“Oh SHERLOCK, SHERLOCK, he’s in town again,

That prince of perspicacity, that monument of brain.

It seems he wasn’t hurt at all

By tumbling down the waterfall.”

So went the ditty that P.G. Wodehouse published in 1903, upon Holmes’s miraculous resurrection from his deadly descent down the Reichenbach Falls. This return from the dead was a momentous event indeed, and well worth Wodehouse’s commemoration: when Arthur Conan Doyle had killed off his famous detective almost ten years earlier, in “The Final Problem,” the news was, to put it mildly, not well received. The Strand—for many years, Holmes’s home—was inundated with letters from jilted readers. Conan Doyle found himself the target of angry mail and vitriolic attacks. It’s even said that City of London clerks wore black armbands to mourn the detective’s passing.

Still, Holmes’s creator stood firm. Holmes was gone for good. “Killed Holmes,” he wrote simply in his diary the December that “The Final Problem” appeared in the Strand Magazine. And Holmes would not be back.

But in the end, despite all his reservations, Conan Doyle succumbed. The pressure was just too great. And so, Sherlock Holmes was born anew. It may not have been the most graceful of returns—as Wodehouse wrote, “The explanation may be thin”—but you know what? Nobody much cared. Just as long as their hero was once again alive, any explanation, however vague on detail, would do. (“But bless you! We don’t care a pin, / If he’ll just give us back our SHERLOCK,” continues Wodehouse’s poem.)

Arthur Conan Doyle had learned a valuable lesson: you cannot kill Sherlock Holmes. It simply isn’t done.

Why did the public react as they did? Why couldn’t they let Mr. Holmes rest in peace at the bottom of a Swiss waterfall, and his creator, be at peace in his retreat at Undershaw, focusing on Spiritualism and other matters that were now closer to his heart than his detective hero? Many factors were at play, surely; but in the end, I think it all comes down to a simple psychological phenomenon: the contrast effect.

In 1989, psychologist Timothy Wilson came up with something known as the Affective Expectation Model (AEM). According to the AEM, the way people felt as a result of an experience wasn’t just a function of the experience itself. It also depended heavily on any prior expectations they might have had. In other words, if we expect that something will make us happy, we are more likely to feel happy after the experience; anything that doesn’t match will be assimilated into the overall expectations. So, for instance, if we know we’ll be looking at cartoons, and some aren’t quite as funny as we’d hoped, we’ll still likely smile if the rest are up to par (even though we might have dismissed them on their own). But what if we expect those same cartoons to make us happy and find instead that there is a great big gulf between the images themselves and what we’d thought we’d find? In other words, what if the cartoons are incredibly unfunny? What happens then? As it ends up, we feel far more frustrated, upset, and overall peeved than if we’d had no expectations to begin with.

The contrast effect is, in simple terms, a mismatch between expectations and reality—that makes the reality all the more disappointing. Take, for instance, a rainy day. If you’re visiting Seattle or London, no big deal, right? You knew to expect it. You might not be ecstatic, being on vacation and all, but you won’t be despondent, either. But what if you’ve just arrived in Santa Barbara or the Bahamas? Suddenly, it’s a much bigger deal. It may seem like the same rain, but it’s really not. You were expecting sun. And because of the contrast between your expectations and the actual weather, the rain is way, way worse. Even a milder shower as you head for the beach in Hawaii can mean a far greater disappointment than a downpour over Piccadilly.

The same thing holds, broadly speaking, for our experience of literature – or any entertainment, really. No longer do our feelings about the rain depend on geography. Now, they depend on that thing known as genre—in the broad sense of the word. What type of thing am I reading (or listening to or watching, as the case may be)? And what does that imply for my expectations?

Simply put, different genres breed different expectations. Anna Karenina may be one of the most beloved heroines of all time—but while many readers may have cried over her fate, few are likely to have blamed Tolstoy for throwing his leading lady under a train. It is completely acceptable, even expected one might say, in a literary novel. Many an Achilles has been thrown to the metaphorical wolves—even adventure heroes can be tragic, if you’re in the realm of Greek drama—but could the three musketeers have gotten away with a loss in their number? Hardly. Alexandre Dumas would have never heard the end of it.

When we’re watching film noir, we know that the hero is going to end badly. I start crying for Humphrey Bogart at the beginning of each of his noir adventures, knowing that every choice is pushing him closer to the unfortunate end. And could Jean-Paul Belmondo please, pretty please live to see the end of at least one of his ill-fated action films? It would make me ever so happy. I know, of course, that he won’t. But I don’t blame Jean-Luc Godard or Jean-Pierre Melville. I knew what I was in for. On the flipside, while Philip Marlowe may get beaten up on more than one occasion, even in this most literary of criminal turns, he always emerges alive in the end, no matter how many guns he’s had held to his head or how many knives have poked him in the ribs. Raymond Chandler knows just how much Marlowe can take.

The list of possible examples is endless. What would have happened to an Ian Fleming who sent James Bond spiraling into frothy waters? Or even an Isaac Asimov who let the Foundation fall into ruin mid-way through his series—and then devoted the rest of the books to a counter-reality? I’m willing to bet the stories would have lost much of their satisfaction. For, that is not their point. That’s the point of other genres and other literatures.

Oftentimes, the ending is what determines the genre in the first place, turning action adventures or detective thrillers into noirs and dramas at the turn of a scene. Does the bullet hit or miss? The police or villain get there in time or just a second too late? Take Shakespeare. Marjorie Garber—and other Shakespeare scholars—have put forth the case that Romeo and Juliet was originally conceived as a comedy—and A Midsummer’s Night Dream, that most comic of comic masterpieces, was to be a tragedy. But then, the endings went in a different direction—and the genres were switched. If you look at the structure of the two plays, you see remarkable similarity. The same mistaken identities, the same meetings of lovers, the same ribald lines and laugh-filled scenes. How easy it would have been for the messenger to have traveled just a bit quicker, or for the magically confused lovers to have remained in the thrall of their delusions. One small change, and you get an entirely different type of story.

Shakespeare had the flexibility of writing both comedy and tragedy. No one came into Romeo and Juliet or Midsummer Night knowing what to expect. But writers don’t always have that luxury—especially not with a recurring character, or a theme that has been set ahead of time. When readers opened to the pages of the Strand to read of Sherlock Holmes’s latest adventure, it wasn’t like reading or watching Shakespeare for the first time, not knowing whether you were in for comedy or tragedy and being willing to go in either direction. The expectations had already been set. You knew you were dealing with detective fiction—just like viewers who saw Bogart of Belmondo go to their deaths knew ahead of time that they were about to see a film noir. It had already been advertised as such.

When you highlight your genre ahead of time—however you’ve chosen to do so—you have to remain faithful to the parameters you set—or risk the wrath and black armbands of your loyal fan base. When you betray the expectations of the genre, it’s like pulling the rug out from your audience. It’s a far worse crime than if you’d not set any expectations to begin with. Even Shakespeare, I’m sure, would not have fared well had some impish or perverse part of him decided to advertise Romeo and Juliet as comedy before opening night. What would the Queen have thought?

It’s safe to assume that on some level, Conan Doyle knew that killing Holmes may not have been the best thing for him to do. He just couldn’t help himself. “I have made up my mind to kill Sherlock Holmes; he is becoming such a burden to me that it makes my life unbearable,” he told Sir Henry Lunn (and Lunn, in a story that may or may not be apocryphal, even suggested the Swiss locale).

But after “The Final Problem,” he’d learned his lesson for good. The next time he said goodbye to Holmes, it was to leave him to a peaceful retirement, tending bees in the country. And when Holmes has reappeared in post-Doylean adaptations—including the two most recent incarnations, in the guise of Robert Downey, Jr. and Benedict Cumberbatch—it has been overwhelmingly clear that the Reichenbach Falls wouldn’t be the end of him. In the film version, we see Holmes emerge in camouflage from the chair in Watson’s study, ready to show the world that he didn’t die that night in Switzerland—and in the BBC telling, we see him watch from the shadows as his friend sheds tears on his supposed grave. Lest the viewer be unfamiliar with the Holmes canon, there is no ambiguity in these new versions: Holmes lives on. And he will be back. It’s not for the author to kill off someone so great.

Conan Doyle didn’t know how to kill his hero. He struggled with multiple endings and even wrote several alternate stories that sent the great detective to his death. But none—not even the eventual winner—felt right. Something always bothered him. Nothing he came up with felt like his best effort. Why? On some level, he must have felt that the enterprise was, in some profound way, doomed. For sometimes, the answer to “how do you kill your hero” is simple: you can’t. You have created a hero who, happens what may, must live.

Mastermind: How to think like Sherlock Holmes is now available for pre-order.



Wilson TD, Lisle DJ, Kraft D, & Wetzel CG (1989). Preferences as expectation-driven inferences: effects of affective expectations on affective experience. Journal of personality and social psychology, 56 (4), 519-30 PMID: 2709307