My reason for writing stories is to give myself the satisfaction of visualising more clearly and detailedly and stably the vague, elusive, fragmentary impressions of wonder, beauty, and adventurous expectancy which are conveyed to me by certain sights (scenic, architectural, atmospheric, etc.), ideas, occurrences, and images encountered in art and literature. I choose weird stories because they suit my inclination best—one of my strongest and most persistent wishes being to achieve, momentarily, the illusion of some strange suspension or violation of the galling limitations of time, space, and natural law which forever imprison us and frustrate our curiosity about the infinite cosmic spaces beyond the radius of our sight and analysis. These stories frequently emphasise the element of horror because fear is our deepest and strongest emotion, and the one which best lends itself to the creation of Nature-defying illusions. Horror and the unknown or the strange are always closely connected, so that it is hard to create a convincing picture of shattered natural law or cosmic alienage or "outsideness" without laying stress on the emotion of fear. The reason why time plays a great part in so many of my tales is that this element looms up in my mind as the most profoundly dramatic and grimly terrible thing in the universe. Conflict with time seems to me the most potent and fruitful theme in all human expression.
While my chosen form of story-writing is obviously a special and perhaps a narrow one, it is none the less a persistent and permanent type of expression, as old as literature itself. There will always be a certain small percentage of persons who feel a burning curiosity about unknown outer space, and a burning desire to escape from the prison-house of the known and the real into those enchanted lands of incredible adventure and infinite possibilities which dreams open up to us, and which things like deep woods, fantastic urban towers, and flaming sunsets momentarily suggest. These persons include great authors as well as insignificant amateurs like myself—Dunsany, Poe, Arthur Machen, M. R. James, Algernon Blackwood, and Walter de la Mare being typical masters in this field.
As to how I write a story—there is no one way. Each one of my tales has a different history. Once or twice I have literally written out a dream; but usually I start with a mood or idea or image which I wish to express, and revolve it in my mind until I can think of a good way of embodying it in some chain of dramatic occurrences capable of being recorded in concrete terms. I tend to run through a mental list of the basic conditions or situations best adapted to such a mood or idea or image, and then begin to speculate on logical and naturally motivated explanations of the given mood or idea or image in terms of the basic condition or situation chosen.
The actual process of writing is of course as varied as the choice of theme and initial conception; but if the history of all my tales were analysed, it is just possible that the following set of rules might be deduced from the average procedure: 1. Prepare a synopsis or scenario of events in the order of their absolute occurrence—not the order of their narration. Describe with enough fullness to cover all vital points and motivate all incidents planned. Details, comments, and estimates of consequences are sometimes desirable in this temporary framework. 2. Prepare a second synopsis or scenario of events—this one in order of narration (not actual occurrence), with ample fullness and detail, and with notes as to changing perspective, stresses, and climax. Change the original synopsis to fit if such a change will increase the dramatic force or general effectiveness of the story. Interpolate or delete incidents at will—never being bound by the original conception even if the ultimate result be a tale wholly different from that first planned. Let additions and alterations be made whenever suggested by anything in the formulating process. 3. Write out the story—rapidly, fluently, and not too critically—following the second or narrative-order synopsis. Change incidents and plot whenever the developing process seems to suggest such change, never being bound by any previous design. If the development suddenly reveals new opportunities for dramatic effect or vivid story telling, add whatever is thought advantageous—going back and reconciling the early parts to the new plan. Insert and delete whole sections if necessary or desirable, trying different beginnings and endings until the best arrangement is found. But be sure that all references throughout the story are thoroughly reconciled with the final design. Remove all possible superfluities— words, sentences, paragraphs, or whole episodes or elements—observing the usual precautions about the reconciling of all references. 4. Revise the entire text, paying attention to vocabulary, syntax, rhythm of prose, proportioning of parts, niceties of tone, grace and convincingness of transitions (scene to scene, slow and detailed action to rapid and sketchy time-covering action and vice versa... etc., etc., etc.), effectiveness of beginning, ending, climaxes, etc., dramatic suspense and interest, plausibility and atmosphere, and various other elements. 5. Prepare a neatly typed copy —not hesitating to add final revisory touches where they seem in order.
The first of these stages is often purely a mental one—a set of conditions and happenings being worked out in my head, and never set down until I am ready to prepare a detailed synopsis of events in order of narration. Then, too, I sometimes begin even the actual writing before I know how I shall develop the idea—this beginning forming a problem to be motivated and exploited.
There are, I think, four distinct types of weird story; one expressing a mood or feeling, another expressing a pictorial conception, a third expressing a general situation, condition, legend or intellectual conception, and a fourth explaining a definite tableau or specific dramatic situation or climax. In another way, weird tales may be grouped into two rough categories—those in which the marvel or horror concerns some condition or phenomenon, and those in which it concerns some action of persons in connexion with a bizarre condition or phenomenon.
Each weird story—to speak more particularly of the horror type —seems to involve five definite elements: (a) some basic, underlying horror or abnormality—condition, entity, etc.—, (b) the general effects or bearings of the horror, (c) the mode of manifestation—object embodying the horror and phenomena observed—, (d) the types of fear- reaction pertaining to the horror, and (e) the specific effects of the horror in relation to the given set of conditions.
In writing a weird story I always try very carefully to achieve the right mood and atmosphere, and place the emphasis where it belongs. One cannot, except in immature pulp charlatan-fiction, present an account of impossible, improbable, or inconceivable phenomena as a commonplace narrative of objective acts and conventional emotions. Inconceivable events and conditions have a special handicap to over come, and this can be accomplished only through the maintenance of a careful realism in every phase of the story except that touching on the one given marvel. This marvel must be treated very impressively and deliberately—with a careful emotional "build-up"—else it will seem flat and unconvincing. Being the principal thing in the story, its mere existence should overshadow the characters and events. But the characters and events must be consistent and natural except where they touch the single marvel. In relation to the central wonder, the characters should shew the same overwhelming emotion which similar characters would shew toward such a wonder in real life. Never have a wonder taken for granted. Even when the characters are supposed to be accustomed to the wonder I try to weave an air of awe and impressiveness corresponding to what the reader should feel. A casual style ruins any serious fantasy.
Atmosphere, not action, is the great desideratum of weird fiction. Indeed, all that a wonder story can ever be is a vivid picture of a certain type of human mood. The moment it tries to be anything else it becomes cheap, puerile, and unconvincing. Prime emphasis should be given to subtle suggestion —imperceptible hints and touches of selective associative detail which express shadings of moods and build up a vague illusion of the strange reality of the unreal. Avoid bald catalogues of incredible happenings which can have no substance or meaning apart from a sustaining cloud of colour and symbolism.
These are the rules or standards which I have followed— consciously or unconsciously—ever since I first attempted the serious writing of fantasy. That my results are successful may well be disputed— but I feel at least sure that, had I ignored the considerations mentioned in the last few paragraphs, they would have been much worse than they are.