Some time this century, the Sherlock Holmes stories became part of the literary canon: 'Joyce, Kafka, Baudelaire, a dash of Conan Doyle. In the late Twenties these were fashionable names.'⁽²⁾ When Dr Watson reported the death of Sherlock Holmes in Arthur Conan Doyle's 'The Final Problem' (December 1893), he ended by talking of: 'him whom I shall ever regard as the best and the wisest man whom I have ever known' echoing another report written long before: 'Such was the end, Echecrates, of our friend, whom I may truly call the wisest, and justest, and best of all men whom I have ever known'⁽³⁾ in Phaedo, Plato's account of the death of Socrates in 399 BC. In a new medium of publication, a new genre of literature and a new style of writing Doyle was asserting a continuity with an ancient higher tradition, almost the opposite of the populist means he was using. Yet this was not a literary tradition.

In a standard philosophy textbook, John Hospers⁽⁴⁾ lists the sources of knowledge as Sense Experience, Reason (which he sub-divides into Deductive Reasoning and Inductive Reasoning), Authority, Intuition, Revelation and Faith. With the publication of the Sherlock Holmes stories the role of reason⁽⁵⁾ might have appeared to have become supreme and unchallengable, if measured by the readership of the vehicles in which it was expressed. The Strand magazine was first published in January 1891 with an immediate circulation of 300,000⁽⁶⁾. In July the first Holmes short story⁽⁷⁾ appeared and within a few issues raised the circulation to 500,000. In the USA magazines featuring Holmes achieved circulations as high as 700,000 or even 1,000,000 (implying far higher readerships), and in those periods when Holmes stories did not run circulation did not fall. (Serialisation of The Hound of the Baskervilles in 1901, the first Holmes story for eight years, increased The Strand's circulation by 30,000). Editors realised that the appeal of Holmes was immense but that other authors in the new genre could supply an equivalent, which meant that despite the appearance of other magazines and detectives (The Rivals of Sherlock Holmes as Hugh Greene called them when he anthologised them)⁽⁸⁾ it was thirty or forty years before circulations began to fall, and by then detective novels were filling the bookshops and circulating libraries.

In 1890 George Gissing, a man intent on staying within the literary canon, considered the literary environment in his three volume novel New Grub Street, published in that next, seminal year 1891. He failed to see how ideas and economics would merge. The Strand was launched on the profits of Tit-Bits (first published in 1881), which Gissing parodied as Chit-Chat: 'what they want is the lightest and frothiest of chit-chatty information - bits of stories, bits of description, bits of scandal, bits of jokes, bits of statistics, bits of fooolery... their attention can't sustain itself beyond two inches. Even chat is too solid for them: they want chit-chat'⁽⁹⁾. In the year that both must have been considering the future of publishing, Gissing did not see, as Alfred Harmsworth the publisher could, of what such a readership might be capable: that people who bought Tit-Bits (and its rivals) would want The Strand (and its rivals). However, Harmsworth and his editor, Greenhough Smith, could not have been aware of the serendipitous workings of Doyle in making logical investigations into bloody murder such an overwhelmingly attractive and available form of reading, as Doyle sat in the consulting room to which patients never came, giving him time to write.

So what is a detective story? Staying within the genre, it is explained in this interview between Inspector French and a suspect in Freeman Wills Crofts' novel Enemy Unseen (1945).

'Your books are in my line, I imagine?'

'Yes and no,' Crane answered. 'I write thrillers. Detective stories are more about your kind of work, but I distinguish very clearly between the two, and I fancy detective-story writers do too.'

'That interests me, Mr Crane. I'm afraid I hadn't properly appreciated the difference.'

'A very serious error.' Crane was now smiling almost genially. 'The detective story is the story of the elucidation of a problem. The solution is reached by inference and deduction from the given facts. In any story worthy of the name all the facts are given to enable the reader to find out the truth for himself. If he fails and continues reading he can watch the detective succeed by the reasoning he should have employed himself.'

'He always does succeed. Unhappily that's where one departs from real life.'

'Oh, yes, he must succeed or the else the story has no ending.'

'I wish we could say the same of our cases.'

'I dare say. Well, the thriller is quite different. Here the object is thrills. Premise and deduction take a second place and conflict is in the forefront: the struggle of the criminal and the police, or of the evil gang and their righteous pursuers.' ⁽¹⁰⁾ (Crofts uses 'premise', which is usually spelled 'premiss' when used in logic)

Mr Crane's description of his own job takes only a quarter or a third of the space required to describe the detective story, but look at the keywords in the account of the detective story: elucidation, solution, inference, facts, truth, reasoning, premise and deduction, and in those words and phases we are are thrown back into the world of scientific reasoning - they are practically all head words we would expect to find in a standard book of philosophy. (In fact, only elucidation, solution and reasoning, do not occur in the index to Bertrand Russell's Human Knowledge⁽¹¹⁾, for example, while Hospers' includes reasoning). The language of the theory of the detective story is the language of those who write as heirs of Socrates⁽¹²⁾ (and, also, clearly from an early date a self-conscious form of fiction).

Doyle's keenness on this claim lead him to use the phrase 'The Science of Deduction' both as the title of chapter two of A Study in Scarlet and of chapter one of The Sign of Four even though he must have been aware that the foundations of scientific reasoning were still open to doubt. David Hume had opened it in the eighteenth century, and John Stewart Mill took it much further in his 1843 A System of Logic, a best-selling volume amended and reprinted throughout the century, though Mill was unable to produce any definite way that scientific advance might progress straightforwardly. His rival, William Whewell, argued that science tends to advance by making hypotheses and then testing them, and the American C.S. Pierce seems to have advanced a similar idea of using hypotheses, which was neither induction or deduction - he called it ab-duction (Crofts' Mr Crane seems to be referring to the use of hypotheses in the phrase 'premise and deduction', as Holmes did in 'The Yellow Face' (February 1893) - 'When new facts come to our knowledge which cannot be covered by it (my theory), it will be time enough to reconsider it'). It is through an example he takes from Holmes in 'The Boscombe Valley Mystery' (October 1891) that Patrick Shaw arrives at his discussion of the alternate positions of Mill and Whewell in Logic And Its Limits:

'(Holmes') argument seems to be: if the stone were the murder weapon then it would not have lain there long and it would correspond with the injuries; it hadn't lain there long and was consistent with the injuries; hence it was the murder weapon.

'What are we to make of this: that a form of argument which is apparantly widely used and even held up as some sort of paradigm of rationality should be an invalid form of argument?' ⁽¹³⁾ (Shaw suggests a schoolboy having been throwing stones is just as likely a source of the misplaced rock).

Shaw's objections seem irrelevant and Holmes' suggestion far more probable, than asking how many schoolboys are throwing stones at all, let alone that in the ambience of a murder how many schoolboys are throwing stones? The most likely source of a displaced stone that fitted the murder wound would be the stone that had been used in the murder - this as Shaw points out eventually, in a discussion of 'arguments with probable conclusions', is a question of probability. (It is unspoken in what Holmes said - I think all we readers infer it).

Doyle again uses this when introducing an example of Holmes' thinking at the beginning of 'The Blue Carbuncle' (January 1892) - 'Tell me what it is that you can infer from this hat?' he instructs Watson, and gives a description of the owner when Watson cannot. This description can be broken down into a series of syllogisms, such as:

loving wives ensure their husbands' hats are brushed

this hat has not been brushed for weeks

> the wife of the owner of this hat has ceased to love him

> Mrs Henry Baker does not love Mr Henry Baker⁽¹⁴⁾.

'These are the more patent facts which are to be deduced from his hat. Also, by the way, that it is extremely improbable that he has gas laid on in his house.' Holmes says, then explaining why he thinks it is 'improbable' in another example of syllogistic reasoning. This suggests that Doyle/Watson was always aware of the role of probablility in scientific proof but did not mention it in 'The Boscombe Valley Mystery', rather than that in the four months between the two stories he discovered the concept in some academic volume and slipped it in as some sort of veneer of philosophic respectability⁽¹⁵⁾.

Throughout the canon Holmes makes statements about his work: 'There is nothing like first-hand evidence'; 'nothing is too little'; 'Eliminate all other factors and the one which remains must be the truth'; 'An exception disproves the rule'; 'You can, for example, never foretell what any one man will do, but you can say with precision what an average number will be up to'; 'There is nothing more deceptive than an obvious fact'. And one can infer statements about his working from his observations, such as the dog that did not bark in the night-time, so how can the apparant paradox of their being both rational and irrational be reconciled, if reconciled they can be?

For, taking the story which has been rated over and over by author and readers their favourite and looking at its scientific basis we would find that that story is 'The Speckled Band' (February 1892), published just a month later, and is a work with no scientific truth in its construction or solution at all.

A worried woman arrives at 221B Baker Street (though she could only have known Holmes' previous address), her sister having died in horribly mysterious circumstances. Holmes will identify the threat sent by her stepfather, and eventually save her by driving back the snake sent into her bedroom, killing the would-be killer instantly. The snake is a biological fantasy, its training invalid and its movements impossible. Later, Doyle corrupted the story even more when he adapted it for the stage and changed other information (such as the hair-colour of Watson's fiancee and the chronology of the stories)⁽¹⁶⁾.

Is the existence of a story so complete in its fiction, evidence of Doyle's disrespect for the logic he claimed? Does it disprove the hypothesis that Doyle's method of construction depended on a line of thought stretching back to Socrates, as these following examples tend to do?

Doyle was prepared to kill Holmes. In Memories and Adventures he wrote: 'After I had done two series of them (the short stories) I saw that I was in danger of having my hand forced, and of being entirely identified with what I regarded as a lower stratum of literary achievement' (Page 99), and yet having intended to kill him, Doyle did not write a death that could be described by an eye-witness, instead he wrote one from which Holmes could return if he had not disappeared over the Reichenbach Falls but had climbed upward instead.

Watson met Holmes after returning from India recovering from a bullet through the shoulder, except that the wound moves to his leg in the first chapter of The Sign of Four. The subtitle of A Study in Scarlet is 'Being a reprint from the Reminiscences of John H. Watson M.D., Late of the Army Medical Department', but in 'The Man With The Twisted Lip' (December 1891) Mrs Watson calls her husband 'James', while she seems to come and go (or at least, a wife), regardless of any attempt at internal chronology, so that it seems that Watson was married three times. It is said that of the fifty-six short-stories and four novels, in only one do the given year, month, day of the month and day of the week agree. Professor James Moriaty has a brother, Colonel James Moriaty, and a third brother, whose name is never given - perhaps it is James - who is a station master in the West Country. Try to take a journey on almost any train you find mentioned and you would find yourself sitting on the station with a long wait ahead of you. In some cases when a train finally arrives, it will drop you at an unexpected destination.

Doyle admitted his secret in Memories And Adventures: 'The first thing is to get your idea. Having got that key idea one's next task is to conceal it and lay emphasis upon everything which can make for a different explanation'⁽¹⁷⁾ (pp106-107). The ultimate logic of the stories is their fictionality, and within that world in which 'The Speckled Band' is set, their deductions follow naturally and validly⁽¹⁸⁾

. Shaw the philosopher could be called in defence of this - 'The primary concern of logic is validity rather than truth' (page 32).

Today, Doyle has been accepted so far into the mainstream, that not only do the Oxford University Press publish a Holmes selected stories in the Worlds Classics series, but a critical edition of the works as well. Volumes such as John A. Hodgson's aimed at undergraduate courses combine the stories with contemporary critical essays, showing that both in subject and their opportunities for methodological investigation the Holmesian canon has been accepted into the mainstream of literature available for study.

The Holmes stories reflect the puzzled path of literature: at first written for immediate gain, though expounding a theory only just coming to be accepted, Doyle surpassed his medium, as the Elizabethan dramatists surpassed the groundlings who provided their mass audience. Though having been able to recognize how to satisfy and increase demand among a growing readership, Doyle did not see the final significance of his work. Yet as he attempted to escape the creation of the canon by killing Holmes his explicit quotation from a line of philosophers into which his fiction merged, given his medical background, must have suggested that Plato, in his inaccurate and therefore fictionalized account of the death of Socrates (hemlock poisoning being painful and nauseous), provided an early precedent for his fictional mix of death and philosophy.

1. Murder Will Out: The Detective In Fiction by T.J Binyon (Oxford: O.U.P 1989) page 9.
vide:
"Constitution
Article I
The name of this society shall be the Baker Street Irregulars.
Article II
Its purpose shall be the study of the Sacred Writings."

From the Articles of (the American) Baker Street Irregulars Inc, reprinted in The Annotated Sherlock Holmes: The Four Novels and Fifty-six Short Stories by Sir Arthur Conan Doyle, edited by William S. Baring Gould (New York: Wings Books 1992), page 38 column 2.

2. Ian Hamilton in a review of Edward Upward in London Review of Books Vol 17 No 2 26 January 1995, page 19.

3. The Works of Plato translated by B. Jowett (New York: Tudor Publising Co n.d.) Volume 3 page 271. Baring Gould and John A. Hodgson (see below) quote other versions of this speech.

4. An Introduction to Philosohpical Analysis rev. ed. by John Hospers (London: RKP 1967)

5. Wilkie Collins' The Moonstone (1868) was described by T. S. Eliot as 'The first, the longest and the best of modern English detective novels' but Collins himself was regarded by his contemporaries as a 'Sensation novelist' (sensation being an antonym of reason) and Collins himself continued to be interested in crime and investigation, but wrote no other detective novel. See In the Secret Theatre of Home: Wilkie Collins, sensation narrative & nineteenth-century psychology by Jenny Bourne-Taylor (London: Routledge 1988)

6. Julian Symons gives a history of the magazine in Bloody Murder: From the Detective Story to the Crime Novel rev. ed. (London: Viking 1985). Another appears in the Introduction to Detective Stories From The Strand Selected and Introduced by Jack Adrian (Oxford: O.U.P 1991).

There was a considerable difference between the publications of the 1890s and the magazines (such as Thackeray's Cornhill) and serial works (of Dickens and others) of the mid-Victorian period. The magazines had circulations of up to 100,000 (a third of The Strand), but one of Dickens' parts sold only in tens of thousands, and his complete novels had print runs of only 1500. See Victorian Novelists and Publishers by J.A. Sutherland (Chicago: University of Chicago Press 1978) pages 37 -38.

7. Doyle had written about Holmes in two short novels. His decision to use him in complete, self-contained short stories was expicitly commercial: 'a single character running through a series ... would bind that reader to that particular magazine ... a character which carried through, and yet instalments which were each complete in themselves, so that the purchaser was always be sure that he could relish the whole contents of the magazine' he wrote in Memories and Adventures (London: Hodder and Stoughton 1924) pages 95 -96. Doyle was recognising that the nature of Victorian publishing had changed - Charles Reade had summed up the earlier style of construction as 'Make 'em laugh; make 'em cry; make 'em wait' (quoted in The Penguin Book of Quotations).

8. The Rivals of Sherlock Holmes Edited by Hugh Greene (London: The Bodley Head 1970)

More Rivals of Sherlock Holmes Edited by Hugh Greene (London: The Bodley Head 1971)

Further Rivals of Sherlock Holmes: The Crooked Counties Edited by Hugh Greene (London: The Bodley Head 1973)

The American Rivals of Sherlock Holmes Edited by Hugh Greene (London: The Bodley Head 1976)

9. New Grub Street by George Gissing (Harmondsworth: Penguin 1968) page 496 -497.

10. Enemy Unseen (London: Hodder and Stoughton 1945) by Freeman Wills Crofts pages 123-124

11. Human Knowledge: Its Scope and Limits by Bertrand Russell (London: Allen and Unwin 1948)

12. Holmes calls himself a Consulting Detective - that is independently available as Consulting engineers, chemists or surgeons were available, but rarely does he stay in his room and reveal to a visitor the solution to a problem brought to him by the visitor, all the clues to which are unrecognised in the dialogue. So his method was not one of Socratic dialogue - however, one of his rivals used this method almost exclusively, this was The Old Man In The Corner, created by Baroness Orczy.

13. Logic And Its Limits by Patrick Shaw (London: Pan Books/Heinemann Educational 1981) page 37

14. 'Sherlock Holmes - The Series' by Martin Priestman, reprinted in Sherlock Holmes The Major Stories with Contemporary Critical Essays edited by John A Hodgson (New York: Bedford Books/Macmillan 1994) page 320

15. In Diagnosis and Detection: The Medical Iconography of Sherlock Holmes (London: Associated University Press 1987) Pasquale Accardo points out that the Holmes stories continue to be used as examples of scientific reasoning: 'In his aha! Insight Martin Gardner uses "The Problem of Thor Bridge" to distinguish the two stages of a scientific problem: a hypothesis generated by intuition or insight (the discovery process) is then tested by looking for deduced consequences of the proposed theory (logic of verification)'. (page 130)

16. Baring Gould gives three foolscap pages to the problems of the snake in his Annotated edition. In his Life of Sir Arthur Conan Doyle John Dickson Carr (New York: Vintage Books 1975) discusses the corruptions to the canon introduced through the stage play (pages 304 - 307).

17. In a letter dated August 9, 1846 Edgar Allan Poe, wrote 'These tales of ratiocination owe most of their popularity to being something in a new key. I do not mean to say they are not ingenious - but people think them more ingenious than they are - on account of their method and air of method. In the Murders in the Rue Morgue, for instance, where is the ingenuity in unravelling a web which you yourself have woven for the express purpose of unravelling? The reader is made to confound the ingenuity of of the suppositious Dupin with that of the writer of the story' quoted in The Mystery Lover's Book Of Quotations Compiled by Jane E Horning (New York: Mysterious Press 1988) page 173

18. 'The Recoil of "The Speckled Band": Detective Story and Detective Discourse' by John A. Hodgson, reprinted in ... Contemporary Critical Essays (1994).