One of the most refreshing things about detective fiction is the high prevalence of women within it, both inside and outside of the pages. By nature, detective fiction tends to pass the Bechedel test, even when it was written at a time when fiction generally didn’t. Think of Golden Age fiction, and there’s a good chance that you’re thinking of works by women: Christie, Allingham, Sayers, Marsh, Heyer, Orczy (creator of the first armchair detective), Highsmith, et al. Under the name Carter Dickson, John Dickson Carr was the fourth Penguin mystery author to have ten books specially released in a year: the first three were Agatha Christie, Ngaio Marsh and Margery Allingham. And now Carr is virtually out-of-print – two of the first three are not. The detective-fiction-cum-thriller genre, the sort that characters buy in railway station bookshops, hasn’t aged well. Its books are out-of-print, its authors forgotten. You may wander into a bookshop and find a Crispin or two, but he’s an author praised for his writing not his plots. Conan Doyle is still around – but is he? He’s Sir Arthur Conan Doyle the creator of Sherlock Holmes, not Sir Arthur Conan Doyle the detective novelist, and there’s a key difference. But Christie, Sayers and Allingham survive. Why?
The interesting thing, that becomes clear if you go into a second-hand bookshop, is that of all the books written at the time, some date very badly (even Carr can be a culprit) and some do not. Yet, naturally, the ones that we remember are the ones that are timeless or, in the case of Sayers, ahead of their time. And of those, women dominate. Why?
There is a good rule of thumb – and, like most, it comes with an uncountable number of exceptions – which is that men write plot and women write character. It’s tempting, at this point, to start crying ‘sexist!’, but take a moment to consider it. Broadly, mystery fiction is split into two genres: the ‘whodunnits’ and the lesser-known category of ‘howdunnits’. And, though the two often merge and there are several exceptions, there’s a clear gender divide between the two.
‘Whodunnits’ are self-explanatory, and they’re what we think of when we think of traditional detective fiction, and would perhaps be better termed ‘whydunnits’. Read an Agatha Christie, or a Dorothy L Sayers, and you’ll be weighing up character’s possible motives more than their opportunity. The ideal Poirot story, as said by the character himself, is Cards on the Table: a crime without evidence, where the deduction is purely psychological. Any of the four suspects had equal chance to commit the crime but, in terms of character, only one of them would commit that crime. Marple takes this to a further extreme: essentially deductions according to Myers-Briggs types. Dorothy L Sayers has always been more of an author of novels that included crime rather than a crime novelist, but consider Whose Body. Initially, this seems like a ‘howdunnit’: an unidentified body turns up in a bathtub in the home of a couple who swear they’ve never seen the man before (a not to dissimilar beginning to The Body in the Library). But its conclusion, Lord Peter realises who the criminal by adding up all the clues till he thinks of the only person who wants to do it. The solution to ‘Gaudy Night’, meanwhile, comes from identifying the type of character and experience that would want to send poison pen letters: unthinkable in Holmes, unthinkable in Carr.
‘Howdunnits’, on the other hand, are exactly the opposite: focussing not on the murderer but on the mechanics of the crime, and are better-known as locked-room mysteries (strictly speaking subcategory). Think not just of Carr, but of Jonathan Creek, or even of Sherlock Holmes. In these, there will be a murderer but the character and motivations of the murderer are never the focus. Because the driving factor of the story is the crime itself.
the lead villain in, say, Conan Doyle’s ‘Silver Blaze’ did it is entirely spurious. How the crime was committed takes up the entire story. The early Poe stories – The Murders in the Rue Morgue, The Purloined Letter – are the earliest examples of detective stories based upon mechanics. We wonder how two people are brutally murdered in a locked room, how the purloined letter is concealed. And, once you know how, there’s little pleasure in reading them again. It’s exactly the same in almost all Holmes: we wonder why there’s a Red-Headed League, what is the speckled band, what is the meaning of the Dancing Men. And, without an explanation, there’s the thrill of not knowing, the hint of the supernatural. And, like with ghost stories, the eventual explanation always feels like something of a let-down: for the same reason that the explanation of a magic trick is never as interesting as the trick itself.
The reason why The Hound of the Baskervilles bears re-reading, why it alone is the story you connect with the name ‘Sherlock Holmes’, is because it combines both the who and the how, because you can only imagine the story working with that murderer and when you reread it you can see the clues – almost uniquely among the Sherlock Holmes stories, it’s closer to Christie than it is to Carr.
And once you understand this, you understand why, generally, the best format for Sherlock Holmes was short stories. After the how? question was asked, the audience were intrigued but didn’t have time to get bored (and bear in mind that while we call A Study in Scarlet a full-length novel now, it’s damn thin at 50,000 words and was originally published as part of an anthology). It’s also why, at the other end of the scale, the best format for Agatha Christie was full-length novels, as it gives us time to fully explore all the characters – the Poirot and Marple short stories are spurious at best.
Wikipedia defines ‘surprise ending’ as ‘a plot twist occurring near or at the conclusion of a story, an unexpected conclusion to a work of fiction that causes the audience to reevaluate the narrative or characters’, and that pretty much sums it up. The twist endings of howdunnits are the literary equivalent of pulling rabbits out of hats, and once we know how the trick was worked – and there’s seldom any way we could guess – there’s nothing more left to see. A character-based ending, on the other hand, is something that the writer has to build up to: pointing at Character X and saying ‘oh yes, it was them, what a surprise!’ doesn’t work if it could equally have been Characters Y, Z and Q. By the end of a good character-based mystery, you know that the murderer could only have been that character and not anyone else: but you’re still surprised. To go back to the example of Whose Body: with regards to the ‘howdunnit’, any explanation as to how the body got into the bathroom would make sense: but it would take a complete rewrite to change the murderer and the reason for the crime. With almost Conan Doyle or Dickson Carr, you could change the murderer’s identity completely and the story wouldn’t be affected, as it genuinely isn’t the point of the novel. A story with a twist ending works best with the suspense that comes before the trick is explained; a story that’s character based leaves you genuinely satisfied with the revelation because you knew that the signs were there.
And this is why some detective fiction survives and some doesn’t. For Sherlock Holmes the format and character survive: the man solving mysteries based on clues no-one else can perceive. The two-pence thrillers sold in railway stations haven’t disappeared, they’ve just been replaced by the next generation: still sold by Smith’s, they can now get away with slightly more sex and slightly more violence, but people read them for the same reasons. John Dickson Carr, master of the locked room mystery, may have gone from our shelves but his spirit lives on in Jonathan Creek.
But Dame Agatha and Lady Dorothy live on, still in print, still adapted: and it’s their stories themselves that remain, not just the lead detective or the format. Because they wrote character-based stories and those stories remain timeless. Which brings us back to gender, and explains why we have many Queens of Crime, but never yet a King.