The science Cognitive Fingerprint™ was built on
Sixty years of research, five separate fields, one conclusion: your most valuable knowledge is the knowledge you can't put into words on your own. This page shows the evidence.
The named rule underneath the repeated move.
Five fields, working apart, one answer.
Between 1966 and today, five fields studied expert judgment without reading each other. They all reached the same conclusion: the best expert judgment runs on knowledge the expert cannot state.
Michael Polanyi in philosophy: "we know more than we can tell" (1966).
Hubert and Stuart Dreyfus on how people climb from beginner to expert (1986).
Harry Collins inside working laboratories (2001, 2007).
Ikujiro Nonaka and Hirotaka Takeuchi inside Japanese firms (1995).
Gary Klein with firefighters making life-or-death calls (1998).
Thinking is invisible until it is made visible.
A second body of work backs the fix. Engelbart, Victor, Ritchhart and Perkins, Klein, and Polanyi, working in different traditions, agree that thinking becomes directable only once it is made visible as an object. Cognitive Fingerprint™ is built on both findings: read the work, surface the pattern, hand it back as something you can use.
What the research supports.
Each claim below carries its sources. Nothing on this site goes beyond them.
Expert judgment is pattern recognition built over years in one domain, not generic intelligence. (Ericsson et al. 2006; Glaser 1989; Gobet 2015)
The more fluent you get, the less of your judgment is available as explanation on demand. (Polanyi 1966; Dreyfus 1986; Feldon 2007)
Asking someone to explain themselves is unreliable evidence. Explanations have to be checked against real decisions and real work. (Feldon 2007; Shanteau et al. 2001)
Structured extraction beats casual conversation: real incidents, repeated cases, evidence over introspection. (Hoffman 1996; Ericsson et al. 2006)
Tacit knowledge can be translated into usable form, but not flattened into rules. (Polanyi 1966; Collins 2001; Nonaka and Takeuchi 1995)
A rigorous reading separates what you know, how you act, when you switch strategies, why, and how you watch your own thinking. (Schraw 1998; McGuinness et al. 2016)
Experts differ in what they notice, what they ignore, and which standards they protect, not just in skill level. (Chabris 1999; Shanteau et al. 2001; Mieg 2009)
Once a pattern is named and evidenced, it becomes teachable, priceable, and hireable-against. (Nonaka and Takeuchi 1995; Davenport and Prusak 1998)
AI can read and cluster more of your work than any human ever could. Human judgment still decides what the patterns mean. (Lam 2022; Collins 2025)
What I refuse to claim.
Most tools in this space overclaim. The research draws real limits, and I keep them:
Not everything can be written down
Some of your skill transfers only by demonstration, and when I find a pattern like that, the deliverable says so.
Intuition is not always right
Patterns get validated against repeated real cases. That is why the evidence standard exists: three or more instances, across contexts, with an observable effect.
Never "scientifically proven"
I point at the research and let it argue.
Not a brain scan, not an assessment
Assessments ask what you think you are. I read what you have already done.
How an expert gets read.
The raw material is your real work: calls, decisions, writing, the way you solve problems out loud. The reading finds moves you repeat without noticing, names each one, and attaches the evidence. Every named pattern reaches four layers deep: the behavior, the mechanism, why you can't see it, and the belief driving it.
Keep moving through the method.