What the Enlightenment Got Wrong about Computers

The way we think about computing is fundamentally shaped by the nineteenth-century civil service. In his brilliant book, The Government Machine (MIT Press, 2003), Jon Agar shows how the computer was the product of a technocratic vision of government that developed from the late eighteenth century and sought to manage a rapidly changing world by gathering as much statistical and other information as possible.

The civil service that emerged in the Victorian period had at its heart an administrative and intellectual division of labour. Thousands of clerks were employed to carry out routine set tasks, such as copying letters and making arithmetical calculations. Middle managers controlled the flow of this routine work and dealt with any hitches. Senior civil servants (the ‘First Division’) dealt with complex matters of policy advice and development.

The computer arose from the dream that the routine clerical work of government could be mechanised. The vision of the nineteenth-century scientist Charles Babbage in developing mechanical engines whose functions anticipated aspects of our digital computers was to produce a machine that could perform the routine mathematical calculations carried out by clerks in offices such as the Admiralty. Much later, the aim of the scientists and technicians at Bletchley Park building early electronic computers was similarly to speed up the decryption of military cypher traffic.

The organisation of the civil service into hierarchies that split routine work and decision making reflects the principle of the division of labour, most famously described by the Scottish philosopher Adam Smith in his 1776 book, An Inquiry into the Nature and Causes of the Wealth of Nations. Smith describes how manufacturers increase their output by dividing the process of manufacture into different components, with each worker being a specialist in just one part of the process.

Adam Smith illustrated the division of labour by referring to the example of the manufacture of pins: ‘one man draws out the wire, another straights it, a third cuts it, a fourth points it, a fifth grinds it at the top for receiving the head … it is even a trade by itself to put them into the paper; and the important part of making a pin is, in this manner, divided into about eighteen distinct operations, which, in some manufactories, are all performed by distinct hands, though in others the same man will sometimes perform two or three of them’.

Smith calculated that by these means, ten people could make 4,800 pins a day, whereas if they had each made the whole pin, they would only have produced about 200 pins a day. Smith saw this principle of the division of labour as fundamental to efficiency and thus to prosperity, and as manufacturing became more and more automated, the aim became to identify those elements of the division of labour what were susceptible to automation.

Faster, cheaper, more efficiently: these are the words we are accustomed to associate with automation and computers. Computers are a means of reducing costs, eliminating the routine, enabling governments and companies to perform their business quickly and cheaply. We judge automation by its ability to cut unit costs and increase speed. Businesses look to keep IT costs as low as possible, to minimise support costs.

But is this vision of automation, rooted as it is in the Industrial Revolution, the right one? Can it be true that computers are just a means of streamlining the division of labour? When I started work as a librarian in the civil service in 1979, if I wanted to answer a letter, I drafted a reply in longhand. I put my draft reply in the outbox, and a messenger would take it to the typing pool. A professionally trained typist would type the letter, and a messenger would bring it back for me to check and correct.

The typing of a letter was a small example of the division of labour. However, with the emergence of personal computers and word processing, I started to type the letter myself. The typing pool disappeared, even though I am a very inefficient typist. If I needed to have a draft of my letter checked, I sent it by email, rather than office messenger, and this job also disappeared.

Far from reinforcing the division of labour, the personal computer has dissolved it. The production of a letter has become a job involving one person, rather than the three or four who were involved in 1979. Since this involves my very poor and inefficient one-finger typing, it is by no means clear that this is an improvement in efficiency, even if it may involve some reduction in staff costs. The networked computer is an all-purpose machine that challenges the idea of the division of labour. I can use a networked computer to do a host of things that previously required specialist equipment and facilities: make phone calls, send text communications, write, calculate, read books, watch films, listen to music, book travel, take photographs, make videos— the list is endless. This is the antithesis of the division of labour.

Because the networked computer is an all-purpose machine acting as an extension of my mind, giving me new opportunities for creativity and imagination, it is much much more than something that just enables routine jobs to be done more quickly and cheaply so as to promote economic growth and prosperity. The networked computer is a powerful means of reimagining the world and reshaping human understanding in more beautiful ways.

When I was a teenager, I was entranced by the music of John Cage and Karlheinz Stockhausen. I was excited by the way Cage had experimented with audio frequency oscillators, wire coils and variable speed turntables in the 1930s and 1940s to produce some of the first electronic music. Stockhausen said of his pioneering 1955 electronic piece, Gesang der Jünglinge (Song of Youths), “I often dream that I can fly, and I want to make music fly. I put my hands on the faders and I send sounds through the air.” The attraction of electronic music is not that it is cheaper or more efficient, but rather that it opens up new creative visions. The most exciting explorations of digital technology are those which investigate its potential to generate innovative sounds and visions. David Hockney did not produce iPad art because it was quicker or more efficient but rather because of the new affordances of the iPad. Artists exploring algorithmic art do so because of the different form of creativity it offers.

To make music fly—to make all human knowledge and creativity fly into completely new and unexpected dimensions—to create unexpected visions of greater beauty and complexity, pushing our imagination further than before—these are what the computer is capable of doing. These are more worthy and compelling ambitions than that of simply performing routine tasks more quickly and cheaply. We need to jettison the idea of the computer as a business machine and instead start to embrace more fully the idea of the computer as a dream engine and extension of our imagination. Adam Smith, writing at the beginning of the industrial period, saw the division of labour as one of the chief explanations of the apparent difference in wealth between advanced European countries and what he saw as more backward agricultural states. The division of labour, and our assumptions about automation promoting efficiency, are deeply enmeshed in these Enlightenment ideas of the nature of progress. Our visions of computing are profoundly linked to Enlightenment conceptions of the arc of progress.

The computer will free us from repetitive labour, promote education and bring continual growth and prosperity. The Enlightenment vision of progress is a linear one. Our assumptions about the value of computing are closely tied to such a view of progress.

The computer as a dream engine and extension of our imagination

the oceans and increasing global inequalities, does the optimistic Enlightenment view of progress still hold? Maybe we want to think about other shapes in imaging history and the future. Our view of history remains dominated by Europe which presents itself as the apotheosis of human achievement—a process Jack Goody has called ‘the theft of history’. That needs changing, and we need to decolonise history. In breaking away from Enlightenment historical models, China, India and Africa all offer potential alternative perspectives. And we can turn to other historical periods—how did Babylonia engage with the emergence of writing? How do the middle ages suggest alternative approaches to the division of labour? We must embrace the beauty and excitement of the networked computer for its own sake and explore how this all-purpose machine can reshape and reignite our imaginations. But for this to be successful, it needs to be a process which breaks away from commercial constraints. This means abandoning cost and efficiency as yardsticks for computerisation. Forget management life cycles; think instead of tinkering, experimenting, wrangling and failing (best of all, failing expensively). Don’t get Google or Microsoft to run your corporate systems; build and run them yourselves. This is another reason why the fight for the health of the internet is so important. A healthy internet means an untrammelled human imagination.