A False Facade, Dr. Watson's Agency & AI

A little, wizened man darted out.

Sidney Paget's drawing from "The Adventure of the Norwood Builder" at the point when Sherlock Holmes's trick pulls the despicable Jonas Oldacre out of hiding  sometimes makes me think of The Wizard of Oz, and the revelation that the mighty wizard is just another man not telling the truth. Perhaps the wizard is not as despicable as Oldacre, but he still is a fraud--a fraud hiding behind a facade. 

This fraud behind a facade has been on my mind of late due to the information I read in the news and on social media about AI creation and use. The revelation that the work of many writers was stolen to train AI is maddening. Another particularly sad truth: publishers are selling author's work to the AI companies without the author's consent or compensation. The entirety of the situation is overwhelming to an individual like me but I still care very much about the writers whose works are appropriated.

My writing is on a very small scale, mostly niche, and often pastiche particularly inspired by Arthur Conan Doyle. In a recent group discussion about the use of AI in fiction writing I was asked if there are any real differences between using another writer's characters to write a story and using AI to write a story. I think there are many differences, mostly having to do with permissions and legalities, but in the most basic sense the difference between the two is that a pastiche writer may be writing a story about known characters but she is actually writing the words and hopefully writing with intent, with skill,  and with respect for the creator of the characters. AI is gluing together words from many sources, words that may or may not make sense.

As for respect, I would never use another writer's characters if the characters were not in the public domain, or if I did not have specific written permission from the copyright holder if needed. Some people seem to dismiss the idea of public domain and the rules surrounding the use of literature in the public domain. They seemingly see little difference between the use of public domain characters for pastiche and the AI harvesting of language from everywhere and everyone. The nuances of the issue keep rolling around in my head. I can't help but wonder what Arthur Conan Doyle would think about it all.

I look at the preceding paragraphs and I'm surprised at what I've written. I didn't sit down here to talk about the AI travesty; I planned to write about something else that has been taking up space in my head for a long time. (I use this blog to think out loud, if you will.) Today, I thought I was going to think about something Dr. Chris Pittard said in a Zoom meeting about a year ago as part of a discussion with Oxford Worlds' Classics editors for the new Sherlock Holmes volumes. 

He mentioned the ending of "The Adventure of the Norwood Builder" with an interesting thought about Sherlock Holmes's last statement:

"I fancy that for some few years you will find your time very fully occupied," said he. "By the way, what was it you put into the wood-pile besides your old trousers? A dead dog, or rabbits, or what? You won't tell? Dear me, how very unkind of you! Well, well, I dare say that a couple of rabbits would account both for the blood and for the charred ashes. If ever you write an account, Watson, you can make rabbits serve your turn." (Emphasis mine.)

 Dr. Pittard talked briefly about how the relationship between Holmes and Watson had evolved by the time of NORW, and how Holmes's consideration of Watson's writings had changed. He noted Holmes makes the writing of the account to be Watson's choice, and Watson can decide to end the story as he desires, including simply stating the bones were those of rabbits, even if he did not know for sure. Dr. Pittard discusses how this statement indicates how much Watson's agency within the relationship had changed from the earlier days.

I have read that paragraph many times and this thought about Watson's agency never crossed my mind. Now. I'm interested in combing through the time period where NORW falls and see if I can find more examples of Watson's role growing larger. The change is interesting; Dr. Pittard's thoughts are interesting. It is, as Christopher Redmond wrote in Canadian Holmes in 2009 (Vol 31, Number 4, p.2), the "...sort of thing that in a Canonical context would give us the opportunity to spend many happy hours creating Sherlockian papers."

When thinking about Watson's ways overall, I think it is safe to say he does not put on a false facade. Watson tells us when he has changed something, and he tells us when he can't remember something. He is honest about any deliberate changes or ambiguity. In an era of writing disarray, it is comforting to know we can trust Dr. Watson.

As for the AI thieves, I can only quote Sherlock Holmes, "Dear me, how very unkind of you!" To be honest, I would like to say something much harsher. Much harsher, indeed.