
language. How came Juan Lopez, then, to speak those significant words in Spanish!' Or else, he could
have his detective whip out an odd device and say, 'As you know, Watson, my pocket-frannistan is
perfectly capable of detecting the hidden jewel in a trice.'
Such arguments did not impress me. It seemed to me that ordinary mystery writers (non-science-fiction
variety) could be just as unfair to the readers. They could deliberately hide a necessary clue. They could
introduce an additional character from nowhere. They could simply forget about something over which
they had been making a great deal of fuss, and mention it no more. They could do anything. The point
was, though, that they didn't do anything. They stuck to the rule of being fair to the reader. Clues might
be obscured, but not omitted. Essential lines of thought might be thrown out casually, but they were
thrown out. The leader was remorselessly misdirected, misled, and mystified, but he was not cheated.
It seemed, then, a matter to be taken obviously for granted that the same would apply to a science fiction
mystery. You don't spring new devices on the reader and solve the mystery with them. You don't take
advantage of future history to introduce ad hoc phenomena. In fact, you carefully explain all facets of the
future background well in advance so the reader may have a decent chance to see the solution. The
fictional detective can make use only of facts known to the reader in the present or of 'facts' of the
fictional future, which will be carefully explained beforehand. Even some of the real facts of our present
ought to be mentioned if they are to be used—just to make sure the reader is aware of the world now
about him.
Once all this is accepted, not only does it become obvious that the science fiction mystery is a thoroughly
acceptable literary form, but it also becomes obvious that it is a lot more fun to write and read, since it
often has a background that is fascinating in itself quite apart from the mystery.
But talk is cheap, so I put my typewriter where my mouth was, and in 1953 wrote a science fiction
mystery novel called The Caves of Steel (published, 1954). It was accepted by the critics as a good
science fiction novel and a good mystery and after it appeared I never heard anyone say that science
fiction mysteries were impossible to write. I even wrote a sequel called The Naked Sun (published,
1957) just to show that the first book wasn't an accident.
Between and after these novels, moreover, I also wrote several short stories intended to prove that
science fiction mysteriescould be written in all lengths.
These shorter science fiction mysteries (including some boarderline cases) are included in this volume in
order of publication. Judge for yourself.
The Singing Bell
Louis Peyton never discussed publicly the methods by which he had bested the police of Earth in a dozen
duels of wits and bluff, with the psychoprobe always waiting and always foiled. He would have been
foolish to do so, of course, but in his more complacent moments, he fondled the notion of leaving a
testament to be opened only after his death, one in which his unbroken success could clearly be seen to
be due to ability and not to luck.
In such a testament he would say, 'No false pattern can be created to cover a crime without bearing upon
it some trace of its creator. It is better, then, to seek in events some pattern that already exists and then
adjust your actions to it.'
It was with that principle in mind that Peyton planned the murder of Albert Cornwell.
Cornwell, that small-time retailer of stolen things, first approached Peyton at the latter's usual