The outline is the most widely taught structure for organising thought, and one of the most reliable ways to prevent it. Hierarchies require categories before understanding is complete; they force ideas into trees when ideas actually grow in networks; they reward the appearance of order at the cost of the associative, generative, non-linear movement that thinking genuinely requires. The problem is not that outlines are useless. The problem is that we reach for them too early, and that reaching for them too early changes the kind of thinking we can do.
I used to write detailed outlines before writing anything substantial. Roman numerals, indented bullet points, sub-sub-headings, the works. I was taught this method formally, reinforced in it by years of education that rewarded structured submissions, and I had come to regard the outline as the responsible first step — the evidence that you had thought before you wrote, the scaffold on which good writing would be built.
Then I had to write something that I did not yet understand.
I could produce an outline — I always could — but the outline, in this case, was not a record of thought I had already done. It was a projection of structure onto territory I had not yet mapped. The categories looked plausible. The headings had a logical sequence. And each time I tried to actually write within those categories, I discovered either that the category did not contain what I thought it contained, or that what I was writing wanted to belong in a different category than the one I had assigned it, or — most often — that the connections between the ideas I was trying to articulate cut across the hierarchy in ways that the hierarchy made it impossible to express.
The outline was not helping me organise my thinking. It was requiring me to perform organisation I had not yet achieved, and the performance was preventing the thinking from which the organisation might eventually have emerged.
What Hierarchies Demand
The fundamental operation of an outline — of any hierarchical structure — is categorisation. To place something in a hierarchy, you must decide where it belongs: which node it falls under, which level it occupies, which items are coordinate with it and which are subordinate. This decision requires that you already have a classification scheme — that you already know, at least provisionally, what kinds of things these are and how they relate.
This is not a problem when the categorisation is the output of prior thinking. A writer who has already worked through a topic, who understands the distinctions that matter, who knows which ideas are subordinate to which other ideas — that writer can produce an outline that genuinely reflects their understanding. The hierarchy, in that case, is earned. It maps real conceptual structure that exists prior to and independent of the outline.
The problem is that outlines are almost universally taught and used as inputs to thinking rather than outputs from it. The student is told to outline before they write. The knowledge worker is taught to structure before they begin. The implicit model is that the outline is where you do the organisational thinking so that the writing can focus on expression. But this model assumes that the organisational thinking can be done before the ideas themselves are fully developed — which is precisely the assumption that breaks down when the material is genuinely difficult.
With difficult material — which is to say, with the material that most benefits from systematic thinking — the organisational structure is not known in advance. It is discovered through the process of engagement. You do not know which category an idea belongs in until you have developed the category far enough to see whether the idea fits. You do not know which ideas are coordinate and which are subordinate until you have worked out the relationships between them. Imposing a hierarchy before this work is done does not accelerate the work; it forecloses it.
How Ideas Actually Connect
Niklas Luhmann, the German sociologist who developed the Zettelkasten system and used it to write more than seventy books and hundreds of papers over forty years, described his note-taking practice in terms that are conspicuously non-hierarchical. His system did not use folders, categories, or any form of hierarchical organisation. It used connection — each note linked to other notes by reference, forming a network rather than a tree.
Luhmann's insight was that knowledge does not have a natural tree structure. Trees are imposed on knowledge; they are one way of organising it that privileges certain relationships — containment, subordination, classification — at the expense of others: association, analogy, contrast, application. Real understanding involves all of these relationships, and a system that can only represent containment relationships is going to distort whatever it is used to represent.
The philosopher Gilles Deleuze and his collaborator Félix Guattari used the concept of the rhizome — the underground root system of plants like grass, which grows laterally and has no centre, no fixed origin, no single trunk — as a metaphor for how knowledge actually structures itself. Unlike a tree, which has a root, a trunk, branches, and leaves in a fixed hierarchy, the rhizome grows in all directions, forms connections at any point, and cannot be represented by a single unified structure. Ideas, they argued, grow rhizomatically. The tree is always an imposition; the rhizome is always the reality underneath.
This is not mysticism. It is a description of something most people have experienced: the way that reading one thing calls to mind something else you read years ago; the way that a concept from one domain suddenly illuminates a problem in a completely different domain; the way that the most interesting intellectual progress often happens not within a category but across categories, through connections that no hierarchical system would have placed near each other.
The Map Drawn Before the Territory
There is a metaphor that I find more practically useful than the rhizome, because it describes not just how ideas are but how the process of understanding them works.
An outline is a map drawn before the territory has been explored. It tells you, in advance, what you will find and where you will find it. This can be useful if you are following a route you have already travelled — if the terrain is familiar and the map represents real knowledge of what is there. It is actively harmful if you are exploring terrain that is genuinely new, because the map you draw in advance will not match the terrain you encounter, and the commitment to the map will prevent you from seeing the terrain clearly.
Explorers in unmapped territory do not produce maps before they travel. They produce maps from what they find — from notes taken in the field, corrected as understanding develops, revised when the territory turns out to be different from what the previous day's observation suggested. The map is the output of exploration, not its prerequisite.
The structural equivalent of this in intellectual work is writing before outlining — or more precisely, writing as a mode of thinking rather than as the execution of a plan produced before writing began. This is what the practice of freewriting, or what Ahrens calls "writing from the slip-box," is actually doing: using the act of writing to generate discovery rather than to execute a plan. The constraint of the outline, applied early, prevents discovery by requiring you to stay within the categories you have already defined.
What Hierarchical Tools Do to Hierarchical Thinkers
The tools we use for thinking are not neutral. They afford certain kinds of cognitive operations and constrain others, and which operations they afford and constrain shapes the thinking we are able to do with them.
This is an application of the observation, associated with Marshall McLuhan but older, that the medium of thought is not separable from the thought itself — that choosing a particular representational structure is not merely a choice about how to express ideas but a choice about which ideas are expressible and which connections are available.
A hierarchical outlining tool affords classification and containment. It makes it easy to place ideas into categories, to create subcategories, to reorder items within a level. What it does not afford — what most such tools actively obstruct — is the representation of connections that cut across levels: the relationship between an item at level three of branch one and an item at level two of branch four, the observation that two things filed in completely different parts of the hierarchy are actually expressions of the same underlying principle.
The thinker who works primarily in hierarchical tools gradually becomes a hierarchical thinker — a thinker for whom the available cognitive operation is classification, and for whom ideas that do not fit cleanly into a classification scheme are either distorted to fit or dropped. This is not hypothesis. It is the consistent observation of people who move between hierarchical and network-based tools: the experience of working in a network-based system (Roam, Obsidian, a well-maintained Zettelkasten) surfaces connections that were invisible in the hierarchical system, not because the connections were new, but because the hierarchical system made no provision for them and the thinker's habit had adapted accordingly.
The Premature Certainty Problem
There is another cost to outlining early that is distinct from the structural problem, though related to it. The outline creates a premature sense of certainty about the shape of what you know — and that certainty discourages the exploration that would reveal its own gaps.
When you have an outline, you have something that looks like a complete account. The headings are there; the sub-points exist; the structure implies that you know what goes under each heading and how each section connects to the next. The appearance of completeness is, precisely, an appearance. But appearances of completeness, even internal ones, reduce the motivation to look further, because the mind registers something like this is taken care of and redirects attention to what remains.
This is related to the Zeigarnik effect discussed in the context of the Collector's Fallacy — the tendency of completed or apparently completed tasks to release the cognitive resources that uncompleted tasks hold. An outline that looks complete functions like a completed task, even if the underlying thinking is full of gaps. The gaps do not announce themselves; the outline conceals them under the appearance of structure.
A practice that deliberately resists this premature closure is what some writers call the shitty first draft (Anne Lamott's term) or what others call exploratory writing: writing continuously without commitment to structure, following the thinking wherever it goes, explicitly refusing to impose organisation until the thinking has run its course. This produces material that looks disordered and is often disordered, but that contains the raw output of genuine engagement with the ideas — material from which organisation can be extracted rather than imposed.
Writing as Thinking, Not Execution
The most productive reframe, once you have accepted that outlines belong at the end rather than the beginning, is a shift in what writing is for.
Most instruction about writing treats it as execution: you have the ideas, and writing is the process by which you put them into communicable form. The outline is the plan; the draft is the execution of the plan. On this model, the hard work happens before writing begins, and writing is essentially a transcription process — a skilled one, involving choices about word and sentence and emphasis, but fundamentally a matter of expressing pre-formed thoughts rather than forming them.
The alternative model — associated with people as different as E.M. Forster ("How do I know what I think until I see what I say?"), the philosopher Peter Elbow, and the cognitive scientist Bereiter and Scardamalia — treats writing as constitutively thinking rather than as the expression of prior thought. On this view, you do not write to communicate what you have understood; you write to understand it. The page is where the thinking happens, not where the thinking is reported.
This reframe has practical implications that are more radical than they initially appear. If writing is where the thinking happens, then a first draft that is disordered and contradictory is not a failed execution; it is early-stage thinking, and early-stage thinking is supposed to be disordered and contradictory. The disorder is not evidence of insufficient preparation; it is evidence that the preparation is happening. The contradiction is not a mistake to be hidden; it is the identification of a tension that needs to be resolved, which is exactly the kind of information you need in the early stages of engagement with difficult material.
The outline, applied at the beginning, eliminates this kind of information before it can surface. If you have a heading and you write under the heading, you produce material that fits the heading, because the heading is a commitment that shapes what you notice and what you write. The contradictions that would have appeared — the places where the heading does not fit, where the idea wants to go somewhere else, where two things that seemed separate turn out to be the same thing — are suppressed by the premature structure. You get coherence before you have earned it, and the coherence conceals the understanding you have not yet achieved.
Where Outlines Belong
I do not want to leave the impression that outlines are never useful, because that is not the argument. The argument is more specific: outlines are useful at the output end of thinking, not at the input end.
When you have worked through a topic sufficiently — when you have written exploratorily, made connections, revised your understanding, resolved the main tensions — an outline of what you have come to understand is genuinely valuable. It allows you to see the overall structure, identify what is missing, check the logic of the sequence, and plan the communication of ideas to a reader who has not been through the same process. This use of the outline is legitimate and useful.
The problem is the reflexive use of the outline as the first move — as the starting condition rather than the eventual output. The question is: what do you actually know at the moment you sit down to outline? If the answer is "I have already worked through this material substantially, and I understand the structure of what I want to say," then outline away. If the answer is "I am about to begin engaging with this material," then the outline is not a tool for thinking; it is a substitute for it.
This distinction requires a kind of self-honesty that is uncomfortable to maintain, because the outline always looks like thinking. It produces something visible. It has the structure of intellectual organisation. It can be shared, evaluated, and praised as evidence of a clear mind. The exploratory mess that actually precedes it — the freewriting, the false starts, the connections noted and then revised — is invisible and unpresentable and cannot be mistaken for finished thought.
But the mess is where the thinking is. The outline is, at best, a record of it.
An Honest Personal Practice
My writing practice, now, starts with a rule I break constantly and return to repeatedly: do not outline until you have something to outline.
What this looks like in practice is a period of writing that is genuinely exploratory — that does not know where it is going, that follows ideas rather than directing them, that produces material which is often unusable but which contains, somewhere in it, the actual movement of thinking that the eventual piece will be built from. This period is uncomfortable because it does not produce anything that looks like progress. The document fills with false starts and unresolved tensions and things that seemed important and turned out not to be.
After enough of this — after enough engagement with the material that the shape of what I actually think has begun to emerge — the outline becomes useful. Not as a structure to fill in but as a way of seeing what has already been produced: which ideas are doing real work, which can be cut, where the argument needs to be strengthened, what the sequence should be.
This is a slower process than outlining first. It produces more waste. It is frequently frustrating in ways that structured approaches are not. It also, in my experience, produces significantly better thinking — not because the process is inherently superior, but because it prevents the single most common failure mode in intellectual work: the substitution of structure for substance, and the confident production of organised ideas that were never actually understood.
The outline's great deception is that it looks like the result of thinking when it is not yet thinking at all. Structure is the output of understanding, not its precondition. Build it last, from what the thinking has actually produced — and resist, for as long as you can bear to, the temptation to impose order before the territory has revealed its own.