Named by Gilbert Ryle (The Thinking of Thoughts, 1968) and developed into an ethnographic method by Clifford Geertz (Thick Description: Toward an Interpretive Theory of Culture, 1973).
Records what people do together with what their actions mean inside their own setting, producing an account an outsider can read and act on.
Field notes from one handover meeting, kept in three layers
Second-shift operators walk out of the daily handover early. The first note is placed as an example. Select each other note and pick its layer: what anyone could see, what it meant to the crew, or the observer's reading. One note cuts against the reading. Illustrative.
- Not placedPlacedSeenMeaningReading
- Not placedPlacedSeenMeaningReading
- Not placedPlacedSeenMeaningReading
- Not placedPlacedSeenMeaningReading
- Not placedPlacedSeenMeaningReading
- Not placedPlacedSeenMeaningReading
- Not placedPlacedSeenMeaningReading
- Not placedPlacedSeenMeaningReading
Field notes
Layer of the account
Your sort matched 0 of 8.Your sort matched 1 of 8.Your sort matched 2 of 8.Your sort matched 3 of 8.Your sort matched 4 of 8.Your sort matched 5 of 8.Your sort matched 6 of 8.Your sort matched 7 of 8.Your sort matched 8 of 8.
- Seen
- Behavior anyone in the room could record, with times and counts. No motive in it yet.
- Meaning
- What the action means to the people doing it, in their terms, from what they say and do.
- Reading
- The observer's interpretation, kept apart so it can be tested against the other two layers.
The sort
- Out at 2:52: Seen. Thin description: true, countable, and silent about why.
- A nod at the door: Seen. Still seen, not interpreted. It shows who sets the timing.
- Sheet already filled: Seen. Seen. It cuts against the idea that the crew skips the handover's content.
- The pour comes first: Meaning. Their meaning, in their words: leaving is how they show up for the work.
- Staying is disloyal: Meaning. Their meaning: the walkout is a crew norm, not a set of private choices.
- Timed against the pour: Reading. The reading the other two layers support. The fix is the schedule, not the crew.
- "They don't care": Reading. A reading built from the thin layer alone. The meaning layer contradicts it.
- The tech leaves too: Seen. The case that cuts against the reading: he leaves to beat shift-change traffic. Keep it in the account; the reading covers the crew, not everyone.
An example
The attendance report at a steel castings plant showed the second shift leaving the daily handover meeting early, every day, for three months. The plant manager read it as a crew that had checked out and asked HR for an engagement program. The HR partner asked for a week on the floor first.
She kept her notes in three columns. What she saw: six of eight operators standing at 2:52, after a nod from the furnace lead, with their handover sheet already filled in. What it meant to them, in their words: the 3:00 pour is when the shift starts, a late crew loses a whole heat, and staying behind while the crew leaves looks like siding with the office. Her reading: the meeting ran on the day shift's clock and collided with the crew's own. One case did not fit: the maintenance tech left too, to beat the traffic, and she kept him in the account. The meeting moved to 2:15, and attendance was full the next week.
Why it works
A report of behavior alone, people left at 2:52, invites the observer to supply the meaning, and the observer supplies their own. That is how a crew protecting the pour becomes a crew that does not care. Thick description adds the layer the thin record leaves out: what the action means to the people doing it, inside their own setting.
Keeping the layers apart is what makes the account usable. The seen layer can be checked by anyone. The meaning layer is sourced to what people say and do. The reading is the observer's own claim, so it can be tested against the other two and revised when a case does not fit. A manager can then act on the reading and know what it rests on.
The thinking behind it
Gilbert Ryle Gilbert Ryle, an Oxford philosopher of mind known for The Concept of Mind (1949), drew the distinction in The Thinking of Thoughts: What Is 'Le Penseur' Doing? (1968). A thin description of an action reports only the observable behavior. A thick description adds the context and intention that make the action what it is. Two acts can look identical in a thin description and be entirely different acts once the thick one is written.
Clifford Geertz Clifford Geertz, an American cultural anthropologist, treated culture as a web of meanings to be interpreted. In Thick Description: Toward an Interpretive Theory of Culture (1973), the opening essay of The Interpretation of Cultures, he took Ryle's term and made it the aim of ethnography: describing an action together with the context that gives it meaning, so that a reader from outside can follow what the people involved understood themselves to be doing.
How to run it
- Go where the behavior happens. Watch the action in its setting, more than once, before reading any report about it.
- Record what anyone could see. Write times, counts, who moved first, and what objects were in play, with no motive attached.
- Ask what it means to them. Talk to the people involved about the moment itself and write down their exact words.
- Keep the layers in separate columns. Seen, meant, and your reading go in different places so no one confuses one for another.
- Write your reading last. State the interpretation as a claim that the first two layers would support or contradict.
- Hunt for the case that does not fit. Look for someone who does the same thing for a different reason, and keep them in the account.
- Hand over all three layers. Give decision makers the evidence with the reading, so they can see what the reading rests on.
Where it goes wrong
The first failure is a meaning layer written by the observer. Notes that say the crew feels disrespected, with no quote or act behind them, are the observer's reading in the wrong column. Source every line in that layer to something said or done.
The second is a single visit. One day on the floor produces an anecdote. Return until new visits stop changing the account.
The third is dropping the case that does not fit because it spoils the story. The maintenance tech's reason is evidence that the reading has limits, and leaving him out makes it look stronger than it is. Report the exception with the reading.
When to use something else
- Empathy Map. When interview notes already exist and a team needs one shared picture of what a customer says, thinks, does, and feels.
- Theoretical Saturation. When the open question is how many observations or interviews are enough.
- Phenomenological Analysis. When the aim is to describe one kind of lived experience, such as being new on a team, from the inside.
Where it sits in SOLVE eX
Go to the source
- Thick Description: Toward an Interpretive Theory of Culture, The Interpretation of Cultures, 1973
- The Thinking of Thoughts: What Is Le Penseur Doing?, University Lectures, University of Saskatchewan, 1968