Gonna drop my working frame for "free will" this morning, because why not?
Get coffee first, I went LONG on this one.
(In a hurry for my answer? Scroll down to section V.)
I. Two Types of Systems
Some systems (factory machines, cello playing, solar eclipses) are deterministic.
Some (weather, turbulence, stocks) are chaotic.
Also, some chaotic systems (waterfalls, standing waves, mobs) exhibit a sort of emergent order at scale.
Okay so far?
Good.
II. On Measurability, and Scale
What matters, from an OPERATIONAL perspective, is whether or not a system's behavior can be USEFULLY predicted. Right?
If your predictions work, the system looks deterministic.
If not, it looks chaotic.
Here's the fun bit:
SMOOTHNESS IS A FUNCTION OF SCALE.
(and measurability is a function of smoothness)
Smoothness is a function of scale.
Measure a coastline in [whole-unit] miles, no fractions or bends. Then, measure it again in meters.
The shorter metric will give you a longer length.
Why? The answer is intuitively obvious: a meter-stick will measure more of the coastline's smaller indentations than a mile-stick.
One is longer than the other, yet both measurements are real.
Which measurement should you use?
I don't care, it's up to you.
Use whichever one is USEFUL.
Whichever one best fits your operational needs.
Now then,
III. Extending the Principle
Similar measurement effects also happen with emergence.
But here, we focus not on differences in UNITS, but on differences in behavior at SCALE.
Like this:
Individual gas molecules behave chaotically, right? Measurements of small-enough individual bits caught in a crowd become uncertain. As collisions mount, trajectories change. Prediction becomes impossible.
But at scale? Things start to change.
(Hang in there, Free Will-ists, we're getting there)
At scale, we can make precise, systemic, testable predictions about the emergent behavior of gas molecules in large [stochastic] groups.
These systemic predictions work so well that we call them "Laws."
The Noble Gas Laws are so consistent, so highly predictive in their ability to measure the emergent behavior of gas molecules at scale, that we can use them to design large machines.
Large machines that work.
Predictably.
Deterministically.
At scale.Machines that give consistently measurable results, at the emergent scale at which they were designed to do so.
Smoothness is a function of scale.
IV. Squaring the Circle
"What," you ask, "does all of this have to do with Free Will?"
Do people have Free Will? Or do they not?
["Substack as notepad" reminder, I'm gonna finish roughing this out here and clean it up later]
Thinking out loud, then,
People mean different things by "Free Will," don't they?
Most commonly, I think they mean these three:
do people have decision-making agency, Y or N?
are their decisions deterministic, Y or N?
are their actions predictable, Y or N?
The answers to these questions are bound up in our previous discussion of smoothness, measurability, and scale.
So, I'll sketch how I think the decision-making machinery works in those terms, and then give direct answers to the above questions.
I define decision-making as "the entire mindspace between inputs & actions."
This is a user-friendly definition meant for daily use on the street.
So defined, the machinery of human decision-making has four main components:
Habit
Instinct
Reason
Emotion
which are bound by the common thread of Identity, and then surfed by the spotlight of Limited Attention.
Okay so far?
Good.
V. Free Will
Within the 4-part HIRE model, then,
ONE part (instinct) grants agency;
ALL FOUR parts interact deterministically;
but NONE of their deterministic interactions can be measured.
Plugging that back into the triple question of Free Will,
do people have decision-making agency? YES
are their decisions deterministic? HELL YES
are their individual actions predictable? NO
do predictable results emerge at scale? YES
Three questions, four answers.
"What the hell," you ask, "am I to do with this?"
Short answer:
Operationally speaking, YES, treat people as if they have Free Will.
Do this as if your life depends on it, or society will crash. It is THAT useful.
Longer answer below.
VI. Nuance
(longer Free Will answer in 7 parts)
YES
(1) People make agentic, deterministic decisions.
BUT
(2) The precise inner workings of decision-making (as mapped by the HIRE model of the decision-making machinery) CANNOT BE MEASURED, even in principle!
SO
(3) You cannot MAKE reliable deterministic predictions at the level of individual behavior. Neither can anyone else.
Cannot.
Sorry.
This is not some metersticks-and-milesticks problem, where the same distance is being measured to two different standards.
In individual human decision-making, the deterministic interactions in question are far too chaotic even to BE measured, even in principle.
Deterministic, yes! But forever locked behind a haze of impenetrable, computationally intractable measurement-fog.
You cannot reduce individual decision-making to deterministic predictions.
If you try, you'll just crush the life out of everything, and crash society.
BUT
(4) You CAN make emergent-scale, deterministic predictions about how all human societies rise and fall.
A ray of hope! If thriving civilizational outcomes DO exist, and they ARE stable, then we CAN find them on the standing wave surfaces of the emergent solution space.
That sloping terrain IS universal, it IS fixed, it DOES exist, and it CAN be mapped.
Hari Seldon was right.
HOWEVER,
(5) Thanks to instinct, you cannot REACH those thriving outcomes you might have found, without first traversing the intervening societal solution space together, AS A SOCIETY OF INDIVIDUALS.
This necessarily means harnessing the agentic power of INDIVIDUAL decision-making (because individual instinct is the source of the motive power we need).
WHICH MEANS
(6) You must find a way of PERSUADING others to align their interests with yours.
Since you cannot harness the precise outputs of their decision-making power by DIRECT control, (which you MUST harness, in order to get where you want to go), you must work within the HIRE decision-making machinery as it actually is.
Which means persuading them to join you.
AND THEREFORE
(7) The easiest OPERATIONAL solution, the best way to DO that, again according to the universal logic mapped out by my HIRE decision-making model,
...is to decide, consciously, to act as if other people have Free Will.
Congratulations, you made it! What a ride.
(rough draft, minimal edits, hit send, why not)
> Smoothness is a function of scale
I first thought of calculus, which has the opposite law from a fractal: it assumes that no matter how complex the function is, at a small enough scale every function is smooth (straight line).
As opposed to the physical examples of Gas Laws, and coastlines, which are more like statistical models using the law of large numbers to smooth out. Another great example of stat smoothing is the second law of thermodynamics, which is like a summary of all possible physical systems.
Interesting that calculus works so well with describing physical systems.