The Figure Of the Detective

Rationality and the Detective Artist

With Arthur Conan Doyle's Sherlock Holmes all of the elements of the Detective (and with him the detective story genre) come together in a single character: Sherlock Holmes, the echt-Detective.

We first encountered Holmes in 1887, in Dr Watson's account A Study In Scarlet.¹ In 1887 the situation was this: writers such as Dickens, Dostoevsky, and Collins had given the crime novel an expressive power and range of character and incident which greatly enlarged its literary possibilities. Poe had suggested a set of principles around which a formula could coalesce and Braddon and Trollope had demonstrated the narrative interest which might be attached to the victim equally as well as to the detective or the perpetrator. Green and Zangwill had pared the formula down to its essentials by tight plotting and a focus on the crime and its solution by emphasizing the dramatic tension between the puzzle and the infallible detective.

All this Doyle inherited. What he did himself was to bring these elements together in a single character and a unified and fully extensible mythology.² This would be an accomplishment, but Doyle was also able to tap cultural ambivalences and antagonisms which he built into Holmes's character, giving him a depth and lifelikeness which is the source of his enduring attraction.³

We first encounter Holmes beating a corpse. This tells up everything we need to know about what is to ensue. The peculiar behavior is explained (he wishes to know how far injury can be sustained after death) but the explanation itself includes an empirical engagement with the sensual world and the desire for evidence does not preclude but may even require curiosity and pleasure. Holmes is not a Dupin, a thinking machine. We notice also the very name of the tale in which Holmes makes his appearance. The study is a term of art, and scarlet is an artist's color loaded with emotional significance and not to be confused with the layman's color red. Blood is red. Scarlet is something else. In this first incident we see the detective artist at work, and encounter the essential quality which is the making of Sherlock Holmes.

Let us return for a moment to the plot summary given in the discussion of the pre-Classical detective (a category within which Holmes must be included) of Conrad's The Secret Agent. The contemporary fear of “bomb-throwing anarchists” exposed the connection between what we most need, which is safety, with what we must acknowledge, the existence of the unseen, the primal, the arbitrary, the ineffable and mysterious. By situating his tale within Verloc's condition Conrad does not expose us to an existential threat to our safety. This is why The Secret Agent is a crime novel but not in the detective genre. It lacks the frisson of the Holmes formula which proceeds by first scaring us with the possibility that rationality is no defense, that there is no complete defense, against an irruption of the unseen world. The artist acknowledges this and confronts it, demonstrating thereby a different strategy; Holmes, as a detective-artist, embodies both these strategies.

Consider what Holmes actually does when he solves a case. The popular view, one supported by Holmes's own claims, is that he is proceeding by deduction. In fact, as C.S. Peirce shows, this process is actually abduction.⁴ Abduction is a probabilistic method to be used when one of the facts to be reasoned with is uncertain.

Abduction

Abduction is a noun, but unlike the other two terms has no verb form. That is, one cannot “abduce” but only engage in or perform an abduction. Note that the verb forms ‘abduct’, ‘deduct’, and ‘induct’ have entirely different meanings and fields of application. To deduct is to take away from, as in accounting and finance, whereas to induct is to add to, in the sense of taking in a new member of a group. (‘Induce’ also has a sense irrelevant to logic, meaning to start up, as in to induce labor in a pregnant woman.) To abduct, from the Latin abducere, meaning to lead, is to kidnap or carry off by force, or in physiology, to draw a bone or muscle away from an adjacent part (as a surgeon would do to expose an area to be operated on). There is, however, a metaphorical relationship between ‘abduct’ and ‘abduction’ as will become clear.

The pronunciation and spelling of abduction are straightforward. It being a technical term there are no synonyms, nor antonyms except phrases indicating some sort of faulty thinking. Peirce used ‘retroduction’ as an alternative but this never caught on.

Broadly speaking, abduction is a method of reasoning under conditions of uncertainty. A syllogism is the simplest way to demonstrate this property. In the case of both deduction and induction both the major and minor terms are known absolutely. In the case of abduction one or both of these terms is only probably true. Since statistical (probabilistic) thinking is at the root of most of modern science and many other contemporary activities, abduction may now be considered the dominant mode of reasoning.

An exact grasp of the concept of abduction as a mode of reasoning used by fictional detectives requires first of all the understanding that detecting and finding out are not the same thing. One may find something out by any means; to detect indicates that what is sought has been discovered or revealed by thinking about it — that is, by rational means. Holmes’s typical procedure is to observe and record data, which ‘facts’ he then associates together or classifies. (Strictly speaking these data are not facts until they have been proved to be empirically true, but we may regard this as a quibble.) Holmes then advances a hypothesis which accounts for the facts in order to identify the possible causes for the observed state of affairs.

Deduction reasons from observations (postulated hypo-theses) to their necessary consequences. Given the facts, what are the results? Thus: “Baldness is an absence of hair, you have no hair, therefore you are bald.” Beginning with a definition and an observation, we conclude that the observable is (or is not) an instance of the general definition. This is the classificatory procedure of taxonomy.

Notice that this is not the same as: Men are bald, you are bald, therefore you are a man. The truth of this conclusion depends on the correctness of the initial generalization that men are bald. This may be true or false. There is no amount of evidence which can prove it true, whereas one observation of a hairy man will prove it false. Karl Popper’s contribution to our understanding of how science works is this concept of falsifiability. Science cannot prove anything to be true, only false. (This is the bone of contention in the cultural argument over evolution, that, being a theory, evolution is only probably true. To someone who accepts only absolute truth, probable things are not true.)

Induction reasons in the other direction, from the observed consequences to their necessary causes. Given these results, what are the facts? Given this state of affairs, what observable facts could have produced such a result? Notice that the facts we require must be observable. This is the problem with séances and divine intervention. Such things may be, but they are not observable, and so cannot be used in reasoning. (There is a dispute about observable, unless you regard an experience as also communicable; there is also now a dispute about the possibility of communication which is one of the legs on which postmodernism stands.) Detectives are always debunking supernatural explanations. Agatha Christie’s stories contain many séances and other occult things because these were such a popular fad at the time but the occult is fatal to rational detection. G.K. Chesterton’s Father Brown often invokes divine intervention, but that always has material means and consequences which can be reasoned about. Thus: “I have met a lot of bald persons, all of these were men, therefore baldness is probably a male attribute.” Notice that this sort of statement is not falsifiable either by encountering a bald woman or a hairy man. The truth of its conclusion rests on an agreement about what frequency is required to constitute an attribute. Induction is only contingently true. Holmes reasons this way when he decides that ash of a certain appearance is from latakia tobacco because he has seen a great many cases where this was so and concludes that the probabilities are in his favor. Note that the more observed characteristics contribute to the conclusion the greater its probability.

Or: “I have met a lot of bald persons, all of those were men, therefore all men are bald.” In this case also the conclusion is probable rather than determined, but here the probability is introduced by the reasoning process itself. General laws are derived from observations in this way, before the proposed law undergoes falsification tests. But just because the proposed law is only probable does not invalidate its immediate usefulness. Such reasoning would not be accepted in court (in detective stories anyway) so that wherever the guilty person must be induced (!) to confess, this sort of reasoning is implicated. The outcome of many stories is stymied by the refusal of the criminal to confess. In these stories justice is usually dispensed in another way.⁵

Abduction, then, or reasoning when one or more terms of the argument is uncertain, is inherently probabilistic.

The second example above, correctly phrased as “Some men are bald, you are bald, therefore you may be a man” is one type of reasoning by abduction. The statement is true as phrased, but requires additional evidence to rescue it from triviality. Abductive statements of this type then become one term in a new chain of reasoning. Holmes reasons this way when he identifies a certain ash as deriving from (say) latakia tobacco. (Notice the implied prior chain which made it probable that the ash was tobacco in the first place.) This initiates a new chain of reasoning as to the significance of latakia specifically, and so on until the criminal is identified. When he is caught and his guilt is conceded then the whole chain of reasoning from the ash to the confession is validated.

If we rephrase the case of the bald man to read “I have met a lot of bald people, I suppose most of these people were men, therefore it is likely that baldness is a male attribute” it may be impossible to say whether a particular bald person is in fact a man. In this case we turn to statistics and argue (probabilistically) on the basis of such standards as ‘margin of error’ and ‘degree of confidence’ and ‘representative sample’ and various tests of the data such as measures of correlation or frequency distribution. I would guess (!) that hardly anyone understands this or can be brought to count as true anything demonstrated only statistically. This is the basis for a major cultural division.

“Pure” abduction occurs when both of the terms are only probable. Action under conditions of uncertainty may be the contemporary challenge, because we are uncertain about so much and can so seldom afford to duck the question. The ability to do this, the self-confidence he has in his powers, and the courage he shows in acting on his conclusions are important components of what makes Holmes fascinating, an adventurer, and a hero. He guesses.

What makes Holmes successful is not that he never guesses but that he guesses so well. Abduction is the first step in scientific reasoning and a guess (a ‘prior’ in Bayesian terms) is the first step in abduction. One wants to begin with the best hypothesis — that which is the simplest and most natural, the easiest and cheapest to test, and will contribute to our understanding of the widest range of facts. This hypothesis, then, is the most probable. It may not be correct. One obtains proof of it’s correctness by experiment — by seeing if it leads to the murderer. Holmes’s advantage over Watson (and the police) is first of all that he is so much more imaginative — that he can see so many more options and combinations to pick from — whereas the police often fail not only because they are unimaginative (dumb, even) but because their priors do not account for all the facts, or because they begin with a prior and then look for the facts which would validate it (misplaced imagination). When Holmes praises a rival (seldom) it is always for careful observation or (even more seldom) for good reasoning. He never accepts anyone as his equal in imagination. For that he turns to music.

Holmes as the Rational Man

Holmes is the epitome in our minds of the Rational, and of the beneficial results of rational behavior. As we have already noted, Holmes when first encountered is engaged in beating a corpse, recalling the folk unwisdom of beating a dead horse – that is, a pointless and stupid, very irrational behavior. Yet it turns out that appearances have deceived Watson, and thus us. It is we who have not been rational in our thinking. We have made the mistake, in Holmes's famous remark, of hypothesizing in advance of the evidence. Could we have any clearer demonstration of what we are to expect?

And yet, as we shall see, Holmes is not, or not entirely, that rational man which we believe him to be, that shibboleth of penetrating observation and thought which we have made of him. Herein lies one aspect of his complexity, his influence, of the humanity which has helped him to endure.

We have, or sometimes have, the idea that thought and feeling are incompatible. Metaphorically the one is cold and the other warm; they are immiscible humours and people are one or the other, either coldly rational or warm, emotional, impulsive.

Where does this idea come from? I am not asking if it is true. Clearly it is not true. Still, how does it hapen that (cold) thought, planning, intelligence can be characterized as inimical or undesirable? Is that what we really think, when we’re not thinking about it especially?

The notion resembles a prejudice which does not survive face-to-face experience but somehow stubbornly continues to intrude. Even people who make their living by rational inquiry — scientists are the most visible — sometimes let this notion pass. And of course, some will find it agreeable and may even have chosen their life’s work because of it, finding that warmth and emotional impulsiveness are not to their liking.

Rationality is, it is true, usually thought of as a good thing. This good is secured by objectivity, the main requirement. Science is objective, and hence rational, because it is supposed to yield knowledge that is disinterested, impersonal. But here the trouble starts, for impersonal has less happy connotations. The practice of science is usually said to require rigor, which has an ascetic note, or perhaps even austere rigor.⁶ And so, ever tightening the screws of purity, we come to the plight of Sinclair Lewis’s Martin Arrowsmith.

“Social life of every kind, even the social life inside the laboratory, is for Arrowsmith at worst a temptation to fudge his results, at best a distraction from the serious business of science. The novel leaves the hero in the woods of New England, withdrawn from all company, pursuing his laboratory research in splendid isolation. In this way he was protected from the temptations of power and reputation, dedicating his life to the relentless pursuit of stony truth and ignoring the social graces represented by soft and deceitful women.”⁷

This is an excess of zeal, surely. A professional hazard, but avoidable. We are after a stronger claim, that rationality is in itself cold, and will freeze anyone who dabbles with it. In this view, knowledge and truth are inimical to life. Those who traffic in knowledge (scientist, shaman, the scribe who manipulates the king when everyone else is illiterate) cut themselves off from humanity, human warmth and community. These people are not to be trusted, which our experience confirms. The man who introduces Syme to the anarchist society in which he takes the role of Thursday gives as his reasoning that men are needed “whose fears for humanity [are] concerned rather with the aberrations of the scientific intellect than with the normal and excusable, though excessive, outbreaks of the human will.”⁸ Of course, Chesterton is speaking here as a partisan anxious to defend faith, Catholicism in particular. The defense is necessary because the industrial and scientific revolutions of the preceding generations had gathered all the benefits of rationality to forces inimical to faith, leaving religion with no mode (it was felt) other than to accept the irrational and oppose a culture of belief to an increasingly powerful culture of evidence. The confrontation is put more bluntly later in the book. “[T]he tyrranic fear of the Professor had been the fear of the tyrranic accidents of nightmare, and … the fear of the Doctor had been the fear of the airless vacuum of science. The first was the old fear that any miracle might happen, the second the more hopeless modern fear that no miracle can ever happen”⁹ Today it is hard to credit that such claims were ever taken seriously by serious people. It seems cartoonish. Yet we have only to remember the division of the sexes between rational man and intuitive woman¹⁰ which is still a live idea with us, as seen in the many squashed efforts of female scientists, to appreciate the force of this prejudice and the fear which drives it.¹¹

Marjorie Nicolson, in another foundational commentary, “The Professor and the Detective,”¹² attempts to pick apart the appeal of the detective story to educated readers who it would seem have thought it frivolous. Nicolson was the Professor of the piece (she was a dean at Smith College), one of the tribe of intellectual workers among whom the detective story is popular because of the visceral reliefs it offers. (A little too visceral. The university librarian, we learn, lays in a stock of detective novels kept sequestered from the students, who are yet too corruptible to be given a squint.) Chess players, bridge players and crossword puzzlers, physicists, astronomers and mathematicians – students of the new science all – stay up late to read them under the covers because the objectivity and impersonality of the form, its lack of sentimentality and love interests¹³ all recommend it as relaxation. And it is these same rational qualities, she says, which lead the popular mind to associate detective literature with Mr. Einstein’s destruction of the moral standard.¹⁴ Rational and exotic at once, detachedly objective yet thrilling, a literature which deliberately forgoes nobility denigrated by its middlebrow readers (who do not play chess) as, well, being without nobility. What a muddle.

Having thus divided the darkness from the light, the beneficiaries and victims of hegemony strive to tar each other. Fifty years after Chesterton and twenty after Nicolson, Michael Gilbert’s narrator speaks of one Mr Hoffman, an accountant, as “a man who hunted down facts with the passionless pleasure of a butterfly collector and pinned them to his board with the same cold precision.”¹⁵ VN, Modernism’s great impassioned butterfly collector, would certainly have jibbed at that. But both slanders are easy tropes, trollops easy to pick up.

There are two routes to this position. One passes through the public need for experts who can stave off arbitrariness through rational formalism.¹⁶ Public knowledge must be aloof from human agency to be reliable. Partisan subjectivity, that which the existentialists more politely called engagement, creates a responsibility to that with which one has engaged. Politics and management are subjective. These people are moralists. One who would be an expert, or pose as one, accepts cold impartial rationality as the consequence of public responsibility. The detective, an enforcer of truth, is one such expert.

The other route to a view of rationality as cold passes through Romanticism and the separation of mind and body. Descartes is an obvious source of this idea, but the trope is much older than that. Religions often privilege the mind, seeing our essence as being imprisoned in foul, fallible flesh. Flesh needs to be disciplined and subdued. Flesh is dangerous. Plato thought so. Gnostics think so. We are all a little gnostic – that’s why it’s a heresy.

It is easy to see that this dichotomy might be stood on its head (!) to favor instead the body and the pleasures of the body, feeling and emotion. This too is an old idea, and human history appears to cycle through an ascendance of first one and then the other. The contemporary preference for the body over the mind is only the continuation of a complex of ideas which derive from the Romantic upheaval at the end of the eighteenth century.¹⁷ Romanticism was more than the poetic sentiments of a few Germans or Britons, but a gradual reformation of western thought throughout the nineteenth century in favor of the passionate, feeling, natural human being. To think in this way means to accept also the irrational and arbitrary, the horrors of the psyche for which Freud gave us a language and which made themselves known in a string of wars and atrocities. We are perhaps beginning to suspect that there is more to our experience of two hundred years than merely feeling run amok. Unreason has revealed itself everywhere, in the mindless and accidental processes of biological evolution, in the indissoluble bond between civilization and discontent, and in the inscrutability of Sartre’s Being set off against our own nothingness, longing, and anxiety.

Into this stew comes the detective. Is it any wonder that this icon (or is it stooge?) should attract opposite fears, both extremes of the popular notions of what it means to traffic in knowledge and truth?

Of course, members of the educated elite can always put it down, put down, or excuse these things as popular shibboleths, mere folk wisdom (or unwisdom in this case). All media which appeal to the indifferently educated or civilized, such as the movies, or the industrial products of genre factories such as the detective story, are vulnerable to being dismissed as kitsch. This position,¹⁸ which may have been tenable in Leavis’s time, had to be bootlegged into the argument by Susan Sontag (in her famous essay on kitsch) and is now thought to be without merit. The more typical modern reading identifies the import of the detective story as a feel-good plot intended to reassure people that the truth is knowable, and attacks on the detective genre arise from the contention that there is no truth (or, a weaker position of epistemological agnosticism, that it is unknowable). The next move is to suggest, as Pierre Bayard does,¹⁹ that the detective story is based on willful blindness, that we are unable or unwilling to see the truth because of some limitation, inability, or inhibition in ourselves, and that the detective exposes this. In postmodern terms, the truth is constructed by the reader, deconstructed by the detective, and then (falsely) re-constructed by the author claiming to act on the reader’s behalf. And this is not just any re-construction: rather, the correct one, hopelessly tainting the genre with transcendental ideas.²⁰ So we come out at the same place, warm justice facing off against cold reason.

This face-off is built into the form, Todorov says. I am curious, then, what ordinary readers take detective stories to mean. What does the not yet deconstructed reader think knowledge is, what it is like to know things? A detective is a knowing person, an enhanced version of the ordinary reader. Perhaps (s)he knows.

A good way to satisfy this curiosity might be to study a resolutely, even radically popular art form such as the movies. Film imagery responds quickly and with subtlety to changes in popular thinking because that is what makes movies enjoyable and successful to the mass public. If we look at movies, not in themselves or to explicate the detective genre, but rather to use them to winkle out how changes in genre conventions can expose changes in popular attitudes, what might we learn? Do people really think that the mind is cold, that thought does not partake of feeling, that intellectuals are not to be trusted?

The movies offer a particular advantage. If we were to tackle detective fiction we would be obliged to confront criticism and literary theory. Alternatively, if we were to put the question to people directly we would let ourselves in for the vicissitudes of the cultural anthropologist, not knowing who to believe or how the evidence might be tainted. People, after all, will tell you anything they like. But if we should examine the imagery of some detective movies for clues, perhaps this material might (unthinkingly) betray our real feelings in an unguarded moment. If we can recognize the moment.

Who was Doyle's Holmes?

Before we turn to a study of Holmes in the movies we ought, if simply as a baseline, to establish what Doyle, Holmes's creator, thought he was doing. To begin, let's examine a simple story, Doyle’s “The Naval Treaty.”

The story is easily told. A young clerk leaves a sensitive draft treaty on his desk when he is working late, and upon returning finds that it has disappeared. Driven mad by his error and the destruction of his reputation and livelihood, he is brought home in a state of mental collapse and has to be confined to bed, where he languishes many weeks. Holmes is brought in. Attempting to work out why the valuable treaty has not surfaced in foreign hands, he concludes that it is sequestered in some place from where it cannot be retrieved, and settles on the invalid's room as a possibility. He arranges with the young man's fiancée to ostentatiously leave him alone and unguarded. Keeping his own watch, he nabs the would-be brother-in-law. It seems that this man had stopped by the office unannounced, saw the treaty lying there, and impulsively snatched it. Tucking it away in a safe place, he had found to his consternation that it was a little too safe. The treaty found, the young man recovers and the unwavering love of his betrothed is vindicated.

Although superficially threatening, the discovery that no one is attempting to profit by the theft of the treaty reduces the whereabouts of it to a worry. And we find in the end that the thief was an amateur, guided by the same impulsiveness which had also put him into financial ruin. He probaby would not have known what to do with his booty if he had kept his hands on it. Yet Holmes keeps saying it's a “dark business.” Our interest in the story is surely as much in the melodrama of the victim and his rescue, and in the contrast between Holmes and the thief – both mentally quick, both infinitely patient, and both masked. What then is it that marks Holmes out against the “darkness” of both criminal and victim? Surely it is that he alone is capable of the jeu of returning the treaty by bringing it to the breakfast table hidden under one of the covers, disguised as it were as toast and stewed tomatoes. Only Holmes is immune to the weaknesses of both the intellect and the emotions both, and indeed takes pleasure from the strength of both sides of his personality.

In this duality Doyle was working against an explicit tradition. (Not universal, of course; one might mention Collins's Mr. Bruff as another instance combining intellect and emotion.) In William Godwin’s Caleb Williams (1794) there is murder, detection, unrelenting pursuit of the murderer, and the crime story’s distinctive construction from effect to cause, solution to problem.²¹ (The book was, in fact, written back to front.) Godwin showed how the detective could be used in serious literature. Nevertheless, some elements were missing. The story is tragic and anarchistic, and it condemns law and lawful punishment, features of Godwin’s theories of politics and justice. These features are the essence of the story: the inadequacies of law are what produces the tragedy, and the novel could never be redeemed for the detective genre as we understand it.²² Law and punishment are bound, just as the motiveless crime of Raskolnikov requires the persecution of Porfiry Petrovich to expose it. (We will see Porfiry’s like again in the person of Jules Maigret, and as Inspector Slimane in Pépé le Moko.)

Along with drawing attention to this early instance, we might point to the gothic conventions just at that time being assembled, or the earlier criminal interests of Defoe. And during the century following there is Bucket in Bleak House, crime in Bulwer-Lytton, and above all in Wilkie Collins’s bestseller The Moonstone. By 1879, just a few years before the invention of Holmes (it is right that a detective should be not born but invented) Anna Green’s The Leavenworth Case gave us Mr Gryce.

Gryce, we are told, “was not the thin, wiry individual with a shrewd eye that seems to plunge into the core of your being and pounce at once upon the hidden secret, that you are doubtless expecting to see.” Expecting? This suggests that Paget’s iconic drawings of the newly imagined Holmes had their feet in existing popular imagery, a tradition going not so far back as 1852 (Mr Bucket was “a stoutly built, steady-looking, sharp-eyed man in black”) but perhaps to 1868 in Collins’s Sgt Cuff.

“A fly from the railway drove up and out got a grizzled, elderly man, so miserably lean… dressed all in decent black, with a white cravat round his neck. His face was as sharp as a hatchet, and the skin as yellow and dry and withered as an autumn leaf. His eyes, of steely light gray, had a very disconcerting trick, when they encountered your eyes, of looking as if they expected something more from you than you were aware of yourself. His walk was soft; his voice was melancholy; his long lanky fingers were hooked like claws.” [Chapter XII]

When your mother told you to stop reading so much and go outside and play, she probably had Sgt. Cuff in mind.

The inventor of detective Gryce clearly understands the metaphysical as well as narrative elements of the form down to small behaviors such as Gryce’s habit of never resting his eye on you but on some insignificant object, projecting himself as unassuming, unforthcoming, self-deprecating, all so that we should underestimate him, be unguarded in deception. We see this gestural language again and again in the movies.

So then, as to the two sides of Holmes’s personality: Sherlock Holmes is to us the epitome of rationality and deduction, and apparently he was intended by his creator to be so. Doyle pronounced him to be “the most perfect reasoning and observing machine that the world has seen.” (A Scandal In Bohemia) “Detection is, or ought to be,” says Holmes, “an exact science, and should be treated in the same cold and unemotional manner.” He upbraids Watson. “You have attempted to tinge it with romanticism, which produces much the same effect as if you worked a love-story or an elopement into the fifth proposition of Euclid.” (The Sign Of Four) “You are an automaton – a calculating machine,” Watson complains a few pages later after Holmes has made light of Miss Morstan’s attractions. “There is something positively inhuman in you at times.” These are the canonical statements, the ones we remember and which appear in the quotation books.

One of Holmes’s immediate successors, Dr Thorndyke, was shamelessly given the moniker “The Thinking Machine.” Machines, of course, are metaphors of the link between rationality and objectivity which is essential to the detective’s manifestation and social role. Machines are affectless and cold and lend these qualities to the man who imitates them. The detective is thin, maigre, meager both physically and emotionally, with a piercing and dispassionate vulture’s eye because that is what people are like who are driven, consumed by thought. The detective is perfect, infallible in the application of his mind. Holmes says he cares nothing for the world but devotes himself entirely to his art, but by the art of detection Holmes does not mean something intuitive or expressive like painting or music. He is an intellectual craftsman. His methods are scientific, or seemingly so. There is something unpleasant and a little creepy about him, which perhaps accounts for his having no sex life and no friends other than Watson.

Yet Holmes plays the violin. He keenly anticipates an opportunity to attend a concert and he has a painting relative in France. He himself might be a little French, inheriting both French sensibilities and something of the famous French police practice dating from (and popularized by) Vidocq in Napoleon’s time. Holmes disdains French detectives out of jealousy. He is also loyal to Watson and has a strong sense of honor. There are many indicators of a deep emotional life. His gallantry, his treatment of Mrs Hudson, superficially cavalier, and his admitted fascination with Irene Adler (A Scandal In Bohemia) testify to his essential humanity toward women. With Holmes, both the warm and the cold are fused into the definition, the essence, of a detective. In him they are necessary to each other. They are found always together.

And yet the one side prevails. There is a hegemony of intellect. The detective genre exists for the purpose of defending truth and rationality against the forces of dissolution and its values are necessarily compromised on the warm side. It is inevitable that lesser men than Holmes should be either more emotional²³ or enslaved by the puzzle trope to the exclusion of anything else. This was given final form in S.S. Van Dine’s rules.²⁴ A successful warm detective would have to await a new sensibility.

Both Christie and Sayers were reaching for something like this in their creation of female detectives, Miss Marple and Harriet Vane. And of course as a genre ages one begins to muddle its attributes to keep from being tiresomely repetitive (which does not sell books). Early in the Classic tradition (1926) there was The Murder Of Roger Ackroyd, combining the roles of murderer and detective.²⁵ More recently one thinks of the detective in Robert Altman's Gosford Park, played there with delicious irony by an addled Jeeves. Though Altman’s imposter meets the minimum standards of craftish reasoning, he's terrible at it and the crime is actually solved by someone else. All formulas eventually become moribund, to be replaced by variations which in the detective universe will permit exploration of a more generous psychology.

If anyone says that Holmes is coldly rational, the texts themselves say otherwise. Using a rudimentary set of word pairs which correspond to the opposite ends of the think feel dichotomy, (think/feel, cold/warm, scientific/art-istic, intelligent/imaginative, enlightened/enthusiastic) a count of the incidence of these, related words, and their roots (e.g. intelli , art ) in five books²⁶ of the Holmes canon finds the two groups about evenly divided.²⁷ A more careful inquiry would go on to ask how often these words are found in proximity to the word “Holmes”.

Perhaps the method is too crude. For example, no account is taken of these words as descriptors of Holmes; what we have measured, if anything, is the proclivities of Doyle. If we require that the word "Holmes" appear in some proximity to one of these words what do we find?²⁸ Again, the results are not entirely conclusive, but suspicion grows that Doyle's Holmes displays the usual blend of opposite qualities characteristic of us all (except detectives), and that he is not an exclusively rational man with no pride in his work, no sense of humor or of the absurd, and no fellow feeling.

With Watson, our Chronicler, things are different. A reading of the first few pages of A Study In Scarlet, where Watson is first introduced to Holmes and to us, shows without a doubt that Watson is a man of feeling. If Doyle's craft were good, this would color our perception of Holmes since we see him refracted through Watson's lens. Before we actually encounter him, Holmes is presented as "queer in his ideas" and an "enthusiast." He communicates only when the fancy seizes him. This sounds warm.²⁹ Eccentric. Passing through the outer parts of the hospital, familiar to Watson but referred to as dun and bleak, we enter the laboratory — spacious, charmingly cluttered, a haven from the colder outer world. Holmes springs up with a cry of pleasure, his features suffused with delight at the discovery he has just made and eager to talk about it. This first impression is decidedly one-sided, and it is not the side of coldness and unfeeling. Rationality is valued, but for its pleasure as much as for its effectiveness.

This is an inauspicious beginning for admirers of cold thought. The conventional wisdom about Holmes, it seems, is simply wrong. Doyle did not portray him as cold and unfeeling and there is plenty of evidence for this in the Canon.

If we turn to the movies can we can find an evolving Holmes responsive to different attitudes rather than one pinned to the views of 1887? Most people probably read the Holmes stories early in life. This experience, which may have become dim without reinforcement, has probably been overlayed since by the impression of Holmes conveyed in the movies. Older people will remember Rathbone's Holmes, younger ones Brett's, those in between a raggle of incomplete realizations. Is Holmes warmer now?

Holmes has been portrayed on film often, possibly more often than any detective. And how exactly are we to recognize a warm, emotional Holmes? The most frequently filmed of the Holmes tales is The Hound Of the Basker-villes. If we focus on this series perhaps we can spot a shift in polarity of the think/feel axis. This in turn may reveal something of how Holmes became the epitome of the aloof, detached researcher in the laboratory of crime, so unlike such portrayals of contemporary scientists as the forensic wizards of CSI with their awful puns, cheerful obsessions, and love interests.

Sherlock Holmes On Film

I. The Hound Of the Baskervilles

This story is well suited to a confrontation between science and art. It traffics (dishonestly, but that is another topic) with the occult and the inexplicable. There is the frightening and spectral hound itself, of course, and a haunting of the Baskerville clan as a consequence of past misdeeds, a spirit needing exorcism by hard minded detection. There is fear of the man on the moor and his suspicious connection with the Baskerville servants, only two of them in this big, creaky and cold house. There is fear of the moor itself which, significantly, will yield to study, but this science produces a somewhat eldritch knowledge acquired at risk. This and many other aspects of the story provide a preponderance of opportunities for Romance which must be dispelled if the detective is to succeed by rational means. And finally, the cast contains a sympathetic woman (the typist L.L.) who starts the machinery of the plot by means of a love affair, a woman in peril, and an evil scientist who powers the action.

It was an original premise of the genre that the detective will prevail, will find the right path to the destination. There remains, of course, at least for dramatic reasons, always a possibility that there will be a wrong turning toward Romance. It would be a disaster if Holmes were to lose his way. Everything is predicated on him. If there are to be Romantic temptations (and in The Hound there certainly are) they must not ensnare Holmes. He must remain apart. It is not whether thought is privileged in these movies, which it is by the nature of the genre, but whether feeling (and thus activities such as art which are assumed to have a special relationship to feeling) will be relegated as we wish it to be, or will break out like some supernatural hound and overwhelm us. This is the reason for the odd sequestration of Holmes during the middle of the story, so as to give full play to these dangerous elements without compromising Holmes.

The Hound Of the Baskervilles has been filmed 16 times in sound, nine of these for television. Four of the five silent versions were German and one British; one of the German versions consisted of six parts made between 1914 and 1920 with different actors playing Holmes before and after the war. Of the talkies there are one each in Russian, Italian, German, and French; the rest are in English. Here is the tally.³⁰

Films of The Hound Of the Baskervilles

We will examine four of these which are readily available: the 1939 version starring Basil Rathbone, the 1959 version with Peter Cushing, and the 1988 and 2002 versions with, respectively, Jeremy Brett and Richard Roxburgh.

To begin, in the movies one can for the first time see a living Holmes. Can we tell if he is warm or cold just by looking at him? The gestural language of taking thought is probably not going to be of much help. The possibilities are too few: an inward stare, a finger to the lips, the contemplation of some object or the sudden breaking off of some activity are about all the actor has to work with.

Here is a series of images of Holmes in thought.

1: Sydney Paget's draw-ing for “The Man With the Twisted Lip” 1891; 2: Rathbone 1939; 3: Cushing 1959; 4: Brett 1988; 5: Roxburgh 2002.

Women, the car-riers of emotion (or perhaps the Typhoid Marys) are kept well away from Holmes in the Baskerville films. One, referred to as L.L. (Laura Lyons), is expunged from three of the four versions here. Her original purpose was to serve as a red herring by suggesting that she was the now dead Sir Charles's mistress and as such threatening to divert the plot into sexual byways. In any case, this sexual diversion is better suited to the scientist Stapleton's ambiguous companion, variously wife, sister, or daughter. In Doyle’s version this woman is Stapleton’s wife but presented by Stapleton as his sister, a duplicity explained and revealed by Stapleton’s prior duplicity toward L.L. Thus, to leave her in the story is to constrain the use of Mrs. Stapleton as a lure to the young heir Sir Henry as earlier L.L. was to Sir Henry’s uncle Charles. If Mrs. Stapleton seems to be a sister this frees Sir Henry from the onus of flirting with a married woman. Released by Stapleton’s probable death in the Grimpen Mire, the two are free to consummate a happy ending. Rathbone follows this narrative, made more decorous by leaving out any suggestion of love affairs and L.L. altogether. Brett follows the original but counteracts the un-Romantic truth of the divorced and exploited penurious typist L.L. by including a thrilling rescue of Mrs. Stapleton from her husband’s abuse. In the Roxburgh version Mrs. Stapleton hangs herself. This is necessary because of the suggestion that Stapleton has gotten through the moor and escaped, thus blocking with his inconvenient life any rapprochement by Sir Henry and Stapleton’s wife. Sir Henry’s virtue is saved. Cushing’s version transforms the two Stapletons into a crofter and his daughter. The girl, who is complicit in the plot against Sir Henry, vamps him in the style of an Italian peasant from a Fellini movie and pays by dying in the Mire herself.

This is ridiculously complicated. The fact that it is so shows the lengths to which these films go to separate emotion from reason, distorting a quite straightforward plot.

Mrs. Barrymore, the third woman in the story, is a simpler case altogether. She tries to do the right thing, fails, weeps, feels sadness and regret, and finally soldiers on as the exemplar of wifely good sense, keeping female emotionality under control. There is also the contrast between science amok in the person of Stapleton, a reproach to rationality and a caution to Holmes, and the good scientist in the person of Mortimer. This dichotomy is a large topic but a few points can be made.

In the Rathbone version Mortimer is an unexceptionable mature man. The Brett version is more conflicted, more chiaroscuro. This is the only version in which Stapleton is portrayed as a genuine scientist. Mortimer is allowed to be a young man as he was in the original text but his science is tainted by a ludicrous interest in phrenology. This would not have been a taint or a point of fun a century before ours (Doyle included both this and a séance³¹) but in 1988 it served to heighten the contrast with the more serious, grave, and older Holmes. In 2002 that whole tension was simply ignored in the preference for a purer melodrama whereas in 1959, a significant period in the public perception of science, Stapleton was no scientist at all and Mortimer was turned into an elderly and irascible countryman.

This is something of a muddle, but it can be said that in these movies rationality and the life of the mind take a beating. Rathbone is as accommodating to both thought and feeling as Doyle was and not especially partisan, but in 1959 and 2002 thoughtfulness is completely rejected as dangerous or boring and the character of Holmes degenerates accordingly. In the hysterical 1959 Cushing version Holmes is manipulative rather than clever and in the matter of the exposure of Sir Henry to the Hound he treats the man as a dupe to be bullied by some ugly remarks about unredeemable peasants. Cushing’s Holmes really is cold, but in a way that disparages both the Neoclassical and Romantic ideals. Between 1959 and the more contemporary Roxburgh (2002) we have Brett's ambivalence. Intending to be faithful to the original, Brett's Holmes can be said to be actually conflicted. He is alert to the siege of the rational by the irrational and does not know whether the Next New Thing might be the haunting, spectral, jealous Hound. Doyle too was, famously, drawn by the occult. Roxburgh's Holmes is supercilious and short-tempered. He may be clever -- he is certainly not irrational -- but he is no representative of the mind calmed by reason.

Finally, between Cushing and Brett we also have Ian Richardson’s version (1983). In the same year this actor made The Sign Of Four, the aborted start of a television series. By contrast with that movie, Richardson’s Hound is restrained, almost intellectual. A half-hearted attempt is made to create a foreboding atmosphere, but the village bustles and the whole place resembles a New England autumn. The plot is for the most part authentic, with one notable exception: the painter Lyons is left on the scene. This means that his obviously married and virtuous wife Laura cannot be duped by a supposedly bachelor Stapleton into an affair with himself. This hashes the charitable motive for Sir Charles’s being out on the moor despite his fear of the hound. Laura and Stapleton are now in cahoots, and when Holmes gets close to the poor woman, who is in this version a crone with dangerous knowledge, she has to be done away with. Lyons himself is portrayed as a bearlike and cruel boor driven by the realization of his failure as an artist. Yet when his wife dies he breaks down and cries his heart out.

Why this melodrama, and at the expense of an economy of plot? (The Sign Of Four is a good deal more so, approaching Roxburgh’s Hound in gothicism.) This is clear if one regards Lyons as the representative of warm, intemperate feeling whose evil counterpart is the Hound. Stapleton is positioned as cold rational intellectuality, offsetting Holmes’s worrisome dual aspect, warm and cold, muting the threat of each by contrast with the other. The 1959 film is seriously hostile to rationality. Evidently by 1983 audiences did not yet feel easy about the life of the mind and still needed some powerful balancing forces. Somewhere in the 80s there was a shift of sensibility which made Brett’s authenticity possible, a redistributed tension between mind and heart more like that of the calmer Rathbone films and what we earlier found hints of in the late Victorian period when Holmes was created.

Richardson later (2000) played the part of Dr. Joseph Bell, the man on whom Holmes was modeled, in a film whose title — Murder Rooms: The Dark Beginnings Of Sherlock Holmes — belies its stout defense of rationality. Indeed, thought and evidence are here a bastion and a salvation, the only source of a moral life. They are also, significantly, the right and the protection of women. This is very different from Roxburgh’s Hound of the same time.

In 1985 Young Sherlock Holmes utilized the character in a Tom Brown public school romp and in 1988, in a seeming jab at Brett, Michael Caine played a slapstick Holmes Without a Clue, a fake who is actually Watson’s puppet. And then we have the 1976 Seven-Percent Solution, in which Holmes, disabled by his ungovernable cocaine habit, is returned to mental life by Alan Arkin’s delicious Freud and goes off at the end on a toot with the rescued heroine. Finally, in 2010 a contemporary character purporting to be Holmes appeared in a few television episodes which owed little or nothing to the canonical stories and were not even in the detective genre but thrillers. (I discuss this distinction in a later chapter.)

The trajectory suggested by the sequence of Baskervilles, and other portrayals of Holmes, gives us, in 1939 at the advent of the noir tradition, a Holmes with gravitas, one who still assumes the wisdom of a preference for intellect but who takes this with the new moral seriousness of noir. At the height of the Cold War, this seriousness hardened into something rather desperate and cynical. By the 1980s we were again ready to tolerate some commingling of warm and cold, but this proved unstable, and we found ourselves (as it seems we now do) in a position opposite to that of 1959. Now the flesh dominates and has given us the gothic melodrama of 2002 and the action-driven farce of 2010.

Even this rudimentary social history suggests how Holmes tracks the zeitgeist, but it does little to explain the attribution of coldness to Holmes's rationality. In fact it is the reverse; thought is warm or cold as the times require, and the Detective likewise. It may be that by now Holmes is rather a vehicle for nostalgia³² for an imaginary time when this was not so.

Regarding the question of whether the Holmes texts present him as coldly rational, the answer is no. The incidence of words in the rational science group, expressed as cases per thousand words of text, with the incidence of words in the warm emotions group, are

This is no support here for the hypothesis that Holmes is a coldly rational man, nor for the opposite.