Well, that about wraps it up for Sherlock Holmes

Embed from Getty Images

I have my doubts about Sherlock Holmes’s method of reasoning. Take what is probably his most famous formula:

“When you have eliminated the impossible, whatever remains, however improbable, must be the truth.”

It sounds impressive, but it really isn’t.

Every event, I would suggest, has an infinite number of possible explanations. And if you don’t buy that, you still have to agree that every event has an enormous number of possible explanations. Which means that eliminating impossible explanations is not a practical way of investigating a crime.

Perhaps Holmes meant it more pragmatically – something like, if you have several suspects then once all but one has been eliminated then the person left must be guilty. But that won’t do either. Your original suspects may all have been innocent.

I can see Holmes’s approach working in a closed system like formal logic, mathematics or chess, but not in solving messy, real-world problems.

I saw another of Holmes’s lines the other day, and I think it’s equally suspect:

“It is a capital mistake to theorize before one has data. Insensibly one begins to twist facts to suit theories, instead of theories to suit facts.”

Again, there is an infinite amount of data. In order to gather it for an investigation, you must already have theories about what is likely to be useful. As Karl Popper says somewhere, if you tell someone to “observe”, they will ask “Observe what?” Observation, like data-gathering, is necessarily laden with theory.

And while twisting facts to suit theories is obviously wrong, twisting theories in the face of facts is dangerous too. What can happen is that once you have a suspect, you think of ingenious tweaks to your theory to account for new information that appears to exonerate them. There comes a point where, to be rational, you have to abandon your theory rather than twist it.

I was encouraged to write this post when I heard Ron Warmington interviewed the other day. He was one of the two authors of the Second Sight report that revealed that the Post Office was persecuting and prosecuting innocent sub postmasters.

Before working for Second Sight, he had been a fraud investigator with multinational companies. And he said that his approach if someone was suspected was to assume they were guilty and then look for ways of proving they were innocent.

If this approach had been adopted by Post Office investigators, they would have found that none of the supposed thieves had thousands in the bank or had just bought a flash new car. They would have found that the missing money had never existed.

So maybe Sherlock Holmes was not as clever as he thought?

Of course, he may have been misquoted by Dr Watson or Conan Doyle. I'd like to give my thanks to Oolon Colluphid and Karl Popper for the inspiration.