Charles asked:
Alan Turing in his 1950 paper Computing Machinery and Intelligence (Mind 49: 433-460) posited that to verify the proposition Machines can think one must use an The Imitation Game, instead of trying to (philosophically) define the terms machine or think, since this would only lead one to, reflect so far as possible the normal use of the words, and that therefore this, attitude is dangerous, given that there is a certain amount of semantic alteration in any term over time and therefore no definitive verification or answer to our question, Can machines think?
However, can one make a philosophical case that such statements as x can y are statements of ability, and that therefore the Turing test is not a substitute for philosophical investigation, since this can must be decided in and for itself? Therefore the question to which I would like some advice is, Can statements of known ability, i.e. x can do y, be verified using an imitation test which may confuse like with identical with?
Answer by Jürgen Lawrenz
To solve this question, you only need to look at the way we humans communicate via books, letters, emails, radios etc. The possibilities for hoaxes are unlimited. You can never be sure that the author you are reading actually wrote the book (maybe it was ghost writer or a parodist). In some sectors of society, the fashion for expressing oneself are so pervasive that everyone seems to have the same voice and idiom, e.g. argot or scientific and academic literature. I am sure that any programmer worth his salt could get his computer to write up a perfect imitation of a paper on quantum gravity and maybe even imitate such voicing.
But this ignores the human specificity. When a writer or speaker or actor or singer imposes individuality on their tone of voice, their manner of delivery, their syntax and grammar (not to mention sentence fillers like ‘just’, ‘you know?’, ‘eh’, ‘um’ and so on) they are constitutive of individuality. Such inflections carry their own communicative value and intention, and intuitive creatures are capable of discerning them. The phrase ‘reading between the lines’ or ‘understanding what is not being said’ indicate something of this.
The crux of the matter is, that the book you are reading, the speaker or singer you are hearing on radio, are not in direct communication with you, but at a step removed which permits all sorts of technical intervention. The only means of foolproof evaluation of authenticity is your actual presence, so that you can watch the facial motions, eyes and eyebrows, stance and comportment, when you receive much more than just the words (or the song).
Turing’s proposition about testing for the quality of imitation does not meet these indispensable criteria. This is because he and his cronies misunderstand the nature of both imitation and intelligence. It does not take a very deep understanding of ‘human intelligence’ to know that it cannot be reduced to mechanised mimetic resources. Turing literally grabbed the stick by the wrong end, proceeding from the limitations of imitation to the origination of behaviour. This only succeeds in obscuring the difference which pertains to a living creature’s intelligence, of which the most significant content is the initiation or origination of behaviour. Further, it completely ignores that intelligent learning is mimetic only to a limited degree and on the whole considerably individualised, even among children. Every human being knows without the slightest expenditure of deep thought that the only way of ensuring consistent imitation in a learning environment is drill. People who have been drilled are the counterparts to Turing’s machine intelligence, in that they are not required to be persons, but on the contrary expected to perform a quasi machine-like response to the situations which the drill expects them to master (without necessarily involving the specific form of human intelligence).
I might usefully remind you in this context of the dispute around IQ tests that have erupted upon the realisation that these tests do not measure intelligence, but the skill of passing IQ tests. Students sitting repeatedly for IQ tests gain familiarity and improve their results. The logical conclusion for designers of these tests is that they must either accept that the students’ intelligence quotient rises after repeated exposure, or else that their definition of an IQ rests on a defective appreciation of intelligent adaptitivity. However, the predicament for them is that changing the test immediately changes the definition of an IQ. Therefore tests cannot not reveal the intrinsic IQ of the subjects, as the whole concept of an intrinsic level of intelligence is faulty.
The relevance of these comparisons is clearly, that in both cases a kind of ‘absolute’ criterion is posited – in IQ tests for a reliable measure of intelligence, in Turing’s test for a reliable measure of discernment. Both are stale exercises, because they under-appreciate the resourcefulness of human intelligence.
Much more could be brought into this argument, but if you will carefully think about it, you should come to the conclusion that all such tests possess only the limited value of revealing the extent to which subjects can be made willing to suppress the nature of their authentic intelligence to fit into the mould of a mechanical model. These conditions, moreover, reveal that the designers of Turing tests, IQ tests and others of the same ilk simply misunderstand the nature of philosophy as something which it is not. They seem to believe that philosophy is logic, whereas logic is merely a tool for the achievement of consistency in thinking. They seem to believe that philosophy is basically the hoarding of knowledge, whereas in reality philosophy is intrinsically about understanding the knowledge which we acquire by experience and research. Most importantly, however, such tests are about the suppression of human individuality. They reveal almost nothing intrinsic to human intelligence, but everything about minds trapped in a deterministic framework that seeks to trim down the living impulses from which intelligence arises.
I’ll leave it to you to work out whether the existence of such tests implies prejudicial authoritarian pressure on us to stop thinking as individuals and accept that society is better off when we all behave like robots.
Ask a Philosopher 
