"Sherlock Homes and Intelligent Design" is the title of an actual article published in the September 2012 edition of The Quarterly Review of Biology. It traces how the two sides of the Intelligent Design/Evolution debate each invoke Sherlock Holmes in a proxy war over reason.
When he was doing research on Holmes for an unrelated project, Professor Brian McCuskey noticed a "constant pattern" of references to the character by both sides of the debate. McCuskey says:
The two teams are trying to recruit Sherlock Holmes to their team. Obviously the team that has Sherlock Holmes on it will be seen as the most credible. The reasoning follows: the more I am doing what Sherlock Holmes does, the more scientific I must be.
The article McCuckey wrote as a result of his findings traces the invocations of Holmes on both sides and then dissects why appealing to a fictional authority does neither side any favors. As to why everyone's trying to "recruit" Holmes, McCuskey says in the article's introduction:
As they clash in the court of public opinion, both sides ask Holmes to testify on their behalf: he is the enduring pop-cultural icon of reason, whose celebrity endorsement counts a great deal with an audience whose votes and tax dollars are at stake. Acting as a star witness at every stage of the debate, Holmes first allows us to cross-examine the arguments for and against design; however, because he seems equally friendly to both sides of the case, we must then question the scientific authority of the detective himself.
McCuskey found that Intelligent Design advocates tend to rely on two Holmes quotes in particular: one about circumstantial evidence ("Circumstantial evidence is a very tricky thing. It may seem to point very straight to one thing, but if you shift your own point of view a little, you may find it pointing in an equally uncompromising manner to something entirely different.") and the always-classic "When you have eliminated the impossible, whatever remains, however improbable, must be the truth." The first is used to argue that evolution is supported by circumstantial evidence, which could be interpreted to support any theory, including Intelligent Design. Therefore, its deserves equal time. The second is used to say that design is the only theory that remains standing, making it the "truth."
McCuskey gives a reason why Holmes is so often used by Intelligent Design advocates:
One reason we find Holmes's fingerprints all over the design inference is that its advocates explicitly describe themselves as sleuths, doing the same "historical detective work" (Hartwig and Meyer 1993:155) as evolutionary biologists, anatomists, geneticists, and paleontologists: observing present effects, inferring past causes, and testing possible hypotheses until only the best explanation remains.
For scientists explaining arguing the evolution side of the debate, McCuskey has found them to be more cautious in their Holmesian references because:
with bad information and unlimited options, the detective's reasoning quickly devolves into what critics of ID now call the "Sherlock Holmes fallacy," otherwise known as the explanatory filter: a special extended edition of the argument from ignorance that leads creationists to "arrive at explanations that are easy to find but difficult to falsify, such as gods and other supernatural entities with unknown properties" (Evolution Education Wiki 2007).
In the race to win the war of public opinion by proving that their argument is more rational by casting themselves in the role of Holmes, scientists drag the character to their side by partially casting the other side as an enemy of reason:
From this point of view, advocates of ID commit nearly every crime against logic in the book—setting up false dichotomies and straw men, making arguments from ignorance and incredulity, moving goalposts and reasoning in circles—whereas Holmes properly belongs on the team of Darwinian investigators.
McCuskey ultimately concludes that using Holmes this way forces both sides to ignore a lot about the character and his creator. The problem Holmes presents for Intelligent Design is that he is searching for a suspect, he's not starting from all possibilities. As McCuskey points out, Holmes does poorly when confronted with mysteries with anything other than a person as the cause:
the late-arriving detective does not direct his reason toward the end of detecting design; instead, he begins by presuming design—somebody threw this harpoon at this victim—and then reasons from that premise to determine what person had means, motive, and opportunity to commit the crime. The premise holds in most stories, but tellingly it does lead Holmes astray in those cases that involve natural accident
For the side of evolution comes the problem that Holmes says that deduction proves the existence of the creator in "The Adventure of the Naval Treaty."
Finally, McCuskey points out that both sides have to ignore the fact that Sherlock Holmes isn't real. Part of the problem comes from Sir Arthur Conan Doyle, who was famously spiritualist. In order to use Holmes as the paragon of rationalism, everyone has to pretend that Doyle's personal views have nothing to do with Holmes. Says McCuskey:
Until then, however, it matters so much—not just philosophically but politically—that both believers and nonbelievers must wash their hands of Doyle in order to keep Holmes clean: because both parties employ the detective to illustrate the same purely scientific method—even if that method leads them in opposite directions—neither party can allow the eccentric author's spiritualism to contaminate his iconic character's rationalism. [Evolution], backed by mainstream science, can afford to mention Doyle's fancies and then defer or disavow them, keeping Holmes's reputation intact; without that official backing, Dembski and his colleagues on the fringe avoid the association altogether, lest it make their inference seem fanciful. At the same time, and for the same reasons, neither party wishes to call too much attention to Doyle's authorship at all, lest it make his detective's example seem too literary to be material: alas, Holmes never existed, but the allusion works because everyone makes believe he did. Both parties blur the line between fact and fiction as vigorously as the Baker Street Irregulars, the Sherlockian literary society famous for pretending seriously that the detective was an actual person rather than a character, which likewise requires them to refer to Doyle as little possible—and when they do, not as Holmes's author, but as Watson's agent. However, the fact that scientists are playing the same Sherlockian game as their opponents should give them real pause, long enough to wonder whether they should continue to use Holmes to symbolize their materialist principles and logical methods.
The other problem with ignoring the fact that Holmes is fictional is that his deductions work because of that fact. Doyle starts from the conclusion and works his way back, which makes Holmes look brilliant but borders on breaking through the fourth wall. What Holmes does is so impressive, it's clear how much work Doyle did to write it.
Doyle's detective gives a false sense of knowing something, which is the very last thing anyone in American higher education needs to encourage—and so let us eliminate him from both evolutionary and design arguments and see what remains.
Oh no. What will the world look like if we can't argue about what fictional characters would hypothetically agree with us?