Blind man’s bluff and the Turing test

It seems plausible that under the conditions of the Turing test, congenitally blind people could nevertheless, with sufficient preparation, successfully represent themselves to remotely located interrogators as sighted. Having never experienced normal visual sensations, the successful blind player can prevail in this test only by playing a ‘lying game’—imitating the phenomenological claims of sighted people, in the absence of the qualitative visual experiences to which such statements purportedly refer. This suggests that a computer or robot might pass the Turing test in the same way, in the absence not only of visual experience, but qualitative consciousness in general. Hence, the standard Turing test does not provide a valid criterion for the presence of consciousness. A ‘sensorimetric’ version of the Turing test fares no better, for the apparent correlations we observe between cognitive functions and qualitative conscious experiences seems to be contingent, not necessary. We must therefore define consciousness not in terms of its causes and effects, but rather, in terms of the distinctive properties of its content, such as its possession of qualitative character and apparent intrinsic value—the property which confers upon consciousness its moral significance. As a means of determining whether or nor a machine is conscious, in this sense, an alternative to the standard Turing test is proposed. 1. Cutting the Turing test down to size I propose to consider the question, does the Turing test provide a valid criterion for the presence of consciousness? This should begin, at least, with a definition of the term ‘Turing test’ (and I promise to return to the more difficult ‘consciousness’ before the end of this paper). Alan Turing’s celebrated ‘imitation game’ test—in which a computer is programmed to pass itself off as a human being—was originally devised as a means of addressing the question, ‘can machines think.’ (Turing, 1950). In the standard version of the game, one player, a (human) ‘referee’, uses a communications terminal of some kind to exchange typewritten messages with an remotely located opponent—and seeks to determine whether this entity is human, or rather, a computer designed to simulate a human personality. If, over a number of trials, a succession of expert referees are consistently unable to make this distinction, the computer is deemed to have won the game—and passed the test. With respect to Turing’s initial question, however, there are at least two distinct affirmative interpretations which might be placed upon such a result: (1) Yes, the machine ‘thinks’—at least, in the limited sense that it can match the formidable information-processing abilities which allow humans to seem human to other humans, under the restricted conditions of the Turing test. (2) Yes, the machine thinks—in the strong sense of being fully sentient and conscious; that is, possessed of sensations, feelings, beliefs, hopes, fears and self-awareness etc., just like a human being. It seems to be widely agreed that a positive result at least entails (1). Somewhat more controversially, a considerable proportion of contemporary philosophers and cognitive scientists are inclined to suppose that (1) implies (2). This claim is based upon a behaviouristic form of functionalism—the view that mental states may be fully defined and understood, at least in principle, in terms of their objectively measurable causes and effects. For the functionalist, consciousness is as consciousness does. Of course, a facility for conversational fluency and repartee, expressed via typewritten messages, does not exhaust all that consciousness does— © 2004 Andrew Clifton