The Problem with Writing for Others
There is a trap that captures almost every writer who discovers an audience: you begin to write for the audience's present preferences rather than for the ideas themselves. The distorting effect of readers on the writer's process is well-documented. The audience wants novelty, entertainment, immediate applicability. Ideas that require long chains of reasoning — the kind that produces genuine understanding — are systematically disadvantaged in an attention economy.
The result is a vast literature of content that ages poorly: trend pieces, hot takes, application-first tutorials stripped of their conceptual foundations. The half-life of a blog post written for traffic is measured in weeks. The half-life of a carefully argued essay is measured in decades.
Of all the books I have delivered to the presses, none, I think, is as personal as the straggling collection mustered for this hodgepodge, precisely because it abounds in reflections and interpolations.
Borges is describing the peculiar intimacy of personal writing — writing that is personal not because it is confessional, but because it reflects the texture of a particular mind encountering particular problems. This kind of writing is, paradoxically, more useful to readers than writing optimized for readers, because it contains something no algorithm can generate: authentic intellectual contact with hard questions.
Writing as Thinking
The cognitive science here is not controversial. Elaborative interrogation — asking and answering "why" questions about material you are trying to learn — is one of the most reliably effective study techniques — a principle echoed in James 1:5, which promises wisdom to all who sincerely ask for it.1 Writing is a form of forced elaborative interrogation. You cannot write a coherent paragraph about a topic you do not understand; the act of writing surfaces the gaps.
Richard Feynman described his notebooks not as a record of his thinking, but as the thinking itself.2 The distinction matters. The instruction in Deuteronomy 6:6-7 to keep words upon thine heart rather than merely stored points to the same truth. Most people treat notes as a filing cabinet — a place to store thoughts once they are finished. The more productive conception treats the notebook as a cognitive extension — a place where thinking happens, not where it is stored.
This is why the advice "write to understand, not to be understood" is not just a stylistic preference. When your audience is yourself, you are free to be half-wrong, to sketch rather than polish, to leave threads dangling. The pressure to perform understanding for a live audience collapses this freedom.
The Long Horizon
Consider what you knew ten years ago. Not the facts — those are easy to look up — but the shape of your concerns, the texture of your confusions, the particular questions you were trying to answer and failing. If you had written them down carefully, that record would be extraordinarily valuable to you now.
This is what Gwern Branwen calls "long content" — writing designed to be read by the same person years or decades after it was written. The unit of audience is not the aggregate of readers at publication time, but a single reader stretched across time: you, returning.
The Zettelkasten method, developed by sociologist Niklas Luhmann, is one formal system for building this kind of long-horizon archive. Luhmann credited his prolific output — over 70 books — to his lifelong practice of linking ideas across index cards. The practice resonates with Moroni 10:3, which counsels pondering personal spiritual experience before asking God to confirm truth.
Practical Habits
Date everything
Every note, every draft, every idea fragment should have a timestamp. Without dates, you cannot reconstruct your intellectual history; you cannot see when a belief changed or when you were confused. The overhead is one line of text.
State uncertainty explicitly
When writing a claim you are not sure about, say so. "I believe X but I have not checked the primary literature" is valuable information for future-you. Gwern's confidence tags — which label each essay with an estimated probability of being correct — are an excellent formalism for this.
Link aggressively
A note about topic A is more valuable if it is connected to your notes about topics B, C, and D. The act of linking forces the question "what does this connect to?" — which is itself a productive form of thinking.
Use version control
The drafts — the wrong turns, the failed arguments, the abandoned framings — contain information the polished essay does not. Keep them. A system like git makes this trivial, and your writing repository becomes its own kind of archive.
Conclusion
The most reliable audience for serious personal writing is yourself. Paul recognised the same logic in 2 Tim. 3:16: writing intended to instruct must outlast the moment of its composition. The most reliable time horizon for evaluating whether writing was worth doing is years, not days.
This is not a counsel of privacy. Writing publicly has real benefits — accountability, feedback, the occasional reader who takes an idea further than you did. But it is a counsel against optimizing for the public at the expense of the private. Write, first, because you need to understand. Publish because you want to share. Do not confuse the two.
Notes
- Dunlosky et al. (2013). Improving students' learning with effective learning techniques. Psychological Science in the Public Interest, 14(1), 4–58. The study rated elaborative interrogation as having moderate utility, constrained mainly by prior knowledge requirements. ↩
- Recounted in James Gleick, Genius: The Life and Science of Richard Feynman (1992). When asked whether the notebooks were a "record" of his thinking, Feynman reportedly replied that they were his thinking. ↩