From the beginning of the computer age, there is no question that smaller machines were not built and could have been. But, given available valve technology, of what use could they have been? And, second, were the solid state electronics, essential to the contemporary personal computer, simply not available; indeed, not yet ‘invented’?
As to the first of these questions, differently configured supervening necessities—the electronic office and the non-nuclear science lab rather than the cold war—could have produced applications for which developments of the Baby Mark I, for example, would have been perfectly viable. But, more than that, a different research agenda would have yielded solid state devices earlier. My contention that the ‘law’ of the suppression of radical potential operates in computer history depends finally on denying the supposed ‘synergy’ of the technologies of computing and solid state electronics, and the idea that the micro-computer had to wait for the microelectronics to come on stream.
Transistors, it must not be forgotten, were ‘invented’ late in 1947, so every machine we have so far discussed came on stream in the semiconductor age; yet, contrary to received opinion, transistors did not become the norm in computers. In fact, the failure to exploit solid state electronics constitutes a fourth element, apart from indifference, language provision and size, in the operation of the ‘law’ as it applies to the computer. Received opinion of the relationship between microelectronics and computers simply ignores this:
It all began with the development thirty years ago of the transistor: a small, low-power amplifier that replaced the large power-hungry vacuum tube. The advent almost simultaneously of the stored program digital computer provided a large potential market for the transistor. The