is supreme: not a thriller, which depends on conflict and thrills, nor yet a crime novel, which is the history of some particular crime, usually from the criminal’s point of view.
Before we begin we must settle one or two points about our detective. Is he to be a gifted amateur, a professional private detective, or a man from the C.I.D.? Is he to be a ‘character’ or an ordinary humdrum citizen? Is he to work alone or to have a Watson? Suppose you settle these points? You have? Then let’s get down to it.
If we’re lucky we shall begin with a really good idea. This may be one of five kinds. Firstly, it may be an idea for the opening of our book: some dramatic situation or happening to excite and hold the reader’s interest. The standard way of finding a body in the first chapter, if hackneyed, is hard to beat.
Secondly, our idea may be for the closing or climax of our book. This must also be dramatic. As an example I suggest the well-known situation in which Tom, who thinks Jack is dead and has impersonated him, is unexpectedly confronted with Jack in a police office or court of law.
Our idea, thirdly, may be for a good way of committing a crime, probably a murder. It should be novel and ingenious—but not too ingenious—and if possible concerned with things with which the man in the street is familiar. This is probably the most usual way of starting work on a book. Every detective fan will think of dozens of examples.
A fourth kind of idea on which to build a book is that we shall write about some definite crime, such as smuggling, gun-running, coining, arson, or frauds in high finance.
Lastly, our idea may be simply to place the action in a definite setting, such as a mining setting, or a golf or fishing setting, or to lay our scenes in a certain place: a bus or an office, an opium den or Canterbury Cathedral.
We may of course build our book on some idea which does not fall under one of these heads. For instance, Dr Austin Freeman’s book, The Red Thumb Mark, was probably built on the idea that a fingerprint is not necessarily convincing evidence.
This then is the first stage in our work: getting the idea to start on. Our second stage is more difficult: we have to build up the plot on our idea.
We do this in a very simple, but very tedious way: we ask ourselves innumerable questions and think out the answers. One question invariably leads to another, and as we go on our plot gradually takes shape.
Suppose we have decided on a murder by antimony poisoning. We shall ask ourselves questions such as: Where does the murderer get the antimony? How does he administer it? What is his motive?
Suppose in answering this last question we choose greed: that he inherits money from the man he kills. At once new questions suggest themselves. What was the relationship between the two men? Why had the deceased left money to the other? And so on.
As we continue propounding and answering these questions, we shall have the happiness of finding a story gradually growing out of nothing. We continue the good work ’til the whole happening is built up, from the first thought of the crime right down to its completion, together with the subterfuges the criminal adopts to secure his safety. A rough synopsis is then made, together with sketch maps of the important localities, short biographies of the principal characters, and a chronology of the main events.
It should be clearly understood that this synopsis is of the actual facts which are supposed to have happened: It is not a synopsis of the book. We don’t get to the book ’til the third stage, for which, however, we are now ready.
In this third stage we reconsider the whole circumstances from a new viewpoint, the viewpoint of the person or persons through whom we are going to tell the story. What is the first thing that would have become known? Would it have been the finding of the body? If so, begin with that. What would then be done? The police would be sent for. What would they do? They would make certain enquiries, they would look for motives, they would find out who was in the neighbourhood when the crime was committed.
We continue working in this way ’til we have completed a second synopsis of the case, this time describing the gradual revealing of the details to the detective. As we do so, we find that we have to supply a good deal of fresh material. That means of course a new set of questions to be answered. There is, for instance, the very important problem of how the detective discovers the truth. He could if possible do so through some flaw inherent in the criminal’s plans, unperceived ’til now by the reader. If, however, this can’t be arranged, the necessary clues must be planted for the detective to find.
This second synopsis which, let us suppose, we have now completed, gives us the sequence of events right from the discovery of the crime up to the arrest and conviction of the criminal. It is, in other words, a précis of our book. We probably have to make another chronology giving the movements of the detective, as well possibly as more sketch maps. Then, having estimated the length of our various scenes and satisfied ourselves that our book is going to run to the required 80,000 words, we can proceed to our next stage.
The fourth stage is the actual writing, and there is nothing to be said about it except that we take the advice of the King in Alice in Wonderland and begin at the beginning, go on ’til we come to the end, and then stop.
When writing we invent the minor episodes. For instance, our synopsis may read: ‘Detective finds paper in X’s room.’ We have now to think out how the detective obtains access to X’s room, whereabouts the paper is hidden, and how the detective comes to look in that place.
The writing of the passages which give the necessary clues to the reader requires a lot of thought. All the clues must be given which he needs to enable him, by the use of his intelligence, to reach the truth. At the same time they must not be easy to pick up.
There are many tricks for concealing clues. The chief is perhaps to invert the sequence of events or to alter their connection. Suppose we want to tell the reader that the murderer is a good shot. If his skill be mentioned in connection with the shooting of the victim, the story is given away. But if it be brought out in relation to a shooting competition in another part of the book, the reader will probably miss its significance.
Let us now pause for a moment to consider our climax. In this we shall try to clear up as suddenly as possible what has been up to now a complete mystery. If on reaching the climax the reader says: ‘Of course! Why didn’t I think of that?’ we shall have done our job well.
Well, we go over our manuscript, checking and cutting and patching and re-writing. Then having typed a fair copy, we try it on the dog: we get as many of our friends to read it as we can. We incorporate the more useful of their suggestions, and at last our book goes off, carefully registered, and with a magic name on the cover. Whereupon we settle down to wait.
FREEMAN WILLS CROFTS
1937
JOSEPH ASHE, signalman in the employment of the Union of South Africa Government Railways, stood in his box at the west end of Middeldorp station, gazing meditatively down the yard to the platforms beyond.
It was his week on night duty, which he
