The Liar

'Heuristic?'

'Capable of learning by mistakes, of operating, like human beings, through trial and error. My interest in all this was not mathematical, nor especially social. I was not frightened of machines becoming cleverer than human beings, nor of their in some way 'taking over'. I was however very interested in the development of new languages.'

'On account of your having learnt all the existing ones and being in danger of growing bored?'

'You exaggerate charmingly. Bela returned to Hungary after the war, Humphrey married Lady Helen, as you know, and became a schoolmaster. I stayed on at Cambridge. But we continued to work, where possible, on our idea for a perfect high level language that could be spoken by both machines and human beings. The dream, you see, was to invent an inter- national language, like Esperanto, that would also serve as a lingua franca between man and machine.'

'But surely the ideal solution would be to teach a machine to speak English?'

'Well I'm very much afraid that this is what will happen. We had no way of predicting the arrival of the microprocessor, or perhaps I should say; we lacked the imagination to predict its arrival. The cost of computing has been reduced by a factor of a million in ten years. It is simply astonishing. This means that you can now buy for one pound processing ability that would have cost you a million pounds in nineteen-seventy-one.'

'But isn't that good?'

'Marvellous, simply marvellous. But of no use to me. There are now dozens of languages at work in computing. Cobol, Forth, C, Lisp, Superlisp, Fortran, BASIC, Pascal, Logo, simply scores of the wretched things. We have a new Babel. This will sort itself out as soon as computing power comes still further down in price. Before the end of the century we shall have computers that recognise existing human languages.'

'So what's the problem?'

'Oh, there's no problem. None at all. We spent thirty years mining what turned out to be a barren seam, that's all. Nothing wrong with that. That's academe-biz, as they say. I tell you this to give you the background of my relationship with Szabo. We stayed in touch, do you see? He inBudapest, I in Cambridge.'

Adrian said that he saw.

'Two years ago Szabo made a curious discovery. He had shifted the focus of his attentions over the years from pure mathematics to electronics, acoustic engineering and any number of invigorating related fields. Hungary is very good about that sort of thing. That coloured cube that everyone is playing with at the moment is Hungarian, of course. I suspect that it is the advantage of speaking a language understood by so few that has turned the Magyars into such experts in numbers and shapes and dimensions. There is even a Hungarian mathematician at the moment who is close to achieving what was once thought to be the impossible. He is on the brink of squaring the circle. Or is it circling the square? Whichever.'

'A Hungarian is the only man who can follow you into a revolving door and come out first,' quoted Adrian.

'Exactly. Szabo is just one such hornery cuss. He had been working, during the seventies, on cures for the common stutter, experimenting with ways of playing back the speech of a stutterer into their ears as they spoke. Apparently if a subject hears his own voice back a split second after he has spoken and while he is moving on to the next thing he wants to say, his stutter will be eliminated.'

'How baroque.'

'Baroque? If you say so.'

'But not very practical to go about the place in headphones, I should have thought.'

'Quite so. Far from a feasible cure for the affliction. Experience in this area, however, did lead Bela towards what turned out to be immensely fruitful researches into the speech centres of the brain. The subject that most interested him was that of lying or, as it were, saying the thing which is not. He wanted to find out what happened in the brain when people said things which were not true; to see, for example, whether there is any difference between telling a lie, making a mistake in memory and inventing a fiction, all of which involve, in one way or another, saying the thing which is not. Thus a man might say: 'I have to work late tonight, darling,' or 'The German for a chive is ein Zwiebel,' or 'Once upon a time there was a fabulous trouser-eating dragon called Geoffrey.' These might all be taken to be examples of a lie. The speaker is in fact not going to be working late that night, he is instead going to the flat of his mistress there to conjoin with her in carnal riot. That is an Alpha-type lie. In the second case the man's brain knows full well that Zwiebel is in fact the German for 'onion' and that the word he is groping for is Schnittlauch, but his mind is unable for the moment to gain access to that information. His statement that ein Zwiebel is the German for 'a chive' is therefore a Beta-type lie. And lastly, there never was a fabulous trouser-eating dragon called Geoffrey, once upon any time and what is more the speaker knows it: a Gamma-type pseudology. The Alpha-type, the first kind of lie, the moral lie, if you like, the lie that disturbs the conscience of the speaker, might well be detectable using a polygram machine, the other two most certainly will not be.'

'Your friends are leaving,' said Adrian.

The BMW couple had stood up and were making for the exit.

'Excellent!' said Trefusis. 'That means we really are being followed.'

'How do you mean?'

'If Nancy and Simon leave the rendezvous first it is a sign that we are not alone. If they let us leave first, it means that we go unobserved.'

'Moscow Rules, George. Moscow Rules all the way.'

'I beg your pardon?'

'Nothing. So who is following us?'

'I dare say we shall find out. Drink your tea precipitately. We must not lag too far behind.'

Out in the car park the BMW had gone. Trefusis opened the driver's side door of the Wolseley, while Adrian looked around for signs of other cars preparing to start up in pursuit.

'Can't see any likely looking candidates,' he said.