The Writing as Thinking Test
You do not know what you think until you try to write it down
The Writing as Thinking Test is Paul Graham's framework establishing that writing is not the act of expressing pre-formed ideas but the primary mechanism through which ideas become fully formed. Graham argues that putting ideas into words is a severe test -- the first words you choose are usually wrong, you have to rewrite sentences repeatedly, and half the ideas in a finished essay are ones you discovered while writing it.
The key insight extends further: if writing always makes ideas more precise and complete, then anyone who has not written about a topic does not have fully formed ideas about it. And someone who never writes has no fully formed ideas about anything nontrivial. Ideas can feel complete inside your head -- it is only when you try to put them into words that you discover they are not. If you never subject your ideas to that test, you will not only never have fully formed ideas, but also never realize it.
This framework matters because it transforms writing from a communication activity into a thinking activity. Writing is not how you share what you know -- it is how you discover what you know.
- Writing is not expressing pre-formed ideas -- it is the process through which ideas become fully formed.
- Half the ideas that end up in an essay are ones you thought of while writing it.
- Anyone who has not written about a topic does not have fully formed ideas about it.
- Writing is a stricter test than talking because you must commit to a single optimal sequence of words.
- Choose a Topic You Think You UnderstandSelect a subject you believe you know well and commit to writing a thorough explanation of it. Graham deliberately writes about subjects he knows well -- Lisp hacking, startups -- because the learning that emerges from writing about familiar topics proves the framework's power. If you can learn new things by writing about something you already know deeply, imagine how much you can learn by writing about something you only partially understand. The goal is not to produce polished prose but to discover where your understanding has gaps.Pro tipStart with your most confident area of expertise. If writing reveals gaps there, it will convince you of the method's power for less familiar topics.
- Apply the Stranger TestAfter writing, pretend to be a neutral reader who knows nothing of what is in your head -- only what you wrote. When this stranger reads your writing, does it seem correct? Does it seem complete? Graham describes this as the core test, and notes that when you make the effort to read as a complete stranger, the news is usually bad. The stranger is rational and will identify missing explanations, unqualified claims, and logical gaps that felt complete in your head but are incomplete on the page.Pro tipRead your draft aloud as if you are reading someone else's work for the first time. The gaps become obvious when you hear them rather than see them.WarningThe stranger test will cost you nice sentences. You have to resign yourself to making them as good as possible while still satisfying the stranger's demands for completeness and accuracy.
- Iterate Until the Stranger Is SatisfiedRewrite repeatedly until the neutral stranger reader would find your explanation complete, correct, and clear. Graham often spends two weeks on an essay and rereads drafts 50 times. Each cycle of the stranger test reveals new gaps, new imprecisions, and new ideas that emerge from the discipline of making your thinking explicit. The process is not efficient -- it is thorough. The payoff is that by the end, your ideas are not just expressed but genuinely formed in a way they never would have been without writing.Pro tipSet a rule: do not share or publish anything until you can read it as a stranger and find no significant gaps. This discipline produces dramatically better thinking.
Paul Graham has written extensively about two subjects he knows deeply: Lisp programming and startups. Despite his deep expertise in both, he reports learning significant new things from the process of writing about them. Things he did not consciously realize until he had to explain them emerged through the discipline of putting ideas into words. If even domain experts discover new insights through writing, it demonstrates that no one's understanding is fully formed without writing.
Graham uses the movie trope of a character who claims to have a plan for something difficult. When questioned, the character taps their head and says 'It is all up here.' Graham notes that everyone watching the movie knows exactly what this means: the plan is vague and incomplete at best. Very likely there is some undiscovered flaw that invalidates it completely. At best it is a plan for a plan.
Paul Graham developed this insight through decades of writing essays on startups, programming, and philosophy. He has written about Lisp hacking and startups -- two subjects he knows well -- and in both cases learned a great deal from the process of writing about them. Things he did not consciously realize until he had to explain them emerged through writing. Graham observes that a great deal of knowledge is unconscious, and experts have an even higher proportion of unconscious knowledge than beginners. The 'stranger test' -- pretending to be a neutral reader who knows nothing of what is in your head -- became his core editing tool. He often spends two weeks on an essay and rereads drafts 50 times, noting that doing this in conversation would seem evidence of mental disorder.