Ramble About Writing And Deduction

[kerravonsen]

One of my favourite things to do in writing is to allow the characters to jump to logical but incorrect conclusions. I find it mischievously delightful. I have an evil smile while I am doing it. No, I do not toss in deliberate red herrings in order to confuse people. That isn't fun at all, that's cheating. What is delightful is putting all the clues there right in front of them, and have them not notice them, or misinterpret them, while not being stupid; making the characters stupid is no fun at all.

It's one of the things that fall under my maxim "remember what your characters don't know" but that isn't all there is to it. I've been pondering it a little, and I realised that one could also describe it as "demonstrating the limitations of logic when applied to incomplete data". One could almost say that I have a campaign about it, a sandbox I stand on. Because having your heroes come to the correct conclusion in a straight line without any detours is:

• unrealistic
• misleading
• too easy

Some authors do realise that it is too easy, but they compound the problem by adding in unrealistic roadblocks to prevent Our Heroes from solving the mystery too quickly -- when they don't need to do that. The author is still behaving as if problem-solving is a straight line.

It's unrealistic because in real life, mystery-solving does not happen that way. And I don't mean just Mystery mystery solving, I mean all kinds of situations and professions that have to figure out things and find answers: doctors with their diagnoses, programmers with their bug fixes, scientists with their hypotheses, and so on.

So how does mystery-solving/problem-solving get done? Speaking as a computer programmer with decades of experience, I'm confident that the process is similar no matter what the specific problem-domain is.

1. Gather the initial data. Define the problem.
2. Form a hypothesis, first considering (a) problems you yourself have seen before, then (b) problems your colleagues have seen before, then (c) problems that The Research has seen before, then (d) when every other possible lead is exhausted, propose something novel and improbable but not impossible.
3. Test that hypothesis by gathering more data.
4. If it doesn't match, go to step 2. If you are certain you know what the problem is, go to step 5.
5. Take the actions you need to do to solve the problem you think you have.
6. Check whether the problem is actually solved.
7. If the problem is not solved, go to step 2.

What makes me the most irritated about the no-detours-problem-solving trope is that it misleads people into thinking that that is the way it is in Real Life, that the first answer is always the correct one, that there aren't other possibly-correct answers, and that they don't have to check whether the answer they found is incorrect. I find it most irritating in computer bug fixing because that's my profession; oh those yahoos who think that because they have a hypothesis and made a change, therefore the bug is fixed, even though they didn't test whether the bug was fixed by having a before-and-after (before being "yes, I can see the bug happening here" and after being "lets see if what I did before to trigger the bug happening will trigger the bug or not").

Yeah, a sandbox.

A related trope is "This weird thing is completely unlike anything I've ever seen before!" Dude, if you're a human being, you're never going to say that, because humans find patterns, and you are going to make whatever-it-is connect to something you've seen before. And if it does actually not resemble anything you've ever seen before, that makes the analysis easier because there are a whole slew of possibilities you don't have to spend your time eliminating.

So I find a mischievous delight in uprooting people's expectations about problem-solving. I like to subvert the dominant paradigm, that linear trope. So beware, for I may make you trip over your assumptions! Here is my blow in the war against confirmation bias! (And if there isn't a war against confirmation bias, there ought to be.)

---

So, writers, what are your favourite things to do in your writing?

Comments

[reynardo]

Subvert the tropes.

Arranged marriage? He respects her, she is looking forward to life away from her controlling parents and he's going to give her the opportunities she craves, they're both committed to the relationship.

Girl disguised as a boy? She's not tomboyish - she still loves dresses and sewing and pretty things, even if she has to keep them hidden.

Historical setting illegitimate child? The mother is given an allowance, the child is acknowledged, and even if it can't inherit, it's sent to good schools and given every opportunity to do well in life.

(I gotta get back to the editing.)

[kerravonsen]

(high fives) Yeah, subvert the tropes!

[watervole]

Use the canonical abilities that often get overlooked.

Put the characters in a situation where both choices are bad.

[kerravonsen]

Use the canonical abilities that often get overlooked.
Yeah, that's a good one. Digging into the overlooked worldbuilding that is actually canonical.

Put the characters in a situation where both choices are bad.
Ouch. But also good.
I remember one writing maxim that says: Think of the worst possible thing that could happen to your characters, and then do it to them.