From my Notebook: Complexity and Constraints


“The only model of a complex system is the system itself.” – Gell-Mann

The problem with human-based physical therapy research is that they keep using humans for subjects. – Me


Humans are complex adaptive systems, so the system is modified by behaviors. (Newton’s 2nd Law puts us at a different starting point at every moment – keep this in mind)

There is a non-linear relationship between cause and effect (Chaos)

Complex adaptive systems (humans) sometimes demand unique and previously untested  methods to solve problems. That cannot be done without innovation.  Safe-to-fail (= do no harm = small strategic influences/changes) experiments are essential to solve problems and allow best practice to emerge (See Cynefin; Snowden for more on complexity and problem solving).

We are dealing in the context of complexity = unknown unknowns.  Many times we really have no idea why a treatment works. It just does. Get comfortable with that.

Many theories exist in solving complex problems because complex problems may have multiple successful solutions.  This is why we have unnecessary arguments over best practice. Many answers to the same question are possible but experience influences understanding and decisions.

Retrospective Convergence = What got you here, won’t get you there

Premature Convergence = Thinking you have a solution too soon and not leaving your options open

Evaluate and Treat
We must limit the constraints to identify the needs or elements of change.

1. Describe the present

2. Identify what can be changed

3. Determine where you can monitor the impact of the intervention

4. Where will the intervention produce a beneficial result

5. If the result is not beneficial, how can I learn from it

6. Move between the complex and complicated domains (constraints are more relaxed here to allow a new state to emerge – see Cynefin)

7. As the new state emerges, increase constraints to exploit and magnify the result

Is PT using a recipe book and creating cooks when they actually should be training students to be chefs? (modified for my use from Dave Snowden quote).  Maybe that’s my job. On it.

Impact of Constraints

Structural constraints (think anatomy) are relatively fixed and more difficult to change.  Change certainly occurs over time but as a result of the manipulation of functional constraints. We cannot focus efforts here as again they are a result.

Functional constraints are malleable, changeable, and adaptable. Energy, ANS, Immune, Endocrine, Respiration, Circulation, Emotion, Experience, Understanding, Memories, etc.  These constraints are relaxed.  This is where we impact the system.  If you think you’re somehow treating a pathology (structural) as a PT, you’re probably wrong or you’re thinking like an orthopedic surgeon. Herein lies the disconnect between the surgeon and the therapist. PT’s live and work in the complex domain almost exclusively (throw in some complicated for good measure and comfort when we can) where input for the desired outcome is often unknown or inconsistent or idiosyncratic (n=1). The surgeon lives more in the complicated domain (more structural) where they do “fix” things (repair, reattach, realign, etc.). It’s okay if they don’t understand us. It is the result of experience and environment. Remember that.


The motor system and other sensory/cognitive systems, either intentionally or unintentionally, are the inputs to impact the functional constraints and gain the often unpredictable output.

Once the source of change is identified, narrow the focus to magnify the desired result or dampen the undesired result (Cynefin).

It’s always n = 1.




Please note: I reserve the right to delete comments that are offensive or off-topic.