Previous Month | RSS/XML | Current

WEBLOG

October 2nd, 2018 (Permalink)

New Book: The Sherlock Effect

There is no question that Conan Doyle was a great writer of fiction. Indeed, he was so good that he made the methods of Sherlock Holmes plausible not only to a general readership but also to a wide variety of forensic doctors, academicians, and scientists. People seem to have forgotten that Sherlock Holmes is make-believe. It is both sad and terrifying to note that professionals from the Victorian Era to the present day apply fictional methods to true-life happenings.1

I'm on record in a previous "New Book" entry as skeptical about Sherlock Holmes being treated as a model of logical reasoning2. I mentioned that I like the stories, and I've read every one of the originals written by Doyle more than once3. However, anyone trying to imitate Holmes would soon find themselves frequently mistaken. Holmes is never wrong only because the stories don't take place in the real world, but in a fantasy world created and controlled by Doyle.

This brings me to the current New Book by Thomas W. Young, a forensic physician, subtitled: "How Forensic Doctors and Investigators Disastrously Reason Like the Great Detective". Clearly, this new book differs from the previous one, Mastermind, in being critical of the effect of Holmes' example.

I'm also skeptical of the notion that forensic scientists have been so influenced by a fictional character4. However, assuming that Holmes has indeed had such an influence, I'm less skeptical that it would be at least partly negative, though "disastrous" is a strong word.

In a review of the book in Psychology Today, Katherine Ramsland, a forensic psychologist, writes:

Young begins with Holmes’ explanation to Watson of “reasoning backwards.” It works like this: Holmes learns about or observes a result…and then uses intuition to describe the steps required for the incident to have occurred. “The reader is tricked into thinking that backwards reasoning is brilliant,” says Young, but it doesn’t actually work. Here's why: “For any result, any set of clues, there may be numerous possible ‘trains of events’ that could explain the result.” … So, any given “result” might have numerous possible routes, and one cannot be certain with intuition alone which one to pick.5

Intuition alone doesn't work because it's only part of the first half of the scientific method: forming an hypothesis to explain a set of evidence. There's nothing wrong with using intuition or imagination to form hypotheses; in fact, it's an essential step. What's missing is the next step in which the hypothesis, or intuition, is tested. In Doyle's fictional world, Holmes' intuitions always turn out to be correct; but in the real world, it's often necessary to go through many hypotheses before hitting upon the correct explanation.

Hopefully, this book not only critiques Holmes' methods, as portrayed in Doyle's stories, but also provides better methods. As a logician and fan of Doyle's Holmes stories, I'm looking forward to reading it.


Notes:

  1. Thomas W. Young, The Sherlock Effect: How Forensic Doctors and Investigators Disastrously Reason Like the Great Detective (2018), Chapter 2: "Sherlock and His Successors".
  2. See: New Book: Mastermind, 2/14/2013. I never reviewed this book, though I have read it. I was not favorably impressed.
  3. There are also so many Holmes pastiches of varying quality that I've only read a small fraction of them. However, one that I can recommend for those who like Holmes, and are also interested in logic and probability theory, is Colin Bruce's Conned Again, Watson: Cautionary Tales Of Logic, Math, And Probability (2008).
  4. Katherine Ramsland, a forensic psychologist who reviewed the book for Psychology Today, is also skeptical of this so-called "Sherlock Effect". See the next note.
  5. Katherine Ramsland, "Sherlock's Curse", Psychology Today, 5/30/2018. Young's response is here: "A CRC Press author reviews and critiques The Sherlock Effect in Psychology Today", CRC Press, 5/31/2018.

Previous Month | RSS/XML | Current