Cedric Chin, (Chin 2020a)


We can acquire the Tacit knowledge of experts through Naturalistic decision making: Cognitive task analysis and Recognition-primed decision.



Let’s say that you believe tacit knowledge exists. Let’s also say that you agree that tacit knowledge is mostly learnt through emulation, apprenticeship and osmosis.

The question that follows naturally is this: are there ways to get better at it? Is there research that tells us how to get better at emulation, or to be more effective at apprenticeship? I mentioned in my previous piece that the field of Naturalistic Decision Making (NDM) [Naturalistic decision making] is the most promising subfield of expertise research I’ve found; I also mentioned that learning ideas and methods from NDM has made the biggest difference in my own pursuit of mastery.

Two caveats

Let’s take a quick tour of all the ways ’tacit knowledge’ is confusing, just so you know that I am skipping over large parts of the philosophical literature:

  • Tacit knowledge may be ’embodied’ — that is, activities like riding a bike or swinging a tennis racquet have to do with the body, but not necessarily the conscious mind. Some philosophers believe that such skills are a type of tacit knowledge that is different from the tacit knowledge in more cerebral fields like programming and designing and writing. (I do not believe this is true, but the philosophical argument for why they believe this is the case will take around 50 pages of writing, so we’re not going to go there).
  • Many types of tacit knowledge may be made explicit: for instance, bike riding or ball kicking may be turned into a collection of physics equations in a research paper. This is a form of explicit knowledge, and it may help you build a bike-riding or ball-kicking robot. But from a pedagogical perspective, this form of explicit knowledge is useless: it won’t help you learn to ride a bike or kick a ball. So is this really tacit knowledge? Or is it explicit? Or is it a bit of both?
  • What about the kinds of tacit knowledge that can be communicated via video, or through pictures? I may not be able to explain to you how to do a complicated series of yoga poses through words alone, but if I record a quick video, or draw a series of diagrams, you might be able to learn to do so. Is such knowledge tacit, or explicit?
  • And how about organisational tacit knowledge? In 1988, Taiichi Ohno published The Toyota Production System, which laid out the ideas behind lean manufacturing for the first time. He did so after resisting codification of those principles for many, many years; Ohno feared that the publication of TPS would destroy Toyota’s competitive advantage. But it turned out that the publication of Ohno’s book by itself did not lead to widespread adoption amongst Toyota’s competitors. More bizarrely, competitors who poached TPS experts away from Toyota were not able to replicate Toyota’s system in their company. It appeared that TPS required an organisational culture to develop alongside the implementation of the system itself. This phenomenon sparked off an entire subfield of research into organisational tacit knowledge.

I’ll skip ahead here and tell you that none of these questions are particularly interesting to me — at least, not in the context of this blog. This blog is a practitioner’s blog: it is interested in what is useful, not what is ’true’. […]

So here’s my way of side-stepping everything above: when I say tacit knowledge in this series, what I am really interested in is ’expert intuition’ or ’expert judgment’. This is the “expertise, full of caveats” that I talked about in Part 1 (a) [ (Chin 2020b) ]. Expert intuition is tacit because it is incredibly difficult to investigate, incredibly difficult to make explicit, and incredibly difficult to teach.

How the naturalistic decision makers do it

  1. […] ‘Cognitive Task Analysis’ (CTA) [Cognitive task analysis]

    […] a set of interviewing techniques designed to explicate tacit mental models of expertise. This is not an easy task. An earlier version of this was called the ‘Critical Decision Method’ [Critical decision method], which focused on ’tough cases’ of expertise-driven performance, in order to elicit whatever it is that was going on in the practitioners’s heads. Today, the set of techniques is much larger, and they are all lumped together under the umbrella of CTA.

    If you study the development of psych approaches, CTA is simply an extension of older techniques used by the behaviourists to analyse psychological effects under experimental conditions. (Paul Harmon, for instance, talks about this evolution here (a)). NDM researchers typically perform cognitive task analysis at the outset of their projects — this is how they extract tacit knowledge from the experts they study.

  2. […] a model of expert intuition called the recognition-primed decision making model (RPD) [Recognition-primed decision making model].

    This model, developed by Gary Klein and his collaborators in the 90s, lies at the heart of their understanding of expert intuition. RPD explains how expert intuition works, and it tells us why experts have so much difficulty when it comes to explaining their expertise. NDM practitioners use the RPD model as a guide during their interviewing process, and they also use it as a guide to the development of their training programs.

  3. NDM researchers develop training programs designed to let less experienced practitioners acquire expert judgment of their own. NDM practitioners typically have some familiarity with pedagogical techniques, but in this, they, too, are guided by the RPD model.

It’s a common misconception to look at this and think: “oh, when you can make tacit knowledge explicit, then you can just explain things to people and they will get it.” This is not the right conclusion to have. As we shall soon see, large portions of expertise is tacit, and for certain types of expertise, the training methods must also be tacit in nature — that is, built around copying, and emulation, and scenario training, not explicit instruction. What NDM gives us is a rigorous method to reason about such things.

[formatting mine]

The Recognition-primed decision making model

See Recognition-primed decision.

How do we use RPD?

See How to improve at Recognition-primed decision making?

How do we know this model is accurate?

One question that you might reasonably ask is: how do we know that the RPD model is true? I’ve described NDM’s methods to you, and you might have caught on to the fact that none of what I’ve described is experimental. NDM researchers do not use hypothesis testing; there is no concept of falsifiability in their work. Klein himself readily admits that NDM is “nearer to anthropology than psychology.”


In the two decades since Sources of Power was first published, NDM has only become more established as a field of research. More militaries around the world are using NDM methods to extract tacit models of expertise from their best operators, in order to turn that knowledge into training programs; Nasdaq uses (a) NDM researchers to design user interfaces to augment the expertise of experienced fraud investigators on their staff. Klein himself wrote a famous paper (a) [Daniel Kahneman, Gary Klein | Conditions for Intuitive Expertise: A Failure to Disagree] with Daniel Kahneman on the situations where expert intuition was valid, and where it might not be valid. Most interestingly, Robert Hoffman, an NDM pioneer and one of the co-authors of the Cambridge Handbook of Expertise and Expert Performance (a), [K. Anders Ericsson, Robert Hoffman, Aaron Kozbelt, A. Mark Williams | The Cambridge Handbook of Expertise and Expert Performance] has spent a good part of the last decade working on an epistemology (a) [Naturalistic Decision Making | Robert Hoffman] for the entire field.

These are all signs, I think, that NDM captures something true about the world. But be that as it may, I believe that this shouldn’t be of much consideration to you, if you simply want to use this for your career.

I have argued elsewhere that if you are primarily a practitioner, your epistemology should be different than if you were a scientist (a) [Cedric Chin | A Personal Epistemology of Practice] If you are a scientist, you will want to know what is true by the standards of scientific truth — this typically means successful replication over a number of years, culminating in meta-analyses of multiple randomised controlled trials with sufficient statistical power. But if you are a practitioner, what you want to do is to learn what can be useful to you today, in the pursuit of your goals. Your evaluation of an idea should be structured around whether the idea works for you — which is a different bar for truth compared to that of the scientist’s.

NDM lies at the highest level in my hierarchy of practical evidence (a) [Cedric Chin | The Hierarchy of Practical Evidence]: I have tried their methods and their ideas in my life, and nearly every one of them have led to improvements.


Chin, Cedric. 2020a. “Copying Better: How To Acquire The Tacit Knowledge of Experts.” Commoncog. https://commoncog.com/how-to-learn-tacit-knowledge/.
———. 2020b. “Why Tacit Knowledge Is More Important Than Deliberate Practice.” Commoncog. https://commoncog.com/tacit-knowledge-is-a-real-thing/.