The nice questions of literary derivation cannot be finally answered for the tale, any more than for other literary forms, without large citation and analysis in particular. But, pending fuller discussion, a general survey of the typical late Greek, late Latin, and mediæval forms is full of suggestion. Stories being primarily for pleasure and the pleasures of decadent Greece being largely carnal, it can give no long amazement to find that the tales popular along the Mediterranean of the Seleucids and the Ptolemies were erotic and often frankly obscene. Known as Milesian23 tales, doubtless from the bad eminence of some collection in the Ionian city of pleasure, they set a fashion for those Roman studies in the naturally and the unnaturally sexual of which the Satyricon24 of Petronius may stand as a type. The famous tale of the Matron of Ephesus, which has more consistency than most of this collection, reveals at once how far such pieces went in narrative form. Clearly a capital plot for a short story, it is just as clearly not a short story, but only a plot. It is as it were a narrative sketch or study, like the scenario for a play. And in this it is like many other tales of its class. The rest, the majority, are simply anecdote.25 They are such stories as men of free life and free speech have in all ages told after dinner. That is their character of subject; that is their capacity of form. Speaking broadly, then, the short tales of antiquity are never short stories in our modern sense. They are either anecdote or scenario.
Of the longer tale of antiquity a convenient type is the Daphnis and Chloe ascribed to Longus. A plot no less ancient than that of the foundling reared in simple life and ultimately reclaimed by noble parents receives from the Greek author the form of a pastoral26 romance, with episodes, complications, and a fairy-tale ending. Its form, then, is essentially the same as the form of Aucassin and Nicolette, Florus and Jehane, Amis and Amile, and other typical short romances of the middle age. Between such short romances and the modern short story there is the same difference of form as between Chaucer’s tale of the Man of Law, which is one of the former, and his tale of the Pardoner, which foreshadows how such material may be handled in the way of the latter. For Chaucer, as in his Troilus and Criseyde he anticipates the modern novel, so in his Pardoner anticipates the modern short story. The middle age and the Renaissance, even antiquity,27 show isolated, sporadic instances of short story, whether in prose or in verse; but these are apart from the drift of the time. Aside from such sporadic cases, the longer mediæval tale or short romance, though often in length within the limits of short story, is typically loose as to time and place, and as to incident accumulative of marvels. It is to the long mediæval romance what the modern tale—not the modern short story—is to the modern novel. And it is a constant form from Greece—even from India and Egypt,28 down to the present. In form the Alexandrian Daphnis and Chloe, the mediæval Aucassin and Nicolette, and the whole herd of modern tales, such as Miss Edgeworth’s, are essentially alike. The modern time has differentiated two forms: first, the novel, in which character is progressively developed, incidents progressively complicated and resolved; second, the short story, in which character and action are so compressed as to suggest by a single situation without development. The former is as it were an expansion of the tale; the latter, a compression. In both cases the modern art of fiction seems to have learned from the drama. Meantime the original, naïve tale has endured, and doubtless will endure. To employ the figure of speech by which M. Brunetière is enabled to speak of literature in terms of evolution, the tale is the original jackal. From it have been developed two distinct species; but their parent stock persists. Indeed, for aught we can see from the past, posterity may behold a reversion to type.
The significance of a division of ancient and early mediæval tales into anecdote and scenario or summary romance becomes at once clearer by reference to the greatest mediæval collection, the Decameron (1353) of Boccaccio. More than half the tales of the Decameron may readily be grouped as anecdote—all of the sixth day, for instance, most of the first and eighth, half of the ninth. Of these some approach consistency of form. Having long introductions, unnecessary lapse of time, or other looseness of structure, they still work out a main situation in one day or one night; they sometimes show dramatic ingenuity of incident; less frequently they reach distinct climax. Where the climax, as in the majority of cases, is merely an ingenious escape or a triumphant retort, of course the tale remains simple anecdote; but in some few the climax is the result of the action, is more nearly a culmination. This is the character of the seventh day. Another class in the Decameron rapidly summarises a large plot, the action ranging widely in time and place. A narrative sketch, usually of a romance, it corresponds essentially to the Aucassin and Nicolette type,29 and includes nearly one half. Here was an open mine for the romantic drama of later centuries. The Decameron, then, is almost all either anecdote or scenario.
But not quite all. Besides those tales which seem to show a working for consistency, there are a few that definitely achieve it. The fourth of the first day (The Monk, the Woman, and the Abbot) is compact within one place and a few hours. All it lacks for short story is definite climax. Very like in compactness is the first of the second day (The Three Florentines and the Body of the New Saint). Firmer still is the eighth of the eighth day (Two Husbands and Two Wives). Here the climax is not only definite, but is a solution, and includes all four characters. If it is not convincing, that is because the Decameron is hardly concerned with characterisation. The action covers two days. It might almost as easily have been kept within one. Finally there are two tales that cannot, without hair-splitting, be distinguished from modern short story. The second tale of the second day (Rinaldo, for his prayer to St. Julian, well lodged in spite of mishap) is compressed within a single afternoon and night and a few miles of a single road. The climax is definitely a solution. The movement is largely by dialogue. In a word, the tale is a self-consistent whole. Equally self-consistent, and quite similar in method, is that farce comedy of errors, the sixth tale of the ninth day (Two Travellers in a Room of Three Beds), which Chaucer has among his Canterbury Tales. Both these are short stories. If the other three be counted with them, we have five out of a hundred.30
The middle age, then, had the short story, but did not recognise, or did not value, that opportunity. Not only does Boccaccio employ the form seldom and, as it were, quite casually, but subsequent writers do not carry it forward. In fact, they practically ignore it. Les cent nouvelles nouvelles (1450–1460), most famous of French collections, shows no discernment of Boccaccio’s nicer art. In form, as in subject, there is no essential change from the habit of antiquity. True, here and there among the everlasting histoires grivoises is a piece of greater consistency and artistic promise. That delicious story (the sixth nouvelle) of the drunken man who insisted on making his confession on the highway to a priest unfortunately passing, who had absolution at the point of the knife, and then resolved to die before he lapsed from the state of grace, is not only a short-story plot; it goes so far toward short-story form as to focus upon a few hours. Yet even this hints the short story to us because we look back from the achieved form. After all it remains anecdote; and it has few peers in all the huge collection. Bandello (1480–1562), in this regard, shows even a retrogression from Boccaccio. His brief romances are looser, often indeed utterly extravagant of time and space. His anecdotes, though they often have a stir of action, show less sense of bringing people together on the stage. So the Heptameron (1558–1559) of the Queen of Navarre fails—so in general subsequent tale-mongers fail—to appreciate the distinctive value of the terser form. Up to the nineteenth century the short story was merely sporadic. It was achieved now and again by writers of too much artistic sense to be quite unaware of its value; but it never took its place as an accepted form.
Thus the modern development of the short story in France has both its own artistic interest and the further historical interest of background. When Charles Nodier (1783–1844), in the time of our own Irving, harked back from the novel to the tale, he but followed consciously what others had followed unconsciously, a tradition of his race.31 Some of Nodier’s legends are as mediæval in form as in subject. But when he wrote La combe à l’homme mort he made of the same material something which, emerging here and there in the middle age, waited for definite acceptance till Nodier’s own time—a short story. The hypothesis that Nodier was a master to Hawthorne is not supported by any close likeness. Yet there are resemblances. Both loved to write tales for children; both lapse toward the overt moral and fall easily into essay; both use the more compact short-story form as it were by the way and not from preference. Smarra (66 pages, 1821), acknowledging a suggestion from Apuleius, is an essentially original fantasy, creating the effect of a waking dream. The nearest English parallel is, not Hawthorne, but De Quincey, or, in more elaborate and restrained eloquence, Landor. Smarra, as Nodier says in his preface, is an exercise in style to produce a certain phantasmagorical impression. The clue to the effect he sought is given by the frequent quotations from the Tempest. It is “such stuff as dreams are made on.” Jean François-les-bas-bleus (1836) and Lidivine, on the other hand, are almost documentary studies of character. La filleule du Seigneur (1806), legendary anecdote like Irving’s, shows where Nodier’s art began. He carried his art much further; but his pieces of compactness, like La combe à l’homme mort, are so rare that one may doubt their direct influence on the modern development of form.
For the bulk of Nodier’s work is not conte, but nouvelle. These two terms have never been sharply differentiated in French use. Les cent nouvelles nouvelles are not only shorter, in average, than the novelle of Boccaccio; they are substantially like the Contes de la Reine de Navarre. Some of the nouvelles of Nodier, Mérimée, and Gautier are indistinguishable in form from the contes of Flaubert, Daudet, and Maupassant. But though even to-day a collection of French tales might bear either name, the short story as it grew in distinctness and popularity seems to have taken more peculiarly to itself the name conte.32 Correspondingly nouvelle is a convenient name for those more extended tales, written sometimes in chapters, which in English are occasionally called novelettes, and which have their type in Aucassin and Nicolette. In this sense Nodier’s writing is mainly, and from preference, nouvelle. Taking as his type for modern adaptation the longer mediæval tale, he did not work in the direction of short story.
Nor, oddly enough, did Mérimée. People who assign to him the rôle of pioneer in the short story, on account of his extraordinary narrative conciseness, appear to forget that his typical tales—Carmen, Colomba,33 Arsène Guillot, are too long for the form; and that many of his shorter pieces—L’enlèvement de la redoute, Tamango, La vision de Charles XI., are deliberately composed as descriptive anecdotes. Mérimée’s compactness consists rather in reducing to a nouvelle what most writers would have made a roman than in focusing on a single situation in a conte. Carmen, though compact in its main structure, has a long prelude. Beyond question the method is well adapted; but it shows no tendency to short story. And the habit is equally marked in Le vase étrusque, with its superfluous characters. Evidently his artistic bent, like Hawthorne’s, like Nodier’s, was not in that direction. All the more striking, therefore, is his single experiment. La Vénus d’Ille (1837) is definitely and perfectly a short story. Giving the antecedent action and the key in skilful opening dialogue, it proceeds by a series of increasingly stronger premonitions to a seizing climax. Like Poe, Mérimée intensifies a mood till it can receive whatever he chooses, but not at all in Poe’s way. Instead, the mystery and horror are accentuated by a tone of worldly-wise skepticism. Less compressed, too, than Poe, he can be more “natural.” Withal he keeps the same perfection of grading. Strange that a man who did this once should never have done it again. But the single achievement was marked enough to compel imitation.
That the propagation of the short story in France owes much to Balzac might readily be presumed from the enormous influence of Balzac’s work in general, but can hardly be held after scrutiny of his short pieces in particular. Of these, two will serve to recall the limitations of the great observer. El Verdugo (1829), though it is reduced to two days and substantially one scene, hardly realises the gain from such compression. Instead of intensifying progressively, Balzac has at last to append his conclusion, and for lack of gradation to leave his tale barely credible. Les Proscrits (1831), more unified in imaginative conception, and again limited in time-lapse, again fails of that progressive intensity which is the very essence of Poe’s force and Mérimée’s. It is not even held steady, but lapses into intrusive erudition and falls into three quite separate scenes. Others of Balzac’s short pieces, La messe de l’athée (1836), for example, and Z. Marcas (1840), are obviously in form, like many of Hawthorne’s, essays woven on anecdote or character. Some of his tales may, indeed, have suggested the opportunity of different handling. Some of them, at any rate, seem from our point of view almost to call for that. But his own handling does not seem, as Poe’s does, directive. And in general, much as Balzac had to teach his successors, had he much to teach them of form?
The tales of Musset, which are but incidental in his development, and are confined, most of them, within the years 1837–1838, show no grasp of form. Gautier, even more evidently than Mérimée, preferred the nouvelle, partly from indolent fluency, partly from a slight sense of narrative conclusion. Few even of his most compact contes, such as Le nid de rossignols, compress the time. He was garrulous; he had read Sterne34; above all, he was bent, like Sterne, on description. But Gautier too shows a striking exception. La morte amoureuse, though it has not Poe’s mechanism of compression, is otherwise so startlingly like Poe that one turns involuntarily to the dates. La morte amoureuse appeared in 1836; Berenice, in 1835. The Southern Literary Messenger could not have reached the boulevards in a year. Indeed, the debt of either country to the other can hardly be proved. Remarkable as is the coincident appearance in Paris and in Richmond of a new literary form, it remains a coincidence. And whereas by 1837 Poe was in full career on his hobby, Gautier and Mérimée did not repeat the excursion.
The history of the tale in England, however important otherwise, is hardly distinct enough as a development of form to demand separate discussion here. For England, apparently trying the short-story form later than France and the United States, apparently also learned it from them. Perhaps the foremost short-story writers of our time in English—though that must still be a moot point—are Kipling and Stevenson. But Stevenson’s short story looks to France; and Kipling probably owes much to the American magazine. Without venturing on the more complicated question of the relations of Germany, Russia,35 and Scandinavia to France, it is safe to put forward as a working hypothesis that the new form was invented by France and America, and by each independently for itself. Our priority, if it be substantiated, can be but of a year or two. The important fact is that after due incubation the new form, in each country, has germinated and spread with extraordinary vigor. Daudet, Richepin, Maupassant—to make a list of French short-story writers in the time just past, is to include almost all writers of eminence in fiction. What is true of France is even more obviously true of the United States. Our most familiar names in recent fiction were made familiar largely through distinction in the short story. The native American yarn, still thriving in spontaneous oral vigour, has been turned to various art in The Jumping Frog and Marjorie Daw and The Wreck of the Thomas Hyke. The capacity of the short story for focusing interest dramatically on a strictly limited scene and a few hours, no less than its capacity for fixing local color, is exhibited most strikingly in the human significance of Posson Jone. Mr. James, though his preoccupation with scientific analysis demands typically, as it demanded of Mérimée, a somewhat larger scope, vindicates his skill more obviously in such intense pieces of compression as The Great Good Place. To instance further would but lead into catalogue. In a word, the two nations that have in our time shown keenest consciousness of form in fiction have most fostered the short story. For ourselves, we may find in this development of a literary form one warrant for asserting that we have a literary history.