A real science of movement that would have equipment that measures human movement very exactly and then runs machine learning algorithms over those measurements is likely yield a science-based version of Feldenkrais that’s more efficient and where you can diagnose issues much better to be able to say beforehand whether Feldenkrais will help a person.
What would science-based Feldenkreis teaching look like?
How would you get from “machine learning algorithms over measurements”, which sounds like statistical curve-fitting, to actionable conclusions, about how people should use their bodies, and how to teach people how to do that? No-one can follow instructions like “increase activation of the iliopsoas by 3%”, even if you somehow validated a causal model that made that a useful thing to do in some situation. People can barely follow verbal instructions at all about posture and movement.
No-one can follow instructions like “increase activation of the iliopsoas by 3%”, even if you somehow validated a causal model that made that a useful thing to do in some situation.
The problem is that we don’t really know which instructions people can easily follow and which they can’t follow. If the problem is “increasing activation of the iliopsoas by 3%” you can empirically test various interventions.
Without having a casual model of what kind of movement is good, you can’t validate interventions and determine whether the intervention is good.
Apart from that it’s possible to do biofeedback. Good feedback can give humans perception and control over many variables.
On the other hand you can make a similar argument about the useful of understanding how proteins do what they do. Just because you understanding a pathway doesn’t mean you can manipulate it the way you want. Science advances by first mapping the space of phenomena and then hopefully finding a way to intervene.
What would science-based Feldenkreis teaching look like?
How would you get from “machine learning algorithms over measurements”, which sounds like statistical curve-fitting, to actionable conclusions, about how people should use their bodies, and how to teach people how to do that? No-one can follow instructions like “increase activation of the iliopsoas by 3%”, even if you somehow validated a causal model that made that a useful thing to do in some situation. People can barely follow verbal instructions at all about posture and movement.
The problem is that we don’t really know which instructions people can easily follow and which they can’t follow. If the problem is “increasing activation of the iliopsoas by 3%” you can empirically test various interventions. Without having a casual model of what kind of movement is good, you can’t validate interventions and determine whether the intervention is good.
Apart from that it’s possible to do biofeedback. Good feedback can give humans perception and control over many variables.
On the other hand you can make a similar argument about the useful of understanding how proteins do what they do. Just because you understanding a pathway doesn’t mean you can manipulate it the way you want. Science advances by first mapping the space of phenomena and then hopefully finding a way to intervene.