About procedural learning, Matthias says,
"This type of learning requires a lot of practice over time - so it has limited use in treating chronic pain I think.
But: maybe this non-conscious type of learning is one of the causes of chronic pain?!"
I'm sure it must be. How about "just UN-do it?"
I practiced yoga for a couple years (w-a-y back when I was 21, 22.., early 70's), about an hour every day. I was taught the basics in about 15 minutes by a Danish PT who worked at the same hospital and who was a yoga enthusiast. She was the same age then as I am now. I bought a book she recommended and carried on alone. My goal was to become perfect at every pose (a goal that was, of course, never realized, although I did conquer a small number of the easier ones). Lucky for me, she had managed to convey the importance of deliberate slow breathing, watching for and then stopping at the first tiny non-nociceptive tug, waiting, observing and just hanging out, waiting for the 'body' to let go by itself. (There's a lot of mystical clap-trap associated with yoga which never appealed to me; her version and the book I bought skipped all that entirely or I would never have bothered.)
In those days, the knowledge base was completely dichotomous - the 'body' was not regarded as part of the brain's ongoing motor expression. The main sensory-motor homunculus had been mapped but was not common knowledge in PT yet. The brain was considered to be that bump up at the top of the body, not the other way round. The idea of ordinary and trainable "virtual bodies" hadn't come along yet.
Even so, over thousands of years this practical system called "yoga" had evolved, whereby one set oneself up in various positions that recruited the help of gravity, then simply relaxed and waited patiently for it to take over. The most important part was to breathe slow and deep, and allow, to refrain from actively "stretching" anything. The job, in other words, assigned to one's subjective conscious awareness, was to take over breathing, then stay completely out of the way (keeping it busy, rendering it harmless).
Pretty smart idea actually: give it something rhythm-producing to do (an activity like breathing which would happen regardless, but that subjective conscious awareness has little or no practice at), keep it distracted from/less likely to interfere with the real activity, i.e., changes in tone, in motor output, keep that part all non-conscious, therefore non-nociceptive. When I look back at that, now I can see that it is much like the "jobs" people give to their big dogs when they are walking them on the street; get them to carry a small piece of rope or a ball so they feel useful and loved and a necessary part of the "pack." As long as the dog carries an object it will remain focused on its own mouth, it will feel engaged, it won't bark or bite, it will stay connected to its owner posing less threat to pedestrians or small children or other dogs.