LYT and maps of content, structure you discover

Linking Your Thinking, usually shortened to LYT, is a practice and a set of workshops developed by Nick Milo for people keeping notes in linked tools such as Obsidian. It grew out of the Zettelkasten tradition and shares its commitment: the value of a note collection is in the connections, not the filing. Where it departs is on navigation. A few hundred densely linked atomic notes are a genuine thinking tool; a few thousand are a haystack, and Luhmann's answers to that (branching numbers, index cards pointing at entry points) do not translate cleanly to software.

LYT's answer is the map of content, or MOC: an ordinary note whose body is mostly links to other notes, written to be a working surface for one subject. It is not a category and it is not a container. Nothing moves into a MOC. The notes stay exactly where they are, and the map is one more note that happens to point at them.

The philosophical claim underneath is what makes LYT different from the folder-first and project-first systems around it. Structure imposed at the start encodes what you knew on day one. If you create a folder called Pricing before you have thought about pricing, you have committed to a boundary drawn in ignorance, and every note you file into it afterwards confirms that boundary rather than testing it. LYT holds off: you capture, you link, and the shape of your interests becomes visible in the linking. That is a real trade, since deciding late means living with mess in the meantime, and people who need a tidy vault to feel calm find the first few months uncomfortable. What you get for it is an organisation that describes your actual thinking rather than your predictions about it.

A MOC is a note. That is the whole technical description: it lives beside every other note, can be linked to and from, shows up in search, and can be deleted without losing anything except itself. What distinguishes it is the body: links, grouped under a few headings, with a sentence of commentary wherever a link needs one.

A short one might be titled Pricing. Under What we believe sit four links to notes arguing positions; under Evidence, six links to notes recording what customers said and what experiments returned; under Open questions, two links and one line of plain text for a question that has no note yet; under Elsewhere, links to the positioning and churn maps, because the subject bleeds into both.

Three things follow from a map being an ordinary note rather than a special object.

• A note can appear on any number of maps. There is no membership to violate. A note about a customer's reaction to a price change sits on the pricing map and the customer-research map at once, and neither has a claim on it.
• The order is yours. Folders sort alphabetically and search sorts by relevance, but a map is hand-arranged, so the sequence can carry an argument: here is the position, here is what supports it, here is what contradicts it.
• You can write on it. A bare list of links is a directory; a list where each entry says why it is there is a piece of thinking, and it tends to become the first draft of whatever you eventually write on the subject.

Maps also nest, mildly. A large subject grows sub-maps and the parent points at them, and most people keep one top-level note (Milo calls it a home note) pointing at the maps they use often. That is usually as much hierarchy as the system acquires, and because it is built from links rather than folders, rearranging it costs one edit.

The reason LYT tolerates the mess is a claim about how ideas arrive. An outline written in advance can only contain what you already knew, and its branches keep material apart by design. Ideas, on Milo's account, grow laterally: a note written for one reason answers a question raised somewhere unrelated, and the collision is the interesting result, which an outline would have prevented by filing the two in different sections.

Emergence in this sense is not mystical. It is the observable fact that in a well-linked collection you keep finding pairs you did not put together deliberately. Luhmann described his Zettelkasten as a communication partner for the same reason, and Andy Matuschak's evergreen notes make the same bet from another angle, insisting on dense linking and on rewriting notes over time rather than finishing them. Both are covered in Zettelkasten and evergreen notes.

What LYT adds is the map as the place where emergence gets consolidated. Building one means laying twenty notes side by side and arranging them, which is a synthesis exercise disguised as tidying. The groups you invent while arranging are usually the actual insight: you discover that six of your twenty pricing notes are really about willingness to pay and the other fourteen are about packaging, and that distinction was nowhere in writing until the map forced you to sequence things.

The caveat is that emergence is neither free nor fast. It needs volume (a hundred notes is where most people start seeing it), notes written in your own words rather than pasted, and a habit of revisiting them. A collection you only ever write into behaves like a landfill however well it is linked.

The differences are practical rather than semantic. A folder is exclusive: a file sits in one, and moving it changes where it is. A tag is inclusive and flat: it marks any number of notes, but cannot be ordered or carry commentary, and returns a machine-generated list rather than an arrangement. PARA sorts by actionability, a genuinely different axis: Tiago Forte's four buckets ask how soon a note will be used, not what it is about.

The last row resolves most arguments. These devices are not competitors, and a common working setup uses several: PARA folders so that active work is separable from reference, tags for cross-cutting states such as unread or unresolved, and maps for the handful of subjects you actually think about. Milo's objection to folder trees is not that folders are bad, it is that a tree is a poor model for a network: anything organised exclusively by tree becomes unreachable from every angle except the one you chose. A rigidly numbered scheme has the same virtue and the same limit, which is why Johnny Decimal is usually recommended for archives rather than live thinking.

Tags are the closest comparison, since both are non-exclusive and both give you a list. The difference is that a tag list is generated and a map is written, so the map can say read these three first, the rest is background. If your tool supports saved queries (Obsidian's Dataview plugin is the usual example), the sensible pattern is a hand-written map with one generated block inside it rather than a choice between the two.

Four failure modes account for most abandoned LYT vaults.

• Maps before notes. Building empty maps on day one is the folder mistake wearing a new hat, and the most common way to get LYT exactly backwards. A map with three links does nothing a search could not.
• Maps of maps of maps. Nesting is allowed and easy to overdo. Two levels deep and still not at a real note means you have rebuilt a folder tree out of links, now invisible to your file browser as well.
• Bare link lists. A map with no commentary ages badly, is never re-read, and produces none of the synthesis that justifies the exercise. If you will not write the annotations, a saved search is cheaper and just as useful.
• Curating instead of thinking. The seductive one. Arranging notes feels productive and can absorb an evening without producing a sentence anybody will read. The test is whether a map ever leads to something at the creating level; if none do, the system has become a hobby.

A tool question sits underneath all of this. LYT assumes bidirectional links and fast search over plain text files. Obsidian is where the practice grew up and has the largest stock of map templates, but nothing here is Obsidian-specific: Logseq, SiYuan and several others support the same moves, and the trade-offs are compared in Obsidian vs Logseq vs SiYuan. What matters more than the choice is that notes are plain files you can still read when the tool changes, and that links are cheap enough to make casually. If a link costs three clicks, you will not make enough of them for anything to emerge.