Tacit knowledge is more important than deliberate practice

Sadly I'm not familiar with books that focus on self-learning. If you're struggling to apply the knowledge from CLT it suggests that you're mostly missing skills related to curriculum design like sequencing, scoping, and scaffolding. You might consider an introductory book into curriculum design (apologies, I don't have the name of a good one at hand). You might also benefit from reading something like "How I Wish I'd Taught Maths"[0] which is a great example of someone who understands these theories attempting to transform them into actionable ideas in the classroom. Following that example might help you in managing your own learning.

[0] https://www.amazon.com/How-Wish-Taught-Maths-conversations-e...

As an atheist this feels mystical.

Indeed, this seems a little misleading (or perhaps I have too broad a view of deliberate practice). The submission talks about learning to ride a bike by first gliding for short periods, then longer periods, then pedalling. Is this not deliberate practice? Or identifying that you struggle with understanding software design, and then working on create designs and getting feedback to improve them so you can develop your own intuition for how to design good software - this also seems like deliberate practice to me.

I'll assume the best in the author, that he simply interpreted deliberate practice differently than I did such that it was not useful for most things, and now he's trying to share the method that worked for him. But to me this is exactly what I would expect from deliberate practice.

I disagree with this assessment. (Or, more accurately, I’d like to see a more comprehensive counter argument).

The article you link to is essentially a response to 2 pages in the book, where Hoffman et al mention, almost in passing, that CLT is a silly theory when you want to train for real world scenarios (the intuition is that if you’re training marine fire squad commanders to plan on the battlefield, perhaps it helps to simulate shooting at them during training?) Hoffman et al use this as an example of a learning theory that doesn’t seem to map to real world requirements.

This reads like a disagreement over one particular dismissal in the book, perhaps because CLT is a pet theory of the article’s authors. The problem: this argument is not core to the book!

The article does not, for instance,

a) Deal with the many examples of successful real world accelerated training programs with no curriculum design (as is commonly understood; ordering of simulations isn’t really designing a syllabus) in Chapter 9 (some of which were designed by some of the authors)

b) Have a rejoinder to the two learning theories presented in Chapter 11 that the authors claim underpins their training approach (if there were something to attack, this would be it!)

c) Nor have a rejoinder to a more central claim in the book, (and to my mind a more controversial claim) that atomisation of concepts impedes rapidised training.

And, perhaps most surprisingly to me, your claim that

> Reading it I got the distinct impression that the authors did not understand a great deal of the research they cited, either when supporting or dismissing it.

is remarkable, given that one of the authors of Accelerated Expertise is Paul J Feltovich, one of the founders of the field of expertise research, and a contemporary of Ericsson’s.

This reminds me of this post: https://putanumonit.com/2021/09/10/rules-for-noobs/

I don't think the guy is any sort of "certified expert" but it rings true to me. And occasionally I've tried to give people related advice but found I couldn't. For example: How do you know when you you're close enough with someone that you tease them in a way that they'll take well vs badly? I'm pretty good at knowing it, but I'll be damned if I can explain it.

tl;dr of the article is that when people are learning then explicit rules and guidelines are helpful. Once you get better at it, they are unnecessary and even harmful. People in the latter stage tend to give crappy advice like "just be yourself" because it works for them now, they don't really exactly know what they're doing (ie, the knowledge is implicit) and they've forgotten what it's like to be totally incompetent.

Anyway, perhaps the point is that while it's difficult to externalize how to "not be awkward" (and indeed it would likely be different for every social group) you can probably use rules to get along well enough, for long enough, to figure it out the hard way.