I have noticed a number of parallels between the genres of science fiction and mystery in terms of what “rules” make for effective writing. In particular, I’m very interested in the intersection of the two: the relatively few mystery stories told in science fiction settings. This page is an effort to explore that intersection. I’m basically thinking out loud here, so I’m always happy to hear feedback.
I’m an unabashed fan of Golden Era detective stories: you know, the ones with the brilliant detective who goes to an English country house and discovers that the doctor killed the matron of the family even though everyone says it’s either the butler or some mysterious outsider. Agatha Christie is widely regarded as the queen of the format, and remains one of the best-selling English-language authors ever. If you’re going to strike up a literary conversation with a stranger, you’re statistically better off bringing up Hercule Poirot than Harry Potter.
One of the hallmarks of this era of fiction is the notion that the reader ought to have a fair shot at solving the mystery before the detective does. The concept of Fair Play was described by writer and priest Ronald Knox, who (possibly inspired by some other set) gave ten commandments of fair play. His list, as posted in Wikipedia, is a useful guide to the philosophy behind some of the finest works — but should also be read as a reaction against those works that failed as mysteries per se, and an attempt to quantify why. For your convenience, I’ll copy them from the Wikipedia page:
- The criminal must be mentioned in the early part of the story, but must not be anyone whose thoughts the reader has been allowed to know.
- All supernatural or preternatural agencies are ruled out as a matter of course.
- Not more than one secret room or passage is allowable.
- No hitherto undiscovered poisons may be used, nor any appliance which will need a long scientific explanation at the end.
- No Chinaman must figure in the story.
- No accident must ever help the detective, nor must he ever have an unaccountable intuition which proves to be right.
- The detective himself must not commit the crime.
- The detective is bound to declare any clues which he may discover.
- The stupid friend of the detective, the Watson, must not conceal from the reader any thoughts which pass through his mind: his intelligence must be slightly, but very slightly, below that of the average reader.
- Twin brothers, and doubles generally, must not appear unless we have been duly prepared for them.
The Decalogue is very much a product of its time (raise your hand if you winced when you read #5). There are a lot of assumptions here: by the time Knox wrote this list in 1929, the genre had begun to take on the familiar forms, largely mimicry of Sir Arthur Conan Doyle’s work. Hence, the references to the story’s Watson. From a step back, though, you can see that it described broadly a contract between writer and reader. Mystery stories, after all, require the writer to withhold information strategically that is vital to the story, then introduce it dramatically later. The reader allows this to happen out of trust: this withholding is done for the sake of a good story, but enough information remains that the reader can still, in principle, beat the detective and solve the case first. That is the essence of ‘Fair Play’, and all the rules stem from that principle.
Now, I do not only write mysteries, I write (or attempt to write) science fiction as well. I sometimes even succeed in one endeavor or the other! The concept of fair play, to my mind, holds true to mysteries whichever genre they are in, and whether that’s the focus of the main story or not.
For example, one of the more obnoxious tropes in science fiction is the delayed explanation. The Main Character in such a story might go about the whole story preparing for The Procedure without ever explaining what The Procedure is… and then it’s something horrible like death or watching day time television. It’s a strategic witholding of information, sure, but it’s not fair: it’s a refusal to describe some aspect of the world in order to surprise the reader with it later, and it only works because of, well, basic dishonesty. The rules of fair play might not completely prevent this sort of thing, but as a mindset, it helps.
If my arm were twisted, I would say that the fundamental problem with writing mysteries is that the reader’s intellect is inserted further into the story than any other genre (except possibly romance). The reader may choose to compete with your detective, but the reader does not have agency in the world of the story: even if the reader knows for sure that the will is hidden under a brick in the fireplace, he or she must wait patiently for someone in the story to get around to checking there. The reader cannot feel for trap doors or ask that one crucial question, and so is reliant on having a guide in the world, a cat’s paw who does not generally do stupid things. I think that many of these rules, then, are for avoiding frustration stemming from this limitation.
So, here are my thoughts on the individual rules:
1. The criminal must be mentioned in the early part of the story, but must not be anyone whose thoughts the reader has been allowed to know.
This actually has two parts. The first part is essentially saying that the reader should have a shot at figuring out who did it, and introducing the criminal late puts the reader at a disadvantage. The second part is interesting to me: Agatha Christie famously broke this one in a book that a lot of people think was a total cheat. What Knox is getting at, I think, is that the reader shouldn’t be coerced into identifying with the criminal.
For a science fiction story this has an additional reading: it is traditional in these works to have a character stand in for the reader. Sometimes it’s simply an ingenue farmboy, sometimes someone literally zapped from our world into the next, sometimes simply a character put in unfamiliar surroundings. Think Luke Skywalker needing to have everything introduced and explained to him. The reader relies upon the honesty of this character and his reactions in order to feel at home in the world. It might have made for a more interesting story for Luke Skywalker to secretly be a Sith lord, but the reader would then have to seriously question everything presented in that movie about learning to become a Jedi knight: it might turn out to have something to do with microscopic magic parasites or something stupid like that.
2. All supernatural or preternatural agencies are ruled out as a matter of course.
Fundamentally, this means that the reader needs to understand what’s possible in the world the story takes place. Once you have eliminated the impossible, as Sherlock Holmes said, whatever remains (however improbable) MUST be true. Impossible things may happen to you before breakfast, but it is unfair for the reader to have to guess which impossible things those are in the course of trying to figure out your mystery.
This holds true, of course, for science fiction, but I’d go further: any science related to the plot ought to be impeccable (OK, plausible) or the story will feel like a double cheat. Fantasy stories too need to at least follow their own rules, whether Jack Vance’s magic or Terry Pratchett’s million-to-one shots. There needs to be consistency so that the reader understands what’s possible.
I’ve never seen this rule broken in a way that I felt added to a story: it may very well be the only truly unbreakable rule on this list. A number of excellent stories have had the appearance of breaking this rule, right up until the last page in which it was shown that there was a Rational Explanation after all!
These days you’ll just invite unflattering comparisons to Scooby Doo.
3. Not more than one secret room or passage is allowable.
The reader deserves to know the lay of the land as soon as possible. One surprise revelation, argues Knox, is acceptable in the name of an interesting plot: One thing presented to the reader can turn out to be wrong, discovered in the course of the detective’s sleuthing, but more than one wears thin. The reader winds up watching for hints of secret passages instead of paying attention to the text. I’d go so far as to say that any such cartographical twist ought not be introduced after the first two-thirds of the story. The reader does not have agency in the story to do their own search for secret passages, and if they become so commonplace that your readers are sitting there frustrated that your detective isn’t searching for them, then your book may become a projectile.
As far as science fiction goes, I think this can be extended to the laws of physics and the rules of the milieu as well: in the middle of the story, faster-than-light travel turns out to actually be possible? OK, there’s your one allowed “secret passage,” good luck with it. But there’s a caveat…
4. No hitherto undiscovered poisons may be used, nor any appliance which will need a long scientific explanation at the end.
Part of fairness is understanding what’s possible, a point I keep coming back to. But also, these things let the readers off the hook, lets them say, “Oh, I had a good excuse not to get that.” Rubbish! The whole point of knowing what’s possible is for the reader to have a slim chance of guessing the solution. This one’s a lot more important in science fiction, I think: it’s much easier for the reader to know what’s possible in an English country house than on a science station floating in interstellar space.
This suggests a refinement to #3: One twist is fine, but it had better not be the key to the whole mystery! The discovery of the secret passage should not obviously and immediately solve the mystery.
As to the second part, about needing a long scientific explanation: First, such explanations are tedious. Second, if it requires such an explanation, then you’re rules-lawyering, and it was never reasonable to expect the reader to guess it.
5. No Chinaman must figure in the story.
Er, wait. That can’t be right.
I suspect that Knox really means that there should be no “inscrutable foreigner” (a terrible term that assumes much) or other character who the audience can’t (?) relate to and so automatically suspects. Around the time he wrote this list, this was a popular ploy in the crappier mysteries. But really, he’d have been better served to consider that the foreigner who was so alien to the reader, and thus automatically suspected by the reader, was more often of a different class than a different nation! (“The butler did it!”)
Even so, this is kind of an awful rule. I’d rephrase it, “An inscrutable foreigner character, if one must exist, should not be the culprit.” Otherwise the author is pretty much let off the hook for a believable motive, and the reader’s only clue will be simple xenophobia. Then again, the presence of the foreigner can’t be an excuse not to cast doubt on some other character — readers who do not suffer from xenophobia will find such a story thin gruel.
Aliens in science fiction are in many ways the ultimate foreigner. Their motivations are by necessity weird and, y’know, “alien”. I’m not saying, though, that an alien can’t be the culprit. I am saying that to have an alien culprit in a fair way, the writer has a lot of work to do: the reader ought to be able to comprehend motives and capability. The culprit is so central a character to the story that a 2D caricature is a disservice to the reader.
6. No accident must ever help the detective, nor must he ever have an unaccountable intuition which proves to be right.
This one is damned hard to pull off. This commandment is most often partly broken in the “Dimwitted assistant accidentally says something that puts the detective on the right track” trope. Watson did this, Captain Hastings did this, Magersfontein Lugg did this, countless one-appearance befuddled Scotland Yard detectives did this — and then were obliged to feel good about having contributed! In some ways this is almost obligatory.
This trope can be done well, especially if it is a way of giving an extra clue to the reader. But it often is not, and just seems like an unfair insult to the poor benighted Dr. Watson.
Personally, I think of this rule as being just straight-up good storytelling. Any stroke of good luck that saves the day should turn out to be either built up to the whole time or turn out to be not so good after all. Anything else just reeks of deus ex machina to me.
In practice, however, there is frequently a single (even traditional!) caveat to this: There usually is a lucky break at the beginning of the story. One thing goes wrong for the killer to prevent the commission of a perfect crime. One lucky break for Great Justice that causes the detective to be called in at the beginning of the story, and handed a single thread to unravel.
7. The detective himself must not commit the crime.
This is related to the first commandment, as the detective is also often a focus for audience sympathy. This is not to say, however, that the detective must not commit a crime, just not the crime being solved. Otherwise you’re just toying with your readers, making them play catch-up.
Of course, if the detective does not know that he or she is the culprit, that could be interesting.
8. The detective is bound to declare any clues which he may discover.
This is an important one, and so often broken. The clues don’t have to be declared in a way that highlights them ([BLINK]CLUE HERE!!![/BLINK]), but the reader should have access to all the information the detective does, at about the same time (if not more and earlier!) Context is key during an investigation: you’ll read an interrogation differently if you already know that the suspect’s alibi has been blown. This does not extend to the detective’s suspicions and deductions: it is safe to keep them secret, but it must at least be clear in retrospect!
There is a gray area here: is the detective bound to declare any actions that he or she has taken? I recall a particular Agatha Christie short story where Poirot secretly instructed his valet to tail a suspect and collect some piece of evidence that he expected to be there. I’m on the fence with this one. On the one hand, telling the instructions would have given away the culprit immediately. On the other hand, I feel as though it should have been clued in better that he had made provisions in this way. Like I said, I consider it a gray area.
As far as science fiction goes, this rule means that relevant science and facts about the milieu have to be trotted out as early as possible. FTL travel, for example, should be established early if it is to feature in. It’s not fair for a detective (or anyone else) to sit and interrogate someone who obviously doesn’t know how anti-gravity works and not clue in any readers who don’t know either.
The converse situation, where the reader knows more than the characters, is dramatic irony. Nearly every mystery story has this on a re-read, though the inverted mystery genre has the benefit that every read is like this, and represents something of a special case. If the reader knows about a clue, then it seems reasonable for the detective to hide knowledge of it. On the other hand, because the reader witnesses the crime, it is important to account for how the evidence comes about!
9. The stupid friend of the detective, the Watson, must not conceal from the reader any thoughts which pass through his mind: his intelligence must be slightly, but very slightly, below that of the average reader.
If you’re going to have such a character, he needs to be a comfort to the reader, as a companion and guide through the universe being inhabited. Any story with a Watson tends to result in a thousand put-downs: if Watson’s ahead of the reader, those are going to sting. (Not that Conan Doyle refrained from that!)
This is not to say that Watson should not have information in his own right! Just that any information he has (such as resulting from his medical training) should be provided quickly and clearly to the reader. Having a Watson who knows the milieu well can be a real benefit in a science fiction story, but I would say that there’s a serious risk of being ham-handed here, overwhelming the reader with exposition. Tread lightly!
10. Twin brothers, and doubles generally, must not appear unless we have been duly prepared for them.
The reader is at an unfair disadvantage compared to the actual characters: even if a twin or double is suspected, remember that the reader does not have agency in the fictional world, and cannot look for evidence like stray birthmarks or the edge of a rubber mask.
The rule also applies, of course, to clones, holograms, and someone time-traveling from the future back to his own timeframe. However! Remember that as a corollary to rule #3, the writer is allowed one twist. The appearance of a twin or clone can be considered that one twist. Rule #3 will not prevent the reader from considering such a development to be ridiculous or hokey, however.
And that’s ten! If you read through that list thinking, “Argh! I can think of a dozen potential stories that break this rule to result in a good mystery!” then go write those stories! The rules aren’t hard and fast, and I believe that Agatha Christie went out of her way to break nearly every one. It wouldn’t surprise me if she used them as a guide, come to think of it. Like many rules related to writing, the key is not to avoid breaking the rules, but to only do so explicitly and with understanding of why they are there.
Of course, in all of these rules, there is one unspoken: The story must make sense. The murder must be solvable, but not so easy that the drama of solving it does not rely on lucky breaks or the detective merely being stupid. For a science fiction mystery specifically, method may need to be novel and clever indeed to fool the forensics tools available to a detective in the year 20X6! John W. Campbell thought it impossible, only to be proved wrong (repeatedly) by Isaac Asimov.