Sherlock Holmes Was Wrong

by Pierre Bayard

  The Holmes method, the one we see at work in the very first Holmes story and in all the cases that follow, rests on three operations: observation, comparison, and reasoning backward.

  As the detective indicates in his conversation with Watson, these three processes sometimes all occur at the same time (“From long habit the train of thoughts ran so swiftly through my mind, that I arrived at the conclusion without being conscious of intermediate steps. [ . . . ] The whole train of thought did not occupy a second.”25). Nonetheless, it is desirable to separate the three constituent operations of Holmes’s method if one wants to study how they function.

  Thus presented, this method offers all the appearances of rigor, since it rests on logic and relies on the discoveries of science. (We too will come to use it, especially when we resort to animal psychology and backward reasoning.) But can Holmes be sure that his trusted method will lead him to the truth? Nothing could be less certain, as we shall see.

  * Another form of indirect trace is the stain, which does not appear in The Hound of the Baskervilles.

  * The devout Sherlock Holmes reader will recall the central importance of animal psychology in “The Crooked Man,” “The Adventure of the Creeping Man,” “Silver Blaze,” “The Speckled Band,” and, most famously, the “curious incident of the dog in the night-time,” also in “Silver Blaze.”—Trans.

  IV

  The Principle of Incompleteness

  CONSIDERED A MODEL of scientific rigor, even an inspiration for certain procedures taught in police academies, the Holmes method still does not yield the anticipated results every time—far from it. And an uncompromising examination of its results throughout the detective’s years of activity leads to complex conclusions that shatter the Holmesian image of success—as well as the Holmesian image of unfailing self-satisfaction.

  To begin with, it is not insignificant that the weaknesses of Holmes’s method are brought out in the prologue, during the first meeting with Dr. Mortimer, as if the admission of failure were an epigraph to the investigation that follows. Mortimer had stopped by the detective’s flat the previous day and, finding Holmes absent, had absentmindedly left his cane there. As they wait for a return visit from their future client, about whom they as yet know nothing, Holmes and Watson amuse themselves by applying the detective’s method to this unknown object.

  Watson begins, indulging in a whole series of deductions based on clues taken from the cane. From the presence of inscriptions on the cane he deduces that it’s a gift offered to an elderly doctor, and from its poor condition that its owner is a country practitioner visiting his cases on foot. The engraved initials suggest to him that the gift was offered to the doctor by members of a local hunting association.

  Holmes ironically congratulates Watson on his abilities before explaining to him that he was mistaken on most of the points. He acknowledges that the visitor is probably a country practitioner and a great walker, but challenges his friend’s other deductions. A gift made to a doctor is more likely to come from a hospital than from a hunting club; further, it is legitimate to suppose, since the doctor went to the country where positions are less sought, that he is a young doctor.

  But although he questions his friend’s deductions and arrives at a certain number of results, Holmes too is wrong, on one point at least. It was not to settle in the country but to get married that Dr. Mortimer left the hospital. At the beginning of the book, the detective acknowledges his mistake:

  As he entered his eyes fell upon the stick in Holmes’s hand, and he ran towards it with an exclamation of joy.

  “I am so very glad,” said he. “I was not sure whether I had left it here or in the Shipping Office. I would not lose that stick for the world.”

  “A presentation, I see,” said Holmes.

  “Yes, sir.”

  “From Charing Cross Hospital?”

  “From one or two friends there on the occasion of my marriage.”

  “Dear, dear, that’s bad!” said Holmes, shaking his head.

  Dr. Mortimer blinked through his glasses in mild astonishment.

  “Why was it bad?”

  “Only that you have disarranged our little deductions. Your marriage, you say?”

  “Yes, sir. I married, and so left the hospital, and with it all hopes of a consulting practice. It was necessary to make a home of my own.”

  “Come,come, we are not so far wrong, after all,” said Holmes.26

  Unfortunately, Holmes’s initial mistake about why Dr. Mortimer left the hospital is not the only one he makes in the book.

  There are others, with much more serious consequences. First, the slowness with which Holmes catches the killer (assuming we believe that Stapleton is the killer) allows another murder to take place. Faced with Selden’s corpse (which he mistakes for Baskerville’s—another error), as Watson is blaming himself for having lost sight of the man he was supposed to protect, Holmes rather reasonably takes the blame on himself:

  “I am more to blame than you,Watson. In order to have my case well rounded and complete, I have thrown away the life of my client. It is the greatest blow which has befallen me in my career. But how could I know—how could l know—that he would risk his life alone upon the moor in the face of all my warnings?”27

  That Holmes, as we will demonstrate, is completely mistaken about the identity of the murderer does not excuse the casualness with which he accuses himself here. This error is different from the mistake with Dr. Mortimer’s cane, since it is not an error in deduction, but it does rest on a faulty evaluation of criminal psychology—a black mark against the detective.

  Unable to protect Baskerville a first time, Holmes is equally incapable in the final scene, in which he causes his charge to run the gravest risks—in fact, to be almost torn apart by the hound. Even if the mistake here is not strictly speaking intellectual, Holmes has once again let us glimpse his difficulty in taking reality into account and adapting his conduct to it intelligently.

  These blunders should not come as a surprise. In the sixty works in the Holmes canon there are countless mistakes, illustrating all the weaknesses of the Holmes method and putting its ostensibly “scientific” quality very much in perspective. There are two kinds of mistakes: either Holmes is wrong—in his actions or in his reasoning—or he doesn’t arrive at the solution.

  In “A Scandal in Bohemia,” one of the very first stories, Holmes undergoes a bitter failure and lets himself be completely manipulated.* In “The Adventure of the Engineer’s Thumb,” he fails to find the criminals. He is negligent in “The Five Orange Pips” and in “The Resident Patient,” in which he allows crimes to be committed and murderers to escape; in “The Adventure of the Solitary Cyclist” and in “The Greek Interpreter,” in which he cannot prevent kidnappings; in “The Illustrious Client,” in which he fails to prevent an attack against himself and the disfigurement of a criminal; and in “The Adventure of the Three Gables,” in which he does not foresee either a major burglary or the destruction of a manuscript.

  To these errors in tactics a great many errors in reasoning could be added. Holmes acknowledges at the end of “The Adventure of the Speckled Band” that his first conclusion was “entirely wrong”; at the end of “The Adventure of the Lion’s Mane,” where the solution comes to him only when he at last remembers something he has read, he allows that he “went astray” all throughout the investigation. In “The Man with the Twisted Lip,” he wrongly tells a man’s wife that her husband has died. In “The Adventure of the Bruce-Partington Plans,” someone other than the expected suspect falls into the trap. In “The Disappearance of Lady Frances Carfax,” Holmes forces open a coffin lid without finding the person he was looking for and acknowledges that he has suffered the “temporary eclipse to which even the best-balanced mind may be exposed.” In “The Adventure of the Missing Three-Quarter,” he misses the truth completely, and in “The Yellow Face” he is so sorely mistaken from start to finish that he will afterward refer to this affai
r as a model of mistakenness.*

  The idea that the infallible Holmes sometimes makes mistakes† does away with the notion that any superior authority may be entirely entrusted with deciding true and false; it makes the truth inherently unstable. And if the person who is supposed to determine the truth can be mistaken in his first conclusion, he may just as easily, when he thinks he has corrected his mistake, simply have fallen into another one in his second conclusion. Thus all the solutions to Holmes’s cases are open to suspicion.

  Besides the cases left uncertain by Holmes’s mistakes, there are also a certain number that remain simply unsolved, either partially or completely. Far from embracing an unequivocal solution, they leave open multiple hypotheses. They are un-decidable. At the end of “The Musgrave Ritual,” for instance, Holmes acknowledges that one important detail will never be able to be clarified. Likewise, at the end of “The Adventure of the Norwood Builder,” the detectives turn out to be incapable of explaining an essential clue; in “The Adventure of the Dancing Men,” doubt persists about who fired the gun; and in “The Adventure of the Six Napoleons,” the method by which a thief came into possession of the jewel remains obscure.

  In other texts, undecidability does not afflict one detail or another of Holmes’s hypothesis; instead, the hypothesis itself, satisfying as it may be, fails to preclude the coexistence of alternative explanations. In “The Adventure of the Norwood Builder,” Holmes himself points out that half a dozen theories would fit the facts. In “The Adventure of Black Peter,” similarly, he studies several different hypotheses, toying with them happily before settling on one.*

  There is actually nothing surprising about the uncertainty of Holmes’s solutions. In fact, the famous detective’s method contains within it three basic elements that, taken together, open up the range of possible conclusions much more broadly than the detective allows.

  The first of these is the way the clues are collected. In the Holmes method, clues are not indisputably obvious things on which the deductive process is exercised after the fact; rather, they arise largely through an act of creation. For there to be a clue, there must first occur a selection within the infinite field of potential clues that a given scene presents. Clues present themselves only after a twofold process of choice and nomination.

  A clear example is the main “clue” that Dr. Mortimer brings Holmes during their first meeting: the footprints of a giant hound, a bit of evidence neglected by the police.

  I confess at these words a shudder passed through me. There was a thrill in the doctor’s voice which showed that he was himself deeply moved by that which he told us. Holmes leaned forward in his excitement and his eyes had the hard, dry glitter which shot from them when he was keenly interested.

  “You saw this?”

  “As clearly as I see you.”

  “And you said nothing?”

  “What was the use?”

  “How was it that no one else saw it?”

  “The marks were some twenty yards from the body and no one gave them a thought. I don’t suppose I should have done so had I not known this legend.”28

  The investigators did indeed see the dog’s footprints, but paid them no mind. What constitutes a clue for one person may be meaningless to another. And a clue is named as such only when it serves as part of a more general story—of an overall construct at the disposal of the person who has decided to grant it the status of a clue.

  If a clue is a choice, it follows that a number of elements of the book’s reality must form “virtual clues,” ones that the chosen hypothesis ignores. Thus we can suppose that the investigators ignored a multitude of signs that could have become clues but that do not figure in the text, since they weren’t transmitted by Dr. Mortimer. What’s more, even confining ourselves to the signs we know about,we’ll see that the text delivers a whole series of signs that are not officially granted the status of clues—and whose correct interpretation can significantly change the overall solution.

  When a clue is a matter of choice, it is also a matter of interpretation; hence the plurality of possible meanings. The second flaw in the Holmes method stems from the subtly maintained confusion between scientific law and statistical generality.

  Holmes usually bases his deductions on statistical recurrences, but he omits the fact that these recurrences have no constraining power over reality. Rather, they only sketch the shape of probabilities and therefore leave open the possibility for exceptions in every case.

  Thus it is altogether excessive for Holmes to speak of “deduction” (“as Dr. Mortimer . . . deduced from the cigar ash”)29 when he refers to the doctor’s hypothesis relating the size of the heap of ashes in front of the wicket-gate to the length of time that Sir Charles Baskerville lingered there.

  What slips between scientific law and statistical generality is in fact the individual suspect. As an object of investigation, the suspect does indeed come under the jurisdiction of statistics—which can include him within a range of probable perpetrators—but he is not under the rule of scientific law; in every circumstance, the suspect retains a share of free choice that allows him to avoid complete predictability.

  A suspect cannot exonerate himself by saying he fled the scene of the crime by flying away; no one can free himself from the law of gravity. But the deduction that Mortimer and Holmes make is not ironclad. It is true that in many cases a large pile of ash signifies a smoker standing still, but it can just as well signify that he waited for a long time before he tapped his cigar. It is true, too, that prolonged standing in the cold could be motivated by a meeting, but it could also be explained if Sir Charles had been lost in thought, or if he had observed something that interested him.

  It is the human subject in general, and the element of the unconscious in particular, that slips into this gap between scientific law and statistical regularity. In so doing, it escapes the Holmes method, which functions perfectly and with great elegance on abstractions but is not necessarily adapted to resolving the complex individual problems with which the police are confronted.

  The third flaw in the Holmes method is the lack of understanding about the place of the subject in the overall interpretation of the case. The pseudoscientific Holmes method eliminates the factor of individual psychology among the actors in the mysteries it is trying to solve. But it also eliminates this factor in the investigator himself, even though it is a key determinant.

  This psychological factor plays a prominent role in the overall construction that allows the investigator not just to interpret clues, but even to decide what is a clue and what is not. Holmes’s construction of the events, which tallies with Dr. Mortimer’s, has obviously been created by the detective, at least partially, before the investigation has even properly begun.

  The detective’s hypothesis rests on a figure and a motive. The figure is that of the murderer with the dog, a criminal who makes use of both legend and a beast to commit his murders. Note that this construction eliminates other hypotheses, for instance that of accident—defended by the police—or that of a murder committed in some different way. Like a magnetic pole, it quickly attracts a whole series of “clues,” such as animal footprints, which would not necessarily be examined at all in the framework of other constructions.

  It is clear, too, that from the very beginning Holmes places a higher value on one motive in his selection and interpretation of clues—namely, on financial interest. This is obviously not to be ignored when the victim possesses a large fortune. But it is not the only reason for which human beings commit murders, and it has the drawback of leaving other possible explanations for the tragedy of Baskerville Hall unexplored.

  This construction is trapped in an imagined universe, a universe that truly belongs to Holmes, even though it finds sustenance in the imaginations of those close to him. It is an imagined world that, although the detective fights against it—Holmes does not believe in the legend of the hound responsible for carrying out the curse—nonetheless remains clo
sely linked to a fantastic vision of reality; from a hound alone, there is a simple transposition to a murderer with a hound.

  Not only is there an imagined world at work behind Holmes’s hypothesis, then, but there is reason to think that the hypothesis is truly active only because it is animated by a private fantasy world—determined both by the sex and the social standing of the detective—which shapes, and thus perturbs, his way of seeing the world.

  Far from being a closed system, the Holmes method thus allows alternative solutions to subsist, both at the point-by-point level of clues and at the level of the overall construction.*

  * Holmes is beaten by a woman, Irene Adler, who is always a step ahead of him and guesses everything he will do.

  * “Watson,” said he, “if it should ever strike you that I am getting a little overconfident in my powers, or giving less pains to a case than it deserves, kindly whisper ‘Norbury’ in my ear, and I shall be infinitely obliged to you.” (“The Yellow Face,” in The Complete Sherlock Holmes, New York: Garden City Publishing Company, 1930, p. 415.) James McCearney notes that “out of the twenty-four stories published between July 1891 and December 1893, a good half dozen of them result in partial or total failure” (Arthur Conan Doyle, Paris: La Table ronde, 1988, p. 152).

  † Aside from mistakes, major elements sometimes elude Holmes. These range from a key plot point—such as an important link of kinship, in “The Adventure of the Priory School”—to a secret, as in “The Gloria Scott.”