Here’s a full transcript of the ninth episode of Shedunnit.
Click here to listen to it now in your app of choice.
Caroline: A good detective story has a recognisable rhythm. The plot might have unexpected twists and the characters can surprise you, but there are certain structures and tropes that recur through much of the crime fiction from the first half of the twentieth century. Some of them have been parodied to the point of cliche, such as the old ‘the butler did it’ solution, but they are usually there nonetheless, providing the author with some creative constraints and the reader with a frame of reference.
Even if you aren’t a big reader of mysteries, these founding principles of the genre are so familiar that I expect you’d still be able to name a few: nothing supernatural, no secret twins, no springing clues or suspects on the reader in the final chapter — the list goes on. But how did these precepts come to be woven through the books from the golden age of detective fiction between the two world wars? And what happens when you break the rules?
Welcome to Shedunnit. I’m Caroline Crampton.
My obsession with what is and isn’t allowed in a detective story began with A. A. Milne. Although he is best known now for creating The Hundred Acre Wood and its residents Winnie the Pooh, Piglet, Eeyore and co, Alan Alexander Milne was also a journalist, playwright and novelist publishing work both before and after the First World War. He also had a passion for detective stories and in 1922 published one of his very own, The Red House Mystery, in which the host’s long lost brother is found shot during a country house party, and two of the guests turn to sleuthing to solve the puzzle. It was a great success, being reprinted many times, and is still read today.
Four years later, he wrote a new introduction for the 1926 edition, and in it Milne set out his own “curious preferences” about what a detective story can and can’t be. He wanted his whodunnits written in plain English, without the intrusion of a romance plot, starring an amateur detective who works just with logic and reasoning rather than specialised scientific knowledge or equipment. There must also be a ‘Watson’ character, via whom the detective can narrate his sleuthing progress, who must be neither too quick to catch on nor a total fool (in fact, a lot like the original Watson in Sherlock Holmes, perhaps).
Milne hated the final chapter reveal, in which the detective proudly unveils the solution to a throng of other characters, which is invariably based on a whole load of clues the reader had never heard of before. He wanted readers to feel that they had a fighting chance of solving the mystery for themselves; that the important clues had been dropped like breadcrumbs through the whole text, if only the reader was smart enough to work out which ones mattered and which ones didn’t.
There’s plenty I agree with here – I too like the illusion that I could outwit the detective — and some that I don’t (I think some of the best detective stories have a romantic element, as Dorothy L Sayers later proved). But what first captivated me about Milne’s essay wasn’t the specifics of what he outlined, because after all he made clear he was just addressing his personal preferences, but rather the seriousness with which he had considered the formal structures of detective fiction as a form. So-called “genre” writing has always suffered from the perception that it isn’t as important or worthy as highbrow literature, but here was a respected author really getting stuck into the tropes and conventions that underpinned this kind of writing.
And he was far from the only one. Plenty of other writers, both at the time and later, have tried to lay down “rules” for detective fiction. T. S. Eliot was one such — he was a big fan of the whodunnit too, and used to review new detective fiction in his literary journal The Criterion (although not under his own name). He even described himself in a letter to his friend Virginia Woolf as a “person who specialises in detective stories and ecclesiastical history”, and I don’t even think he was joking, although by that time he’d published Prufrock, The Waste Land and other major poetic works.
Eliot’s own favourite story was The Moonstone by Wilkie Collins, which although it was published in 1868, long before the advent of detective fiction’s golden age in the early 1920s, Eliot considered to be the first and best example of the form. (This is also an opinion that was shared by G. K. Chesterton, author of the Father Brown stories and the first president of the Detection Club.)
From the work of Wilkie Collins and his familiarity with the novels that followed it, Eliot evolved a few rules of his own about how a whodunnit should be put together. He banned elaborate disguises, supernatural incidents and bizarre coincidences, and insisted on a clever detective who is not so brilliant that it comes across as some kind of superpower. Most of all, though, he wanted the criminal’s motive to be normal or logical, and for the reader to feel that they had a sporting chance at finding the solution.
On the other side of the Atlantic, the American art critic Willard Huntington Wright published in 1928 under his regular pseudonym of SS Van Dine an essay titled “Twenty Rules for Writing Detective Stories”, which has often been quoted and cited since. He has many of the same preferences as Milne and Eliot as far as romance, transparency and rationality, but he also had some more specific (and funny) hang ups. No secret societies, no murderers who are also domestic servants, no long ‘atmospheric’ passages of writing, no professional criminals, no fake seances, no code letters, no knockout drops or hypodermic syringes, no cigarette butts as evidence. . . The list goes on and on and on.
As listeners will have no doubt have spotted, Van Dine’s rules were broken left, right and centre. Just from those few points I listed, I can think of popular and acclaimed detective stories that include those specific elements, from the scene setting in The Moonstone itself, to the professional criminals that Margery Allingham’s Albert Campion encounters, the fake seance in Dorothy L Sayer’s Strong Poison, to the possibly coded letters in Agatha Christie’s The ABC Murders, the hypodermics in Sayers’ Unnatural Death, and finally Sherlock Holmes’s own preference for using cigarette ash as a major element of detection.
This is the key point, I think — these rules were never written to be taken very seriously. Ronald Knox, himself a detective novelist as well as a Catholic priest, wrote a rather more tongue in cheek list of ten ‘commandments’ for detective novelists that were published in the introduction to the Best Detective Stories of the Year anthology in 1928. Several things we’re now very familiar with are prohibited by his list too, such as twins, ghosts, multiple secret passages, overly smart Watsons and concealed evidence, but he also banned detectives who are also murderers and Chinamen.
This last was a reference to the racist stereotypes prevalent in the popular thrillers of the time, where mysterious Oriental villains from smoky Limehouse opium dens abounded. Knox himself said in the same piece that too many rules could cramp an author’s style — he clearly never intended detective fiction to become some kind of tick box exercise. Putting down these ideas was just another way of recognising the popularity and legitimacy of the form.
An idea that all of these rule-makers had in common, though, was that of “fair play”. Indeed, it was such a foundational part of the style in this period that the first item in the constitution of the Detection Club, to which Christie, Sayers, Marsh and others all belonged, says that “it is a demerit in a detective novel if the author does not ‘play fair by the reader'”. This comes back to that sense that T. S. Eliot reference of wanting the reader to have a “sporting” chance at solving the crime for themselves; indeed in his history of this time, The Golden Age of Murder, Martin Edwards links this desire to the idea of being honourable on the cricket pitch, of that old fashioned English idea of “playing the game” and not deceiving anyone.
The kind of detective novel that most closely adheres to the notion of fair play is the pure puzzle, where the whole setup of the crime scene is described to the reader so they can form their own deductions. Dorothy L Sayers didn’t often write this kind of book, since she was usually breaking rules all over the place with her seances and her romances and her elaborate disguises, but in Busman’s Honeymoon she did have a go at it. This story actually started life as a play that she later turned into a novel, and is the last full-length appearance of Peter Wimsey and Harriet Vane. It features that classic murder plot — the locked room, or house in this case — in which the victim seems to have had his head bashed in somehow while completely alone in a secured dwelling.
Pretty much everything you need to know to solve this one is there in the first few chapters, but I’d be very impressed if anyone manages to do it without any prior knowledge of the twist at all (if you do, write in and tell me how, I’m on firstname.lastname@example.org). That said, when I’ve reread this book since, some of the clues do seem a bit obvious and clunky, so maybe it isn’t as impenetrable as I think.
Another writer who was very interested in maintaining a sense of fair play was John Dickson Carr, who even included a meta discussion of it in relation to locked room murders in his 1935 novel The Hollow Man. It’s the most extraordinary scene in which Carr’s protagonist Dr Gideon Fell delineates all the different ways a supposedly ‘locked room’ mystery can be engineered, with commentary about which ones are more or less common or fair. He even says “we’re in a detective story, and we don’t fool the reader by pretending we’re not”. Carr was a great admirer of G.K. Chesterton, who was himself a great proponent of fair play — it’s even been suggested that his description of Fell was meant to suggest that he looked like the great creator of Father Brown.
J. J. Connington, a favourite of T. S. Eliot’s, even went so far as to include a “clue finder” appendix in his 1929 novel The Eye in the Museum, which gave the page numbers of all the major clues so that after they’d read the solution, the reader could go back and check up on all the hints that they missed. He, and the handful of other authors who used this device like Ronald Knox, really wanted people to know that they were trying to stick to the rules.
But what about when authors threw the rule book out of the window? I think there are two good examples of this, the first a single book and the second an entire career. The first is, of course, Agatha Christie’s 1926 novel The Murder of Roger Ackroyd, which is perhaps the finest example of the exact opposite of fair play (although some critics maintain that it doesn’t completely break the rules). When it was published, it caused a bit of a stir for its rule-breaking structure (I’m not going to say anymore, because I don’t want to ruin the first time shock for anyone who hasn’t read it yet, but please do seek it out) and its disgruntling effect on some readers became well-known enough that when the American critic Edmund Wilson wrote a grumpy essay about how he didn’t see the point in detective fiction in 1945, it was headlined “Who Cares Who Killed Roger Ackroyd?”.
I think Christie enjoyed experimenting and causing a fuss; lots of her books are certainly full of smaller examples where she deviates from the rules, not least because of Hercule Poirot’s love of keeping the workings of his little grey cells a secret until he’s absolutely sure of his solution. Five Little Pigs is another good example of inverted fair play, because that’s a novel in which Poirot is called into reinvestigate a case from decades before, and he has no evidence to go on beyond the psychological inferences he can make about the people involved. This style of story is sometimes referred to as that of an “armchair detective”, because there’s no need for any energetic sleuthing to crack the mystery. I’m not sure that T.S. Eliot would have been madly keen on that as a way of telling a detective story, but I personally think it’s one of Christie’s stronger books — she apparently created it as a challenge to herself to see if she could pull off a plot without the more conventional elements of an investigation like fingerprints to help her.
The second major contradiction of the fair play doctrine I think comes in pretty much all the work of Margery Allingham. Her books do have many of the trappings of the conventional golden age mystery, such as the singular detective in Albert Campion, the country house and upper class settings, and so on, but you only have to read the first Campion novel, The Crime at Black Dudley from 1929 to see how quickly she chucks all notions of fair play out the window. Campion disappears at a crucial moment in the plot, and then reappears several chapters later, and doesn’t even then really explain what he’s been up to until the very end.
Allingham said that she wrote this book via what she called the “plum pudding” method, in which anything can be stirred into the mixture to enhance its richness. That’s certainly how the novel feels, as more gangsters and ancient curses turn up. There’s something of the P.G. Wodehouse style romp to some of her books, and she did also really like to hint at the supernatural, such as in 1931’s Look to the Lady. This is something that she had in common with Gladys Mitchell, who also liked to make reference to witchcraft and folk customs in her detective novels, and didn’t particularly trouble herself about whether her sleuth Mrs Bradley’s methods were always completely fair and transparent to the reader.
Part of the focus on fair play at this time stems from the other kinds of puzzle games that were popular, like crosswords, mahjong, treasure hunts and so forth — the classic, truly honest fair play detective novel should be as easy to solve as a crossword with all the clues listed underneath it. And while there are novels that manage to do that while also creating a story that’s exciting and enjoyable to read, plenty of authors clearly struggled under the constraints and ended up prioritising their puzzle plot over everything else.
Allingham described the construction of a mystery story as a process of building a box with four sides, made up of “a Killing, a Mystery, an Enquiry and a Conclusion”, and said that the box could be both “a prison and a refuge”. I think what she meant by that was that the restrictions of the form could both prompt her to be more inventive, but also curtail some of her more outlandish ideas. I’m glad to say though that she didn’t allow it smother many of her stranger ideas — with authors like Allingham and Mitchell particularly, it’s the moments when they break free of the rules that I most enjoy their work.
I don’t think anybody other than a few extremely grumpy critics has ever put down a detective novel and refused to read further after discovering it doesn’t completely adhere to the idea of fair play, but I do still sometimes observe the traces of this attitude when it comes to contemporary television adaptations of these stories. There’s an element of the audience who want to judge a TV version of a story by how faithful it is to the source book, and I have seen people post on social media about turning off an episode in disgust halfway through because an element of the plot has been changed or approached in a different way.
Of course there’s nothing wrong with having strong feelings about the quality of how something was made, but when I see that kind of view expressed, it does make me think back to the rules of SS van Dine and others, and wonder how differently detective fiction would have developed if everyone had always coloured inside the lines, rather than extravagantly slopping paint everywhere just to see what would happen. I suspect that everything would have been much flatter, more conventional and less captivating if they had resisted experimentation in favour of obedience.
After all, everybody knows that rules are made to be broken.
This episode of Shedunnit was written, narrated, and produced by me, Caroline Crampton. You can find more information about all the books that I’ve mentioned in the show notes for this episode at shedunnitshow.com/therules. There, you can also read a full transcript.
Something both really surprising and lovely has been happening over the past few weeks — I keep getting these emails from Paypal telling me that listeners have sent the show money. Thank you very much to everyone who has donated, it’s really kind of you and all helps keep the wolf from the podcast’s door. If you’d like to join in, you can head to shedunnitshow.com/donate to send me your loose change, or as per lots of requests, I have now set up a wishlist so you can buy me books to help research future episodes. You can find that at shedunnitshow.com/wishlist..
I’ll be back on 20 February with a new episode.
Next time on Shedunnit: The Other Detectives.