Dorothy Sayers (essay date 1932)
SOURCE: Sayers, Dorothy. Introduction to The Second Omnibus of Crime, 1932, edited by Dorothy Sayers, pp. 1-16. New York: Coward-McCann, Inc., 1932.
[In the following essay, Sayers elucidates the defining characteristics of the detective story and the virtual disappearance of the detective short story genre.]
Ten years ago, or rather more, a friend and myself, being stranded in London on a wet Sunday with nothing particular to do, went down to one of the great termini, to see if we could find a detective-story to read aloud. I remember that we started with no very high hopes. The book-stall was more likely, we thought, to be stocked with “ordinary novels.” We were lucky. We found Cleveland Moffet's Through the Wall, and we congratulated ourselves. Today we should be surprised if we could not choose between a score or so of detective-novels, all reasonably well written and issued by first-class firms.
Within the last decade, the detective-story has, in fact, sprung into a popularity and a respectability almost alarming to its friends. It is pleasant, of course, to feel oneself respectable and respected, but all privilege carries with it its own responsibilities. Even within the few years which separate this volume of Great Short Stories of Detection, Mystery, and Horror from its elder companion, an increased solemnity has come over both the writers and the critics of detective-fiction. Mystery-mongering has become self-conscious; monographs and critical studies are appearing. Here and there a really advanced and intelligent reviewer will even make a distinction between the detective-story proper and the “blood”; although his efforts are usually frustrated by the caption-artist, who contrives to lump together a mathematical problem by Freeman Wills Crofts, a curious study of murderous psychology by C. S. Forester, a spy-story by “Sapper,” and a book about Chicago gunmen by Edgar Wallace, under the generic title: “Four Holiday Thrillers for Murder Fans.”
Up to a point this serious approach is all to the good. The literary technique of the detective-story has improved out of all knowledge. To-day, when I hear somebody say that he never reads detective-stories “because they are so abominably written,” I know that he has simply not been trying, and is probably basing a generalisation on some remark made by Mr. Gladstone in 1882. Here and there a writer, otherwise careful and conscientious, still clings to the old clichés: the clues are “examined with meticulous care,” the case is still “one of the most baffling and sinister that Detective Jones had ever investigated,” the criminal still “slips through the meshes of the police-net,” the discoverer of the corpse is still “stunned by the shock of this horror,” and while our old friend the “blunt instrument” is now seldom mentioned except within the inverted commas of implied ridicule, our yet older friend the “sickening thud” still occasionally makes its solemn and melodramatic appearance. But actual bad grammar and bad English are becoming rare in the serious detective-story; the worst fault of style is now mere flatness and lack of distinction, while in many of the more modern tales there is a genuine attempt to create real atmosphere. Occasionally this results in over-writing or, what is still more dangerous to the detective-story, in elaborately “clever” writing, but, in spite of failures here and there, it is evident that there is a genuine eagerness to bring the detective-story into line with the traditions of the English novel and to make it increasingly a real part of literature.
A year or two ago, it was confidently predicted that the detective-story was going to slump heavily. It was pointed out that all the possible combinations and permutations would shortly be exhausted. This point of view was rather strengthened than otherwise by the appearance of a number of critical studies in which the genre was analytically examined. It is painfully true that, generally speaking, things are not analysed until they are dead, and the friends of the detective-story began to feel that perhaps this sudden critical awareness was merely a symptom of approaching decay.
So far, however, publishers' sales do not support this rather depressing theory. In spite of unfavorable world-conditions, they show, in this country at any rate, a solid and satisfying steadiness, and this is a healthy sign. Setting aside that hardy perennial, the book-stall shocker, which is skimmed through and left in the hotel, there is a “firm demand” for the detective-mystery of the more rational type, the figures of whose sales vary little from month to month. Authors of established reputation are able to go quietly on, perfecting themselves in the manner expected of them, without being obliged to produce at too quick a rate in order to take advantage of a fleeting fashion.
It may be said that this circumstance tends to the production of the machine-made story. That is, of course, true to a certain extent. The detective-story, however, suffers less than most from being produced in this practical and unromantic manner. It is none the better for being thrown off at a white heat. In fact, it cannot be so produced. The central idea may be the result of inspiration, but the details must be worked out in a mood of calm and almost scholarly leisure. To get the central idea is one thing: to surround it with a suitable framework of interlocking parts is quite another. To judge by the questions asked from time to time of mystery-novelists about their methods of work, it is not always realised that idea and plot are two quite different things. Frequently a story with a brilliant central idea is ruined by failure in plot-construction, and it is here that over-production is apt to have fatal effects upon a writer's work. Perhaps a brief outline of the way in which the average detective-novel is put together may be of interest, and may help to throw a little light forward upon the future possibilities in store for the detective-story.
The central idea, as a rule, is the first thing that presents itself. It may be a new method of killing somebody; it may be an ingenious method of constructing an alibi, or outwitting the law; it may merely be the fascinating mental picture of a corpse found in an odd situation. For example, in Haydn Talbot's It is the Law, the central idea is that a person who has once been wrongly condemned to lifelong penal servitude for the murder of a certain man (afterwards discovered to be still alive), comes out of prison and murders the real man with perfect impunity, relying on the fact that he cannot legally suffer twice for the same offence.1 In John Rhode's The Davison Case, a somewhat similar legal point is exploited: there, the murderer kills a man by method A, leaving a set of false clues to prove that he murdered him by method B. He is tried on the charge, the false clues are exploded according to plan, and he is acquitted. Subsequently the detective discovers the method by which the crime was really committed, but the murderer is safe, being able to plead autrefois acquit. In one of my own books, Unnatural Death (which I mention only on the grounds that I know better than anybody how it came to be written) the central idea was the medical fact that it would be possible to kill someone by injecting a bubble of air into an important vein, while in my first book, Whose Body? I rather rashly started with the entrancing situation of a naked dead body planted in an innocent person's bath. In Agatha Christie's Murder of Roger Ackroyd, the central idea must, I think, have presented itself thus: We have had murders by this and that Unlikely Person and even by the detective himself; now how about a murder by the unlikeliest person of all—by the harmless Watson who tells the story?2 With Mr. Freeman Wills Crofts the central idea is usually a new and ingenious kind of alibi, as, for example, in Sir John Magill's Last Journey, though in, for instance, The Box-Office Murders the story probably started from the invention of a bright and novel system of uttering forged coins.
The “idea,” whatever it is, usually presents itself in a flash of insight. It may be suggested by a conversation with a doctor or lawyer, or by the account of a curious case in the newspaper columns. For a few days the mystery-monger goes about wrapped in a happy glow of murderous enthusiasm. Then he has to sit down, with more or less reluctance, to do a good deal of solid thinking. It is imperative that he should not be forced by the exigencies of his publisher's lists to start writing at once. He has still to construct the plot.
The idea usually helps to some extent in determining the outline of the plot and the characters. If the idea involves the use of some obscure poison obtainable only by a medical man or chemist, the murderer must either be a doctor or chemist, or have access to a surgery or hospital or chemist's shop; if the body has to be discovered floating in mid-Atlantic, some kind of ship or boat, or aeroplane, must be introduced to bring it there; if some legal point about the succession of an inheritance is involved, a suitable family must be brought into being to account for the circumstances of the inheritance. This seems obvious, but it is just here that the great difficulty of being plausible makes itself felt. If the work is scamped at this point, there will be nasty gaps in the continuity of the story, which can only be filled up by wresting something out of shape. That something, I am sorry to say, is usually the psychology of the characters. Human nature is so odd and various that it is tempting to make the characters behave in obliging accordance with the exigencies of the plot. Nevertheless, too much improbability is fatal. It irritates the reader to find himself asking: “But why did Algernon behave like a boob? Why, when he found the bloodstained dagger in his left-hand boot, did he not take it to the police, instead of burying it in the garden and catching the next train to Sheffield?” If the answer is that Algernon was afraid of being suspected, the reader mutters sourly that Algernon might have known better; if, that Algernon suspected his own best girl, the reader replies by damning this imbecile love-interest; if, on the other hand, Algernon's conduct is dictated by the silly vanity of wanting to do all the detecting himself, the reader's suffering is still more acute; every time the police detective comes along he feels he wants to take him by the arm and say, “Look here, old man—that dagger, you know—that mutt Algernon took it—it isn't fair to treat you like this.” Mr. Temple-Ellis, in The Cauldron Bubbles, takes an even more exasperating way out of the difficulty. His hero, who, at half-a-dozen points during the story, might have cleared up the whole imbroglio, and brought the book to a premature close by informing the police and remaining quietly at home, persists in getting into every kind of unnecessary danger, merely remarking on each occasion, “I know that I ought to have done thus-and-thus, but some impulse which I cannot explain prompted me to do the exact opposite.” Such behaviour is mere perversity and a determination to make the plot occur in the teeth of all commonsense.
Yet the plot must occur as planned. That is where the detective-writer has a much more difficult task with his psychology than the author of the “straight” novel. He cannot afford to follow Mr. Galsworthy's plan of letting his characters “kick free of swaddling clothes and their creators.” Many a hard-working mystery-monger must have gnashed his teeth with envy on reading of Mr. Galsworthy's method of work. “I sink into my morning chair, a blotter on my knee, the last words or deed of some character in ink before my eyes, a pen in my hand, a pipe in my mouth, and nothing in my head. I sit. I don't intend; I don't expect, I don't even hope. … Suddenly, my pen jots down a movement or remark. … When the result is read through it surprises me by seeming to come out of what went before, and by ministering to some sort of possible future.”3
In the case of the detective-story, the future is fixed, and the characters must conform to it, with the author's deliberate intention. This does not mean that the psychology need be artificial or bad. It merely means that proper care must be taken at the beginning to choose the right plot-character combination. Wherever a detective-story fails in plausibility, it may be safely said that the failure is due to a wrong choice of plot to carry the idea or to insufficient care in so visualising and presenting the necessary characters that their actions may appear self-consistent. Usually the fundamental mistake is in the plot-construction. If the plot demands that a character should appear at one point as a careful, methodical business man and at another as a happy-go-lucky, hand-to-mouth hedonist, it is as a rule useless to attempt to reconcile the incongruities. The plot must be scrapped, If, however, it is only that the plot calls for a somewhat elaborately drawn character, full of human contradictions—such as, for example, an artistic person, intensely sensitive to natural beauty but selfish and peevish in the routine of daily life, then it is only necessary to see that this contradictoriness is plainly displayed to the reader from the start. It is useless to carry on with the mechanics of the plot in the happy hope that the characters will “all come right on the night of the performance.”
Naturally, this kind of “fundamental brainwork” takes time and thought, and—what is even more important—space. But there is very much less excuse for carelessness in this respect than there was ten years ago. At that time publishers were naughty enough to insist upon a limit of 70,000 or 80,000 words for a detective novel. Naturally, within such rigid limits, it was almost impossible to do anything but set out the mechanical details of the plot and reduce the characters and their motives to an unintelligible summary. As a result, the said characters and motives were of the crudest kind: a murder was committed, let us say, out of revenge for an ancient wrong; but there was no time in which to show us that the murderer was the type of person to brood over an ancient wrong, or, indeed, to tell us much about the wrong, whatever it was. It had to be stated baldly, somewhere about the time of the discovery, that “Smith, who was a man of sullen and revengeful disposition, had never forgiven Brown for robbing him of his affianced bride, the beautiful Mary Jenkins.” The proper way of showing that Smith was sullen and revengeful would be, of course, to show him exhibiting sullenness and revengefulness about some matter extraneous to the plot, but you cannot do that in 80,000 words if your plot is at all complicated. At the present time, however, publishers have come to realise that the public prefers a novel with “plenty of reading in it.” They will accept, without wincing, novels of 100,000 or 120,000 words, so that the writer now really has a chance to explain his people a little, and even to indulge in a few airs and graces in the way of humour, natural description and charm of presentment.
In consequence partly of this fact and partly of a natural reaction against the extremely mathematical form of detective-problem which held the field for so long, there has been of recent years a distinct tendency to produce a type of story with a more elaborate and realistic psychology than was once thought necessary or desirable. Murderers are on the whole less wicked than they were and victims less innocent—there is more light and shade about them both. Several attempts have been made to show the circumstances leading up to the murder, thus producing what is rather a novel with a detective interest than a detective-story pure and simple. Mr. Anthony Berkeley has issued a manifesto on the subject4 in which he makes his own intentions clear:
“I personally am convinced that the days of the old crime-puzzle pure and simple, relying entirely upon plot and without any added attractions of character, style or even humour, are, if not numbered, at any rate in the hands of the auditors; and that the detective-story is in process of developing into the novel with a...
(The entire section is 6780 words.)