1June twenty-third, nineteen-twelve; London, birth;
Alan Turing, a fellow, in his day,
Of King's College, Cambridge; and there the worth
This, done there.
David Hilbert sought a way:
2A definite procedure that, for each
Statement of logic, answers yes or no:
It has a proof, or none; and in its reach
Each one; not cleverness that's sometimes so.
3Turing defined a number one could write
By finite means: computable, its line
Of decimals, calculable outright;
The means, a table, finite; the design:
4A tape, a strip of squares; the machine sees
One square; one of a finite list of states;
The table says, for state and symbol, these:
What to write, which way to step, and what waits;
5He called it computing machine; the name
That bears his own came later. They can be
Listed, the machines; the reals can't; the same
The list of methods leaves the line still free:
6The numbers ordinary work will need
Are on the list; the line is more than work.
Fed one's table, another can proceed
As that one does; on paper, he would work
7One for machines: the universal; it
Knows not the work ahead; it knows the way
To follow tables; and the table, writ,
Is the work. Then, the twenty-eighth of May,
8The learned society in London took
1
His paper; he was twenty-three; and read
November twelfth, its line says; and the book
On Computable Numbers: titled, spread,
9Applied to the decision problem; made
In two parts, thirtieth November, then
December twenty-third; a fix, delayed,
Came in the next; no flawless text. And when
10Alonzo Church, at Princeton, had defined
Another method: no tape; and, in spring
Of 'thirty-six, the same refusal, signed:
No general procedure for the thing
11Decision.
Max Newman saw
Church's; so
Turing wrote an appendix: the two meet,
The definitions; he did not, though, show
The limit as his own alone. The feat:
12Hilbert's question has no procedure; none
Takes every statement, gives the verdict; more,
A nearer limit, of machines: some run,
Write digits on; some don't; and none, before,
13Given a table, says, in every case,
Which future it will have. A method made
For worlds of questions may, within its space,
Be no method for its own future, laid
14Before it. That September, Princeton; there
The Graduate College named him, in print;
In 'thirty-one,
Kurt Gödel had shown: where
A system strong enough for sums, a stint
15Of true sentences it can't prove; one set
Of rules; and
Turing's, of the method whole:
Does any method always answer? Yet
The two are not one fact; not one dark hole.
16In 'fifty-two, the state tried him; and he,
Convicted, for a man's love, chose to take
Hormones, not prison; 'fifty-four: and he
Was found dead, cyanide; and, for his sake,
17This book keeps him among its honored dead;
A finite means: the calendar, which runs
Names to its days, by rule; yet more, it's led
By more than human hands, as stars and suns;
18Some questions have a method; some, though asked
In method's language, have none settling all;
Count what the table writes; and keep, unmasked,
The questions no table can close at all.