Monthly Archives: July 2014

Science and Sherlock

I’ve never quite been able to decide how I feel about BBC’s “Sherlock.” I’ve been even more unsure how I feel about Sherlock himself. By all accounts, I should like him. I should find him cool, as I find many douchebags. Somehow, though, the character has failed to speak to me. I’m starting to warm up to him, as he seems to be less of a douche in season 3, but I’m still confused as to how I, a person with legions of douchey favorite characters, did not like him from the beginning. Picking things apart, it can’t really see why I’m not a huge fan of the character. One instance, however, sticks out in my mind.

Bizarrely, it wasn’t something classically douchey that he did. It was something illogical. By which I don’t mean that he acted emotionally or anything like that. I mean that he used faulty logic. More precisely, he never recognized that he was not using logic and proceeded to lecture others on its proper use.

In the first episode, Sherlock determines that the woman whose corpse he has briefly examined was an adulteress.He does this by observing that all of her jewelry is clean, except for her wedding ring, which means she’s in an unhappy marriage of at least ten years. He further notices that the inside of the ring is clean, which means that she takes it off, leading him to the conclusion that she’s been cheating on her husband. Thus, we are all shown of Sherlock’s brilliant reasoning.

But is this actually reasonable? The thing is, when considering human actions, the consideration of multiple interpretations of a data set is critical. While Sherlock turns out to be right, because the show is written in such a way that Sherlock is right, this represents a failure of reasoning. Sherlock makes the induction (and detective work is induction, rather than deduction, despite what Sir Arthur Conan Doyle may have written) that since the woman took off her ring frequently and adulterers take of their rings frequently, this woman is guilty of adultery. Yet, surely this isn’t the only interpretation of the fact of a woman taking off a ring. I’m not married, but I assume a number of people take their wedding rings off to shower or to sleep. Perhaps she works somewhere where wearing jewelry is unwise or even dangerous. Similarly, maybe she doesn’t clean the wedding band because she doesn’t care about her marriage, but maybe it’s because the ring is everyday jewelry and the other pieces are only for certain occasions.

I could ignore this. I should ignore this. Ultimately, a show that depicted detective work as tedious as it is in real life would be boring. Besides, some of the magic behind creating a hyper-smart character is making him see and realize things that other people wouldn’t and he will often have to be right simply because they need him to be right to show his genius.

And yet. And yet in episode two, Sherlock goes on a big rant to Detective-Inspector Whatshisface about how he, the actual detective with an actual job in this stuff, is ignoring all the evidence that disagrees with his hypothesis. Sherlock then goes on to explain his hypothesis that includes all the facts, still ignoring that it’s not impossible to have multiple hypotheses that include all the facts or alternate interpretations of facts.

My problem with this, you may have guessed, is that Sherlock did this exact thing the episode prior. And, while this same thing is less blatant in other episodes, Sherlock is never actually made to fully explain how he gets to any of his conclusions. Sherlock therefore, as far as the viewer is shown and not told, makes snap judgments, rather than reasons things out. Even when he does take a longer time to reach a conclusion, we rarely know if he’s considered other options.

Of course, Sherlock is usually right. I’m actually okay with this. The hero must ultimately save the day and in a detective show this means that he must solve the case. I would just like my improbably smart detectives to not lecture me on the proper use of “deductive” but actually inductive reasoning if they aren’t going to ever actually use it.