Page archived courtesy of the Geocities Archive Project
Please help us spread the word by liking or sharing the Facebook link below :-)


Chapter 1


My name is Sherlock Holmes. It is my business to know what other people don't know.

—— The Adventure of the Blue Carbuncle

I propose to devote my declining years to the composition of a textbook which shall focus the whole art of detection into one volume.

—— The Adventure of the Abbey Grange

What does the art of detection aim at?

What does Sherlock Holmes do?

Why do people call in detectives?

These are, all of them, one and the same question. If we can answer one of them we can answer all of them. And it is easy to answer at least one of them. We all know that we call in detectives when there are things we want to know, which nobody would tell us. Take one of Sherlock Holme's cases for example. In The Naval Treaty an important document had been lost——a secret treaty between Italy and Britain. The treaty was left on a desk unattended only for a short time. It was impossible for anyone to have come into the premises and left without being noticed. Yet it was gone; it had vanished; no trace of it could be found. It had disappeared so completely that there was not even a word of it in that shadowy world populated by spies, in which secret documents were regularly bought and sold. As usual Scotland Yard was at its wit's end, so Sherlock Holmes was called. Through his exertion the thief was caught and the document recovered.

What happened in this instance? How was the treaty stolen? Who stole it? Why would any one steal a secret treaty and not sell it or do anything with it? These are questions for which we would like answers. Now someone knew the answers to these questions from the very start, even before Sherlock Holmes was brought into the case. That someone was the thief. But thieves have a habit of keeping their thievery to themselves, the reason why we need detectives. What does Sherlock Holmes the detective do? As he himself tells us, his business 'is to know what other people don't know.' By 'other people' it is understood that they do not include the thieves themselves. The thieves know, but unfortunately for us they do not tell; they hide what they know. But this does not mean Sherlock Holmes cannot find out! Things that have happened can be known. They can be known without ourselves being there. They can be known without ourselves being told by those who were there. We could be completely absent from the scene; yet we could still find out. This some people find hard to believe, but those familiar with Sherlock Holmes know it can be done. In fact, there is an art for doing it. We call this art the art of detection.

We exercise the art of detection not only in crime investigations. We can exercise it, or at least attempt to exercise it, whenever there are things to be known to which we are not directly present. What was Watson doing that morning when he was out by himself? Sherlock Holmes can tell without Watson first informing him. How does he know? In the same way that he knows who stole the naval treaty, that is, by exercising the art of detection.

Sherlock Holmes hunts down criminals; scientists hunt down the causes of diseases. How do scientists do it? They do it the same way Sherlock Holmes would do it, by exercising the art of detection.1

Because of the popularity of detective stories most of us nowadays know something about the art of detection. Indeed, when we are reading these stories we often pitch ourselves against the detectives themselves, to see if we can't beat them at their game. The simple premise here is, if they can do it, so can we.

Among detectives Sherlock Holmes is universally acknowledged to be the best. He is the world’s only consulting detective, the one who takes on the most difficult cases, cases that other detectives are unable to solve. But while the cases Sherlock Holmes takes on are difficult, when they are solved Sherlock Holmes always has an explanation. He can explain why things must have happened as he has described, or if not completely as he has described then at least 'in essentials'. This is to say, when a case has been solved Sherlock Holmes is always able to take us rationally from a state of ignorance to a state of knowledge where that particular case is concerned. Detection is an art but this does not mean reason plays no role. That something is an art does not necessarily mean there is no place in it for reason. Discovering the proof for a mathematical theorem is very much an art but the person who makes it cannot avoid a lot of reasoning.

Sherlock Holmes was planning on writing a textbook on the art of detection when he retired. For many years it was assumed he never got around to it. Things have changed since 1991. In that year, in the farm to which Sherlock Holmes retired, a manuscript was discovered under a false ceiling. It contains a Preface which begins this way:

Having retired, I now have the leisure to write my long-promised magnum opus, in which I pull together into a manual the various principles I have derived during a long career as a consulting detective.

So it seems Sherlock Holmes did write his textbook after all, even though it was put in the form of a manual. Professor Gardiner, who made this discovery, has given to this manual the title, The Whole Art of Detection (for the history of the discovery and the text itself see

This book I am writing in these pages is not the textbook / manual Sherlock Holmes wrote, which Professor Gardiner discovered. Mine has a more humble name. Conscious that I am a lesser mortal than the Great Detective, I am calling it The Art of Detection, hoping that readers will understand thereby that I lay no claim to knowing the whole subject.

In respect to form, The Art of Detection is different from The Whole Art of Detection. The Art of Detection, the book I am writing here, is not a manual, but more like a regular textbook. Do not get me wrong, I have no objection against manuals. Indeed, I treasure them; manuals serve important functions; because of their compactness, they are useful for quick reference, besides being good for other purposes. The regular sort of textbooks are useful too. Because they are longer, in writing them one can provide the kind of explanations beginners usually look for and need. This book I am writing is a textbook of this sort. For reasons that will become clearer presently I am interested in providing the kind of explanations that beginners look for and need.

To me a textbook on the art of detection has one very important explanation to provide. It has to explain in a general way why in certain situations we can go from a state of ignorance to a state of knowledge when both the ignorance and the knowledge are about the same things, that is, things which are hidden from us, things to which we are not directly present. Why is it that at one point we are ignorant of these things and then at a later point we have knowledge of them (even if that knowledge is to some degree imperfect, for example, missing in some details)? If we are ignorant at first, why do we not remain equally ignorant later on? Indeed, why do we not remain ignorant for ever? Notice that when Sherlock Holmes succeeds in a case he will have found out what he wants to know before the criminals confess. Indeed, that he is able to find out so much on his own, unassisted by those who already know, is often the reason why criminals confess. When he knows so much already what point is there in keeping the rest from him?

That Sherlock Holmes can find out, therefore, cannot be because the things he wants to know, while at one time hidden, are hidden no longer because those responsible for these things, those who have direct knowledge of them, have passed on to him, Sherlock Holmes, what they know. Even less can it be because Sherlock Holmes was able to transport himself back in time and make himself present to the misdeed. Sherlock Holmes can no more travel back in time than any of us. When he is called in, the event that we call the crime is hidden from him by time: it is past. At the time the crime was committed Sherlock Holmes was not there to witness it; he was somewhere else. At that time the event was hidden from him by space. Now none of this is altered after Sherlock Holmes has solved the mystery. By the time he has solved the mystery, the event is still in the past, at a point in space at which Sherlock Holmes was not present. Without altering the hiddeness of the event; without breaking the bounds of space and time; Sherlock Holmes is nevertheless able to tell us what happened. How can this be? How can Sherlock Holmes find out things that are hidden while these things remain hidden? Why indeed can he be so certain about them? When a thing was hidden but is no longer, it is not surprising that we can change from a state of ignorance to a state of knowledge about this particular thing. If I were to hide an object in my hand by wrapping my fingers around it, you will not know what it is because you cannot see it. But the moment I open my hand, the moment I unclasp my fingers, you will know. But this is not what Sherlock Holmes does. He does not ask that you open your hand so that he can tell you what it is that you are hiding. As it were, he lets you keep your fingers closed and then proceeds to describe to you and any one else who cares to listen the object in your hand. Now he will not be completely accurate, but he will be close, so close that most people will be prompted to exclaim, 'How does he do it?!'

Why can Sherlock Holmes do what he does? Why can anyone at all? What general explanation can we provide for this kind of accomplishment? What general explanation can we provide that calls for nothing supernatural?

If Sherlock Holmes had supernatural powers we would not have been surprised at his ability. People with supernatural powers are able to do all kinds of things ordinary mortals cannot do. But we know Sherlock Holmes does not have supernatural powers. He may have been able to do certain things better than most of us, but he is still mortal. How can mortal beings with their limited powers come to know things that are hidden? We have to remember, although Sherlock Holmes is the best, there are many others who can do the same kind of thing that he does. Indeed, to different degrees we can all do the same kind of thing that Sherlock Holmes does. We all play Sherlock Holmes at times.

The textbook that I want to write will have to explain why under certain circumstances human beings by themselves can come to know things that are hidden while these things remain hidden; that is, why they can know even when they have no direct access to these things and those who have would not talk.

Of course, this question, while difficult, must be answerable. After all, Sherlock Holmes can explain to us how he solves each particular case. Now if there is an explanation for each particular case, there must be (there is likely to be) a general explanation why this kind of thing can be done at all, that is, why under certain circumstances the hidden can be known while remaining hidden.

And in such an explanation it is to be expected that the concept of a clue will have to play an important role. For Sherlock Holmes solves crimes by following clues. All detectives do. We all know that in solving crimes, if we cannot find any clues we cannot get anywhere. In looking for things hidden while they remain hidden clues play a crucial role. It is by following clues that we come to know these things. A textbook on the art of detection, therefore, should tell us how we should follow clues. Following clues, we should not assume, is necessarily easy. There is no reason for thinking that whenever we follow clues we will do so correctly. A textbook on detection should tell us explicitly how we should follow clues.

In practice when we follow clues we often rely on intuition. But intuition is fallible. This is why we have textbooks. Textbooks in making things explicit explain to us when our intuition is correct and when not.

And this textbook on the art of detection should explain why by following clues we can under certain circumstances uncover knowledge of the hidden while the hidden remains hidden. This explanation should be the centrepiece in such a textbook. For that we can uncover knowledge of the hidden while the hidden remains hidden is counterintuitive. Indeed, we often equate the hidden with the unknowable, and limit the knowable to the not hidden. Even as well-respected a philosopher as Emmanuel Kant would say, things-in-themselves cannot be known because we can never experience them. Things-in-themselves are hidden. What is hidden cannot be known. This is what intuition tells us.

In ordinary speech we often say we see when we mean we know. But Sherlock Holmes knows even when he does not see. He sees clues; he does not see, literally, what the clues point to. But he knows eventually what the clues point to; he knows without being told by those who already know. In a textbook on the art of detection we should be able to find an explanation how this is possible.

To uncover the hidden while it remains hidden we have to follow clues. Before we can follow clues we first have to find them. How do we find clues? How do we recognise them? How do we know that this thing we have found is a clue? These are also questions that a textbook on the art of detection will have to answer.

In looking for clues we sometimes make mistakes. We think the ash tray is a clue; later on we find out it is not. How is this done? How do we correct mistakes when following clues? How do we correct our own mistakes? Now without doubt this is what we do. In following clues we correct our own mistakes. For, who else is there to do it for us? When Sherlock Holmes makes mistakes, and he sometimes does——for example, in The Man with the Twisted Lip, whom does he ask? Does he say to the criminals, 'Please point me back to the right track.'? No; he has no reason to expect the criminals will co-operate. When Sherlock Holmes makes mistakes he corrects them himself.

How do we correct mistakes by ourselves when looking for things hidden while these things remain hidden? If a thing is not hidden, it is easy to correct mistakes. I think the tower is round. On closer approach I discover it is hexagonal. I have made a mistake but now I can correct it. But when we are looking for things hidden, we are looking because we do not know what they are. But if we do not know what they are, how can we correct mistakes? If I do not know that the tower is hexagonal, how would I know that I have made a mistake when I say it is round. And if I don't even know that I have made a mistake, how can I correct it? What shape should I say it is instead of round?

How to correct mistakes when following clues seems like a difficult question, but it is one that a textbook on the art of detection also has to answer.

In recent years more and more people have come to realise that in science we correct our own mistakes too, and some philosophers have asked how this can be done without God's help.2 Now scientists are detectives: they also follow clues; some of them (forensic scientists) even participate in investigating crimes. Now if we know how we correct mistakes when looking for things hidden, we know why we can correct mistakes in science without God's help.

This book I am writing is a textbook on the art of detection. It will explain why under certain circumstances things hidden can be known while remaining hidden. It will point out to us how we should follow clues, that is, what those things are which are right for us to do, and what those things are which are wrong. And it will tell us how to find clues and how to determine that something we have found is, or is not, a clue.

Since I am not Sherlock Holmes; since I am not as talented as he; this book has not been easy to write. It took me some time to satisfy myself why under certain circumstances things hidden can be known while remaining hidden. And even after I had done this I had great difficulty in arranging what I had understood in a systematic way. Somehow the various bits simply would not fall into place. But I am happy to report, by this time, to a large extent this difficulty has been overcome as a result of a clearer understanding on my part that the art of detection is not a form of reasoning, but distinct from it. This will shock a lot of people, so I should explain.

Most people expect the art of detection to be a form of reasoning. It is not deductive reasoning, they say; it is not the kind of reasoning that we meet with ordinarily, say, in the syllogism; but something similar. Sometimes they point out the difference by saying deductive reasoning is reasoning forwards whereas in the art of detection, we are reasoning backwards. They are both forms of reasoning; it is just that they move in opposite directions.

What evidence is there to support the view that the art of detection is a form of reasoning?

It would seem there is a great deal. Watson calls Sherlock Holmes the world's greatest reasoning machine. Would he do this if he did not know that the art of detection is a form of reasoning? Look at all the difficult cases Sherlock Holmes has solved. Did he not solve them by reasoning? And what does Sherlock Holmes do after he has solved every case? Does he not give you excellent reasons as to why the case can be solved? From the very important role reason and reasoning play in crime detection it seems it has to be the case that the art of detection is a form of reasoning.

Moreover, people will sometimes tell us, reason itself tells us that the art of detection has to be a form of reasoning. The art of detection leads to knowledge of things hidden. Now there are two ways to know: either we know directly or by inference. Obviously, things hidden cannot be known directly. Therefore they can only be known by inference. And this is what we do when we practise the art of detection: we reason, so that we can arrive at knowledge of things hidden. If we don't reason, how else can we arrive at this kind of knowledge?

The case that the art of detection is a form of reasoning seems very strong. The facts seem to point to it and reason itself seems to compel it. When there is such a strong case, what can I, who is on the opposite side, say?

To make my case that the art of detection is not a form of reasoning, I first plead for patience: a new idea requires time if it is to be examined carefully. Second, let me proceed in the following manner.

Suppose I reason in the following way.

We know what a person looks like either by having met the person or by reasoning.

I know what Osama looks like without having met him.

Therefore I must have known what Osama looks like by reasoning.

Now if I did reason this way readers will quickly point out I have made a serious mistake. Even if reasoning is one way by which we can know what a person looks like, it will still not follow that I know what Osama looks like by reasoning. There is at least one other way by which we can come to know what a person looks like. Because of photography, we do not have to have met Osama, and we do not have to reason; we can simply look at a picture of Osama. (After saying all this they will then ask me, how do we know what a person looks like by reasoning? How can reasoning tell us what a person looks like? I beg not to have to answer this question.)

But suppose in reply I say, 'But photography is a form of reasoning!'

Now if I did say this, readers will start scratching their heads. How can photography be a form of reasoning? Most obviously, it is not.

I have to agree with my readers that photography is not a form of reasoning; I spoke too soon when I said it was; I was not thinking clearly. But, having recognised my mistake, I now want to raise the question, why is photography not a form of reasoning? How can you point out to me that it is not?

In the case of photography it is not difficult to point out why it is not a form of reasoning. One simple way is to say, to take pictures we need light; we can't take pictures in a dark room. But reasoning we can do anywhere, even in a dark room.

Everybody can see that photography is not a form of reasoning, but if one needs a reason, a reminder, one can be quickly found.

Now I say the art of detection is not a form of reasoning, but in this case I cannot say it is obvious that it is not, since whether it is, is precisely the matter in contention. What reason can I provide to show that the art of detection is not a form of reasoning?

For photography to be possible, we need light. No light, no photography. To practise the art of detection, what do we need?

If we think a little about it, we find there is an answer, an answer that will show us that the art of detection is not a form of reasoning. The art of detection needs … not light, but clues! If there were no clues, even if you were Sherlock Holmes you would not be able to find anything. In the field of crime detection, everybody knows the importance of clues. No clues, no solution.

But clues are important not just in crime detection. Whenever we are interested in knowledge of things hidden, we need clues. Take science for example. Has anyone not heard of the apple falling on Newton's head? (The story about the apple turns out to be false: the clue was not the apple but something else; still, the story reminds us that even in science we follow clues.)

Can we reason without clues? Certainly, we can. We can reason in dark rooms and we can reason whether there are clues or not. In reasoning we need premises from which to draw conclusions, but we do not need clues. Take a common example of a piece of reasoning.

If pigs had wings they could fly.

Pigs had wings.

Therefore they could fly.

In carrying out this piece of reasoning no one will be thinking of, or looking for, clues.

No light, no photography. No clues, no detection. No premises, no reasoning. Photography, detection, reasoning——these are three separate and distinct arts. The art of detection is no more a form of reasoning than photography is.

But, readers will ask, what about the argument employing the premise, 'Either we know directly or by inference.'? Does it not lead to the conclusion that the art of detection is a form of reasoning? Obviously, what is hidden cannot be known directly!

We can now see that this argument is fallacious, as fallacious as the argument about knowing what Osama looks like by reasoning. In the case of Osama we pointed out that it is not true there are only two ways to know what a person looks like; there is a third way: photography. So it is not true that we know what a person looks like either by having met the person or by reasoning. Similarly it is not true that we know either directly or by inference. There is a third way: we can know by employing the art of detection. The art of detection is not a form of reasoning. If the art of detection were a form of reasoning, there would have been just two ways. But since it is not; since it is distinct from reasoning; there is a genuine third.

What about the facts then? Does Sherlock Holmes not reason? Is he not the greatest reasoning machine in the world?

Certainly Sherlock Holmes reasons. We have to reason when practising the art of detection. But when we do this we are practising one art within another. It is common to practice one art within another. I am writing a book; I am practising the art of book-writing (at which, I am afraid, I am not very good). In doing this, I am also practising the typing art (ditto). I am practising the typing art within the art of book-writing. That I practise the typing art when writing a book does not make the art of writing a book into a form of typing. Typing schools do not teach writing skills.3 We have to reason when practising the art of detection, but this does not make the art of detection into a form of reasoning.

Neither reason nor the facts compel us to take the view that the art of detection is a form of reasoning. The art of detection and the art of reasoning are distinct one from the other. To think that the art of detection is a branch of the art of reasoning is a mistake. In this regard, I should point out a difference between I myself and Sherlock Holmes. Sherlock Holmes, I should caution readers, shares in the common but mistaken belief that the art of detection is a form of reasoning. He calls what he is good at, reasoning backwards. Deductive reasoning is reasoning forwards. The art of detection in this mistaken view goes in the opposite direction (as I have alluded to earlier).

In solving a problem of this sort, the grand thing is to be able to reason backwards. That is a very useful accomplishment, and a very easy one, but people do not practise it much. In the everyday affairs of life it is more useful to reason forward, and so the other comes to be neglected. There are fifty who can reason synthetically for one who can reason analytically.

—— A Study in Scarlet

I think Sherlock Holmes has made a mistake here. He solves his kind of problems not by reasoning backwards, but by practising the art of detection, which is not a form of reasoning.

Am I being unkind to Sherlock Holmes in saying that he has made a mistake? Does Sherlock Holmes ever make mistakes?

Those familiar with Sherlock Holmes will know that Sherlock Holmes is quite prepared to admit that he makes mistakes. Indeed, he thinks it is important that the public should know that he is not immune from mistakes. On at least one occasion, he openly chastises Watson for not informing the public of his mistakes more often.

I made a blunder, my dear Watson——which is, I am afraid, a more common occurrence than anyone would think who only knew me through your memoirs.

—— Silver Blaze

Why did I have such a hard time arranging what I have understood of the art of detection in a systematic way? I now know the reason. At one time, like Sherlock Holmes and many others, I too made the mistake of thinking that the art of detection was a kind of reasoning. Worse, I could not shake off the influence of this belief even after I had recognised it to be a mistake. I had a hard time in arranging what I had understood of the art of detection in a systematic way because I did not treat the subject as really distinct from deductive reasoning even though I knew it was. Talk of putting new wine into old bottles! What I was doing was like treating photography as though it were a branch of logic. No wonder things would not fall into place.

Because the art of detection is distinct from reasoning and not akin to it, I now find it useful to remind myself from time to time that I should organise what I have understood about this art in the way that this art itself demands, and that in this process I should not be distracted by pre-occupations those have who think of it as a form of reasoning akin to deductive reasoning. The art of detection is distinct from reasoning. In thinking about it, in talking about it, in writing about it, we should treat it as such. It is true that in following clues we cannot avoid reasoning, but neither can photographers.

In the following pages we will have many occasions on which we will want to speak about deductive reasoning. This cannot be avoided completely since readers will want to know about the relation between deductive reasoning and the art of detection. But when we do, readers should keep in mind that the art of detection is distinct from deductive reasoning.

Since the art of detection is distinct from reasoning; since it is not customary to treat the subject in this way; it has occurred to me that a good way to present the material I have accumulated on this subject is to do it in the form of a textbook. Textbooks are written for students, for people who are new to the subject. In a textbook, one starts ‘from the beginning’, and presents the material in such a way that people new to the subject will find it easy to understand. This I want to do. I want to start from the beginning so that we can all see clearly what kind of creature the art of detection is.

In saying that I will start from the beginning, I do not mean I will assume nothing at all from my readers. Certainly I do not mean we should all go back to the Garden of Eden, to the time of Adam and Eve. By 'starting from the beginning' I merely mean I will assume no more than what I think I can reasonably assume from a general audience interested in the art of detection at the present time (that is, at the beginning of the 21st century or thereabouts). For example, I will assume that such an audience will have heard of Sherlock Holmes and may even know something about him.

A moment ago, I said textbooks are written for students. Perhaps I should have said textbooks are written mainly for students. For textbooks sometimes can be useful even to those who are experienced in the subject in a practical way. Experienced people can sometimes find themselves in a quandary: they can find themselves in a situation different from what they have been used to, a situation in which they are puzzled as to what they should do or think. But when their practical experience is backed up by the kind of systematic understanding available from a textbook, the number of this kind of situations can be reduced. A textbook has to be systematic to be useful to students: systematicity facilitates understanding. But a systematic understanding is useful not only to students; it can be useful also to people who have practical experience. A systematic understanding often allows us to extend what we already know from practice to situations which we have not met with in practice in the past. Experience is precious; we benefit from it; but we can benefit from it even more when it is combined with systematic understanding.

I should not omit mentioning one other important function a textbook can sometimes perform. By treating a subject in a systematic way a textbook can sometimes lead to new discoveries, and thus to an expansion and enlargement of the subject. When we deal with a subject in a systematic way, we sometimes notice gaps and openings, gaps and openings which we can then profitably explore. Remember how fruitful the periodic table in chemistry has been in the discovery of new elements.

Most of us are not detectives or scientists by profession, but most of us will meet with situations in which we have to evaluate results produced by detectives and scientists. If we know more about the art of detection it will only help us in this kind of situation.

Theseus went into the Labyrinth to slaughter the Minotaur. To lead himself out of the Labyrinth he followed a ball of thread, which he had unwound when going in. Sherlock Holmes finds clues and follows them to uncover things hidden. Following clues is often compared to following a thread.4 In view of this I am coining a new word, ‘theseology’, to stand for ‘the art of detection’, that is, the art of finding and following clues, an art distinct from the art of reasoning. Theseology leads us towards knowledge——and not just any kind of knowledge, but knowledge of the hidden while the hidden remains hidden. The process in which we practice this art I call the theseological process, and those who practice this art, theseologisers. Sherlock Holmes is a theseologiser, the greatest the world has ever known.

There is one minor point I should like to make before I conclude this chapter. I have repeated many times, and I will repeat many times more, that the art of detection is not a form of reasoning. In doing this, it should be clearly understood, I am not implying that the art of detection is somehow superior to reasoning, nor that it can replace reasoning. We need the art of reasoning: in the theseological process we have to reason; we practice one art within another. The art of reasoning cannot be replaced by the art of detection. The art of detection can no more replace the art of reasoning than it can replace the art of photography. Different arts serve different purposes (and some arts are ends in themselves). Because we have many things to do, we need many arts. It would be foolish for anyone to single out one of these, declare it 'superior', and discard the rest.


1Sherlock Holmes is a scientist himself. When he and Watson first met, he had just discovered a highly sensitive test for blood stains (see A Study in Scarlet).

2Many philosophers have argued that because we cannot see things from God's point of view, we cannot know anything, including what is a mistake and what is not.

3Some logic textbooks contain a section called Scientific Method. Sometimes they would even illustrate scientific method by reference to Sherlock Holmes. Significantly logic textbooks do not teach us how to find clues and how to follow them.

4In Chinese the two characters that stands for the word clue means the thread that allows us to find what we seek.