354 | 32 TURING AND THE PARANORmAl
be expected on a probability calculation, so that the interrogator might be unable to make the
right identification. On the other hand, he might be able to guess right without any questioning,
by clairvoyance. With E.S.P. anything might happen.
In my view this is among the most puzzling passages in ‘Computing machinery and intelli-
gence’. For here it is the interrogator, not the percipient, to whom Turing is attributing the capac-
ity to rig the imitation game, either by using his telekinetic power to de-randomize the guesses
made by the random number generator, or by determining which of the two players is the human
by the covert use of telepathy or clairvoyance. To make matters stranger, in the first of these
scenarios the interrogator appears to be exerting his telekinetic influence unconsciously—why
otherwise would the result be that he ‘might still be unable to make the right identification’?—
and in the second he is cheating outright. (I presume here that the interrogator, in making the
identification by clairvoyance, is fully conscious that he is doing so. If not—if he makes the
identification by clairvoyance, but thinks he is making an educated deduction—the scenario
is the same.) But why would the interrogator want to cheat? In order to deprive the computer
of its opportunity to win? In this case, is the interrogator someone so threatened by the idea of
machine intelligence that he is willing to sabotage the game in order to stop it?
Cheating humans and cheating machines
One facet of Turing’s thinking into which argument 9 gives us some insight was his willingness
to allow for cheating, deception, and lying as viable strategies in the playing of the imitation
game. In this he followed once again the example of the ‘psychists’, for whom cheating was more
of the order of an occupational hazard than an unpardonable offence. As the SPR’s members
had discovered early on, most mediums—even authentic ones—cheated occasionally, out of
pique, laziness, or fatigue. A good example was the famous Italian medium Eusepia Palladino,
in whose psychic powers William James (among others) believed utterly, writing in 1909:^33
Everyone agrees that she cheats all the time. The Cambridge experts . . . rejected her in toto on
this account.
Was this justified? James thought not:
Falsus in uno, falsus in omnibus, once a cheat, always a cheat, such has been the motto of the
English psychical researchers in dealing with mediums. I am disposed to think that, as a matter of
policy, it has been wise. Tactically it is far better to believe much too little than a little too much . . .
But however wise as a policy the S.P.R.’s maxim may have been, as a test of truth I believe it to
be almost irrelevant. In most things human the accusation of deliberate fraud and falsehood is
grossly superficial. Man’s character is too sophistically mixed for the alternative of ‘honest or
dishonest’ to be a sharp one. Scientific men themselves will cheat—at public lectures—rather
than let experiments obey their well-known tendency towards failure.
I find this a fascinatingly perverse paragraph. In an essay whose ostensible aim is to lend
legitimacy to psychic research, the esteemed Professor James of Harvard ended up, if not con-
doning cheating outright, then exhibiting a far more lenient attitude toward it than the institu-
tion with which he was affiliated. He even went so far as to admit that he himself had ‘cheated
shamelessly’—and all this toward the goal of rehabilitating the reputation of an Italian medium
considered by most to be a charlatan. Or was James up to something more here?