The complexity of explaining the intuitive.

I’d like you to conduct a little thought experiment with me.

Assume you and I are standing 10 feet apart, and I’m holding a soft NERF ball, approximately 25 cm in diameter.

I toss the ball to you, underhand, in a soft, lazy arc.

You have two options: You can either eyeball it, and catch the thing; or you can attempt to calculate the speed, angle, impact of wind, and gravitational effects to predict both where the ball can be caught, and how much grasping force you’ll need to hold on to it.

Why yes, I was intending to be absurd. Glad to know I pulled that one off.

What I’m trying to say is, often proving that a thing can be done, is much more difficult than actually doing the thing.

This is only a problem because, and I’ve only realized this in the last couple of years, I’ve spent my entire life in pursuit of intuitive understanding of systems and effects. I can explain the rules in broad strokes, and point to research, but inevitably I solve problems by looking at as much data as I can that is directly or tangentially involved in the problem, and then I arrive at a solution.

The next step, irritatingly, is to back-rationalize the solution, picking out every assumption I’ve made along the way to the insight or approach I’m running with, and then looking for evidence or supporting theories that I can use to get other people on the train with me.

I find this frustrating for an obvious reason: it doubles the workload, without actually adding anything (of value comparable to the effort expended) to the outcome.

In Cory Doctorow’s book For the Win, he mentions a concept called fingerspitzengefuhl, or fingertip-feel. In literal terms, it’s the idea of having the world resting against the nerve-dense tissue at the end of a finger, and being able to sense every little tremor on that globe. It’s an artful word, and it’s a concept that I think, in part, pairs very well with that of systems thinking.

To understand something so well that you can LOOK at it and see the problem, that is magic to me, in large part because I’m an intuitive person. I do poorly with rules, but I do very well with perceiving flaws, and seeing where the potential for correction exists.

At the end of the day, it’s substantially simpler to understand something than it is to explain it.

Understanding can come simply from observation, immersion, experience. You need to delve into how something works, but you can use your unconscious mind to fill in some of the blanks.

Explanation requires breaking down your understanding into a digestible model, finding a way of explaining that digestible model, and then providing evidence for every non-obvious element of that digestible model.

You also generally need to explain things in a way that doesn’t require any pre-existing understanding of the subject matter. Which, when dealing with truly complex (rather than just complicated) issues, can be nearly impossible.

There are clear issues with what I’m saying here, foremost among them being the arrogance of stating that ‘just getting it’ should not be interrupted by a need to justify every decision, but it comes from the same central point as the defence of creative arts - some things are too complex to explain well consciously. And in the attempt to work in reverse (or in the accepted direction), to move from data to assumptions to justifications to an idea or plan, all true understanding and complexity is removed.

Ask someone to build an algorithm that will correctly intuit human emotion, or the details of a moving biological body, or even music, and they will be hard pressed to generate something that passes an initial inspection. But we ask that people develop plans and strategies to influence human behaviour on a mass scale, and we ask that they do it by trying to reduce their understanding of a system to a spreadsheet.

I understand the rationale, but I admit I also find it exhausting, at times. I’d rather pursue the ability to feel what is and isn’t working like it’s on the tip of my finger, than pursue an approximate, ever-more-obsolete, model of behaviour.