Recently I have been writing columns about "junk science." I was talking about this subject to my next-door neighbor when he asked a question that stopped me cold. He said, "I read the newspapers and a few popular science magazines, but half the time I don't know what to believe. How do you decide when you are reading an article if it is some kind of bogus rubbish, or if it's real new science that will prove to be important?"

What makes his question so challenging is the sheer volume of scientific papers being published today. A couple of thousand years ago it was possible for someone like Aristotle to learn and write about all the science of his day. Now, according to estimates that I have seen but have never tried to verify, more than half of the scientists who have ever lived are alive - and every one of them seems to be producing papers and giving speeches at an increasing rate.

No one today can read more than a tiny fraction of the scientific papers being written. It is as much as most people can do to keep up with the flood of results in their own specialized field. Outside that field, even the smartest and most hard-working scientists in the world are forced to rely on summaries and overviews provided by others. Hence, the very valid question: "What should I believe, at least provisionally, and what should I put on the heap of discards?"

I have no foolproof answers, but over the years I have developed my own set of rules and questions. I am now setting them down on paper for the first time.

Rule No. 1: Be wary of dogmatism. Theories and results are usually tentative in their initial phases. You need to be particularly cautious when two different works appear at about the same time and offer contrary conclusions. I have a fine example of this sitting on my own bookshelves. In 1998, "Probability 1: Why there must be intelligent life in the universe," appeared. This book, by Amir Aczel, "proved" using statistical arguments that not only life, but intelligent life, will be found all over the galaxy and all over the universe when we get around to looking for it. Two years later, the book "Rare Earth" was published. Written by Peter Ward and Donald Brownlee, this work argued that the development of complex life in general, and of the intelligent life form known as humans in particular, had a vanishingly low probability of happening anywhere except on Earth. Based on the evidence of these two books alone, it is reasonable to view both conclusions with skepticism.

Rule No. 2: Watch out for hidden agendas. After I had read "Rare Earth" I went back and looked again at the arguments being made. A number of them were circular, as though assumptions had been set up at the outset to reach a desired conclusion. It is easy to imagine that the book was written with the hidden agenda of proving that human life could not have come into existence without some supernatural agency being at work. I am not saying that the authors are "scientific creationists," and I should write to them to confirm or deny that idea. However, the book is certainly consistent with such an assumption.

Rule No. 3: Ask the proponent of an idea for tests by which it might be proved false. Scientific ideas live or die based on experimental evidence. If you see an idea put forward which cannot be falsified by any test whatsoever, you have a right to be vastly suspicious. People offering unconventional theories often go to great lengths providing evidence that supports them. Far more valuable is a suggestion for a simple crucial experiment by which a theory might be shown false.

Rule No. 4: Look at the source of the idea. If it comes from a Francis Crick, a Murray Gell-Mann, or a Harold Kroto, each a Nobel Prize winner, then it is appropriate to take the idea seriously and look for ways to test it. On the other hand, should the idea derive from someone in their eighties who has never previously produced anything of scientific validity or interest, then suspicion is an appropriate reaction. Actually, without wishing to discriminate against older people, radical suggestions from anyone in their seventies or eighties ought to be treated with caution. Many great scientists in their later years tend to unorthodox viewpoints. Oliver Lodge embraced spiritualism, Julian Schwinger became an advocate of cold fusion, Linus Pauling was convinced that vitamin C had powerful curative properties for almost everything, and Fred Hoyle argued that many diseases originate not on earth but in space. Yet each of these men was without a doubt a great scientist in his own field.

Rule No. 5: Ask how contrary the new idea is to all existing theories and evidence. To quote an old scientific principle, "Extraordinary claims require extraordinary evidence." If someone tells you that he or she is a direct descendant of Napoleon, it is hardly worth your while to invest any effort in determining whether or not the statement is true. On the other hand, if a man tells you not only that he is the reincarnation of Napoleon, but that reincarnation itself is a genuine phenomenon for which he can offer scientific proof, then it is reasonable to ask him to show you evidence of his claim (and also ask him to tell you of a way in which the claim might ever be proven false).

Even when you have applied all the above rules and answered every implied question, something may still slip through the net. Just as great truths tend to begin as scientific heresies, many things now believed by all of humanity may turn out to be wrong. Working on the borderlands of science is a messy business, and probably it always will be.

So here, then, is my final rule: If you want to pick a scientist to rely on, choose the one who most questions the validity of his or her own work. A doctrine of infallibility has no place in authentic science.

"Borderlands of Science" is syndicated by: