The sixth commandment of detective fiction is:
No accident must ever help the detective, nor must he ever have an unaccountable intuition which proves to be right.
In his 1939 commentary on his decalogue, Fr. Knox said:
That is perhaps too strongly stated; it is legitimate for the detective to have inspirations which he afterwards verifies, before he acts on them, by genuine investigation. And again, he will naturally have moments of clear vision, in which the bearings of the observations hitherto made will become suddenly evident to him. But he must not be allowed, for example, to look for the lost will in the works of the grandfather clock because an unaccountable instinct tells him that that is the right place to search. He must look there because he realizes that that is where he would have hidden it himself if he had been in the criminal’s place. And in general it should be observed that every detail of his thought – process, not merely the main outline of it, should be conscientiously audited when the explanation comes along at the end.
This may be the commandment in Fr. Knox’s decalogue with which I agree most strongly (with a few caveats). Curiously, along with Commandment #8 (the detective must not conceal evidence from the reader), this may be the commandment which is most often broken in detective fiction.
I agree with it because the whole point of a detective is to detect, not merely to be the recipient of pure luck. Pure luck is the domain of comedies or, in some curious cases, of tragedies. It is not the domain of detective fiction. And I should note that this is true whether one is talking about play-fair detective fiction or not. Even if the reader has no earthly way to guess the solution to the problem, the detective should.
This really gets to the question of what a detective story is. A Franciscan friar to is a good friend of mine suggested that the fundamental structure of a detective story is that some villain, through the misuse of reason, has disturbed the natural order of things and that the detective, through the right use of a superior reason, restores the right ordering of things. It is, fundamentally, the Christian story—humanity has messed up the world and God condescends with us to restore it.
This structure to the detective story only works if it is the detective whose right use of reason restores the natural order. Luck and unaccountable intuitions are to detective fiction what the Gnostic and later Arian heresies were to Christianity. Just as those heresies turned Christ into a creature and thus from a savior into a mere conduit of information, luck and unaccountable intuitions turn the detective from a savior into a mere conduit of information. If Christ is a creature, and the detective merely a lucky fool, neither is capable of saving himself, let alone us.
This rule is violated so often precisely because writing detective fiction is hard. This is especially true of mystery novels. As I’ve mentioned elsewhere, mystery novels and mystery short stories are fundamentally different creatures. The short story is a puzzle at the end of which—preferably on the next page—is a solution. The novel must be the tale of the assembly of all the clues necessary to solve the problem or the author must pick whether he wants the story to be unsolvable or to drag on long after it should have finished.
But to make a mystery novel the tale of the assembly of all the necessary clues, there must be some reason why the clues satisfy two opposing conditions:
- They are not readily available
- They are available
There are a variety of ways to satisfy these two—if not, there would be no detective fiction—but they tend to boil down to a few generalities:
- You have to figure out where to look.
- The clue doesn’t exist yet.
- You already have the clue but it doesn’t mean anything until you find additional evidence satisfying some other condition on this list.
(To be clear, how one goes about doing these things are what make the story, and there are endless ways to do these in fresh and interesting ways.)
The second type of clue—the clue only coming into existence later in the story—requires a certain type of story to work; specifically, one where the murderer is still active. It’s a great type of story, but if it’s not the story one is writing, it’s not an available option. And in that case, we’re left with #1 and #3. And the obtaining of clues which make other clues significant in #3 will resolve into #1, most of the time, because otherwise it’s a police procedural, not a detective story.
So the big trick to writing a mystery story is really figuring out where to look for clues. And the mystery writer is hung, to some degree, on the horns of a dilemma: if the location of the clues are obvious, anyone could follow them up. If the location of the clues are not obvious, why on earth would the detective think to look for them where they are?
There are, of course, ways to unhook oneself from the horns of this dilemma; this commandment forbids the author from simply waving the problem away.
This post is already long enough, but I will mention the ways to unhook oneself from this dilemma, if briefly (and, I should note, they generally work best when combined):
- Specialized knowledge—this runs the risk of being simply esoteric, rather than mysterious, but in a detective story which is more educational than adventurous, the detective giving away the requisite knowledge when he comes across the clue can make for an enjoyable story. Detectives are, after all, supposed to be not merely brilliant, but also learned. And mystery readers do, as a rule, enjoy learning things. They do need to be real things, though (see rule #4, poisons).
- Psychological insight—Probably the best example of this is Chesterton’s Father Brown mysteries. If the detective can think like the criminal he will be able to predict what the criminal did and therefore where to look for clues.
- Gaining the trust of people who have clues—this can easily be done badly (chiefly where the person shouldn’t need his trust to be gained), but it is an extremely workable way to withhold clues for a time. This all too often is an occasion for the detective to become a liar, though; gaining trust under false pretenses is distressingly common in detective fiction.
- Legwork needed—This is a case where the insight of the detective leads to knowing, not exactly where a clue is, but a small number of places to look. It will take time to explore all of the possibilities until finding the clue. (This works best when there is some reason the possibilities must be investigated by the detective himself, generally to be found elsewhere on this list.)
- Labwork, police work, etc—chemical analysis of substances, the interviewing of every lawyer in London or everyone living within three blocks of the murder, etc. all take a lot of time. When this is done off-screen, it produces space in which the detective can be doing something interesting. N.B. That interesting thing that the detective does while the poor off-screen laborers toil in the clue-mines should be something that either explains the clue when it comes in or else the arrival of the clue should have no value other than to explain what the detective found while he was occupying the reader’s attention. Anyone can cavort with an attractive member of the opposite sex for a few days until the solution to the problem they have not advanced a wit falls into their lap.
- Delayed clues—this needs to be used with extreme care because it can easily become cheating, but clues which only show up once the post office has delivered them, or after some device with a timer reveals them, can work. The clue absolutely has to be mysterious on its own, and require everything the detective did in the interim to be meaningful, or the author has wasted the reader’s time until the clue shows up. The reverse could, in theory, work; but it doesn’t work. That is, the previous work being mysterious until the clue shows up in the mail will always feel like a cheap deus ex machina. In either case, there has to be a very good reason why the clue was intentionally delayed, and moreover, it absolutely cannot be the clue which solves the mystery.
There are probably other ways, too, though they’re likely to be some sort of variant on the methods above. And I want to stress again that giving a general description of the structure in no way implies that the stories must be formulaic or un-creative; the beauty of any story is in its specifics. It is no more saying that a story is formulaic because there are only so many workable structures for their plots than it saying that all people look alike to say that all men have the same bones in their skeleton. A man without a skull is not bold and daring, and a man with a skull is not boring and repetitive. It’s what’s inside his skull that really counts.