Though today they are often published without the standard prefix, I think it’s important that so many of the fifty-six Sherlock Holmes stories bear the word “Adventure” in their title: “The Adventure of the Speckled Band,” “The Adventure of the Solitary Cyclist.” It has become commonplace to view the Holmes tales, and the detective story tradition that they engendered, as fundamentally conservative.1 In this reading the detective, while technically independent of the law, is in truth the dedicated agent of the prevailing social order—a static, hierarchical structure, in which murder is an aberration. This was the view Raymond Chandler took of “murder in the Venetian vase,” against which he famously posited his “mean streets” theory, in which the autonomous if not anarchist detective operates in a disordered and fluctuating world that can never hope to be restored, in which social position is transient, the law a hopeless fiction, and morality flexible at best.

This view of the Holmes stories as reassuring fables of the fixed values and verities of the Victorian order contains an element of truth. Especially after the first two Holmes novels, A Study in Scarlet and The Sign of Four, and beginning in 1891 with the first great short story, “A Scandal in Bohemia,” Conan Doyle gradually abandoned most of the louche, Wildean touches with which he had initially encumbered the character of Holmes. The outré personal habits, the vampiric hours, the drug use, the willfully outrageous ignorance of “useless facts,” such as the order of the solar system or contemporary politics, gave way to a more conventional and cozy sort of eccentricity.

While Holmes is curt with toffs and colonels, he can be a suck-up to royalty, and beneath the surface of the tales glides the majestic shadow of Victoria, emerging only at the end of “The Adventure of the Bruce-Partington Plans,” when Holmes, having saved the Navy by helping recover the stolen plans for a submarine—“the most jealously guarded of government secrets”—returns from a visit to “a certain gracious lady,” wearing an emerald pin in his tie.

Holmes’s veneration of methodology, his love of rank and classification (we are informed that Moriarty’s henchman, Colonel Sebastian Moran, is “the second most dangerous man in London,” and the blackmailer Charles Augustus Milverton is “the worst man in London,” and John Clay, who conceived of the Red-Headed League, is “the fourth smartest man in London”), his systematic approach to cataloging the minutiae of crime (as in his monograph on the ashes of 140 different varieties of pipe, cigar, and cigarette tobacco) all partake of the Victorian passion for taxonomy, for hierarchies and progressions, for articulating new, rational systems of control. But to read Sherlock Holmes, regardless of his frequent service to Queen and Empire, as a prop and agent of the dominant social order, to regard the function and effect of the stories as characteristic of industrialized, imperialist, Darwinistic, bourgeois, nineteenth-century Britain, the literary kin of Bentham’s panopticon or the proposed Cape-to-Cairo railway, misses the point.

In fact, the classic detective story is a device that, with all due respect to Poe and Chevalier Dupin, Conan Doyle invented. This is less a matter of intent, ideology, or effect than of technique. Stories have always manifested a twofold nature, deriving their impact and pleasure in part from the difference between the chronology of the story to be told and the ordering and presentation of that chronology.2 Conan Doyle took those two elements—in the form of the crime and the reconstruction of the crime—and completely reengineered them. Like the builder of Skidbladnir, the sailing ship of the Norse gods that could be folded up to fit into your pocket, or an engineer packing an extra million transistors onto a 3-mm chip, Conan Doyle found a way to fold several stories, and the proper means of telling them, over and over into a tightly compacted frame, with a proportionate gain in narrative power. “The Speckled Band” and “The Adventure of the Dancing Men” are storytelling engines, steam-driven, brass-fitted, but among the most efficient narrative apparatuses the world has ever seen. After all these years they still run remarkably well.

The classical Holmes story is framed with Holmes and Watson at home, in their rooms at 221B Baker Street. The frame forms part of a larger, ongoing macro-story of the two companions’ lives and long career together, seeing each other through the vicissitudes of, in Watson’s case, at least one tragic if vague marriage and, in the case of Holmes, a cocaine habit, several black depressions, and a self-imposed witness protection program known to Sherlockians as the Great Hiatus, when, after the death of Professor Moriarty, the vengeful henchmen of the Napoleon of Crime made England too hot even for Sherlock Holmes. Set within this frame is the story of a client who has sent up a card or blustered in to see Mr. Holmes, and has now sat down to tell it.


Nearly all the Holmes stories, therefore, are stories of people who tell their stories, and every so often the stories these people tell feature people telling stories (about what they heard or saw, for example, on the night in question), and if this sounds like a dubiously metafictional observation then we may have forgotten how fundamental such stories within stories have always been to popular art from Homer to Green Acres, and how lightly worn. The new client tells the story of a recent crime, an apparent crime, or an impending crime, or simply, as in “The Red-Headed League,” “The Sussex Vampire,” and “The Six Napoleons,” recounts a strange and inexplicable incident. As the story proceeds, its teeth engage with the works of the next story, which is the story of the investigation conducted by Holmes and Watson, often with the assistance of one or more slow-witted policemen. The investigation in turn produces the story of how the crime was committed, or of the genuine meaning behind the seemingly inexplicable occurrence. In “The Six Napoleons,” for example, what appears to be the story of a rather unlikely anti-Bonapartist serial vandal and murderer bent on smashing busts of Napoleon is found to conceal the story of immigrant Italian artisans trying to recover the “black pearl of the Borgias” that had been hidden in one of them.

But that’s not all. The solution of the crime is typically not the last of the tellings and retellings that Doyle manages to compact within his endlessly flexible frame: often there remains an account of how the malefactor has been pursued, staked out, hunted down, or how a trap has been laid. Once caught, he or she may introduce an entirely new version of the story, by way of pointing out certain flaws in Holmes’s reasoning or confirming his wildest surmises, and then offering reasons for the crime, reasons that can have their roots in yet another story, often one that played out many years before. And then we are back in Baker Street, to be handed over by Watson to the next story.

Writers and storytellers had been nesting their narratives for centuries, of course, in an effort to approximate the networks of story that ramify and complicate our experience of everyday life. But until Conan Doyle, no one had ever hit on a way, or even seen the need, to ensure that the gears of each nested story were fully engaged with those it contained and were in turn contained by. Conan Doyle, in other words, invented a way to tell stories about the construction of stories without the traditional recourse to digression, indirection, or the overtly self-referential. It was a radical step, and it has been paying off for him, and for us, ever since.

But if the Victorian spirit of improvement and efficiency and control—the Cape-to-Cairo spirit—is crucial to the technique of Conan Doyle’s stories, what can we say then about their function and effect, which the conventional view holds to be to reinforce and to eulogize the iron-and-brick social structure of the Empire?

We can repeat that the stories are, as they insist on being called, adventures. Their function is not to reinforce or validate the dominant social order but to transcend it, abandon it, if only for the space of twenty pages.

A familiar lament of adventure fiction of the late nineteenth and early twentieth century, a lament given its fullest expression in the opening pages of Conan Doyle’s novel The Lost World (1912) and its most ironic (but no less wistful) in the opening pages of Heart of Darkness, is the disappearance of what Conrad’s Marlow calls “the blank places on the map.” From the time of Odysseus, literary adventurers have sought to write their names in those blank places, and to fetch back stories from them. In this sense, a key part of the business of Empire was to obtain new zones of adventure in which its writers could lay their tales.

Empires are built, however, by laying the groundwork for their own destruction. Subject peoples are educated, organized, given national identities. Any colony made strong enough to survive and flourish becomes too strong to remain a colony. And by the time that Conan Doyle came to write A Study in Scarlet in the 1880s, the great explorations undertaken by the Empire, the surveys and royal expeditions of the previous few centuries, had done grave harm to the atlas of adventure. In the years to follow, adventure writers were obliged to devise new strategies. Edgar Rice Burroughs resorted to laying his otherwise classic stories not only in a remote jungle but on Mars or Venus, or in the center of the earth. Robert E. Howard and Talbot Mundy reached back to the pre-exploration past, to prehistory and beyond.


After the technical innovation of packing multiple stories into a tight narrative frame, Conan Doyle’s second flash of genius was to find a way to locate the land of adventure within the known world itself, to depict a place beyond laws, where human nature returned to savagery, and where a hero could flourish, right there in the Home Counties. Many of the tales deal with the crimes, misdeeds, and scandals of transported convicts, colonial adventurers, or imperial soldiers returned to England. But these travelers don’t merely bring back their tales; they are, as in “The Crooked Man,” hunted down by them, haunted by them, killed or forced to kill by the adventures that befell them beyond the sea. As Angus Wilson pointed out in his introduction to The Return of Sherlock Holmes, the Holmes stories are replete with imagery of Holmes and Watson as hunters in the jungle, and of men depicted as animals and half-brutes living not on some remote island like Dr. Moreau’s, but ten minutes’ walk from Marylebone Station. In this moral vacuum Sherlock Holmes is as much a law unto himself as Chandler’s Marlowe or Hammett’s Continental Op. Repeatedly, persistently, as a matter of existential necessity, in the absence of any real higher authority, he acts to punish those whom the law would exempt, or to allow the guilty to go free. The prevailing view of the Holmes stories as neat little allegories of Victorian positivism is belied by the concluding lines of “The Cardboard Box,” in a passage which tends to be passed over by both those who love the stories and those who dismiss them:

“What is the meaning of it, Watson?” said Holmes, solemnly, as he laid down the paper. “What object is served by this circle of misery and violence and fear? It must tend to some end, or else our universe is ruled by chance, which is unthinkable. But what end? There is the great standing perennial problem to which human reason is as far from an answer as ever.”

Philip Marlowe couldn’t have put it better.


A third innovative stroke of Conan Doyle’s was to find a new way to play the oldest trick in the book, to revise the original pretense of all adventurers, liars, and storytellers—that every word you are about to hear is true. For at least two hundred years before him, a writer of fiction might employ teasing initials, pseudonyms, and half-censored dates to give the impression that his story had been drawn from some available record, or that the author could personally vouch for its accuracy, but not without harming the innocent, embarrassing the guilty, or defaming the dead. Conan Doyle took this gambit one step further: not only were the Holmes stories presented as factual, with all the necessary names disguised, but their having been published, and subsequently widely read and even enjoyed, was known to their characters. Holmes was not only aware of his status as the subject of Watson’s “chronicles,” he resented it, and mocked it, even as he profited by the fictional version of the very real success that the stories enjoyed, first in the pages of The Strand and Collier’s, then in the many collected editions.

This kind of heroism, aware of its own literary celebrity, was not entirely new. The heroes of the Iliad know that they will one day feature in an epic song. In the second volume of Don Quixote, the knight’s career is distorted by the first volume’s publication (and subsequent piracy). And the opening lines of Huckleberry Finn display the same sort of post-publication self-awareness. The difference, and the innovation, is in Dr. Watson. Unlike the singer of the Iliad, or the narrator of Don Quixote, Watson is at once an active participant in the adventures and their recorder. And unlike Huck Finn, he deliberately presents himself, over and over, as an author preparing accounts for immediate publication, as the man charged with translating Holmes’s notes and his own recollections into stories that will, as Holmes sourly puts it, “pander to popular taste.” He serves as the guarantor of the stories’ “factuality” by experiencing them firsthand, by faithfully transcribing them, and finally by taking the heat for the supposed difficulties and annoyance their publication afterward causes for Holmes. Watson’s repeated insistence on his own active part in the stories’ finding their way into the hands of the reader—fully half begin with some kind of recognition of their own published status—encourages us to confuse the two doctors, Watson and Conan Doyle, who seem physically and even in their appetites to resemble each other.

Conan Doyle’s decision to play this particular fictional trick—to confuse the boundaries between fact and fiction, to write a disguised version of himself into the stories, to have Watson insist on the literal truth of the accounts—had consequences, like all the best tricks, that he did not foresee. On the one hand it produced the thousands of people who have written letters to Sherlock Holmes over the years and mailed them hopefully to 221B Baker Street (where they are read and answered, to this day, by a specially designated employee of the Abbey National Bank, which has offices at that address). It produced a sense of happy confusion in at least one discerning reader over the years: “Perhaps the greatest of all the Sherlock Holmes mysteries,” T.S. Eliot wrote, “is this: that when we talk of him we invariably fall into the fancy of his existence.” The pleasure to be derived from pretending to take fiction as fact was also one of the necessary conditions for the rise of the Sherlockians.

The other necessary condition for the rise of that alternately amusing and tedious tribe (not always alternately; to be honest, not always amusing) was the haste and carelessness that so often attended Conan Doyle’s writing of the tales. The first twenty-four Holmes stories were written in a period of twenty-nine months, and they are replete with all the contradictions, lacunae, and interesting mistakes of inspiration working under deadline. After the first batch there followed a ten-year interval, corresponding to two years in the world of Holmes and Watson (after the “death” of Holmes, in Moriarty’s arms, at Reichenbach Falls), during which time Conan Doyle forgot a lot of things he had already written about Watson and (the now resurrected) Holmes, producing further contradictions and errors. The perennially thorny question of Watson’s wife Mary, who may or may not have been his first or second wife, and may or may not have died, is the best-known example of this kind of authorial haste.

Conan Doyle, who always wrote quickly and claimed not to revise, seems to have been almost willfully careless when writing his Holmes stories, as if the act of disregarding each story’s predecessors and the assertions made therein about Holmes and Watson somehow mitigated their cumulative importance. Or perhaps Conan Doyle simply could not bear to reread them.

Thus the Holmes stories are constructed around a series of gaps. Some of these gaps are introduced only to be filled by the intuitions and inferences of the Great Detective: they are mysteries to be solved, as when the plans for the Bruce-Partington submarine have disappeared. Then there are the gaps deliberately introduced by Conan Doyle and deliberately left unresolved, in order to lend a greater air of authenticity to his stories. Some of these take the form of those famous allusions to other, unpublished or even unwritten cases that remain, in the view of Watson or Holmes, too scandalous, too libelous, or simply too horrifying to see the light of day. The best known of these is probably that of the Giant Rat of Sumatra, “a story,” as we are deliciously warned, “for which the world is not yet prepared.”

And then there are all the tantalizing gaps introduced purely through authorial carelessness into the chronology of the stories and the histories of the characters—the lack of information, for example, about Holmes’s university career; the strange intermittence and obscure fate of Watson’s wife, Mary, who suddenly disappears from the stories, or the oddly migratory wound that Watson received, in his leg or his shoulder, from a Jezail bullet.

Into these gaps has flowed the mock-scholarly tide of the Sherlockians. For the last ninety years, since Monsignor Ronald A. Knox’s essay “Studies in the Literature of Sherlock Holmes,” a vintage work of deliberate, straight-faced English silliness, writers well known and obscure have been devoting themselves, with a silliness that is sometimes deliberate and faces that are always straight, to trying to settle the questions raised by the gaps that Conan Doyle left lying around the Canon. They are playing the game begun by Conan Doyle—the game of pretending that the stories are true, that Holmes and Watson are, or were, real people. In this sense, the Sherlockians, or Holmesians (rhymes with Cartesians), as they are called in the UK, like The New Annotated Sherlock Holmes, are all Conan Doyle’s fault. He asked for them.

In good Victorian style, Leslie S. Klinger, the editor of the comprehensive new Holmes (all fifty-six of the stories in two volumes, with a third volume, of the four novels, to follow), appears to lead a double life. He has a day job as a Los Angeles tax and estate-planning attorney, but at night, one imagines, he dons the Inverness cape and deerstalking cap, and plays the great Sherlockian game. An acknowledged eminence in the field, he is the author of a number of less-whimsical bits of Holmesian research and of The Date Being—?, in which he tackles all the conflicting chronologies of Holmes’s and Watson’s lives and careers that Sherlockian scholars have attempted to construct, over the years, from the mishmash that Conan Doyle left behind. Mr. Klinger maintains an evident command of the vast corpus of Sherlockian essays, papers, and monographs, in which writers have treated subjects ranging from the high incidence in the stories of women named Violet to the angle at which Colonel Moran must have fired his air gun at the wax dummy of Holmes that Mrs. Hudson so diligently turned around in front of the window of the Baker Street flat.

Klinger is eminently qualified, in other words, to have produced a series of annotated editions of the Holmes stories for the fan-oriented Gasogene Books (the name of the press alludes to a kind of seltzer bottle that is one of the beloved furnishings of the flat at 221B), in the margins of which all the great and little controversies are rehearsed and pursued through the stories. Readers who take a tongue-in-cheek but nevertheless passionate interest in the question of why Mrs. Watson, or the first Mrs. Watson (in the event that you believe there to have been a second Mrs. Watson), should call her husband James when his name is John, or the shape and design of the Beryl Coronet, will find that Mr. Klinger has done all the reading and mastered all sides in the argument.

But the general reader, or even the devotee of Conan Doyle’s work, may experience a certain amount of confusion, or annoyance, in reading the enlarged, trade version of the Klinger annotation, published by Norton in a series that will eventually include annotated versions of Alice in Wonderland, Grimm’s Fairy Tales, The Wizard of Oz, A Christmas Carol, and others. Whatever its other virtues or weaknesses, I imagine that threaded through the notes to the new Norton Annotated Huckleberry Finn one will not encounter the running claim being advanced that Huck was a real person who wrote his own memoir, that Mark Twain merely passed himself off as the author when in fact he was just Mr. Finn’s literary agent, or that the reason Pap Finn so abuses Huck is because Jim had secretly fathered the boy on the Widow Douglas.

The Norton New Annotated Sherlock Holmes is, in other words, a thoroughly Sherlockian production, of a piece with, and in many ways the culmination of, the ninety-odd-year history of the curious game of pretending that Watson was a real man who wrote all the stories and that Conan Doyle was no more than “the Literary Agent.” In this regard it is a conscious sequel to the Baring-Gould Annotated Sherlock Holmes of 1967, which supplies extensive background information and commentary on the stories. But even for a reader who was at one time (circa 1973) relatively well versed in the grand minutiae of this game, it is disconcerting to read and enjoy the work of an author—to feel, thanks to the quality of his prose, a direct link to him and his imagination and his gentlemanly soul—while all around the text swarm a host of notes busily engaged in denying the poor man’s existence. After a hundred pages or so a kind of Pale Fire uneasiness settles in, and you begin to sense that the annotations do not always have the best interests of the text in mind.

Klinger, an intelligent, thorough, well-informed commentator, often seems prey to the same uneasiness. He knows that, like Rufus T. Firefly elevated to the position of president of Freedonia, he has a once-in-a-lifetime opportunity to put over the platform of the prankster party, to preach the message of the Sacred Writings to a potentially wide general audience. He writes, for example, that Conan Doyle while doing other things “also assisted Dr. Watson in the publication of The Valley of Fear, the last long story of Sherlock Holmes.” And in a note on “A Scandal in Bohemia,” the first published Holmes story, when Holmes remarks that Watson had chronicled “one or two of my trifling experiences,” Klinger asks, “Why ‘one or two,'” since only one story, A Study in Scarlet, had previously been published. Klinger then makes a characteristic comment: “Watson may have shown Holmes the manuscript of some of his early stories pre-publication.”

But at crucial moments he flinches—as when a discussion of Conan Doyle’s spiritualist beliefs becomes inescapably germane to a story like “The Adventure of the Sussex Vampire,” and Klinger makes a point of Holmes’s view being contrary to Conan Doyle’s; Holmes, rejecting the notion of spiritual intervention, says, “No ghosts need apply.” If Conan Doyle didn’t write the story, who cares what he thought of ghosts? As early as the prefatory pages, Klinger runs into difficulties, presenting a biographical section on Conan Doyle in which Sir Arthur’s authorship is, according to the rules of the game, never conclusively asserted. He refers, for example, to “Watson’s reminiscences [being] published under the byline of Arthur Conan Doyle.” And yet Klinger can’t quite bring himself to deny Conan Doyle’s authorship, either. He is a responsible annotator, who does an admirable job of clarifying points of Victorian life and culture—the tangled nomenclature of the Old Pence, the era’s railroads, the rise and fall of its successive Liberal and Tory governments—in a way that really does assist the reader.

The facts matter to Klinger, even when, as in the case of the fact that the stories show indisputable and frequently fascinating traces of Arthur Conan Doyle’s having written them, they make it hard for him to play his game. Sherlockian scholarship is an unacknowledged source, or at least an early instance, of what became a dominant mode of late-twentieth-century literary criticism: that which proceeded by first killing the author. But Klinger never quite steels himself to throw the final lever that will cut off Conan Doyle forever from his creation. And because Klinger is unable to stick in the knife—maybe he is himself too much of a gentleman for that—aspects of Conan Doyle’s authorship—such as his having served in the Boer War—crop up repeatedly in the annotations, spoiling the effect of the Sherlockian game.

The result is an awkward hybrid, one that soon obliges the reader to choose sides, between the notes that matter and the notes that don’t, between the facts and the “facts,” between fifty-six short works of enduring genius that can still be read profitably without any annotation at all, and ninety-odd years of collective fooling around. After several hundred annotated pages, the reader, however sympathetic in theory to the impulse of literary play that underlies the enterprise, begins to lose patience with the mock-serious, mock-informative, mock-mocking tone of the proceedings, and feels the urge to lay about the assembled throng of Sherlockians with a nice stout Panang lawyer.


It’s an awkward project, this New Annotated Sherlock Holmes. It’s as if the distributors of a Star Trek DVD had invited the leading speaker of Klingon to provide all of the extra features and commentary, appearing in full Klingon garb to explain that the episodes represent actual transmissions from the future, and that Gene Roddenberry was just some shnook who used to write for Have Gun, Will Travel. It’s also, I’m sorry to say, a superfluous project, because as fine as Klinger’s factual and purely explanatory annotations can be, we already have, from Oxford University Press, an excellent, multivolume, annotated Holmes; one that furthermore is a lot easier to read in bed.

But in putting its imprimatur on the Sherlockian project, Norton has implicitly validated a proposition I have long suspected and of which I am, in fact, a case study. Monsignor Knox’s puckish essay was more than a piece of self-parodying scholarship: it was an appropriation, for his own fictive purposes, of the characters, situations, and what would now be termed the “universe” or “continuity” of Conan Doyle’s stories. “Studies in the Literature of Sherlock Holmes” led directly, through works like Baring-Gould’s “biography” of Holmes, Sherlock Holmes of Baker Street (essentially a novel in the form of a biography), Billy Wilder’s film The Private Life of Sherlock Holmes (1970), and Nicholas Meyer’s The Seven-Per-Cent Solution (1976), to the contemporary, largely Web-based phenomenon that has devotees of various television programs, cartoons, and film series presenting their own prose versions of the adventures, histories, and sex lives of characters from Buffy the Vampire Slayer and Xena Warrior Princess. Such efforts are often derided or dismissed for the amateur productions they are, but the fact is that for at least the past forty years—since (take your pick) the French New Wave, or the Silver Age of Comics, or rock and roll’s British Invasion, popular media have been in the hands of people who grew up as passionate, if not insanely passionate, fans of those media: by amateurs, in the original sense of the word.

The first short story that I ever wrote was a tale of Sherlock Holmes, a pastiche written in a clumsy, ten-year-old’s version of the narrative voice of Dr. Watson. I was inspired to write my account of Holmes’s fateful encounter with Jules Verne’s Captain Nemo by having read and adored Nicholas Meyer’s then-popular account of the encounter between the detective and Sigmund Freud, which had in its turn been inspired, like every pastiche and Sherlockian monograph before and since, by those magical gaps, those blank places on the map that Conan Doyle left for us, by artlessness and by design. Readers of Tolkien often recall the strange narrative impulse engendered by those marginal regions named and labeled on the books’ endpaper maps, yet never visited or even referred to by the characters in The Lord of the Rings.

All enduring popular literature has this open-ended quality, and extends this invitation to the reader to continue, on his or her own, with the adventure. Through a combination of trompe l’oeil allusions, of imaginative persistence of vision, it creates a sense of an infinite horizon of play, an endless game board; it spawns, without trying, a thousand sequels, diagrams, and Web sites. In this sense the Sherlockian Game anticipated, and helped to invent, the contemporary fandom that has become indistinguishable from contemporary popular art; it was the Web avant la lettre.

And yet there is a degree to which, just as all criticism is in essence Sherlockian, all literature, highbrow or low, from the Aeneid onward, is fan fiction. That is why Harold Bloom’s notion of the anxiety of influence has always rung so hollow to me. Through parody and pastiche, allusion and homage, retelling and reimagining of the stories that were told before us and that we have come of age loving—amateurs—we proceed, seeking out the blank places in the map that our favorite writers, in their greatness and negligence, have left for us, hoping to pass on to our own readers—should we be lucky enough to find any—some of the pleasure that we ourselves have taken in the stuff we love: to get in on the game. All novels are sequels; influence is bliss.

—This is the second of two articles.

This Issue

February 24, 2005