Agreement (was Re: Gene Gendlin)

CSGnet Archive CSG2008
1

Bill Powers wrote:
[Martin Taylor 2008.09.15.14.44]

[From "bill Powers (2008.09.14.0417 MDT)]

There always seems to be more agreement between us than is evident in our writings.

Best,

Bill P.

Isn't that rather to be expected, if PCT is more than just an idiosyncratic idea you are trying to promulgate? It's not as though you are free to propose just any crazy idea and have it correspond in any reasonable way both with everyday observation and with the main body of science. You have observed nature and refined your ideas by noticing what works. If PCT is the basis of a correct theory (which is something we will never know), then anybody else who starts from the same general conceptual structure is likely to come into pretty good agreement with as to the basics. And remember, I found PCT having independently developed Layered Protocol Theory, which I discovered to be only a special case of PCT with two independent protagonists.

If PCT is correct, then control of perceptions does not usually result in observable overt behaviour if the perceptions autonomously stay near their reference values. Mapped onto e-mail exchanges, this means that we don't tend to write a lot when we think the other already accepts what we would write. We write about what we are not sure the other accepts, and particularly about things with which we disagree but for which we think we might achieve agreement. So, it is natural that "There always seems to be more agreement between us than is evident in our writings."

Having said that, there is a difference in cognitive style between us. It's more a matter of degree than of kind. It might be exemplified by a comment in [Bill Powers (2008.08.31.1145 MDT)] responding to [Martin Taylor 2008.08.31.11.30 ]:

"[MT] That is the basis from which I want to develop further ideas, such as the notion that to increase the likelihood that the legal system provides "a variety of means of control sufficient to assure that if one path to a higher goal is forbidden, another path to it still exists" requires that government should tend to be opposed to power and wealth rather than to be controlled by power and wealth.

[BP]Yes, but one has to be careful about trying to deduce things that one believes anyway, as I've been nagging Rick about. The more I am delighted and reassured by the outcome of my own logic, the less I trust it."

This is a shrewd comment, and one that we all should keep in mind. The difference in cognitive style is in the preferred us of the word "logic". You do tend to want proof of things, and if they are not proven, then they are not useful to you. My preference is more a Bayesian balance of likelihoods. If two possibilities offer themselves, then I tend to follow the more likely, but not to discard the other unless its likelihood is very low.

Each of these approaches has its own danger. The "pro-logic" approach tends to cut off exploration, whereas the "pro-likelihood" approach is liable to follow paths that leave one "delighted and reassured". The Bayesian approach tends in that direction because Bayesian probabilities include the subjective prior -- whatever you thought more likely before getting data has an advantage in the post-data analysis.

Parenthetically, the conclusion that led you to your comment was actually one that surprised me. I had long believed that it is not good for governments to be in the pocket of the wealthy or of a power structure such as that of the Russian government, but I had assumed that neutrality to wealth and its associated power would be the optimum. To come to the conclusion that optimum government should be opposed to the power of wealth was a surprise, and not one that is altogether welcome.

It's not a logical conclusion, because the steps in the argument are based on likelihoods (which are always subjective), not mathematical certainties. It's an exploratory conclusion, open to change based on further evidence, whereas a logical conclusion is open to change only on alteration of the premises or a discovery of fault in the logical process.

Anyway, as far as agreement or disagreement between us goes, I think that there is a very large body of agreement about the core of PCT. Where we often disagree is in this realm of exploration out from the core, particularly in exploring the interactions among many control systems in complex environments. Many of those disagreements seem to revolve around misunderstandings rather than disagreements of substance, though I'm not sure yet where we stand on bottlenecks of degrees of freedom, though, as to whether there's just misunderstanding or disagreement on substance.

When we are dealing with disagreements about a science, we can ask nature to arbitrate, but only if we can define the question.

Martin

2

[From Bill Powers (2008.09.15.1859 MDT)]

Martin Taylor 2008.09.15.14.44]

Isn't that rather to be expected, if PCT is more than just an idiosyncratic idea you are trying to promulgate? It's not as though you are free to propose just any crazy idea and have it correspond in any reasonable way both with everyday observation and with the main body of science. You have observed nature and refined your ideas by noticing what works. If PCT is the basis of a correct theory (which is something we will never know), then anybody else who starts from the same general conceptual structure is likely to come into pretty good agreement with as to the basics. And remember, I found PCT having independently developed Layered Protocol Theory, which I discovered to be only a special case of PCT with two independent protagonists.

I like that reasoning. Yes, it is reassuring that you independently came up with very similar concepts about control of perceptions and hierarchical control. This shifts the center of gravity out of both our heads and perhaps into that inaccessible region where things actually happen.

You do tend to want proof of things, and if they are not proven, then they are not useful to you. My preference is more a Bayesian balance of likelihoods. If two possibilities offer themselves, then I tend to follow the more likely, but not to discard the other unless its likelihood is very low.

Well, I like to think of this difference as my wanting higher probabilities of predicting correctly. Not too different from your strategy, in the long run.

Each of these approaches has its own danger. The "pro-logic" approach tends to cut off exploration, whereas the "pro-likelihood" approach is liable to follow paths that leave one "delighted and reassured". The Bayesian approach tends in that direction because Bayesian probabilities include the subjective prior -- whatever you thought more likely before getting data has an advantage in the post-data analysis.

I don't suppose we can avoid that. But I never start out with the "pro-logic" approach -- I assure you that I'm as confused and dependent on intuition as anyone when trying to work out something new. My image of the way my mind tackles problems is that of a print-developing tank under a safelight. The picture shows up as blurry patches of light and dark that make no sense at first, but then gradually take shape all over the print at the same time until suddenly a pattern jumps into view -- still fuzzy, but sharpening all the time. So I think I, too, go through the "balance of likelihoods", but as a phase, not an end product. My preference is to leave the print in the tank until I'm pretty sure what it shows, before inviting anyone in to see the pretty picture.

Parenthetically, the conclusion that led you to your comment was actually one that surprised me. I had long believed that it is not good for governments to be in the pocket of the wealthy or of a power structure such as that of the Russian government, but I had assumed that neutrality to wealth and its associated power would be the optimum. To come to the conclusion that optimum government should be opposed to the power of wealth was a surprise, and not one that is altogether welcome.

It's not the power of wealth to which I object, but the concentration of power in a few individuals that goes with control over great wealth. I have never met or heard of anyone smart enough to use that power without leaving various kinds of injustice and devastation behind. As soon as people accumulate some amount of wealth, they start imagining (often correctly) that there are other people who want to take it away from them. Then the walls go up, the gates are locked, and the class warfare begins. Rich people are more fearful than poor people.

Anyway, as far as agreement or disagreement between us goes, I think that there is a very large body of agreement about the core of PCT. Where we often disagree is in this realm of exploration out from the core, particularly in exploring the interactions among many control systems in complex environments. Many of those disagreements seem to revolve around misunderstandings rather than disagreements of substance, though I'm not sure yet where we stand on bottlenecks of degrees of freedom, though, as to whether there's just misunderstanding or disagreement on substance.

When we are dealing with disagreements about a science, we can ask nature to arbitrate, but only if we can define the question.

On that we are in definite agreement. At least we should be able to simulate the principle being proposed to see what would actually happen according to the rules we have put into the system. Thought-experiments are good lead-ins to an investigation, but they are never complete until the demonstration agrees with the imagination, or vice versa. Imagination is not very good at leading to correct predictions when the logic gets beyond simple syllogisms.

Best regards, old friend.

Bill P.