Rule Makers and Rule Breakers of the Golden Age

A.A. Milne

S.S. Van Dine

Ronald Knox

Agatha Christie

Dorothy L. Sayers

This article is written from research done for the paper I delivered at Bath Spa University Crime Writing Conference. Captivating Crime 2: Traditions and Transgressions.
WARNING: This article contains some spoilers.

When we speak of the Golden Age of crime fiction the first names that come to mind are Agatha Christie and Dorothy L Sayers, closely followed by Ngaio Marsh, Margery Allingham and Josephine Tey. In the late 1920s Christie and Sayers were both well into their detective fiction careers, but they were not the writers who decided to set down a list of rules for the genre. Of course, the position of women was very different  then it wasn't until 1928 that all women in Britain were allowed to vote, and in 1918, when the first law was passed to enfranchise some women, neither Christie nor Sayers were old enough to vote. My instinct is that neither woman would let the expectations of society prevent them from doing what they wanted, certainly not Sayers, with her strong Feminist principles. Perhaps it was simply that they were too busy writing and stretching the boundaries of the genre to bother about formulating rules.
So who were the men who felt themselves uniquely fitted to the challenge of creating a set of rules for writing crime fiction? Who else could it be but a Roman Catholic priest and an American art critic?
However, they were not the first. In 1922 AA Milne had published his only detective story, The Red House Mystery, and, in the 1926 edition he included a foreword that laid out, with his characteristic humour, the things he felt essential for a detective story.
1. The use of straightforward language.
2. Clues to be revealed throughout the book not kept to the last chapter.
3. To do this he advocated the use of a Watson figure, so that clues should be presented in accessible dialogue form.
4. Milne also preferred no love interest in detection books, because he felt romance detracted from the business in hand.

In 1928 a group of mystery writers came together and formed what, a couple of years later, became known as the Detection Club. One of the members, Ronald Knox, created ten rules of detective fiction, which he ironically called the Ten Commandments. Knox was a brilliant scholar who converted to Roman Catholicism and became a priest. As well as being an eminent theologian, he wrote essays, pastiches, and broadcast on numerous radio programmes.
Knox had a dry, sense of humour and an unfortunate lack of appreciation of the fact that other people didn't always realise when he was joking. In 1926 he made a radio broadcast entitled Broadcasting From the Barricades, which described a scene in which London was being destroyed by a rampaging mob and a Government Minister was hanged from a lamp-post. It should have been clear that this was a spoof: it was announced at the beginning, the names were absurd and so were the government departments, but many people tuning in slightly early for the following programme were alarmed. After all, the BBC was not expected to make jokes and it was earlier in the year of civil unrest that led to the General Strike. And it was only a few years after the First World War, and many people were attuned to expecting loss. Knox's recording career was tarnished by this practical joke.
Knox was also hoist by his own petard when in 1911, he delivered a satiric lecture, followed by an essay: Studies in the Literature of Sherlock Holmes which claimed that Holmes had really died at the Reichenbach Falls and the later stories were the fabrications of an alcoholic and delusional Watson. From that time Knox was regarded as an expert on Holmes and continually appealed to for essays and talks. In The Life of the Right Reverend Ronald Knox (1959) Evelyn Waugh reports Knox's response to one such request: 'I can't BEAR books about Sherlock Holmes. It is so depressing that my one permanent achievement is to have started a bad joke.'
In 1925, Knox started writing detective novels in an attempt to increase his income. In 1927 he introduced his series character, Miles Bredon, an investigator for the Indescribable Insurance Company.
It is generally accepted that Knox had no intention of his Ten Commandments being carved in stone. They show the humour and intelligence typical of Knox, mocking many of the crime writing clichés of the time, such as an over-abundance of secret passages, identical twins, supernatural agencies and unknown poisons.
Despite his rejection of Holmes and Watson, he agrees with Milne about the value of a Watson-like figure.
His rejection of the use of Chinamen is especially interesting. The Yellow Peril was over-used in crime fiction at this time but he is also making a serious social point. As the historian Julia Lovell wrote in 2014: 'In the early decades of the 20th century, Britain buzzed with Sinophobia. Respectable middle-class magazines, tabloids and comics alike spread stories of ruthless Chinese ambitions to destroy the west.'
Knox does keep his own commandments, although I think he may fall into Temptation when it comes to number ten, banning twins and doubles, when the whole basis of one of his plots depends on fraudulent identity and passport photos where one person closely resembles another.
A year later, the American crime writer, S. S. Van Dine made his comprehensive list of Twenty Rules for mystery writing. Under his real name, Willard Huntington Wright, he was a literary critic and journalist and an art critic. Initially quite successful, his career had plummeted, due to his aggressive personality and his support for Germany in World War I.  In 1923 his health broke down; according to his biographer, John Loughery (Alias S.S. Van Dine, 1992), his collapse was due to his secret addiction to cocaine. Confined to bed, he read hundreds of the books he had always claimed to despise, detective stories. As a result he planned his first three detective novels and wrote an essay which explored his opinions of the history, traditions and conventions of detective fiction.
In 1927 he published the essay, The Great Detective Stories, under his own name of Wright. In it he treated the work of Agatha Christie to harsh criticism. Although Christie had, at this time, published six detective novels, only the Poirot novels and short stories are selected for mention in Wright's essay. In it he refers to Poirot as a 'pompous little Belgian sleuth' and claims that, 'Poirot is more fantastic and far less credible than his brother criminologists.... and the stories in which he figures are often so artificial, and their problems so far fetched, that all sense of reality is lost, and consequently the interest in the solution is vitiated.' Wright becomes positively outraged when referring to the solution of The Murder of Roger Ackroyd (1926), which he describes as a 'trick' and not a 'legitimate device of the detective-story writer.' Compared to the rest of the essay, the references to The Murder of Roger Ackroyd ferment with wounded vanity and it seems probable that Christie's device had fooled Wright and he was not going to forgive her 'trickery'.
Now consider this description of a fictional detective: wealthy, cultured, well-born, prone to quotation, educated at Oxford University and liable to use affected mannerisms in his speech, wears a monocle, a bachelor, living in comfort and affluence, waited upon by a devoted English servant, a connoisseur of good food and drink, served with distinction in the First World War. He is fond of interesting himself in police investigations and is allowed an extraordinary amount of access and influence by the official investigators. That sounds like Lord Peter Wimsey, but it also applies to Van Dine's fictional hero, Philo Vance.
Dorothy L. Sayers' first Peter Wimsey book, Whose Body? had been published in 1923, but in The Great Detective Stories, both the author and her creation, are skimmed over in one sentence, 'Lord Peter Wimsey, the debonair and deceptive amateur of Dorothy L. Sayers's Whose Body?'
Of course, I cannot prove that Van Dine based Vance on Wimsey, but I do feel that the points of similarity are too numerous to be a simple coincidence.
Of course,Vance was different to Wimsey in many ways, not least his all-consuming snobbery: 'An aristocrat by birth and instinct, he held himself severely aloof from the common world of men. In his manner there was an indefinable contempt for inferiority of all kinds. The great majority of those with whom he came in contact regarded him as a snob. Yet there was in his condescension and disdain no trace of spuriousness. His snobbishness was intellectual as well as social. He detested stupidity even more, I believe, than he did vulgarity or bad taste.' (The Benson Murder Case, 1926.)
Contrast this with Wimsey in The Haunted Policeman (1939), the penultimate short story that Sayers wrote featuring him. Wimsey has been happily married to Harriet for just over a year and she has just given birth to their first son. All the household has retired to bed and Wimsey has invited a chance-met, off-duty police constable to join him to celebrate. '”Quite right,” said Peter, and filled the glasses again. He found the policeman soothing. True to his class and training, he turned naturally in moments of emotion to the company of the common man. Indeed, when the recent domestic crisis had threatened to destroy his nerve, he headed for the butler's pantry with the swift instinct of the homing pigeon. There, they had treated him with great humanity, and allowed him to clean the silver.'
Van Dine may have intended to be witty but many of the Twenty Rules sound elaborate and pompous, and, in fact there are really twenty-nine rules, as he fits in several extra rules under the heading Rule 20.
Van Dine and Knox agree regarding fair play in giving the readers all the clues, forbidding supernatural agencies, and Van Dine's bête noir, the detective must not commit the crime. However, Van Dine goes much further and decrees that murder is the only crime worthy of a detective novel and it should never prove to be an accident or suicide. He bans secret societies, camorras, mafias and professional criminals from committing the murder. Also he declares that the murderer must be a decidedly worth-while person and no servant or lower-class person need apply.
Knox certainly contravened Van Dine's rule regarding murder: in one of his books the person believed murdered was not dead at all. But did Van Dine keep his own rules? In my opinion, he did not. Rule 16 states that the detective novel should contain no long descriptive passages, no literary dallying with side-issues etc. Van Dine spends pages describing the glories of Philo Vance and his possessions and as long again with Vance showing off with literary quotations of  dubious relevance. Even worse, Rule 5: The culprit must be determined by logical deductions – not by accident or coincidence or unmotivated confession. In The Canary Murder Case (1929) Vance and his companions revisit the scene of a locked room mystery and, wishing to relax, Vance selects a record labelled as classical music to play on the phonograph. The record, by chance proves not to be music but a recording of the victim's last words,which were spoken from behind a locked door. This destroys the main suspect's air-tight alibi. It also damages Van Dine's credibility.
The Rule most frequently broken by other Golden Age writers was the prohibition of a love interest in detective novels. Of the rule makers, only the celibate priest, Ronald Knox did not disapprove of romance. Both Milne and Van Dine felt that love proved a distraction from the main business in hand, although they phrased this objection very differently:
Van Dine: 'There must be no love interest. The business in hand is to bring a criminal to the bar of justice, not to bring a lovelorn couple to the hymeneal altar.'
Milne: 'A reader, all agog to know whether the white substance on the muffins was arsenic or face-powder, cannot be held up while Roland clasps Angela's hand 'a moment longer than the customary usages of society dictate.' Much might have happened in that moment, properly spent; footprints made or discovered; cigarette-ends picked up and put in envelopes. By all means let Roland have a book to himself in which to clasp anything he likes, but in the detective story he must attend strictly to business.'
So many Golden Age writers broke the romance rule that it is impossible to list them: Christie, Sayers, Marsh, Allingham, Tey and Heyer all had romances in many of their books. Sayers, Marsh and Allingham had their sleuths fall in love, eventually marry, and continue detecting. Patricia Wentworth divides the action between her elderly female sleuth and a variety of young lovers who come to her for aid. So did Christie with her lesser known series detectives, Tommy and Tuppence Beresford. And it was not just women writers: Michael Innes had Appleby fall in love and marry, and the love John Creasey's heroes had for their wives dominates his books. Of course, Leslie Charteris turned the matter around: the Saint books started with him meeting and falling in love with Patricia Holm, sharing several adventures with her, then fading her out, so that Simon Templar can enjoy encounters with other women.
In her full-length novels Christie sticks to murder, although several of her short stories feature lesser crimes or no crimes at all. However, Josephine Tey clearly did not feel that a good mystery had to have a corpse and wrote several books, such as Miss Pym Disposes (1946), The Franchise Affair (1948), and To Love and Be Wise (1950), in which either the death was unintentional or no death occurred. As the perpetrator in To Love and Be Wise remarks, 'It's a bit of an anti-climax, isn't it? To set out to kill someone and end with a breach of the peace.'
If, as I suspect, Van Dine did model Philo Vance on the early Peter Wimsey, I would have enjoyed seeing his reaction to the penultimate Wimsey novel, Gaudy Night (1935.) In it Sayers broke: Rule 3, a love story; Rule 7, no corpse; Rule 9, two detectives; Rule 11, a servant did it; Rule 16, exploring side-issues.
Sayers proved to be an equal opportunity author in her final Wimsey novel, Busman's Honeymoon (1937), when again a servant committed the crime. And in Five Red Herrings (1931), the death was tried as manslaughter rather than murder.
Also she bent Rule 20j in Have His Carcase (1932), when Wimsey and Harriet Vane decode a letter which has an important bearing on the plot.
Both Sayers and Christie use the identical twin motif in short stories.
In the 1920s and 1930s spiritualism and seances were popular both as a serious attempt by the bereaved to contact loved ones and as an entertainment. As well as short stories based on spiritualism, Christie uses seances and the ouija board to either set the scene for a murder, as in The Sittaford Mystery (1931) and Dumb Witness (1937); to obscure the truth: The Pale Horse (1961); or to startle a killer into confessing, as in Peril at End House (1932), which definitely breaks Van Dine's Rule 20b. Sayers also uses a phoney séance in Strong Poison (1930), although not to gain a  confession.
As for the Commandment/Rule that Knox and Van Dine agree about: that the detective or official investigator or character whose thoughts we have been in should not commit the crime. Both Agatha Christie and Georgette Heyer had a police officer committing the crime: Christie in Hercule Poirot's Christmas (1938) and Heyer in A Blunt Instrument (1938), but when it comes to breaking the record for contravening this Commandment and Rule, the prize must go to Agatha Christie. At the beginning of her career, with the narrating character in The Murder of Roger Ackroyd (1926); also with the narrating character of her dark, stand-alone novel, Endless Night (1967); and in Curtain (1975), the last Poirot novel, written at the height of her career and stored in a safe until near the end of her life.
I do not believe that any of the Golden Age greats deliberately went out of their way to break the rules, they just had fun playing with the permutations of the genre. Christie was particularly inventive, which meant that she broke the rules with exceptional panache. As Anthony Horowitz said of Christie, 'It is as if she was given five or six basic stitches and somehow used them to create the Bayeux Tapestry.' (Introduction to the Hercule Poirot Novels Collection published by the Folio Society.)
In the end keeping the rules or breaking them does not matter, it's the resulting book that counts. As AA Milne said in his 1926 prologue: 'The only excuse which I have yet discovered for writing anything is that I wanted to write it.'