Sherlock Holmes Called It Deduction

You remember Sherlock Holmes. Deduction, my dear Watson. Except it wasn't deduction. And if we're being precise about it, neither is most of what we call reasoning. Sherlock Holmes is supposed to be the smartest reasoner in the room. There's just one problem. The thing he's famous for, deduction, isn't what he's actually doing. And that mislabel has been hiding something important about how reasoning actually works. So let's fix it. Holmes observes. A tan line. A posture. A callus in exactly the right place. And from those specifics he builds toward a conclusion. You've been in Afghanistan, I perceive. That's not deduction. That's induction, evidence first, conclusion emerging from it, always provisional, always technically open to revision. It's impressive. It's rigorous when done well. But it is not deduction. Deduction is something much more specific. All men are mortal. Socrates is a man. Therefore Socrates is mortal. You cannot accept those premises and reject that conclusion. The certainty is built into the structure. It doesn't come from observation, it comes from logic. And it only works in environments where the premises are agreed upon, the terms are precisely defined, and the rules are shared. Formal logic. Mathematics. Certain kinds of structured philosophical argument. It is a specialist tool. Powerful in the right conditions. Useless outside them. Which is why it almost never applies to politics. Politics doesn't give you agreed premises. It doesn't give you precisely defined terms. It gives you incomplete information, motivated actors, contested definitions, and systems too complex for any single mind to hold. In that environment, which is the actual environment of political life, deduction doesn't work. It can't. The conditions it requires don't exist. What does work, when it's done well, is induction. Evidence accumulates. A provisional conclusion emerges. You hold it loosely. You stay open to revision. You don't mistake the conclusion for certainty just because it feels that way. That last part is where it gets hard. The psychologist Daniel Kahneman spent decades studying two modes of thinking that every human mind runs on. Not two types of people, two modes inside the same person. There's the fast mind, automatic, pattern matching, confident, built for speed. And there's the slow mind, deliberate, effortful, structurally careful, built for complexity. The fast mind is not a flaw. It handles an enormous range of ordinary decisions efficiently and well. The problem isn't that we have it. The problem is that we use it in conditions that require the slow mind, and we usually don't notice when we've made that switch. Because the fast mind doesn't feel fast. It feels certain. And that's where confabulation enters. Confabulation isn't lying. It's not even conscious. It's what happens when the conclusion arrives first, through instinct, identity, habit, or fear, and the mind goes to work assembling reasons to support it. The reasoning looks real. It follows a structure. It cites evidence. It feels, from the inside, exactly like thinking something through. But the direction is reversed. The conclusion wasn't reached. It was assumed. And the reasoning was recruited to defend it. This is what Holmes is actually doing in the reveal scene. Not in his investigation, the stories imply genuine rigor there. But in the moment he announces the answer and walks you backward through his reasoning, the dramatic structure is confabulation shaped. Conclusion first. Justification arranged beneath it. It's satisfying precisely because it mimics the feeling of airtight logic without requiring the process. We find that satisfying because our fast minds recognize it. It feels like how reasoning should look. And so we model on it, without realizing that what we're modeling is the performance of reasoning, not the thing itself. In politics, that gap is where a lot of confusion lives. And it's where a lot of manipulation enters. Because if you can trigger a fast mind conclusion, through fear, through identity, through a sufficiently confident performance, the slow mind often gets recruited to defend it rather than examine it. The reasoning comes after. And it feels genuine, because from the inside, it is. So what does the alternative actually look like? It looks like induction done honestly. Evidence first. Conclusion held loosely. Genuine openness to revision when new evidence arrives. Comfort with uncertainty that doesn't resolve cleanly. The willingness to say, I think this is probably true, based on what I can see, and I could be wrong. That's not weakness. That's the appropriate cognitive tool for the actual complexity of political life. It's what Holmes was doing in his best moments, whatever he called it. The label was wrong. The instinct, when followed carefully, was right.