Tuesday, November 06, 2007

On suffering, evil, the existence of God - NY Times

I don't want to argue about the nature of "evil" here, but I do want to reflect on the nature of arguing. Stanley Fish of the New York Times recently posted a column "On suffering, evil, and the existence of God" which made the top-10 most popular list for on-line readers, and has 150 comments posted covering a wide range of opinions. If you ponder "theodicy", it makes interesting reading.

The University of Aberdeen Philosophy Department's glossary defines this as follows:
THEODICY: An argument which tries to explain how a good and all-powerful God could create a world with suffering and evil in it.
I used to spend much time wrestling with such arguments, but finally decided that they are largely a waste of time, akin to one person I recall who had been struggling for a year with whether the word "satan" should be capitalized.

Most of us are not well trained or equipped to develop chains of solid qualitative reasoning that will stand up to time or peer-review, and for many people I observed these arguments largely served as an excuse to avoid doing the chores or facing problems in their own lives. such as how they themselves treated their neighbors.

For my own part, I have observed that it is difficult, or possibly impossible, to extrapolate from one "level" of this physical universe to even the next higher level, let alone extrapolate across a billion levels. Even within humans, children have little knowledge of what adults spend time worrying about, and I've spend many posts on why "management" and "labor" are effectively blind regarding each other's concerns and perspectives and, frankly, realities.

On a very slightly larger scale, if viruses or cells could reason, I doubt that their certain pronouncements about the nature of "life" would remain valid as seen by people.

So, I would suggest a different approach, and take a page from Science's book here, and ask that an instrument or technique or approach be validated and calibrated before being relied upon.

If your lab equipment cannot correctly measure a simple known sample, I wouldn't think you should rely on it to measure a complex unknown one.

So, I'd suggest people who have logical chains of reasoning they think extend "upwards" lower their sights to simply helping us figure out what's going on, say, at the scale of galaxies or even our own neighborhood in the Milky Way galaxy. Are the the first ones here, or is the evidence that someone was here before us, and is maybe still here? Is what is going on on earth of our own making, or is there a larger agency involved?

In this case, by agency, I mean simply and literally that - some agency, institute, corporation, secret government, the CIA, or Club of Rome, or some boring group of "aliens" that is busy growing, mining, or shaping human affairs.

Does anyone know how to make robust measurements of anything, in any way, that lets them determine, with both certainty and correctness, whether "someone else" is meddling in the affairs of some large group of people? And, if so, can the measurement locate either the route or the mechanism or the "source" of the meddling?

Do we have an advanced compass we can set down in a group of people and have the needle spin and point towards some outside group meddling in their affairs?

Or, these days, maybe it needs to make a list of the hundreds of groups who are meddling in the affairs of everyone else, in order to remove those factors, and see if there is still some huge unknown factor meddling in the lives of all of us.

(If you have such a device, I think it has a very high market value and the CIA would love to talk to you.)

Anyway, to my knowledge, no, no one can do that. And, compared to tracking down "God", that should be an easy task.

I'm not saying it's an impossible task, or one that is not a great thing for Science to work on, from a public health point of view and from a SETI (Search for Extraterrestrial Intelligence) point of view, or from a Homeland Security (or X-files) point of view.

The problem runs immediately into the questions I've posted on numerous times before, and not by coincidence. Can we recognize and describe and measure and reason about the kinds of causality that happen over long distances and long times, instead of being immediate and local? Right now, no. That is the current frontier of public health - multilevel distal causality.

When do you look and how do you look? For example, say there is a cake baking in an oven in my kitchen. (I wish!) What would a scientist "measure" to "see" that the cake is an artificial construct that I am "making", and not some natural phenomenon that is simply occurring and carrying out chemical reactions following physical laws?

If you know the answer, please comment or write me. I can't think of any way to make a local measurement of the cake that would nail down that relatively simple fact.

If you enjoy that sort of thinking, I'm also curious how you'd measure a water molecule and tell whether it was just bouncing off its neighbors blindly, or whether it was, viewed from further back, part of a flow of "water" in a "pipe" on the way to fill a glass of water for me to drink.

Unless I'm mistaken, which I'll allow, Science has no way to make such "local" measurements about "global" realities, even on boring things like water or cakes, let alone on whether a parent is "raising" a child or neglecting them when letting them learn something the hard way, let alone making judgments about impacts or influences on whole societies over thousands of years, etc.

Science has other measurement blind-spots right now, such as detecting that something, exposed to many forces, is changing far less than "it should" or, for that matter, not changing at all. Most statistics has trouble going from "As I vary X, you can see it has no effect on Y" and concluding "AHA! Something's clearly going on here!" So if I launch an anti-smoking campaign and smoking doesn't change one bit, is that because my campaign had no effect, or that it was way too effective and immediately cancelled out by an increase in marketing by the tobacco companies in reaction to my original action?

We are not very good at determining the constants of the motion, and even worse at detecting "loop invariants" of the motion - things that can vary but, if you wait long enough, always end up getting reset back to some original value. Those kinds of relationships fall through the standard statistical net, even though they are well-defined and common in computer programming. We can't even always detect things we are sure are there because we put them there.

Or, you can take all the scientific measurements you want to on my car, and extrapolate all you want to, but you will never predict that next Tuesday it will suddenly go to Chicago, which, by some miracle I can predict. That's a very real, observable, physical outcome that doesn't show up on the radar at all before it happens. What else is that radar missing?

Now, my car does not have a "going-to-Chicago-ness" property hidden inside it somewhere that is making it go to Chicago. Yes, the car burns gasoline, but the aspect of life that makes the car end up in Chicago has to do with the nature of human affairs, not with some property of the car. As T.S. Eliot said, "A thousand policemen directing the traffic / Cannot tell you why you come or where you go." There is more than one "reality" at work. The fact that the car is "just sitting there" tells you nothing, it turns out, about what it will do next Tuesday.

There is, literally, a gap large enough to drive a truck through, in the ability of Science to predict, using proven physical laws, what motion the center of mass of my car will take next week. The laws are correct, but incomplete, and in some ways very misleading.
This is not a small thing. When we are considering the nature of humans and society, this little "gap" is huge. It changes everything.

We are way past the simple case where we can simply neglect interactions between the parts. We are deep into "complex adaptive systems" where you can't understand a part without understanding the whole. Almost everything around humans is a created world, dominated by effects that Science, to date, has generally "left out" of the equations.

The gap is done silently, as an obvious, implicit assumption, that, "well, of course, we leave out the case of someone coming along and messing with the equipment." Fine, unless that is precisely the case you are trying to measure.

Even Science recognizes that it is an error to assume your conclusion. The right way to set up an experiment is to do all you can, in good faith, to try to disprove your conclusion and see if you can do it. You can't take a billion experiments, every one of which assumes no one is messing with the equipment, and conclude anything from them about the case where someone is, in fact, messing with the equipment. You threw out all the relevant data. You have to start over.

As we always ask in computing, when finding a "bug" or flaw in a program, "Well, if that's wrong, what else is wrong? What else slipped through this defect in our mental model?"

So, it seems to me, that taking such inadequate tools and boldly extrapolating to an infinite number of levels over all time and space is, well, a tad unjustified.

It's a fascinating question, but, right now, that method of coming up with an answer we want to rely on is not well developed or reliable. I'm all for developing it further as a way of better understanding all the forces on "me", and all the different worlds and dimensions and time-scales in which "I" exist, whatever that means.

The only conclusion Science can responsibly come up with so far is that this question isn't one that can be answered yet. We can't track down causal forces that take a decade to operate, let alone those that take centuries to operate, let alone larger ones.

The water molecule in the pipe doesn't understand the concept "municipal water supply", and probably never will. It's own little world cannot perceive the new niches and dimensions that come from "nowhere" into view when expanding the scale of the view outward.

If there are symmetries across different scales, it may be able to see an analogy on its own scale that suggests the right direction, but it cannot possibly span the other possible cases.

It's like trying to extrapolate out the room you are in now, and, on the basis of what you see around you, stating definitively what is going on in China right now, or on Mars.

Small steps, though. Let's see if we can solve some smaller, simpler problems first. And let's not neglect the chores or doing our homework or treating our neighbors well or dealing with our own personal short-comings because we are wrapped up in a grand challenge problem about the true nature of everything.

1 comment:

Anonymous said...

two items.
1) Analogy of the "downward" force in a rocket ship might be appropriate; (2) why stop at pain & suffering, which may be transient? Why is there death is probably the bigger question.