Sample of one

Because life exists on Earth, Ken Smith claims, it must exist elsewhere in the Universe (Letters, 18 March). "Probability theory", he says, makes it "not merely improbable, but ... impossible, that this should not be the case."

If probability alone is to be our guide, then it is equally likely that the Eiffel Tower or Marmite will exist "out there". If this sounds silly, it's because we need other grounds for judging the likelihood of life occurring elsewhere in the Universe - organic chemistry, evolutionary biology, even history. All attempts to apply probability theory to the problem founder on the impossibility of extrapolating from a sample of one.

A sample of two would be decisive, but I'll bet Smith £100 that none will be discovered in our lifetimes. In fact, I'm so certain that he may as well pay up now.

Robert Poole, University of Cumbria, Lancaster.