In the May/Summer issue of The Writer's Chronicle, Alice Mattison offers an interesting essay (not available online, but the issue's table of contents is available here) defending the use of coincidence in fiction. Subtitled "An Essay Against Craft," the essay commends the use of coincidence as a way of taking risk, which Mattison feels is discouraged in a literary world dominated by the workshop "rules" implicitly taught in creative writing programs. Writes Mattison: "I don't think directions or rules are available, just terms. . .that undeniably simplify discussions of writing and literature." Such simplification is at times useful, but "the problem arises when we begin to draw conclusions from succesful choices, assuming that what works once will work in every instance."
A few paragraphs after the statements just quoted, Mattison is discussing a Charles Baxter essay in which "Baxter glances at the sort [of stories] that were rejected as old-fashioned by the authors who first made stories turn on insight. He characterizes the stories that Henry James and James Joyce rejected as those with 'plot structures tending to require a set of coincidences or connivances of circumstance.'" Mattison comments: "It hadn't occured to me, before I read Baxter's sentence, that coincidence defines the type of story in which it appears. I hadn't noticed that such stories. . .were helpless without coincidence."
Although Baxter and Mattison don't use the word, what they are both describing is the influence on early novels in English of the "picaresque" narrative. The picaresque story--derived from the term identifying the protagonist of such stories, the "picaro"--was introduced by Spanish writers of the 16th and 17th centuries, and is essentially a journey narrative in which the picaro, usually a rogueish character, embarks on a journey in which, literally, one thing happens after another. There's not really a sense of progression in the picaresque narrative, just a series of episodes, and usually the protagonist remains more or less unchanged, undergoing no transformation or "epiphany." The most famous picaresque novel is undoubtedly Don Quixote, in which Cervantes alters the form by making his protagonist a deluded but not antisocial or rascally character.
The early British novelists of the 18th century were greatly influenced by the picaresque narrative, especially such writers as Tobias Smollett and Henry Fielding. Fielding's Tom Jones is probably the most famous of these British picaresque novels. It adopts the journey conceit, the episodic structure, and adds an element of explicit comedy that exceeds even the kind of doleful humor to be found in Don Quixote. (Tom Jones remains a tremendously readable book, and I would highly recommend it to anyone who wants to know what the picaresque form can accomplish.) Charles Dickens was in turn profoundly influenced by Smollett and Fielding, and his novels represent a further fashioning of the picaresque into a narrative technique of great flexibility and latent aesthetic potential.
But this was indeed the "old-fashioned" kind of storytelling that came to be rejected by later writers more concerned about the "craft" of fiction. Perhaps the first writer to really move away from the picaresque was Flaubert, and he may be the writer most responsible for converting fiction into a more gracefully "shaped" kind of storytelling, and therefore a form that could be taken seriously as a mode of literary art. (I greatly admire Flaubert, and nothing I say here is meant to denigrate his achievement in any way.) Mattison identifies James and Joyce as the writers who came to "shape" their stories around the occurence of an "epiphany," but it was really Flaubert who showed James and Joyce that such an aesthetically intricate effect could be brought off in fiction.
Since Flaubert, the notion of "story" in fiction is thus usually associated either specifically with the kind of dramatic narrative leading to revelation or epiphany pioneered by James and Joyce or more generally with the kind of carefully structured narrative encapsulated in "Freytag's triangle": exposition, rising action, climax, denoument, etc. Most genre fiction probably uses the latter, most "literary" fiction the former. Most best-selling potboilers are likely to use the Freytag-derived narrative filtered through Hollywood melodrama. In this context, the picaresque story almost doesn't seem like a story at all, since it doesn't arrange itself in some shaped pattern, but is instead just a series of incidents strung together.
I go over all this not to offer some kind of lesson in literary history but ultimately to suggest, with Mattison but more broadly than her advocacy of "coincidence" goes, that the picaresque ought to remain a viable option and can provide an alternative to the workshop-reinforced domination of the revalatory narrative. To some extent the picaresque style was revived by postwar American writers such as John Barth and Thomas Pynchon, but in my opinion it still contains much untapped potential. It can free the writer from the tyranny of story--the creation of narrative tension by which too many stories and novels are reductively judged--but at the same time allows for the depiction of external events, provides an aesthetically justified motive for abjuring the directive to probe the psychological depths, and perhaps most of all makes available all kind of other effects--satire, subplots, a larger cast of characters--that the craft-like story discourages. Of course, this is not the 18th century, and writers now would be using the picaresque form in a much more self-conscious way, but that in itself would likely give such fictions a "shape" that would rescue them from mere formlessness. (Although attempting a truly "formless" novel might be an interesting experiment in itself.)
I am not suggesting that the picaresque narrative is superior to the more conventionally shaped narrative most novels employ. The possibilities in "shaping" the latter kind of narrative have by no means been exhausted, although most published novels don't seem much interested in exploring these possibilities. A renewed interest in the picaresque might, however, help demonstrate that there is more than one way to tell a story, multiple ways to "shape" a work of fiction, without sacrificing readibility or even fiction's "entertainment" value. (Both Don Quixote and Tom Jones are nothing if not entertaining.) And in the final analyis using such a narrative strategy wouldn't really involve abandoning "craft"; ideally it would further demonstrate that craft is just as much involved in the breaking of convention as in its repetition.