Getting the Words Right

There is a long-standing tension in how we represent work. One side values local, situated understanding. The other values clarity at a distance.

James C. Scott uses the term métis to describe the first: practical knowledge, built through experience, shaped by context, and constantly adapting. He contrasts it with efforts to make systems legible: simplified, standardized, and easy to read, measure, and control.

Legibility

Legibility has a clear appeal.

It allows:

• coordination across teams
• reporting and planning
• systems to be understood by people who are not directly involved

But it comes with a consistent move: it reduces complexity in order to make things visible.

That reduction is not neutral. It selects:

• what counts
• what gets named
• what gets ignored

Once something is made legible, it tends to be treated as if it is complete.

Scott's critique is not that legibility is bad. It is that legibility often replaces reality with a simplified version of reality, and then operates on the simplification as if it were the real thing.

You see this in:

• standardized categories
• clean hierarchies
• formal plans
• system-of-record artifacts

Métis

Métis works differently.

It lives in:

• judgment calls
• adjustments in the moment
• shared understanding within a group
• things that are obvious locally but hard to explain globally

It does not try to eliminate ambiguity. It works with it.

In practice, much knowledge work sits much closer to métis than organizations often admit.

• priorities shift as new information comes in
• definitions evolve mid-stream
• the meaning of the work changes as it unfolds

Across Traditions

The same distinction appears in several adjacent traditions, even if they use different language.

Across these traditions, the recurring pattern is similar: legibility makes the system easier to see, while métis helps people cope with the reality the simplified view leaves out.

The Mismatch

The systems we use are overwhelmingly optimized for legibility. They assume:

• stable categories
• clear boundaries
• definitions that hold over time

That mismatch is where many familiar problems come from.

When legibility dominates too strongly:

• artifacts start to drift away from reality
• a ticket says one thing, but everyone "knows" what is really going on
• a status is technically correct but practically misleading
• a plan is clean, but the work behind it is anything but

At that point, the artifact is no longer a reliable representation. It becomes a simplified surface that requires interpretation.

Compensation

Teams respond in predictable ways. They do not abandon the system, but they stop relying on it alone.

They:

• restate things in multiple places
• add context in conversations
• maintain parallel documents
• re-explain what something "actually means"

This is not redundancy. It is compensation. They are rebuilding the context that legibility stripped away.

That is the part that is easy to miss. The issue is not just that legible systems are incomplete. It is that the more you rely on them as complete, the more you have to recreate reality somewhere else.

That "somewhere else" is where métis lives:

• in conversations
• in evolving documents
• in shared but unwritten understanding

A Practical Contrast

Both are necessary, but they are not interchangeable.

A Simple Test

The failure mode is not using legibility. The failure mode is treating legibility as if it were sufficient.

When that happens:

• systems look clean
• coordination appears smooth
• but the real work moves somewhere else

And once that gap opens, it tends to grow, because each additional layer of simplification creates more distance from the underlying situation.

A simple test:

• if you removed all conversations, would the artifacts still make sense?
• if you handed the system to someone new, would they understand what is really happening?

If the answer is no, you are relying on métis whether you acknowledge it or not.