Split notes at 800 words or 3 topics — decomposition reveals hidden causal chains
When a note exceeds 800 words or covers three distinct topics, decompose it into 2-4 separate atomic notes and rewrite the connections between them to reveal causal chains invisible in the original structure.
Why This Is a Rule
Long notes hide structural relationships inside prose. A 1,200-word note about "team productivity" contains claims about meeting culture, deep work scheduling, and tool selection — three distinct ideas with causal relationships between them (meeting culture → fragmented time → reduced deep work → compensatory tool adoption). Inside the monolithic note, these causal chains are implicit. Decomposed into separate notes with explicit links, they become visible, testable, and reusable.
The 800-word threshold is a practical heuristic: below this length, a note is likely atomic or nearly so. Above it, the note almost certainly contains multiple ideas. The three-topic rule catches short-but-compound notes that pack multiple ideas into fewer words.
The act of decomposing and rewriting connections is where the cognitive value lives. Splitting a note forces you to decide: what is the core claim? What is supporting evidence? What is a separate argument? These decisions, deferred inside a monolithic note, produce clarity when confronted during decomposition.
When This Fires
- During knowledge base review when you encounter long notes
- When a note feels relevant to multiple different contexts (it's probably compound)
- After capturing a dense meeting or reading session that produced a long note
- Any note that you've struggled to link precisely because "it's about several things"
Common Failure Mode
Splitting mechanically by paragraph rather than by idea. A note split at paragraph boundaries produces fragments, not atoms. The split points must be at idea boundaries — where one claim ends and a different claim begins, even if they occupy the same paragraph.
The Protocol
When a note exceeds 800 words or covers 3+ topics: (1) Read the note and identify the distinct ideas it contains (usually 2-4). (2) For each idea, create a new atomic note containing only that idea, written to be self-contained. (3) Write explicit links between the new notes: "A causes B," "C is evidence for D," "E contradicts F." (4) Delete the original compound note or convert it to a hub that links to its children. (5) The explicit links between decomposed notes are new knowledge — causal chains that were invisible inside the monolith.