In my own lifetime of thirty years, global society has been transformed by the widespread availability of inexpensive computing technology. Indeed, only within the past ten years, a new combination of commoditised hardware, software, and network infrastructure has put this technology within reach of millions of new people. A certain taint of presentism is, therefore, inevitable in any attempt to write the history of computing in our time, as we are positioned at a particular point in a dynamic of ongoing social and technical change. As with earlier historians of the “industrial revolution”, we must assess the historicity of the information or “digital revolution” both as historical narratives and popular common sense.
This presentism presents particular challenges to the historian in his or her craft of framing a coherent narrative of technological development. Here I will consider different approaches to the history of computing which confront both the the familiar challenges of a historian of technology, as well as the unique aspects of computing as an object of historical inquiry. In the introduction to his A History of Modern Computing, Paul Ceruzzi discusses two distinct approaches to the history of computing, what he calls the technological systems approach and the social constructionist approach. What are the objects of inquiry of these two approaches?
Derived from Thomas P. Hughes landmark study of large technological systems, a “systems approach” must emphasise the connectivity and interdependency of an entire industrial ecology on which modern computing practices depends for its effective technical operation and social organisation. Much like earlier large systems such as electrical industry, public utilities and power grids, or the automobile, steel and petroleum industries and the transport network, with computing the systems approach must account for today’s far-flung sites of hardware manufacturing, each of which its own scale and mode of organisation for semiconductors, chipsets, motherboards, displays, peripherals.
Such an approach must also account for the global network of metropolitan and offshore software production which has become the mainstay of present-day capitalism based on finance, insurance, banking and other information services, as well as a vibrant consumer market for commoditised personal hardware and software. Here our focus is not so much on the “computer” as a simple object to “computing” as a complex system. If large technological systems are the object of historical inquiry, it will be evident to the historian that their scale and complexity escapes the efforts of individuals and social groups to master them. Hughes’ systems approach to the history of technology is a narrative of the displacement of individual inventors and technologists in their workshops, by scientific researchers and experts in corporate and state-sponsored laboratories.
The locus of socio-technical change is now with these large bureaucratic and military structures, and not with entreprising and heroic individuals making new discoveries. The systems themselves structure social relations, until such time as the system is superseded or rendered obsolescent by a new technology. These moments become the real ruptures in the systems narrative, the technical “breakthroughs” in inventions such as the silicon chip, the integrated circuit, and the microprocessor. This form of history has a peculiar way of describing the development of technology, focussing its efforts on documenting the technical problems of storage and transmission, processing and speed to which these inventions were a response, and giving a privileged place in the narrative to those individuals and institutions who could overcome these systemic problems. Ceruzzi states it clearly that this approach describe[s] computing’s history as a series of technical problems met by engineering solutions that in hindsight seem natural and obvious. Individual and collective agency is assimilated to the needs of the system, and only given significance by the extent to which the system advances through these diffuse efforts. It is in this sense that large systems constitute structures which effect the relations of the parts to the whole, of fragments to the totality.
As opposed to the systems approach, what Ceruzzi calls the “social constructionist” approach emphasises the social context of and political negotiations over the development of computing, in that technological development takes place within social relations and is subject to the struggles and contradictions dominant in any social formation. The best example of this approach in an academic genre is in Donald MacKenzie’s work on missile guidance and targetting systems. Ceruzzi suggests that most academic histories of computing, however, ignore this approach in favour of the systems approach. Popular histories, on the other hand, inexplicitly adopt a social constructionist approach in their focus on the social and political values underlying the rise of personal computing. These heroic narratives, which begin the garages and workshops of Silicon Valley teenage entrepreneurs and hackers, and culminate in the rise of personal computer companies such as Apple Computer and Microsoft, became common-sense household myths in the nineties, propelled by the speculative mania of the first dot-com boom and sustained by magazines such as Wired.
The problem with this popular social constructionist approach is not only, as Ceruzzi points out, that it ignores the advances in solid-state electronics and the presence of large defence contractors in Silicon Valley, which formed the backdrop for the activities of groups of California hobbyists and entrepreneurs in the seventies, when the personal computer was born. The problem is also that hackers were not hippies. The sixties counterculture was unambiguous in its rejection of the system and the machine, a totalising metaphor for the domination of social life by the dead labour and mechanical routines of mass production.
Technological innovation often takes place in the competition and antagonism of social groups, and the story of the mass computing “revolution” in the eighties and nineties must account for these contradictions. The combination of the countercultural idealism of the sixties with the marketing of personal computers in the eighties and nineties has produced a powerful myth that sees a direct line from the streets of Haight-Ashbury and fields of Woodstock in the sixties, to the garages and start-up shops of Silicon Valley in the seventies, to millions of computer desktops across the world in the eighties and nineties. This is a powerful and seductive mythology that defeats efforts to understand the social antagonisms and struggles that defined the form of contemporary computing.
Situating the development of computing technologies in the social struggles of our time attempts to separate the social and technical aspects of technological change, and assign a primacy to the social, arguing that the technical is an expression of the social. Some of the most critical relationships in the history of computing, as Ceruzzi suggests, are relations between entrepreneurs and inventors, and the popular mythology obscures the contingent outcomes of the negotiation between these groups in the development of actually existing computing. It is in this sense that we can refer to socio-technical formations as the object of inquiry of the social constructionist approach to the history of computing, whereas the systems approach emphasises the complex interdependencies and relations of diverse parts and to the whole of computing as a technological system.
So far it seems as if the challenges in history of computing follow the well-worn debates between the relative primacy of structure and agency in historical representation. What are the implications of these two approaches for writing this history, and what aspects of the history of computing depart from the familar questions of structure and agency, determinism and autonomy? Computers are both fungible commodities as well as self-contained factories. Software is both a set of instructions (code) and a mode of social organisation. These peculiar characteristics of the modern computer assume significance in both the systems approach and the social constructionist approaches, with consequences for our understanding of structure and agency in historiography.
A good history of computing needs to both see computing as a technological system as well as a social formation, and must borrow from both historical traditions of writing about the origins and growth of technology. Fortunately, we have a broadly accepted periodisation of the early history of computing, organised around distinct shifts in the nature of computing as both technical object and social practice. These would be, firstly, the early era of mainframe computing originating in the second world war and until the early sixties; secondly the era of networked minicomputers from the early sixties to the mid-seventies; thirdly the era of personal computing, from the late-seventies to the early nineties. I want to posit that within this sequence of stages, that we identify our contemporary moment as one of social computing.
By social computing I wish to draw attention to the combination of aspects of the second and third stages identified above, the eras of networked and personal computing, with the advent of the public internet in the early nineties. This led to a rapid dissemination of networked computing amongst the earlier generation of personal computer users, mostly in North America, East Asia, and Western Europe, as well as a even larger number of completely new users who had never set hands on a computer before the commercialisation of the internet in the mid-nineties. It is with the advent of social computing that the public inquiry into the origins of computing, another way of describing the history of computing, assumes importance as an intellectual problem.