202 | 20 BABy
The National Physical Laboratory (NPL) had mailed Williams a copy of Turing’s ‘Proposed
electronic calculator’ in October 1946, in the month before Williams lodged a draft application
for a patent on the Williams tube.^13 It isn’t recorded whether Williams read it at that time, but
there is no reason to think otherwise. He need only have opened the envelope and glanced at
the table of contents to see the irresistible chapter title ‘Alternative forms of storage’, the very
field he himself was pioneering. Racing to the chapter, he would have found Turing’s pithy
description of the Williams tube—enough to make his ever-present cigarette pop out of his
mouth with surprise. He might well have objected, though, to Turing’s airy statement that
turning the basic ideas into engineering reality would not involve ‘any fundamental difficulty’.
There is now no way of telling how much or how little Williams learned from Turing’s docu-
ment. All that can be said for certain is that he quickly dropped his original ‘anticipation pulse’
method, in favour of the simpler read–rewrite method for refreshing the data that Turing
described.
Too little credit
Although Williams and Kilburn had brilliantly translated the logico-mathematical concept
of the stored-program computer into electronic hardware, the mathematicians in Newman’s
department gave them too little credit.^14 Peter Hilton, who joined the Manchester mathemati-
cians in 1948, explained that Williams and Kilburn were regarded as excellent engineers but not
as ‘ideas men’. Nowadays, though, the tables have turned too far and the triumph at Manchester
is usually attributed to the engineers alone. Fortunately, Williams’s own words survive to remind
us of the true situation. ‘Tom Kilburn and I knew nothing about computers’, Williams said.^15
‘We’d had enough explained to us to understand what the problem of storage was and what we
wanted to store, and that we’d achieved, so the point now had been reached where we’d got to
find out about computers’, he recalled, continuing: ‘Professor Newman and Mr A. M. Turing . . .
took us by the hand’.^16
Kilburn, in later life, was an important source for what became the canonical view of the roles
of Turing and Newman (or rather, their lack of role) in the origin of Baby. In his first papers on
the Manchester computer, Kilburn gave credit to both Turing and Newman,^17 but in later years
he was at pains to assert the independence of his and Williams’s work from outside influence,
presenting the history of Baby in a way that assigned no role to Turing or to Newman. In an
interview with me in 1997, Kilburn emphasized that Newman ‘contributed nothing to the first
machine’ (Baby).^18 Turing’s only contributions, Kilburn told me, came after the computer was
working, and included preparing what he called a ‘completely useless’ programming manual.
Yet Turing’s logico-mathematical ideas of 1936 had led, via Newman, to Manchester’s stored-
program computer project. Even in the midst of the attack on Tunny, Newman had been think-
ing about Turing’s universal machine: when Flowers was designing Colossus, Newman showed
him Turing’s 1936 paper (‘On computable numbers’) about the universal computing machine,
with its key idea of storing coded instructions in memory.^19 By 1944 (as Newman related in a
letter to von Neumann) he was looking forward to setting up an electronic computer project as
soon as the war was over.^20 The Princeton computer, too, was a physical embodiment of Turing’s
universal computing machine of 1936: Julian Bigelow told me that von Neumann even gave
his Princeton engineers ‘On computable numbers’ to read.^21 The reason why von Neumann