I like this object level review. I think that epistemic spot checks are a good idea overall, and like the focus on extracting the underlying model. However, I am worried that just checking the citations that are actually provided in the book makes this sort of check non-robust against the type of selection bias that Scott describes in part 4 of this post. Do you have any idea's for dealing with that? I'll describe mine after a line break for people who would like to consider the problem.
I think it is worth looking at existing reviews from experts, as well as choosing a small but representative sample of new to you concepts from the book to research. Or possiblity to pick a few claims and look specifically for counter examples.
I agree that "your own citations support your statements" is a low bar. In general I consider a book's own citations being good and saying what they claim to be the bare minimum, and if something fails that I won't go farther. I made an exception in this case because it came so well recommended. For the second chapter I reviewed I happened to have a lot of implicit understanding that would have taken a long time to formalize, via my neurologist and research I did years ago verifying things my neurologist told me.
What you suggest about looking at experts seems like an excellent way to hone in on concepts; I think I'm approaching the same goal, segmented in a different way. I read multiple books in the same vein and run these checks on all of them, and synthesize understanding from there. Segmenting it by book frees up a lot of RAM for me.
I'm torn on looking for counterexamples. On one hand, it's an easy way to disprove wrong things. On the other hand, even the best theory is fairly easy to poke holes in, especially if you consider claims in isolation rather than the system as a whole. I find if I focus on disproving things, it inhibits understanding and learning, and leads to rejecting things that are imperfect but useful.
Possible solutions to both of these are to involve other people. For synthesizing concepts, you can have each person read on related topics and come together to discuss interactions. For looking for counterexamples, one person can focus on that while the other focuses on understanding in a generous way, without worrying about looking stupid for believing something wrong.
Nice, thanks for writing this up.
The prediction is that improving the quality of processing via the principles explained in the book can reduce pain and increase your physical capabilities.
Is there a summary of the principles somewhere?
I want to talk more about the pain model in the book, but I'm deliberately not writing up the principles. I don't think people would get much from a summary that's more than what I have here but less than the whole book. Being walked through it was really integral to my getting the benefits.
This is part of an ongoing series assessing where the epistemic bar should be for self-help books.
Introduction
Thesis: increasing your physical capabilities is more often a matter of teaching your neurological system than it is anything to do with your body directly. This includes things that really really look like they’re about physical constraints, like strength and flexibility. You can treat injuries and pain and improve performance by working on the nervous system alone. More surprising, treating these physical issues will have spillover effects, improving your mental and emotional health. A Guide To Better Movement provides both specific exercises for treating those issues and general principles that can be applied to any movement art or therapy.
The first chapter of this book failed spot checking pretty hard. If I hadn’t had a very strong recommendation from a friend (“I didn’t take pain medication after two shoulder surgeries” strong), I would have tossed it aside. But I’m glad I kept going, because it turned out to be quite valuable (this is what triggered that meta post on epistemic spot checking). In accordance with the previous announcement on epistemic spot checking, I’m presenting the checks of chapter one (which failed, badly), and chapter six (which contains the best explanation of pain psychology I’ve ever seen), and a review of model quality. I’m very eager for feedback on how this works for people.
Chapter 1: Intro (of the book)
Claim: “Although we might imagine we are lengthening muscle by stretching, it is more likely that increased range of motion is caused by changes in the nervous system’s tolerance to stretch, rather than actual length changes in muscles. ” (p. 5).
Overstated, weak. (PDF). The paper’s claims to apply this up to 8 weeks, no further. Additionally, the paper draws most (all?) of its data from two studies and it doesn’t give the sample size of either.
Claim: “Research shows the forces required to deform mature connective tissue are probably impossible to create with hands, elbows or foam rollers.” (p. 5).
Misleading. (Abstract). Where by “research” the Hargrove means “mathematical model extrapolated from a single subject”.
Claim: “in hockey players, strong adductors are far more protective against groin strain than flexible adductors, which offer no benefit” (p. 14).
Misleading. (Abstract) Sample size is small, and the study was of the relative strength of adductor to abductor, not absolute strength.
Claim: “Flexibility in the muscles of the posterior chain correlates with slower running and poor running economy.” (p. 14).
Accurate citation, weak study. (Abstract) Sample size: 8. Eight. And it’s correlational.
[A number of interesting ideas whose citations are in books and thus inaccessible to me]
Claim: “…most studies looking at measurable differences in posture between individuals find that such differences do not predict differences in chronic pain levels.” (p. 31).
Accurate citation. (Abstract). It’s a metastudy and I didn’t track down any of the 54 studies included, but the results are definitely quoted accurately.
Chapter 6: Pain
Claim: “Neuromatrix” approach to pain means the pattern of brain activity that create pain, and that pain is an output of brain activity, not an input (p93).
True, although the ability to correctly use definitions is not very impressive.
Claim: “If you think a particular stimulus will cause pain, then pain is more likely. Cancer patients will feel more pain if they believe the pain heralds the return of cancer, rather than being a natural part of the healing process.” (p93).
Correctly cited, small sample size. (Source 1, source 2, TEDx Talk).
Claim: Psychological states associated with mood disorders (depression, anxiety, learned helplessness, etc) are associated with pain (p94).
True, (source), although it doesn’t look like the study is trying to establish causality.
Claim: Many pain-free people have the kinds of injuries doctors blame pain on (p95).
True, many sources, all with small sample sizes. (source 1, source 2, source 3, source 4, source 5)
Claim: On taking some cure for pain, relief kicks in before the chemical has a chance to do any work (p98)
True. His source for this was a little opaque but I’ve seen this fact validated many other places.
Claim: we know you can have pain without stimulus because you can have arm pain without an arm (p102).
True, phantom limb pain is well established.
Claim: some people feel a heart attack as arm pain because the nerves are very close to each other and the heart basically never hurts, so the brain “corrects” the signal to originating in the arm (p102).
First part: True. Explanation: unsupported. The explanation certainly makes sense, but he provides no citations and I can’t find any other source on it.
Claim: Inflammation lowers the firing threshold of nociceptors (aka sensitization) (p102).
True (source).
Claim: nociception is processed by the dorsal horn in the spine. The dorsal horn can also become sensitized, firing with less stimulus than it otherwise would. Constant activation is one of the things that increases sensitivity, which is one mechanism for chronic pain (p103).
True (source).
Claim: people with chronic pain often have poor “body maps”, meaning that their mental model of where they are in space is inaccurate and they have less resolution when assessing where a given sensation is coming from (p107).
Accurate citation (source). This is a combination of literature review and reporting of novel results. The novel results had a sample of five.
Claim: The hidden hand in the rubber hand illusion experiences a drop in temperature (p109).
Accurate citation, tiny sample size (source). This paper, which is cited by the book’s citation, contains six experiments with sample sizes of fifteen or less. I am torn between dismissing this because cool results with tiny sample sizes are usually bullshit, and accepting it because it is super cool.
Claim: “a hand that has been disowned through use of the rubber hand illusion will suffer more inflammation in response to a physical insult than a normal hand.” (p. 109).
Almost accurate citation (source). The study was about histamine injection, not injury per se. Insult technically covers both, but I would have preferred a more precise phrasing. Also, sample size 34.
Claim: People with chronic back pain have trouble perceiving the outline of their back (p. 109).
Accurate citation, sample size six (pdf).
Claim: “Watching the movements in a mirror makes the movements less painful [for people with lower back pain].” (p. 111). Better Movement. Kindle Edition.
Accurate citation, small sample size (source).
Model Quality
Reminder: the model is that pain and exhaustion are a product of your brain processing a variety of information. The prediction is that improving the quality of processing via the principles explained in the book can reduce pain and increase your physical capabilities.
Simplicity: Good. This is not actually simple model, it requires a ton of explanation to a layman. But most of its assumptions come from neurology as a whole; the leap from “more or less accepted facts about neurology” to this model is quite small.
Explanation Quality: Fantastic. I’ve done some reading on pain psychology, much of which is consistent with Guide…, but Guide… has by far the best explanation I’ve read.
Explicit Predictions: Good, kept from greatness only by the fact that brains and bodies are both very complicated and there’s only so much even a very good model can do.
Useful Predictions: Okay. The testable prediction for the home-reader is that following the exercises in the back of the book, or going to a Feldenkrais class, will treat chronic pain, and increase flexibility and strength. Since the book itself admits that a lot of things offer short term relief but don’t address the real problem, helping immediately doesn’t prove very much.
Acknowledging Limitations: Poor. GTBM doesn’t have the grandiose vision of some cure-all books, and repeatedly reminds you that your brain being involved doesn’t mean your brain is in control. But there’s no sentence along the lines of “if this doesn’t work there’s a mechanical problem and you should see a doctor.”
Measurability: poor. This book expects you to put in a lot of time before seeing results, and does not make a specific prediction of the form they will come in. Worse, I don’t think you can skip straight to the exercises. If I hadn’t read the entire preceding book I wouldn’t have approached them in the correct spirit of attention and curiosity.
Hmmm, if I’d assigned a gestalt rating it would have been higher than what I now think is merited based on the subscores. I deliberately wrote this mostly before trying the exercises, so I can’t give an effectiveness score. If you do decide to try it, please let me know how it goes so I can further calibrate my reviews to actual effectiveness.
You might like this book if…
…you suffer from chronic pain or musculoskeletal issues, or find the mind-body connection fascinating.
This post supported by Patreon.