Flavio Copes has a post on why he thinks more clearly when he types than when he dictates. He points out that writing means two things: capture (get the fleeting idea down before it evaporates) and development (turn it into something that actually holds up). He points out that voice dictation wins at the first and typing wins at the second because the friction of typing forces you to look at a sentence before you can meaningfully move past it.

How to Know If You Care How Fast You Type

WPM as a benchmark comes out of the typing pool: a room of secretaries whose job was to take handwritten (or shorthand, or dictated) pages and turn them into uniform typed copy. That was a commodity clerical job, and the speed mattered because speed was the whole job. There was no composition happening. The only decision left to the typist was how to format the page, and even that wasn’t really a decision. It was dictated by house style, the same way every letter in the pool had to look the same regardless of who typed it.

So WPM was built, from the start, to measure fidelity and throughput on a task with the thinking already removed. Applying it to writing where the thinking is the entire point isn’t just an imperfect proxy, it’s a category error. You’re importing a metric from a job that was explicitly designed to have no judgment in it, and using it to evaluate a job that’s only judgment.

Of course it measures the wrong thing. It was built to.

How to Distinguish between Typing and Voice Dictation

An even more interesting part of Copes’s post isn’t “typing vs. dictation,” it’s that he draws a line at all.

Most of the current advice about writing and thinking points toward dictation, toward talking it out, toward removing friction wherever you find it. The implicit assumption is that friction is at best overhead: something between you and the idea that a better tool should erase. Copes doesn’t argue that. He argues that friction is crucial in one of the two jobs and useless in the other. Capture wants zero friction, development wants exactly the kind of friction that makes a weak sentence visible before it’s finished. That’s a much more useful claim than “typing is better,” because it tells you which tool to reach for based on what you’re actually doing, not on which one is more convenient in general.