Sunday, August 03, 2008

Musings on the nature of prayer

If we simply take the concept that psychological context has a great deal of impact on both the results of thinking and the results of actions, then there is a place in our model for what religions call prayer. It's not a full match, but it is close enough to some aspects of prayer and meditation to be interesting.

Given that a billion people or so have reported success with prayer in some manner, it does seem worth investigating whether they are all deluded, or whether something is going on.

This context model does not require "God" in order to work. It simply says that if you can get yourself in some way into the right reference frame, with the right mental model of what is going on, that your outcomes will be better, as measured by you and as measured by the people around you and your organization, tribe, family, culture, nation.

That much is worth clinical study. There are good solid reasons in simple physics (well, General Relativity) for why context matters and there is already mathematically sophisticated equipment for analyzing and modeling this. At the current time this is not generally accessible, but computer front ends could easily do for this what they have done for statistics, namely, make the arcane suddenly available for use (and misuse) by the normal scientist.

The same is true for feedback effects and modeling situations involving feedback.

I am fairly certain that, if there were tools on the market that made analysis of feedback as easy as SAS or SPSS or Stata make analysis of regression, or Excel makes analysis of budgets, the psychological resistance to the idea would suddenly evaporate.

(Sadly, the field has not yet had the equivalent of Sir R. A. Fisher, to define the basic statistical tests so that everyone else can use them with only vague comprehension of what they do and how they work. And, there is not a stable / pool of feedback-icians one can assess or hire or build into the grant, or a Feedback-Effects-Core one can go down to and ask for assistance. Again, the existence of these would remarkably reduce the resistance to considering whether this is a valid construct or not.)

In any case, in a "weak hypothesis", a person doing a good job of psychologically putting themselves back into a strong social context should be observably different from a person who does not do that, even if this is "all in their head." That would be measurable.

In a "strong hypothesis" there is more going on in the world here than simply what is happening "inside the person's head." Odds of various outcomes are changing.

It is not necessary to buy the strong hypothesis to study the weak one. If there are in fact effects that cannot be explained by it, they will show up.

And, if scientists want to "debunk" the power of prayer, then it would be a marvelous move if they came up with something that worked even better.

So far, they neither accept that challenge nor have delivered something that works better for the average person than prayers of submission and reconnection.

"Reason" is put forward as the answer to everything, but it is remarkably devoid of motivational power or of regenerating a depressed spirit. Maybe psychoactive drugs is the solution suggested?

Saturday, August 02, 2008

Beyond Reason to Action

(continuing a series of posts in the last 3 days on "reason")
===========================================

There is one additional problem with "reason" that I have not mentioned that is immediately relevant, and that is that "reason" is so "academic."

What I mean by that is that reason has, by itself, absolutely zero action component. One could "reason" all day, or a million years, and it would never affect any external thing one whit.

In the real world, the one academics avoid, there is no point in thinking about something if you aren't contemplating action of some kind.

So, I just discussed the idea that the context in which we think or reason has an impact, and alters the outcome of such reasoning. The mechanism for this is pretty obvious, once you think about it.

However, context has a different function as well. We are, after all, cybernetic organisms. We perceive the world, map that into some model, determine some action to initate, initate it, and quickly repeat that loop endlessly, attempting to converge on some goal.

So my point is that context influences, and may alter or determine the outcome of any action.

One very specific context I have in mind here is the intention of the actor, assuming the actor is human. Along with that context, and mixed into it, is the mental model, i.e. framework, i.e. meaning associated with this action and possible outcomes.

So, it matters "what we think we are doing" in addtion to what we are doing. More specifically, what we think we are doing is in fact part of what we are doing, and colors it in some manner, almost epigenetically, as methylation states color genes on DNA.

This effect has anecdotally been reported by a great many people over the past several thousand years.

An example of this was the story I told, probably made up, of two stone-masons working in a Medieval church construction. One was doing poor work and one great work, and the supervisor went to see what was going on. He asked the one doing shoddy work what he was doing, and he replied, as if it was obvious and what a dumb question, "Building a wall."

Then he asked the one doing great work what he was doing and that mason replied, again as if it was obvious and a dumb question, "Building a cathedral."

Clearly, and probably undisputed, the meaning with which a human holds an action influences the outcomes of that action in many ways.

What is surprising is what a powerful effect this seems to be, even in cases where you cannot see how the work itself is being influenced, but you can observe the overall outcome.

Here, like M. C. Esher's waterfall in some cases the large-scale outcome of building a cathedral is a cathedral, and the outcome of "building a wall" is nothing.

The same seems to be true in software engineering. The mechanism is not clear here, and one would think that "Java is Java." Yet, seemingly good code written by angry people works less well than similar appearing code written by happy people.

It is not clear exactly how the mood gets under the skin of the program, but to myself, and to some other senior IT people I discuss this with, this effect is very real, and they know immediately what effect I'm talking about.

In any case, I wanted to open up the door for further discussion about the failings of pure reason, and one of them is that humans are never pure anything, and so being reasonable, even being able to totally prove one's point, is not necessarily a good guide to the outcomes of action. It is an often necessary but also often not a sufficient criteria.

The busy executive who has his secretary buy flowers for his wife, versus one who buys them himself, may find a very different result once he gets home.

It is not my point to analyze all the possible pathways and mechanisms involve here. My point is to raise another dimension in which pure reason, by itself, is insufficent for daily living, and not a very good guide to action.

Certainly, for every one of the cases I mention, something can be defined as the locally causative agent. I am asserting that if it wasn't that agent, it would have been another one, because the actual cause of the outcome is only expressing itself through whatever is handy. We need to look further upstream to actually see what is going on.

The difference, incidentally, is not just in the external world. The meaning of some action or activity greatly influences the ability of someone to carry it out, and affects how they feel afterwards - exhausted or elated.

Again, to understand what is going on we have to look beyond "content" to "context" to get a complete picture. I'm asserting that it matters to more than a court of law. It matters to phsyiological outcomes in the body of the person acting. It matters to the impact of the action, seen from afar, on the outside world.

It is as if, mentally connecting an action and a larger game plan or framework of meaning somehow completes BOTH the action and a little bit more of the framework's agenda than would have occured otherwise. The effect may be "negligible" locally, but over time, it can accumulate to something and even accumulate to totally dominate the resultant, seen from afar.

The key point here is that this effect which I allege, anecdotally, is not within the realm of "reason" for action. Listing the reason for an action is a dormant activity. Having an intent is very different.

Let us suppose for a minute that this is true. The advocates of reason would argue that such "emotional" content is baggage, in the way of a pure life, and should be discarded. I would argue that this "intent" or "will" thing is, in fact very real, and very important, because not only can it be a thing that destroys a desired outcome -- it can be the only thing that can create a desired outcome.

Great victories are never won on the basis of "reason". If that is true, then we should be paying attention as well to something else, or we might redefine (reason + something else) as "reason-2.0".

Wade















more problems with reason

A profound and relevant insight was shown in a Charles Schulz "Peanuts" cartoon strip one day.
Snoopy, the beagle, is lying on top of his doghouse and pondering. He says...

"Did you ever notice,
that if you think about something at 2 AM,
and then again at noon the next day,
you get two different answers?"

Snoopy manages to capture one of the problems with reason, as practiced today - namely, both the meaning of "facts" and the outcome of logical reasoning are context sensitive.
"Reasoning" alone is not sufficient. One has to track not only the math of how the contents relate to each other but also how the contexts relate to each other.
Probably this would require a new kind of computer which was a context processor, not a contents processor.

This is a very familiar problem in Einstein's General Relativity, and it is not insurmountable.

It is also a familiar problem to any mediator or serious negotiator, who knows that the meaning of a phrase or word or step is highly dependent on context.

In Relativity, it is clear that perceptions are a very specific combination of something which might be observed, and some type of distortion of the observer's measuring equipment. This means that the meaning of an observation is relative to the reference frame in which it is being observed.

This effect is terribly misunderstood by people, and misquoted as "Everything is relative" and "no view is right" and "There is no underlying reality." All those are wrong. What Einstein said is that, you need to make corrections to observations that depend on the type of distortion of the observer's world, and, after you do that, any two observers should agree on what they see.

So, there is very powerful mathematical formalism, originally involving "tensors", that allow these effects to be described relatively easily -- compared to the alternative of having to write down 256 components of the curvature tensor every time you said something about it.

This is good news, because people have a very similar problem. What we "observe" is in no way what is actually "out there" -- it is a similar conflation of our current set of biases, prejudices, mental blind spots, stereotypes, and what is "actually there". As in General Relativity, we cannot see our own blind spots, and are unable to detect where our biases have taken over and warped our judgment which, to us, looks just fine.

This, of course, is why "double blind" experiments are good if you can get them, so that the unconscious biases of the experimenter, regardless how hard he or she tries to be careful, do not alter the final conclusions.

Humans are excellent at mentally "papering over" their blind spots with what they think should go there, and not being able to detect that they have done this.

Anyway, "reasoning" alone is not sufficient. One has to track not only the math of how the contents relate to each other but also how the contexts relate to each other.

You could, I suppose, say that,well, yes, this is just reasoning. No, it is not "reasoning" as currently understood and practiced. The current interpretation of reasoning is that space is flat, there is no curvature, scientists are capable of being unbiased, and if A+B=C over here to this scientists, then it will still equal C over there or to a different scientists, without correction for where it was observed.

In fact, a basic assumption of physics is that it doesn't matter where you are, or what time you start, or what direction you are pointed, you should still measure exactly the same physical laws. I'd add, they also assume that you get the same reading regardless how large a ruler you use.

In reality, it seems that there is context dependence, and life is fractal or at least non-uniform, and what is measured at large scales may be exactly the opposite of what is measured at small scales, for the "same" physical event. (See post on "hybrid images" for examples, such as "Marilyn Einstein" or the "happy/angry faces" that look different, depending on how far you stand back from the picture.)

So, at a minimum, it would help if people understand how much of what we think we see is actually a function of our viewpoint, our biases, our mental models, our context, what we had for breakfast, what our spouse had for breakfast, etc.

Responsible observers making careful observations in different contexts will get different answers.

Responsible logical thinkers, starting with the same facts, in two different contexts, will assemble them into different, and possibly contradictory conclusions, regardless how carefully they work.

This is a crucial fact to sort out, and even it is not dealt with well by science, let alone by laymen. All of these effects can be demonstrated, and can be counteracted, but seldom are.

All of these effects can be adjusted for, to remove them from the final conclusion, by different people, but again, this kind of meta-reasoning, or reasoning about the impact of deterministic context in which reason is applied, is not currently the usual practice on Earth in 2008.

Even it is blocked by those who refuse to consider the possibility that there is something that can be improved about the process they use to reach conclusions and to consult with ech other.


Wade

further beyond reason



(note - this post extends the prior post here, "Beyond Reason" )

There are several other dimensions in which this argument can be supported.

The first is models of any kind of collaborating agents, as I mention in my comment to my last post.

For those who insist on purely academic and "scientific" examples, I'd suggest the model of connecting many legacy computer systems together in a company and trying to get that mess to work.

The "tinker toy" model of little balls (legacy systems) connected by little rods (messaging pathways) simply doesn't work, and breaks down pretty rapidly as you attempt to scale it up. Having each agent determine its own version of "truth" doesn't work.

However, having each one chip in to the discussion about accepting each fact does - in what is known as a "two-phase commit."

A true "distributed operating system" is one in which each agent gives up some level of control and self-determination in exchange for the entire thing working at all.

It's a model that works. It's a model close to the way cells relate in the human body, or neurons in the visual system.

It should be explored more as a model for civilized behavior.




A second strong support are the deviations from Turing's hypothesis. Computability is generally determined on the basis of a Turing Machine model, which is complex to explain but basically says any computation can be reduced to a very long tape of ones and zeros and a machine that reads them and decides which way to move the tape.

The problem with applying that model to human beings is that humans have finite energy, bandwidth, and attention span. In fact, often a very small volume in communication space. Any communication that has any longer-wavelengths in it that fit in such a small box will simply be lost and are not possible. No "long-term" issues will "make sense" at all.

"Reason", if by reason we mean the application of logical steps to a series of strings of symbols, is thereby limited for humans to what can be said, via shared symbols, in a short period of time. And, for the symbols to be shared, they have to at least have been learned, which means they can take no more than 100 years to learn. Again, we have huge limits here on what can ever be communicated that way, even neglecting how fast things are forgotten and partial understandings evaporate. In reality something that takes over a day to communicate is highly unlikely, especially if it has high information content, i.e., is very surprising and doesn't fit preconceived notions.

Furthermore, human brains and perception fill in gaps, leap ahead, resolve ambiguous phrases into stereotyped meanings short-circuiting the whole process. The communication is, in sort, noisy, misunderstood, mis-resolved, and decays rapidly.

The power of reason to accomplish synthesis of meaning and action and synergy in such a context is demonstrably poor. And, we have no other context.

On the other hand, if we go to higher dimensionality communication, such as image processing or higher, many of these bandwidth limitations can be pushed back at least an order of magnitude. These communications would be somewhat holographic, conveying many related idea and their relationships in parallel, not in serial order. This works. "A picture is worth a thousand words".

Actually, I think virtual reality and immersive experiences can be even more compelling and complete than pictures, but that's a different post.

I simply don't see any other conclusion than "Therefore, reason is not a good tool to use for stabilizing civilization and getting the mutual destruction to stop."

Reason may be part of a set of tools, but it is not complete, and I can see no empirical evidence that it works at large scale, at all, ever. Our toys grow more sophisticated, yes, but our conversation about them grows less sophisticated, increasing daily the risk that our toys will be misused.

Frankly, we are stymied by the simple problem of making N plus 1 people more powerful than N people at deciding something or perceiving something. We have what is referred to as "committees" and God help us if the committee size is over 5. We are without tools to help. Unassisted appeals to reason do not seem to do the trick, in realistic contexts.

Third, all reasoning assumes, essentially, that the metric space in which we are working is flat, and that the value at any given "point" in space of "truth" doesn't depend on the pathway we used to get there. Or, more precisely, that the value we show in our tabulator of taking the present facts and reasoning with them to that point should be, we think, independent of the pathway we took. We can use B implies C before or after we use A implies B and end up in the same place.


First, this assumes truth is single valued. Even that is not clear.

Second this assumes that the metric we are reasoning in, some kind of Hilbert Space, is flat -- so that, at the end of a day, we do not end with something like M.C. Escher's waterfall -- that is locally reasonable and globally unreasonable.


In reality this appears to be unjustified. Small factors that we see as "ignorable", in a world with millions of cycles of feedback, turn out to be more important that what seem locally to be "big" factors. Tiny differences result in hurricanes and tornadoes out of "thin air". This matters.

In any Hilbert space, I believe there is always a locally flat coordinate system that works locally, but this does not mean that the entire space is flat, nor that it can be covered by a flat metric.
I find this is the strongest argument at all, involving geodesics and "lines of reasoning", but I fear it is the least accessible, unless the reader has experienced metric spaces in some mathematical context.

In point of fact, reality seems to me to be more fractally complex than flat. From higher up in an organization, the myriad of details that lower levels have to deal with are completely invisible, recursively.

And truth seems to be fractal, in that the size of ruler used to measure it, and the scale of time and space used as a framework, seem to change the measured value. (See hybrid images post).

So, between a world way more complex than desired, and a reasoning power collectively that is way weaker than desired, the ability of reason alone to solve humanity's problems seems in doubt.

This is not to say that reason is not "locally valid".

But to deal with larger scale issues, we need some tool that is not bandwidth limited by the way humans think and perceive and communicate.

I'm not sure what it is, but pretty sure what it is not: It cannot be serial messaging in nature.

What we call "reason" and "reasoning" are incomplete, and will never span the problems we are trying to solve, let alone span them in the time we have left to work on them before we annihilate each other and our planet.
Reason and Science have worked very well on problems that could be flattened into a flat space without damage, such as building machines. They have not worked on problems that cannot be detached and flattened, such as every social problem. This will not be fixed by waiting. It will not be fixed by "trying harder". So, it is doubly cursed - even if "there" made sense and covered all cases we care about, you cannot "get there" from here. Reason alone will not work.
Mindless defense of "reason" as the only pathway to our future
hurts more than it helps, because it prevents adequate research into methods of finding truth that are better than reason, and that converge much more rapidly in the actual world with actual humans, messy politics, corruption, legacy hatreds, etc.


Denial that anything could possibly be better than reason
is a religious statement, based on faith, not solid facts or even solid
theory.

No one could possibly know that, for almost any definition of "better."
This is not an argument to go back to superstition, but an argument to go forward to actual mature capacity to face and deal with hard problems together. Such maturity undoubtedly uses reason, but was not and will not be created by reason. There is something else at work.
The astoundingly poor record of science to cure the worlds ills, and, in fact, to seemingly only solve some at the expense of making others far worse, argues that science is not close enough to the right path, or the right speed, to be left unexamined.
If scientists do not recognize this issue, and continue to paint all their critics as some kind of blind, irresponsible, religious zealots, their arrogance will bring about their own extinction, along with everyone else's.
Seriously - if everyone in Congress simply had 8 more years of math and science, do you contend we'd now have a balanced budget? If they all had dual PhD's in math and science, would we be home free? I have to look at the University of Oxford, which can neither balance it's budget nor select an email system, and say, "No."
No, because the problem is way bigger, deeper, and wider than that. Reason itself only works in a social context, and reasoning is not the process by which social contexts have ever been decided or determined. Reasoning about Oxford will not "fix" Oxford, and I can't begin to grasp the solid basis for an assertion that it could, based on the contrary evidence of the last 1,100 years.
We can't even make committees of 5 that work, people.
It's time to address the actual problem, not this straw man. Something else is wrong. Get "unreason" out of our faces so we can see what it is.
Wade

Beyond Reason

(Another letter I sent today to the editor of NewScientist ).
================================
Thank you for the excellent coverage of "The forbidden question - What's wrong with reason?" (July 26)!

I would suggest that our culture has learned a lot from time time of Kant's quote (page 42), "Immaturity is the inability to use one's understanding without guidance from another." If culture is akin to a person, what that describes is perhaps an 18 year old young adult.
What we are sorely lacking is the vision of advancing past that to the equivalent of marriage or civilization. They say that a good marriage is one between two adults who are capable of living separately just fine, but who prefer to live together.

They have gone beyond the need for independence and have moved on to mastering interdependence -- what might also be called "unity with diversity." Neither one dominates and both benefit.

American language has no word for that, as "adult" or "mature" has been converted to mean "able to purchase cigarettes and pornography". Clearly this is not a subject of much discussion.

This is extraordinarily relevant, because it is not in the independent reasoning category that our civilization is breaking down, so much as in this re-linking step into a synthesized larger reasoning, perceiving, and acting unit -- a meta-person.

Our schools still teach the "Rambo" approach to great individuals, but across the board we see the start of realization that we need better group work, or team work, or collaboration, or cooperation, or unity, or something over in that general direction. We need something in which no one person (or corporation, or country, or concept) dominates everyone else to the extinction of the other.

It seems to me that many of the attributes of human beings which appear to be "bugs", so far as pure reason go, will turn out to be "features" that allow this meta-synthesis to take place and work. For this we need empathy, the capacity to be unconsciously moved by the others nearby to leave one's mind and hands free to work on the problem at hand while a subconscious unity is messing and mending whatever it is down there that we lack.

And, we need to be joyous about that activity, and welcome it -- both of which have no meaning at all in pure reason. Reason appears suited for computers that wish to network. Something larger is required for humans who wish to civilize each other.

It is not back to authority we seek, not an abandonment of maturity and reason, but going forward to the joyous interdependence that is the basis of civilized living, and true "maturity."

And, yes, this topic is in Chomsky's category of forbidden subjects in the US -- anything even remotely like extolling the virtues of it would be smeared as socialism or communism or terrorism or the smear of the day.

But we know it's right. We can feel that it's right. We are torn in education and teach both "Do your own work" and "Learn to work well with others." The breakdowns in our teams, families, companies, and societies are in this realm, not in "reason". Reason is the messages but reason does not build the communication lines - something else does that. We need them both.

The prestigious US Institute of Medicine, in Crossing the Quality Chasm, notes the critical role that teams play in health care. In fact, in all critical high-risk industries, it turns out that honest, open teamwork and "psychological safety" are required, not optional. Even the US Army leadership doctrine requires superiors to, gasp, listen to their subordinates and build relationships and trust and integrity and character, as being more important than technology to success in battle. (see links to the US Army Leadership Field Manual in http://newbricks.blogspot.com/2006/12/us-army-leadership-field-manual-fm-22.html as the certificate in the army site seems to be broken.)

Again, if these are part of "reason" they are not very accessible or explicit. There is ample empirical evidence that these human traits are crucial to survival, success, and any plan for thriving instead of mutual annihilation amid fragmentation.

Please stop destroying those values in the quest for the goal you have for "reason."