growing labour problems, and by a general cultural offensive to restore confidence in scientific salvation and technological deliverance following the twin traumas of depression and global war. Often with state initiative and subsidy, industrial application of these new technologies (as well as an intensification of older forms of fixed automation and mechanization) began to take hold, in steel, auto, petroleum refining, chemical processing (and uranium enrichment), and aircraft, machinery, and electrical equipment manufacture, among others (pp. 24–25).

The economy was growing healthily in the postwar era, so the slow and gradual introduction of automation didn’t lead to crises of unemployment or massive worker resistance. It started to by the late 1960s, though. “The increasing displacement, deskilling, and disciplining of workers in industry proceeded apace, largely unnoticed except by the workers themselves until, by the end of the 1960s, the situation exploded in an upsurge of pent-up rank-and-file militancy.” All across the Western world, the ruling class had to put down an epidemic of rank-and-file resistance, often violently, in the 1970s. Layoffs, the degradation of work, and speed-ups resulted in countless wildcat strikes and relatively new forms of working-class direct action such as “shop-floor organization, counter-planning strategies against management, rank-and-file caucuses against union leadership, and systematic sabotage.” North America and Western Europe were riven by industrial conflict.

One of the responses to all this discontent was to partially co-opt it. “The managers of some companies experimented with new methods—so-called job enrichment, job enlargement, and quality of worklife schemes—designed to absorb discontent and redirect energies along more productive paths. Sweden was a centre for such experimentation and became a model throughout the industrialized world.” In the end, most of workers’ victories from these schemes of “participation” and “job-enrichment” were very limited and short-lived. “In the wake of these limited gains, the rebellious [rank-and-file] energies that had brought them about dissipated and all but disappeared. In their place arose committees, rules, agreements, and other formal devices for dealing with the new challenges at the workplace, including the challenge of new technology.” The technology issue was formalized, bureaucratized, regulated in the form, e.g., of new contract language between unions and employers pertaining to the introduction of new technology. “Whatever these gains, however, they were achieved at the expense of removing the technology issue from the shop floor and thus from the realm of direct action available to the workers themselves. ‘With increasing formalization,’ [one sociologist] observed, ‘the spread of sabotage could once again be held in check by pressure from trade union organizations opposed to it.’”

As has often happened, then, union leadership cooperated with business to suppress and defuse rank-and-file discontent. Resistance lasted into the 1980s, as workers in Europe and North America engaged in sabotage of new technology that threatened to put them out of a job, deskill their work, and give ever greater powers of surveillance to management, but ultimately the rolling tide of automation—organized and implemented so as to benefit management, not workers—proved unstoppable. So here we are today, on the precipice.


Noble is an expert on the social history of technology, and he continually insists on the fact that automation has been designed in such a way as to benefit management, not workers. It isn’t some “automatic,” “natural,” or apolitical process by which the best and most efficient technological designs are chosen by engineers, adopted by businessmen, and then tested (for productivity, efficiency, profitability) in the crucible of the marketplace. Technical development “is not some abstractly rational enterprise with an internal logic all its own, but rather a human effort that reflects at every turn the relations of power in society.”

Engineers simply want to do what’s good for society, “yet, consistently, again and again, they turn out [technical] solutions that are good for the people in power (management) but often disastrous for the rest of us (workers). Can this be explained?” Yes, it can! “For one thing, few technical people have any contact whatsoever with workers; in their education and their professional careers, they typically communicate only with management. Not surprisingly, they tend to view the world pretty much as management does, whether they know it or not. They are taught and usually believe that this is simply the most objective way of looking at things, but it is, in reality, the view from the top, the perspective of those with power.”

Noble gives a couple of examples. Here’s one:

For seven years I investigated the history of automated machine tools. Much of the pioneering design and development work of these tools took place at MIT, and I spent many months pouring over the vast collection of documents from the ten-year project. I discovered that the engineers involved in creating this self-professed revolution in metal-working manufacturing had been in constant contact with industrial managers and military officers, who had sponsored and monitored the project. Yet I found not a single piece of paper indicating that there had been contact with any of the many thousands of men and women who work as machinists in the metal-working industry—those most knowledgeable about metal-cutting and…those most directly affected by the technical changes under development… The engineering effort was essentially a management effort, and the resulting technology reflected this limited perspective—the worldview of those in power (p. 74).

In our society, an authoritarian pattern predominates in all institutions and workplaces. “So when an engineer begins to design a top-down technical system, he reasonably assumes from the outset that the social power of management will be available to make his system functionable. Such authoritarian systems are also simpler to design than more democratic ones, since they entail fewer independent variables, and this also makes them more appealing to designers. Finally, authoritarian systems satisfy the engineer’s own will to control and offer the engineer a powerful place in the scheme of things. Thus, for all these reasons, new technical systems are conceived from the outset as authoritarian ones, perfectly suited for today’s world.”

Behind the history of industrial automation, Noble found not merely technical and economic considerations but rather three overriding impulses: “1) a management obsession with control; 2) a military emphasis upon command and performance; and 3) enthusiasms and compulsions that blindly fostered the drive for automaticity.”

So, first of all, managers “consistently solicit and welcome technologies that promise to enhance their power and minimize challenge to it, by enabling them to discipline, deskill (in order to reduce worker power as well as pay) and displace potentially recalcitrant workers. Perhaps more than any other single factor, this explains the historical trend toward capital-intensive production methods and ever more automatic machinery, which have typically been designed with such purposes in mind.” These purposes, and technologies designed in accord with them, are hardly new: they go back to the early days of the first industrial revolution. But let’s focus on more recent times:

In the late 1940s control engineers at MIT (who had just completed a rolling mill control system designed to enable Bethlehem Steel management to eliminate “pacing” by workers) turned their “fertile genius” to the metal-working industry. The ultimate result of their efforts, “numerical control” (NC), reflected management’s [priorities] and set the pattern for all subsequent development of what are now known as computer-aided manufacturing systems. As the very name suggests, control was and remains its essence, not just management control of machines but, through them, of machinists as well.

Noble quotes industry insiders on the nature of numerical control. For example: “With numerical control, there was a shift of control to management. The control over the machine was placed in the hands of management.” “I remember the fears that haunted industrial management in the l950s. There was the fear of losing management control over a corporate operation that was becoming ever more complex and unmanageable. Numerical control is restoring control of shop operations to management.” “Numerical control has been defined in many ways. But perhaps the most significant definition is that [it] is a means for bringing decision-making in many manufacturing operations closer to management. Since decision-making at the machine tool has been removed from the operator and is now in the form of pulses on the control media, [NC] gives maximum control of the machine to management.” Automation, in the ways it has been carried out, is therefore just another stage in the long history of “scientific management” (sometimes called Taylorism, though that makes it too narrow) that Harry Braverman wrote about in the 1970s.

In the history of automated machine tools, alternative designs, more democratic and worker-friendly than numerical control, were proposed, but they never got much funding from military or (later) industrial backers. “Thus, NC became the dominant and, ultimately, the only technology for automating metal-working.”

The perfect compatibility of the military mentality with the business mentality is interesting, and telling. Sorry for the long quotation, but it’s informative:

The military has always played a central role in the technological development of U.S. industry, from mining and metallurgy to shipping and navigation, from interchangeable parts manufacture to scientific management. As the army and navy have been the major movers in the past, the air force has led the way in our time… If we just consider today’s so-called high technology—electronics, computers, aerospace, cybernetics (automatic control), lasers—all are essentially military creations. When some of these war-generated technologies were brought together to automate the metal-working industry, the military was once again the driving force.
From the start in the late 1940s down to the present day, the air force has been and remains the major sponsor of industrial automation. With regard to numerical control, the air force underwrote the first several decades of research and development of both hardware and software, determined what the technology would ultimately look like by setting design specifications and criteria to meet military objectives, created an artificial market for the automated equipment by making itself the main customer and thereby generating demand, subsidized both machine-tool builders and industrial (primarily aerospace) users in the construction, purchase, and installation of the new equipment, and even paid them to learn how to run it.
Numerical control was just the beginning of air force involvement in the automation drive. The air force numerical-control project had global significance; on a recent visit to a locomotive factory in Prague, I was surprised to find the air force NC programming system in use even there. And before long, this single project had evolved into the more expansive Integrated Computer Aided Manufacturing Program. More recently, ICAM became the still more ambitious and diversified MANTECH (manufacturing technologies) programs, designed to promote the computer automated approach to manufacturing not only in industry but also in universities…
The effects of this military involvement reflect the peculiar characteristics of the military world. First and most obvious is the military emphasis upon command, the quintessence of the authoritarian approach to organization. This means, essentially, that subordinates must do as they are told, with no ifs, ands, or buts; the intent is to eliminate wherever possible any human intervention between the command (by the superior) and the execution (by the subordinate). It is easy to understand the military emphasis upon automation, given its potential for eliminating such intermediate steps… In the military outlook, an army of men behaving like machines is readily replaced by an army of machines. This command orientation neatly complements and powerfully reinforces the managerial obsession with control. If the business suit and the uniform are interchangeable in our day, so too are the minds that go with them.

As for the idea that the ostensibly “no-nonsense economic rationality of profit-seeking businessmen” is primarily what guides their adoption of new technologies, it’s largely false. Noble presents fascinating evidence that businessmen aren’t motivated mainly by the prospect of cost savings or higher profits in their adoption of automation. Rather, there are various cultural, social, and psychological pressures to get the newest cool thing, to follow the herd mentality and do what others are doing, to obey the supposedly objective momentum of technological progress even without calculating whether some new machine is profitable or might actually increase costs. It seems that rigorous studies of the cost effectiveness of new equipment—does it increase productivity or not?—are rarely conducted. Actually, “by means of creative accounting and sophisticated use of the tax laws, machines can mysteriously make money for their owners even if they don’t work or are never used.”

So the dominant ideology of the “inevitable laws of technological progress” and the hyper-rationality of businessmen’s economic genius is nonsense. New equipment is frequently unnecessary or costlier than what was used before.

What about the vaunted discipline of the market? Doesn’t this ensure that firms whose acquisitions of new equipment aren’t profitable will succumb to the competition? No. For one thing, the state, which is perfectly willing to spend extravagantly and wastefully, is largely responsible for the advance of automation. “Not only has it subsidized extravagant developments that the market could not or refused to bear, but it also absorbed excessive costs and thereby kept afloat those competitors who would otherwise have sunk.” When your primary customer is the government, you don’t have to worry about market discipline.

Moreover, defense-related industries “are expanding along with the military automation programs, as more and more businesses rush to this state-supported sanctuary to escape the unpredictable vicissitudes of the market. At the same time, the military automation programs are today being matched by those of civilian agencies such as the Department of Commerce, the National Science Foundation, and others. All have now become the publicly funded pushers of automation madness, charting a course and promoting a pace that no self-adjusting market, had it existed, would ever have tolerated.”

Economists tend to focus on “market logic” and ignore the influence of the state, so they write as if it’s the market that is impelling the “inexorable,” “automatic” adoption of automation. But, again, this is either false or highly oversimplified.


Noble dedicates chapter five to an effective critique of market worship, together with the “quasi-religious faith in the automatic beneficence of technological progress” that tends to go along with it. He summarizes the quasi-religious philosophy as follows:

It goes like this—people with money are offered incentives (the chance to make more money) to urge them to invest in a new, improved plant and equipment (so-called innovation). This innovation automatically yields increased productivity and, hence, lower costs and prices, which results in greater competitiveness. Finally, this enhanced competitiveness necessarily brings about what Adam Smith, the great eighteenth-century philosopher of capitalism, called the “wealth of nations”: economic growth, jobs, cheap and plentiful commodities, in short, prosperity.

At every point, this religious faith is questionable.

For example: “To begin with, the assumption that rich people will invest in new means of production if given sufficient lucrative incentives presupposes that they would not do so voluntarily without such inducements. As such, it is itself a tacit recognition of the inadequacy of the market as a stimulus to development.” What if it isn’t lucrative to invest in new means of production, new factories or businesses or whatnot? Then capitalists won’t invest in them. In recent decades, real estate and financial speculation have become more lucrative than productive investment. This is one reason for the deterioration of infrastructure, the decline of manufacturing, and the slow collapse of society. So much for the automatic beneficence of the market.

What about the link between innovation and productivity? “When investment does in fact generate innovation [as it usually doesn’t], does such innovation necessarily yield greater productivity? The assumption here is that the return of profits to the investor will be matched by more and cheaper goods for society.” He remarks that this assumption is “the cornerstone of apologies for capitalism, its central tenet of legitimation.” But it has become highly dubious. Even the business press admits that there’s a lot of investment in labor-saving technology designed to increase profits without necessarily adding to productive output.

In any case, with all the capital investment in recent decades and supposed advances in productivity, have you noticed any declines in prices? Hardly. If there have been gains from productivity, it seems that neither workers nor consumers have captured them. (You can check business headlines from recent years to confirm this.) So there goes another justification of capitalism.

Well, I’ll end the summarizing here, even though I haven’t discussed at all a very interesting chapter on the millennium-old (Christian) origins of the Western (and now global) “religion of technology” that has become hegemonic since the first industrial revolution. It’s a rich book, despite its brevity, and worth reading, in its totality amounting to a brilliant critique of this technology-worship (as if technology itself, regardless of social or political context, holds out salvation). Elsewhere, Noble wrote a cutting critique of the automation and commodification of higher education. It seems that his life’s mission, or one of them, was to deepen our understanding of the human meaning of technology and to critique and contextualize our blind reverence for it. As an antediluvian traditionalist, I can only wish more writers would adopt the same calling.


Recent Posts

See All


Thanks for submitting!

bottom of page