Is the problem posed by trying to understand the way the spine works a complex problem, a complicated one, or both?
Does it matter? Well, to answer that we need to shed some light on the difference between the two.
A complicated problem is one where there are lots of parameters. For example, if we have two members of a set (a, b) there can only be one relationship between the two. For three members, it is three. For 4, it is 6.
We can see that as this set gets larger there is a step-change in the number of interactions between the members – a triangular number sequence. With a complicated problem, the interactions between all members of the set are fixed and unvarying. Chess is like that.
So far, so good. All this means is that a complicated problem is one where it becomes more difficult to see what is going on as the number of parameters rises. But there aren’t, in principle, any computational impediments to understanding the system if we have sufficient time and a powerful enough number cruncher.
That’s why chess programs can now beat the very best human players.
What is a complex system? To understand this, look at weather forecasting. People used to think that forecasting the weather was an example of a complicated problem and that all they needed to do was measure the starting parameters of a weather state and get the right result.
But experience since the mid-’60s and the work of scientists like Lorenz explains why forecasting weather states is so difficult. This seems to be for two reasons.
First, there are non-linear relationships between many of the different weather variables (unlike in chess).
This makes it more difficult to predict a change in the state of the system.
Second, the state of the system at any point is extraordinarily sensitive to the starting position of the variables. Small changes in starting conditions have dramatic effects on the predicted system state.
So we find weather forecasts are generally reliable up to 4-5 days, but then tail off dramatically, despite impressive models and incredibly fast super-computers compared to 20 years ago.
Weather is complicated, but its predictive difficulty is because it is also a complex system.
And what about the spine?
There are a huge number of variables that affect spinal function and healing responses, two of the things we are most interested in.
That just makes it really complicated.
What makes it complex is that the physiology of function and structure is all about non-linearity. For example, the way that collagen can suddenly fatigue, or how homeostatic negative feedback loops can fail.
Further, we know it is impossible to measure accurately any of the variables in the system.
So the spine is at least as complex as it is complicated. Why does it matter if the spine is an example of a complex or a complicated problem?
The reason is that despite the challenge posed by complexity, it is possible to see beyond the complexity and to simplify in a way that adds value.
Extracting insights from complexity
See this TED talk for a neat explanation of this.
What Eric Berlow argues is that you can look at complex systems to extract meaning and make predictions, but you can’t do this with a merely complicated system.
If the spine was just complicated, our attempts to simplify and predict by looking for patterns and heuristics would be futile. We really would be fooling ourselves. We would have to rely on supercomputers and AI like DeepMind.
But because the spine has complexity, our human struggles probably aren’t futile.
It’s likely that, though error prone, a practitioner’s insights and appreciations of what’s going on, gleaned through clinical experience over many years of practice, are not has bad as we sometimes think.