Sherlock Holmes and Philosophy

by Josef Steiff

  Let us hear the suspicions—I will look after the proofs.

  The final facet of the JTB theory is justification. To see why it’s important, let’s look at examples from the 2010 BBC Production Sherlock. In the episode titled “The Great Game,” Ms. Wenceslas (the curator of the Hickman Gallery) displays a recently discovered (fake) Vermeer painting. Since the gallery is presumably reputable, and the forgery is of the utmost quality, everyone who sees it is justified in believing that it is authentic. However, Holmes discovers reasons to doubt its authenticity. This undermines the justification for his belief that it’s real. Thus, even if the Vermeer were genuine, his doubts would mean Holmes no longer knows it. If something legitimately weakens your justification, your belief no longer counts as knowledge.

  But can justification turn a true belief into knowledge? In some cases, yes! An example is found in that same episode when Alex Woodbridge (the security guard at the gallery) contacts Professor Cairns regarding the Vermeer. Mr. Woodbridge believes that the Van Buren supernova shouldn’t appear in the painting (since it was not visible when Vermeer would have painted it). But when he reaches out to Professor Cairns, he needs confirmation of his belief (generally speaking, suspicions and hunches are not well justified). Had Mr. Woodbridge actually reached Professor Cairns, she would have confirmed his suspicion. Confirmation by an expert would provide Alex with stronger justification, and his belief would become knowledge. In short, while hunches and suspicions generally don’t give rise to knowledge, with proper confirmation they can become knowledge.

  There is nothing more stimulating than a case where everything goes against you.

  The JTB model sets up excellent criteria for assessing most cases of knowledge. Even when we feel absolutely certain that we know something, losing any one of those parts (the belief, the truth, or the justification) leaves us without real knowledge. But, strong as it is, the JTB theory isn’t perfect. In 1963, a philosopher named Edmund Gettier argued that justified true beliefs don’t always count as knowledge. There are in fact cases where you have a justified true belief but you only got there because of sheer luck.

  To illustrate the point, Gettier came up with an example much like the following: Suppose that Holmes reveals evidence to Watson that a murder was likely committed by Mr. Jones. Mr. Jones is a former ship hand so Watson rightly forms the belief “the murder was committed by a former ship hand”. Unbeknownst to Holmes and Watson, a different former ship hand actually committed the murder. Now, here’s the trick: It turns out that Watson’s belief—that the murder was committed by a former ship hand—is justified (Holmes showed good reason to believe it), it is true, and it is believed by Watson. Thus, according to the JTB theory it has all three parts and should count as knowledge. But it doesn’t seem as if Dr. Watson really knows that a former ship hand committed the murder. It’s only a matter of luck that he got it right. After all, Watson came to his conclusion by thinking about the wrong chap, so he might technically be right (the murder was committed by a former ship hand) but somehow it’s just a lucky guess.

  Fortunately for the King and Queen, I was on top of my game.

  What exactly is a lucky guess? In epistemology luck tends to be spelled out in terms of the safety and sensitivity of a belief. If a belief is safe then in most nearby possible worlds where you form the belief in the same way, it’s still true. A belief is sensitive if in the nearest possible worlds where it’s false, you wouldn’t form the belief in the same way.

  But what are “possible worlds?” Think of a possible world as a hypothetical situation, some way you can imagine the world as it might have been. For instance, imagine a world just like ours, but where Holmes was a real man; that’s one possible world. Some possible worlds are called “nearby” because they closely resemble our world in important ways (the real-Holmes world you just imagined is reasonably close). Others are considered to be “further away” because they are significantly different to ours (perhaps a world where Earth is populated by robot dinosaurs from Mars). There’s a lot more to possible worlds, but this is enough for our purposes.

  Let’s look again at what safety means now that this “possible worlds” stuff is a little clearer. Essentially, if a belief is safe, then so long as your reasons for the belief don’t change, small changes in the world shouldn’t make it false. For example, if you flip a coin and see that it came up heads, the belief that it did come up heads is pretty safe. It doesn’t matter if you flipped the coin in a kitchen, a den, or on the street. It seems that so long as your reason for the belief (seeing it for yourself) doesn’t change, the belief is still probably true.

  Additionally, assuming your reasons for forming a safe belief remained constant, falsification would have to involve large-scale changes to the world. Again, by way of example: your belief that Arthur Conan Doyle wrote the original Sherlock Holmes stories is pretty safe. Why? Because your belief that he wrote them is based on repeated testimony from multiple sources. All of those people would have to be mistaken for your belief to be false. A world where everyone is mistaken about who wrote a famous story seems pretty far away from ours.

  In contrast, the philosopher Bertrand Russell gave an example of an unsafe belief much like the following: suppose that Watson realizes that his pocket-watch has stopped during the night. So, when he heads out the door, Watson glances at his table clock and sets his watch to match it. However, unbeknownst to him, the table clock also stopped during the night. Luckily for Watson, the clock stopped precisely twelve hours before. Thus, when he glanced at it, it displayed the correct time (because he happened to look at it at precisely the right moment). Assuming that the table-clock is well made and normally accurate, Watson has every reason to believe it is accurate now (he’s justified, and does believe it). Further, it did give him the right time (it’s true). However, it seems that he might not really know what time it is. Since Watson would have formed his belief about the time in the same way even if there were a tiny change that made it false (such as the clock stopping five minutes earlier, or Watson glancing at it five minutes later), this belief is not safe.

  Now let’s look at sensitivity. If a belief is sensitive, then in the nearest possible worlds you won’t form the belief that a fact is true unless it actually is true. For example, your belief that you can read English is probably a sensitive one. If you couldn’t read English, you most likely wouldn’t believe you could. Sure, we can imagine a world where you were raised speaking Dutch and thinking it’s English. However, that would require large-scale changes to the world and would be a comparatively distant possible world. All that sensitivity requires is that in the closest worlds where it’s not true, you stop believing it. In the close worlds you simply can’t read English, and everything else is roughly the same. Since in those worlds you wouldn’t believe that you could read English, your belief is sensitive.

  I have heard your reasons and regard them as unconvincing and inadequate.

  So, does Holmes ever get lucky? While Holmes often forms beliefs before the reader or Watson, most of us don’t think that Holmes is merely making lucky guesses. Holmes himself denies this possibility in The Sign of the Four: “I never guess. It is a shocking habit—destructive to the logical faculty.” But why aren’t his deductions just guesses? How does he succeed where we do not?

  The answer is found in his process of Holmesian deduction. This process is what Holmes is describing when he says: “When you have eliminated the impossible, whatever remains, however improbable, must be the truth.” Holmes starts with simple observations and moves to seemingly incredible conclusions. However, our own Holmesian deductions are not safe. Even if we rightly deduce “whodunit,” we have to acknowledge that things could merely appear that way to us regardless of whether it’s true.

  How does Holmes avoid unsafe beliefs? He relies on his incredible skill at analyzing evidence, finding the truth, and adopting it as his hypothesis. His skills allow him to seemingly unerringly find the right (the safe) j
ustification for any belief he forms. An example of this is found in the 2009 film Sherlock Holmes, starring Robert Downey Jr. Midway through the movie a constable informs Holmes that Lord Blackwood is back from the dead. It turns out the constable’s belief that Blackwood is alive is not very safe; It is based upon a combination of his belief that Blackwood was dead, and testimony that Blackwood was seen walking through the graveyard. However, since one of the beliefs justifying his belief that “Blackwood is alive” is false it seems as if there must be many nearby worlds where he formed this belief about Blackwood in the same way even though it’s not true. Blackwood could really have been dead and the groundskeeper mistaken. Or, he could have never been dead at all, thus the belief that he is “back” is mistaken. Either way the constable’s belief that Blackwood lives seems unsafe: it’s not properly connected to the fact that he is alive.

  Holmes, however, realizes that if the groundskeeper really saw Blackwood, then he must have simply faked his own death. Further, he did so convincingly enough to fool Watson (a fact which Holmes plays upon to draw Watson back into the investigation.) From this he deduces that Blackwood must never have been dead in the first place. Thus, Holmes’s belief that Blackwood lives, based on the second-hand testimony of the groundskeeper, seems safe. Holmes’s superior powers of reasoning allowed him to “rule out the impossible” and thus unerringly determine the real truth by reflection on the facts presented to him. This seems to allow Holmes to zero in on only the relevant facts which can support his beliefs in a “safe” way.

  One should always look for a possible alternative, and provide against it.

  But that only covers safety. Now let’s look at sensitivity. As you may recall, if a belief is sensitive, then you would give it up when the relevant fact turns out to be false. Holmes escapes problems with insensitive beliefs by using his famous powers of observation. The combination of the evidence at hand and the lack of evidence to support alternative theories makes his deductions sensitive.

  A very straightforward example of this can be found in the adventure “The Stock-Broker’s Clerk.” In this story, Holmes notices that Watson’s new slippers are fire scorched but retain a paper wafer, near the instep, which is stamped with a maker’s mark. Holmes correctly surmises that Watson had a cold and kept his feet near a fire for warmth. The other probable explanation (drying after soaking) would have removed the maker’s label, so Holmes rules it out. Any alternate explanation would presumably leave different evidence.

  When a belief is sensitive it means that in the nearest possible worlds where the fact is false, you cease to believe it. We can turn this around and instead formulate it as “How many facts about the case could you change before Holmes changed his belief?” Because Holmes would (presumably) notice any evidence that undermined his theory, we can assume that if almost anything changed, he would notice. In all the nearest possible worlds where his belief is false (Watson didn’t have a cold) we can presume that Holmes would not believe it. He could be deceived, however, doing such a thing would require significant effort (like the earlier example about your ability to read English.) Surely, those possible worlds with huge conspiracies to deceive Holmes are not the nearest worlds where he is wrong. Since sensitivity is only concerned with the nearest worlds where the fact is false, huge conspiracies and other large changes to the world don’t create problems For Holmes’s sensitive beliefs.

  Now, there are cases where Holmes has been wrong. For example, he didn’t realize that Irene Adler knew who he was, a fact which enabled her to elude him. One might wonder if this sort of failure undermines the idea that his methods are reliable. If Holmes’s methods are unreliable we might worry that his deductions can’t give him the justification he needs for a justified true belief, much less safe or sensitive beliefs! It turns out that, despite his failures, the fact that Holmes himself believes something is sufficient justification for him to form knowledge. This might sound like Holmes is “cheating” the knowledge system. Surely he can’t be justified in believing something just because he believes it, can he? In short, yes he can.

  We balance the probabilities and choose the most likely. It is the scientific use of the imagination.

  This “cheating” self-justification is far more common than you might think. It’s actually very similar to the way knowledge is formed in many scientific and academic fields. Suppose you’re an expert in some particular field. If you formulate a theory based on some piece of evidence, you and your colleagues are justified in believing your theory partly because of your expertise. The fact that an expert believes something gives you some justification to believe it, even if that expert is you!

  If this were not the case, observations about physics made by any “average Joe” might be as well justified as those of professional physicists. This idea is crazy; when Einstein discovered relativity he was far better justified in believing it than most of us would be if we thought about the same things. When professionals come to conclusions it’s different from when the rest of us do. The reason for this difference can be found in the reliability of professional methods. Holmes’s self-justification is based on his own history of successfully coming to the right conclusions.

  Every time Holmes successfully deduces the answer he is better justified in believing in his own methods. As strange as it might sound, his skills of deduction can be likened to a professional billiards player. Just as we wouldn’t say “a professional pool player is ‘just lucky,’” so Holmes isn’t just lucky when he correctly deduces an answer. The difference between Holmes and us is the difference between a professional and someone with beginner’s luck.

  When a first-time player successfully makes a difficult shot in billiards, it is rightly considered “beginners luck.” However, if that same player were to go on and continue making difficult shots for years, we would say he or she is a naturally talented professional. Similarly, until we build up a long successful history like Holmes’s, we’re just getting lucky.

  Further, as often noted in various stories, Holmes’s few failures have caused him to redouble his efforts: he learns from his mistakes. Assuming he properly adjusts his methods to shore up any weak spots and always works to form only safe and sensitive beliefs, he can be justified by his own methods.

  Education never ends Watson. It is a series of lessons with the greatest for the last.

  All in all a reasonable case can be made that Holmes doesn’t regularly fall prey to problems of luck. However, if we found ourselves in his shoes and deer-hunter cap, we would be lucky (in more ways than one). We’ve seen that even in very difficult epistemic situations, Holmes somehow manages to keep his head and only form responsible beliefs. Holmes’s amazing powers of deduction and keen observational skills allow him to reliably form safe, sensitive beliefs in cases where we cannot. This, combined with the justification granted by his career as a professional consulting detective, ensures that if anyone knows what’s really going on in a case, it’s Holmes.

  Now, it’s unclear if Holmesian deduction would ever give anyone in the real world actual knowledge. None of us have access to Holmes’s famous powers. Further, even if we assume that we could concoct the sorts of reasonable explanations he comes up with, there’s no real way of knowing that there aren’t a huge number of plausible alternative theories to explain what we see. For now, the job of consulting detective must be relegated to the realm of fiction. While the character Sherlock Holmes might be able to pull off amazing feats of deduction, it’s only because we’re willing to accept as fact that there really are no nearby possible worlds that would spoil the safety or sensitivity of his beliefs.

  It might be the case that Holmes does benefit from occasional lucky strikes (after all, even the best occasionally get lucky). However, when Holmes exercises his due diligence, eliminates all the impossible alternatives, and utilizes his formidable prowess as a detective, he doesn’t arrive at the answer merely by chance. It’s skill, not an accident of fate. This is why we can ri
ghtly say that he’s not just the luckiest detective in fiction: he really is that good.

  Chapter 4

  The Adventure of the Candle and the Dumbbell

  Fiona Tomkinson

  Of course, it was Holmes who, to alleviate an evening’s tedium, first introduced the question. It is possible, though not certain, that the whole affair arose out of one of the solutions to a crossword puzzle which could have been either “object” or “things.”

  At first, Watson expressed bewilderment at the possibility that the terms could be anything other than synonymous. Then he began to think that, though people tend to use the words interchangeably, perhaps a rigorous scientific man such as himself should not. But where does the difference lie? He began to advance tentative theories.

  Is a thing natural and an object manufactured? Is thing a popular term and object a scientific one? Is it a question of value? Is a beryl coronet or a golden pince-nez an object, but a missing three-quarter merely a thing? Yet we always talk of the “object of enquiry”, never the “thing of enquiry,” even when that object is insignificant!