Why doesn't writing it all down work?
Because the knowledge that matters most can't survive the trip through writing. Documentation fails on expertise for three structural reasons: the expert cannot see his own tacit knowledge (so he can't write what he doesn't know he knows), writing captures conclusions while the value sits in reasoning, and documents freeze at their write date while the knowledge keeps moving. None of these are effort problems — more writing time doesn't fix them.
Reason one: the expert is blind to his own expertise
Ask a veteran to document 'how he assesses a supplier' and he will produce a sensible checklist — and apply forty other criteria he doesn't mention because they don't feel like knowledge to him; they feel like common sense. This is the structure of expertise itself: it automates. The philosopher Michael Polanyi's summary — we know more than we can tell — is a limit on self-documentation, not on articulation: the knowledge does come out, but through questioning in context, not through a blank page.
Reason two: writing keeps the what and drops the why
Documents are conclusions: do this, avoid that, use these settings. The expertise is the reasoning underneath — what was weighed, what almost went wrong, when the rule doesn't apply. Writing strips it for honest reasons: the reasoning is long, feels self-indulgent to write, and no template has a field for it. The result is documentation that works until the first situation it didn't anticipate — which is exactly when expertise was needed.
Reason three: documents freeze, knowledge moves
A documentation project produces a snapshot; the expert keeps learning. Within a year the snapshot and reality diverge, silently — the document doesn't know it's stale, and readers can't tell which parts still hold. Maintaining documentation against a moving expert is a race that writing loses structurally, because every update costs the same scarce writing time that made the backlog in the first place.
What works instead
The failures point at their own fix: capture from speech instead of writing (the expert explaining a real case articulates what a blank page never gets), capture reasoning with its context and sources, and capture continuously instead of in a project. This is not an argument against documents — procedures and specs belong in documents — but the judgment layer needs a different instrument. The comparisons spell it out: documentation project vs knowledge graph and SOPs vs a living knowledge base.
Where MentX comes in
MentX is a personality-aware knowledge base: it captures the knowledge of a company's key person — from meetings and a guided intake track — into a living, temporal knowledge graph with full source attribution. MentX is that different instrument: capture runs on speech — the expert's meetings (Connect) and guided intake sessions (Deepen) — so the tacit layer surfaces through questioning in context. The reasoning lands verbatim in a living, temporal knowledge graph with full source attribution: the why stays attached to the what, and the dates keep stale and current distinguishable.
