The thought police Brief Answers To Big Questions: 2. Do butterflies dream?

If people never dreamed, philosophers would certainly lose the most natural way of introducing scepticism. The Chinese Taoist, Chuang Tzu (369BC - ?), dreamt he was a butterfly. When he awoke he wondered whether he was a man who had dreamt he was a butterfly or a butterfly that was dreaming it was a man. Any experience can be explained as either a faithful representation of the world or as a mere figment of a sleeper's imagination.

Rene Descartes (1596-1650) tried to put science on a foundation of absolute certainty by devising a test for whether he was dreaming. He suggested that he could not be dreaming if he had a "clear and distinct idea" of being awake. Descartes' reasoning was that he had earlier "proved" that God exists and that He would not have created a world in which even responsible thinkers were doomed to err. Since God's goodness precludes Him from being a deceiver, the methodologically scrupulous are infallible. Ironically, Descartes' "clear and distinct idea" test was inspired by a dream (while in military service on 10 November, 1619 in Ulm, Germany) in which God revealed the treasure chest of knowledge.

I have never had a dream directly about God. My dreams are generally nightmares about lecturing mishaps. For instance, in one dream a perceptive student asks: "Professor Sorensen, why are you dressed in pyjamas?" Sometimes these anxiety dreams are more erudite. When I first lectured on my hero, David Hume (1711-1776), I was nervous about whether I had correctly interpreted his subtle objection to the following argument:

1. If there are miracles, then God exists.

2. There are miracles.

3. Therefore, God exists.

Hume balked at the second premise, but not on the grounds that it was false. His complaint was that no one could rationally believe the second premise. Miracles are intrinsically incredible. A breach of a natural law is as improbable as the law is probable. Consequently, we should always bet on the lesser miracle that the evidence for the miracle is misleading. Ever the over-preparer, I summarised Hume's reasoning in a detailed hand- out. After distributing it to the class, I discovered a typo. Instead of saying "You CANNOT know that there are miracles" the hand-out read: "You CAN know that there are miracles". I told everyone to correct the error by writing "NOT" after "CAN". We all did so.

Ten seconds later, the "NOT" disappeared from each person's page! The class let out a collective gasp. Resolving not to panic, I announced in a business-like tone: "That correction did not take. Please write 'NOT' again." I mentally counted down: "10, 9, 8, 7, 6, 5, 4, 321. Aha, still there! I wonder what the scientific ... explanation was for, er...." I had counted too fast: the "NOT" made another majestic exit. There were now thirty pairs of eyes focused on me. The brains behind those eyes were not impressed with David Hume. Nor were they impressed with me. Could my hero have led me into a worse pedagogical debacle? I attempted a cheerful face and squeaked, "Gee whiz, isn't that strange! I wonder what would happen if we wrote in NOT again?" The students mercilessly complied.

Yes, the "NOT" again evaporated. As did my composure. The students began to laugh at me. Softly at first, then louder and louder. My voice cracking, I once again asked them to insert the "NOT". This merely amplified their swirling derision. I could hardly hear myself think! Blood rushed to my cheeks. My mouth ran dry. I racked my brains for the most probable hypothesis. Finally, I stammered "I must be dreaming!"

And I was.

Roy Sorensen is Professor of Philosophy at New York University and author of 'Thought Experiments' (Oxford University Press).