Update to
|
 Alan Turing: the Enigma |
Page 59: That AMT bathed naked, I know from a photograph of him taken on the rocky coast of Sark. You will see it on this page of my Alan Turing Internet Scrapbook.
Page 67: I did not labour the obvious fact that AMT read von Neumann's book in German. Reading German was, in 1932, an essential part of serious mathematical work. (The Gödel paper that was his main reference in Computable Numbers was also in German, and Princeton Ph.D. involved a German language qualification.) On the same page 67, it is noted that AMT visited Germany with his father; he was in Austria skiing in March 1934, and back in Germany again in June 1934 (pages 88-90). The incidents reported there mean that he had quite adequate conversational German. He would not have had the slightest difficulty with any German language questions arising in the cryptanalytic work after 1938.
Page 90 and note 2.26 The leading statistician R. A. Fisher was the referee for AMT's dissertation. His correspondence with King's College can be seen here.
Page 92: It is impossible to give a conventional nationality for Gödel. He was born into the Austro-Hungarian Empire, in a German-speaking part of what is now the Czech Republic. But he would not have identified with the Czechoslovak state of the 1930s. The indication of 'Vienna' would have been more appropriate. The question of Gödel's national identity illustrates vividly the difficulty of applying the principle of 'self-determination' as embodied in the Versailles treaty.
Page 94: I suspect that the clerihew about AMT's alluring qualities did not actually circulate until the period of 1948 when he was back in residence at King's.
Pages 98-99: The Turing machines are now illustrated on this website, implemented in JavaScript code. Go to this page.Pages 104-109 and note 2.37: The publication year for AMT's paper is now generally given as 1936, but some references may give it as 1937. The reason is that volume 42 of the Proceedings of the London Mathematical Society was published in parts. The parts containing AMT's paper bore the dates of 23 November and 23 December 1936. However, the complete volume counts as a 1937 volume (see the LMS page, from which a facsimile copy can be downloaded.
My original text gave the publication date of the paper as 1937 for this reason, but I have changed it to 1936 in the 2014 editions for conformity with the general usage. In any case, the writing and submission of the paper all fell within 1936.
Page 109: My reference to Babbage's 'projected universal machine' recurs at various points in the book. This understates AMT's achievement, by making it appear that he had merely rediscovered Babbage's idea. Turing himself led me into this, by claiming in his 1950 paper that Babbage's Analytical Engine would be a universal machine, but I need not have followed AMT's modest self-effacement. And others generally, awestruck at Babbage's farsightedness, have tended to assume that Babbage had all the ideas.
The Analytical Engine did not allow for the unlimited number of conditional branches that would be necessary for universality. Nor did Babbage did not have the modern concept of a program. Turing's 'tables of instructions' are programs in the modern sense, being logical entities in their own right. It is essential to Turing's analysis that tables of instructions can be seen as data for the machine itself to process. Babbage does not seem to have had any idea like this, but it is vital to the concept of the modern computer.
Page 109: I wrote the sentence 'Alan had proved that there was no 'miraculous machine' that could solve all mathematical problems, but in the process he had discovered something almost equally miraculous, the idea of a universal machine that could take over the work of any machine.' The sentence has been picked out by B. J. Copeland for criticism in the pages of the Times Literary Supplement.
Copeland says that my words 'any machine' overlook the 'oracle' concept Turing defined in 1938, (as appears in chapter 3 of the book). But Turing wrote that the nature of the 'oracle' is that it
My usage of the word 'machine' in this sentence simply echoes Turing's own language. In his 1938 work, he made his own summary of what had been achieved here by identifying the concepts of 'effectively calculable', 'purely mechanical process', and 'what can be carried out by a machine.' More comment and references are available in web-page form here.
There is further material in the Alan Turing Internet Scrapbook: The Inspiration of Life and Death and Turing machines.
Update Index | Intro | Parts 1 | 2 | 3 | 4 | BP | 5 | 6 | 7 | 8 | Authors Note
| Quick Links: |
|
|
|
|
|
|
| ||||