Sunday, August 20, 2017

There's something happening here

A passage from The Murder at the Vicarage by Agatha Christie, published in 1930. Doctor is speaking to the Vicar.
“We think with horror now of the days when we burnt witches. I believe the day will come when we will shudder to think that we ever hanged criminals.” [Doctor]

“You don’t believe in capital punishment?” [Vicar]

“It’s not so much that.” He paused. “You know,” he said slowly, “I’d rather have my job than yours.”

“Why?”

“Because your job deals very largely with what we call right and wrong—and I’m not at all sure that there’s any such thing. Suppose it’s all a question of glandular secretion. Too much of one gland, too little of another—and you get your murderer, your thief, your habitual criminal. Clement, I believe the time will come when we’ll be horrified to think of the long centuries in which we’ve punished people for disease—which they can’t help, poor devils. You don’t hang a man for having tuberculosis.”

“He isn’t dangerous to the community.”

“In a sense he is. He infects other people. Or take a man who fancies he’s the Emperor of China. You don’t say how wicked of him. I take your point about the community. The community must be protected. Shut up these people where they can’t do any harm—even put them peacefully out of the way—yes, I’d go as far as that. But don’t call it punishment. Don’t bring shame on them and their innocent families.”

I looked at him curiously. “I’ve never heard you speak like this before.”

“I don’t usually air my theories abroad. Today I’m riding my hobby. You’re an intelligent man, Clement, which is more than some parsons are. You won’t admit, I dare say, that there’s no such thing as what is technically termed, ‘Sin,’ but you’re broadminded enough to consider the possibility of such a thing.”

“It strikes at the root of all accepted ideas,” I said.

“Yes, we’re a narrow-minded, self-righteous lot, only too keen to judge matters we know nothing about. I honestly believe crime is a case for the doctor, not the policeman and not the parson. In the future, perhaps, there won’t be any such thing.”

“You’ll have cured it?”

“We’ll have cured it. Rather a wonderful thought…”
That is a clear example of a determinist.

There's an old joke in science circles "There are 10 types of people, those who understand binary and those who don't." We do have an inclination towards binary manichaeism. One of the traditional bifurcations has been between two classes of thinkers - those who view the world deterministically and those who view it tragically.

Painting in broad strokes, determinists believe that every phenomenon has a root cause that is knowable. Plato and Marx are thinkers in the determinist school. There is no such thing as free will for determinists. Everything that happens, happens for a specific, determinable reason. Religion is simply mumbo-jumbo for weak minds.

The tragedians see the world differently. Some things are not knowable. There is free will. People do make choices. Some things happen for unknown reasons. Religion has a more acceptable role among tragedians because it engages with unknowable mysteries.

I suspect everyone starts from one base or the other and as they grow, learn, and experience, they oscillate between the two extremes and end up settling somewhere in the middle; determinists in some realms of thought and tragedians in others. But being human, there are always the outlying fanatics - either those who insist that everything is knowable or those who insist everything is a perpetual mystery.

In recent years I have begun to wonder whether there isn't some third option around complexity, that it is possible that there is no free will in the long run but in the short run, and in realms of uncertainty, actions do in fact replicate what would be considered free will. That's a whole separate conversation.

But even more recently I have been visualizing the model in a different fashion. This concept is informed by a recognition of the limits of legibility (visible and measurable) and knowability (confirmable and predictable). We cannot measure everything to the degree of precision required and all knowledge is contingent. We live in a world where there are boundaries to the nature of our knowledge and that determinism and tragedianism are simply realms within the continuum of legibility and knowability.

If we accept that the world is constituted of innumerable systems (human behavior, economic activity, geological processes, etc.) which evolve at some rate over time and if we accept that all systems have at least four constituent parts (context, inputs, processes and outputs) and if we accept that time frames are relevant to how much we can understand at a given point (a function of the capacity of our brains) and if we accept that all these systems manifest themselves in some fashion of patterns, then I suspect that the following is true.

Click to enlarge.

In the bottom left are systems which have a very high degree of legibility in terms of their context, inputs, processes and outputs and which are very stable in the pertinent time frame. Manufacturing processes are a preeminent example of a complex system where an enterprise controls the manufacturing environment, all the inputs, manages the process in order to produce a six sigma output. In short time frames, the degree of change and uncertainty are low. Within the specified parameters, everything is highly predictable.

All systems are subject to probabilities but in a well managed manufacturing system, the probabilities are reasonably knowable and manageable and redundancies can be constructed so that the output remains unperturbed.

There are other examples much nearer to home than that. Some people habituate their lives. In order to focus their cognitive functions on the matters of greatest moment, they establish habits that are highly legible and stable. They wake in the morning, make the bed, eat breakfast (the same thing every morning), take a shower, dress, exit home, start the car, drive the same route to work in the morning, park in the same place. Again, the system is highly legible and highly stable.

In general, the shorter the time frame, the more systems meet the criteria of high legibility and low dynamism. If I am a plant manager, I can tell you at the beginning of the morning with a high degree of confidence what the output will be at the end of the day because the entire system is highly legible and highly stable (low dynamism). If I am working with a 24 hour frame - the system is highly deterministic.

If I am working with a 365 day frame, the system is less deterministic. Business goals might have changed, market demand might have changed, technology may have changed. There are as many variables as in the 24 hour frame but they are much more dynamic in the 365 day frame. I can know that it is highly probable that things will have changed that would affect my forecast of what the plant will produce 365 days from now but I cannot forecast with confidence the net change implied by the aggregate of all the changes, particularly as some changes affect others. If there is less market demand, I might simply produce less, or I can change the process in order to produce it cheaper, or I can abandon the product, etc.

The net of this is that in very short time frames with highly legible systems subject to very low dynamism, it is very feasible to view the world as nearly entirely deterministic.

But the more you lose legibility (you don't have the data, the data is hard to get, you don't understand the context or the interdependencies, etc.), the longer is the time frame and the more dynamic is the system (rapidity of system evolution), the more uncertain we are about the resultant patterns and the more difficult it is to predict the future. I cannot use a 1930 roadmap to navigate between Chicago and Los Angeles in 2017.

In this model, I would arbitrarily accord three states of knowledge. In State I, there is high legibility and low dynamism and we can characterize that state as Deterministic.

In State II, there is some loss of legibility either in terms of precision or in terms of not being able to specify the context, the inputs, the process, or the outputs. In addition, the rate of change is greater and so we know that the aspects of the systems we can glimpse are likely to have changed between points of time. We may be able to attach probabilities to an a=outcome but we cannot specify the outcome. We can characterize State II as Tragic.

In State III, we simply do not have enough or any information of the state of things. We do not know the context, we don't understand what might be the inputs and/or we don't know the causative process, and/or we don't know what are the outputs. There is no predictability in whatever patterns we think we see. There is sensory perception without cognition. We can characterize State III as Mysterious or Incomprehensible.
There's something happening here
But what it is ain't exactly clear
There's a man with a gun over there
Telling me I got to beware

I think it's time we stop
Children, what's that sound?
Everybody look - what's going down?


In this model, the degree to which you are a determinist or a tragedian is primarily a function of humility. If you think that we either know everything about all things, or if you think that all things can and will be known at some point, then you are a determinist. If you believe nothing is knowable and cannot be known, or if you take the humble position that we are constantly revising what we believe to be true, then you are a tragedian.

And then there is the pragmatic position which recognizes that the degree of legibility about a particular system can be at different levels and might vary over time. The pragmatist accepts that there are some systems which can be treated as deterministic over specified time frames and under specified conditions and that there are other systems about which our knowledge is vestigial and tragically, we can't even see how we might be able to know the system.

UPDATE: Heh, forget determinists versus tragedians. Tillman suggests that the epistemological divides is between the Eloi and the Morlocks. From Conference Paper, Karl Popper’s Falsifiability: The Foreign Emoluments Clause—A Debate Between Constitutional Eloi and Constitutional Morlocks. by Seth Barrett Tillman.
The reality is that this debate is a conflict between constitutional Eloi and constitutional Morlocks.

The ninety-nine percenters are our constitutional Eloi, our beautiful people, our self assured true believers who regularly assume they understand 99% of the Constitution’s original public meaning. For them, figuring out what a yet-to-be adjudicated clause means is easy: it only requires their selecting the most eligible meaning which already fits in with what they already know. And what’s the danger of that strategy—when you already know (or believe you know) 99% of what there is to know?

On the other side, we have constitutional Morlocks. Morlocks are ugly, or, at least, their theories are ugly. Ugly and dangerous. Morlocks don’t believe they know 99% of what there is to know, and, not surprisingly, they don’t believe the Eloi or anyone else knows 99% either. Moreover, Morlocks believe that fitting what you don’t know into what you (think you) know permanently freezes our constitutional theories even when those theories are entirely wrong.
Deterministic Eloi and Tragic Morlocks.

No comments:

Post a Comment