Why we do that

[MartinT]

[Martin Taylor 2010.04.17.17.20]

[From Bill Powers (2010.04.17.1225 MDT)]

Martin Taylor 2010.04.17.10.06 --

MT: anything about a structure that increases the likelihood its continuing to exist will be more likely to be found in patterns that do exist at any moment in time than are things that decrease the likelihood of the pattern continuing to exist. It's called "survival of the fittest" or "Natural Selection".

In PCT, there is a mechanism for sustaining those "high fitness" structures. It's called "reorganization". Any living entity has many variables, typically biochemical ones, that must be maintained within reasonable bounds of some optimum or the entity dies. If it has died before propagating a reasonable facsimile of itself, its internal structures will not be found in future living entities. So, these biochemical "intrinsic" variables must have some mechanism to support their maintenance, and that mechanism must act on the world outside the entity as well as on its internal world network of chemical reactions.

Why must the mechanism act on the external world? Because every reaction involves an increase of entropy, and the entity needs an input of low entropy and a sink of higher entropy (I'm talking physics, here, so these are simply abstractions, not observations, and therefore are unconvincing to Bill P. They might fit with Bruce G's perception of the world, though). That is equivalent to saying there must be an energy flow through the entity. The energy flow must be regulated, as too much too quickly could be as damaging as too little too late. So any living entity must act on the outer world at least so far as to influence how and when the entity gets an appropriate ration of energy.

BP: I've actually offered an alternative to selection by reproductive success or availability of an energy supply: my proposed basic criterion is accuracy of replication (AoR for short). Species or life forms that survive are those that are most capable of resisting all influences that reduce accuracy of replication.

I have understood your point, but I haven't see it as an alternative. Rather, I see it as a mechanism. There would be no survival value in accurate replication of a structure that did not help the entity to survive to the age of reproduction. But there would be survival value in accurate replication of a structure that had helped the entity to survive to the age of reproduction. Natural selection therefore can use as a mechanism anything that improves the accuracy pf replication.

I don't see where "availability of an energy supply" enters the discussion. I mentioned as a necessity for the deveopment of a perceptual control structure that the controlled perception(s) must serve to regulate the energy throughput, ensuring that it is not too great and not too small, but if there is a problem with availability of the energy supply, it certainly won't be too great!

Other than that I didn't mention the mechanism of accuracy of replication, did I reasonably represent your view on how and why perceptual control structures come about? If not, then I have to say that it is my own opinion, though I thought I got a lot of it from your writings.

Martin

[Martin_Lewitt]

[From Martin Lewitt (2010.04.17.1525 MDT)]

Species survive by being both robust and evolvable. We retain most the benefits of accuracy of reproduction through of redundancy, phenotypic canalization and internal and developmental homeostasis, i.e., control. But the redundancy and sexual recomibnation allow an underlying reservoir of variation and potential for evolvability to be maintained. However similar species in the fossil record appear to species and forms alive today, the similarities may be superficial and the ability to interbreed may not have been retained.

However, I think this discussion has strayed a bit from the relevant elements. I wonder if PCT is falsifiable. The robust but unobservable internal mental mechanisms give it some of the same tautological characteristics that were noted in natural selection. Perhaps PCT is a "metaphysical research program", in the sense that Karl Popper proposed for evolution:

Issue of Darwinism

Karl Popper

-- Martin L

···

[From Bill Powers (2010.04.17.1225 MDT)]

Martin Taylor 2010.04.17.10.06 --

MT: anything about a structure that increases the likelihood its continuing to exist will be more likely to be found in patterns that do exist at any moment in time than are things that decrease the likelihood of the pattern continuing to exist. It's called "survival of the fittest" or "Natural Selection".

In PCT, there is a mechanism for sustaining those "high fitness" structures. It's called "reorganization". Any living entity has many variables, typically biochemical ones, that must be maintained within reasonable bounds of some optimum or the entity dies. If it has died before propagating a reasonable facsimile of itself, its internal structures will not be found in future living entities. So, these biochemical "intrinsic" variables must have some mechanism to support their maintenance, and that mechanism must act on the world outside the entity as well as on its internal world network of chemical reactions.

Why must the mechanism act on the external world? Because every reaction involves an increase of entropy, and the entity needs an input of low entropy and a sink of higher entropy (I'm talking physics, here, so these are simply abstractions, not observations, and therefore are unconvincing to Bill P. They might fit with Bruce G's perception of the world, though). That is equivalent to saying there must be an energy flow through the entity. The energy flow must be regulated, as too much too quickly could be as damaging as too little too late. So any living entity must act on the outer world at least so far as to influence how and when the entity gets an appropriate ration of energy.

BP: I've actually offered an alternative to selection by reproductive success or availability of an energy supply: my proposed basic criterion is accuracy of replication (AoR for short). Species or life forms that survive are those that are most capable of resisting all influences that reduce accuracy of replication. A species cannot continue to exist if its replication process is inaccurate or is affected by variations in the environment. We can define fitness simply as the ability to reproduce accurately; increasing the ability to reproduce will only lead more quickly to extinction if replication is not accurate. Reproductive success alone can't be the essential criterion: it's necessary but not sufficient.

The same would apply to energy ration as the criterion. If an species doesn't reproduce accurately, an easily available energy supply will not help it to survive. Again, the energy supply is necessary, but not sufficient for survival of a species.

If a given organization fails to protect against environmental disturbances in regard to effects on AoR, the result will automatically, without the need for any special mechanisms, be a variation in the organization. If the new organization still doesn't replicate accurately, another change will occur, and so on -- either to extinction, or to a new form that once again reproduces accurately. The reorganization principle is inherent in the very nature of reproduction, and in the very interactions of a species with its environment. This vastly increases the probability that we are not alone in this universe.

The criterion of AoR could, in fact, result in a decrease in reproductive success in cases where overpopulation threatens a species. It could result in a species reducing its size and energy requirements when the energy supply dwindles. Both of those changes could increase the chances of accurate reproduction and survival of the species, whereas an increase in either reproductive success or energy intake without any change in AoR could reduce the chances of survival. The one thing that all species MUST do in order to continue in existence is to reproduce accurately.

The Darwinian focus was on trying to explain the vast variety of species and the obvious effects of natural selection. It was assumed that if it were not for occasional mutations, species would simply continue as they are.

My focus has been on trying to explain the incredible fidelity of reproduction over geological stretches of time. I have concluded that this fidelity can be maintained only by active control of all variables that tend to disturb replication. This active control will arise simply because failure to develop it will lead to extinction, leaving only those lines that can protect themselves the best against the forces that tend to make replication inaccurate, or even repair the damage of disturbances after they have occurred. Think "repair enzymes."

Best,

Bill P.

[rsmarken]

[From Rick Marken (2010.04.17.1550)]

Martin Lewitt (2010.04.17.1525 MDT)--

�I wonder if PCT is falsifiable.

Yes, it is. Every experimental test of PCT (many of which are
described in my book _Mind readings: Experimental Studies of Purpose;
you should get a copy) could produce data that is inconsistent with
the predictions of the theory (and, thus, would "falsify' it). So PCT
_is_ falsifiable, it just hasn't been falsified...yet.

Best

Rick

···

--
Richard S. Marken PhD
rsmarken@gmail.com
www.mindreadings.com

[Bruce_Gregory6]

[From Bruce Gregory (2010.04.18.0800 EDT)]

[From Rick Marken (2010.04.17.1550)]

Martin Lewitt (2010.04.17.1525 MDT)–

I wonder if PCT is falsifiable.

Yes, it is. Every experimental test of PCT (many of which are
described in my book _Mind readings: Experimental Studies of Purpose;
you should get a copy) could produce data that is inconsistent with
the predictions of the theory (and, thus, would "falsify’ it). So PCT
is falsifiable, it just hasn’t been falsified…yet.

BG: It would seem to me that such an experiment can only show that the experimenters had failed to identify a controlled perception, not that such a controlled perception does not exist. Unless the experimenters can demonstrate that they have exhausted the set of possible controlled perceptions, no experiment can be said to falsify PCT.

Bruce

[MartinT]

[Martin Taylor 2010.04.18.11.00]

[From Bruce Gregory (2010.04.18.0920 EDT)]

[Martin Taylor 2010.04.17.10.06]

So I argue that Perceptual Control is an intrinsic feature of life, and that its existence is the only defining feature of life. If an entity can be shown to have perceptual control, it is thereby shown to be living. If it does not, its continued existence is due either to material structural strength (an iron bar, a rock, the skeleton of a skyscraper) or to self-organized structures in an energy flow (a set of which living things are a subset), such as a whirlpool in a river. Whirlpools often spawn "child" whirlpools in the form of downstream eddies, but cannot be reasonably called alive. They control nothing.

BG: You seem to be saying that my thermostatically controlled geothermal system is alive. Is that what you meant to say?

No, that's not what I meant to say, and clearly there's more to defining life than I said. The boundaries of "life" are pretty fuzzy at the best of times, and I don't know whether a precise definition is either possible or useful. Is a virus alive? Some say it is because it replicates (their definition of life), some say it isn't. "Life" is a bit like "intelligence" in that way. We used to define "intelligence" more or less as the ability to do arithmetic, but when the first computers came out they could do arithmetic better than humans could, so that definition had to be scrapped. After various redefinitions, it became "what humans can do that computers can't", but that meant face recognition was intelligent. As far as I can see, the current state of that definition is like the definition of pornography -- "I know it when I see it".

That said, it's clear that my proposed definition covers your thermostat, and that no reasonable person would claim the thermostat to be alive. I then thought of whether "built" versus "grown" might be part of a definition, but even if that worked, how would you define the boundary between those two concepts in an age of self-organized nanostructures? If biochemists build a DNA molecule from stores of carbon, hydrogen, nitrogen and oxygen etc., and develop a sustaining matrix within which the synthetic DNA can express itself in protein, and so forth, would the result be considered "life"?

My next idea (this morning) was that the perceptual control system necessary for life must support the continued existence of some other active energy-using structure. That might work, but it carries the risk of being circular, and the further risk of making it hard to determine just what the energy-using structure might be. In the case of your thermostat, it supports the continued existence of Bruce Gregory.

So, right now, I have to modify my proposal to say that the existence of perceptual control is a necessary but not sufficient condition for an entity to be alive, leaving the specification of sufficiency as "unresolved".

In the tradition of answering a question with a question, could an autonomous software system that used perceptual control to maintain its own structure against software disturbances over a network (in the manner of many Sci-Fi stories and movies) ever, even in the distant future, be considered alive? If so, could a computer worm be the software equivalent to the pro/pre-life biological self-organized structure, from which real software life might evolve? In other words, would it be reasonable to consider the possibility of life in silico, completely parallel to carbon based life? If not, what about the definition of life would deny that possibility other than a reference to the earlier definition of intelligence "its something computers can't do"?

Second question, specifically to Bruce G. Did my essay on "why" that started the thread under its current subject line answer your comment about PCT not addressing why we do what we do?

Martin

[rsmarken]

[From Rick Marken (2010.04.18.1000)]

Bruce Gregory (2010.04.18.0800 EDT)-

Rick Marken (2010.04.17.1550)--

Martin Lewitt (2010.04.17.1525 MDT)--

�I wonder if PCT is falsifiable.

Yes, it is. Every experimental test of PCT (many of which are
described in my book _Mind readings: Experimental Studies of Purpose;
you should get a copy) could produce data that is inconsistent with
the predictions of the theory (and, thus, would "falsify' it). So PCT
_is_ falsifiable, it just hasn't been falsified...yet.

BG: It would seem to me that such an experiment can only show that the
experimenters had failed to identify a controlled perception, not that such
a controlled perception does not exist. Unless the experimenters can
demonstrate that they have exhausted the set of possible controlled
perceptions, no experiment can be said to falsify PCT.

Identifying controlled variables is not really a test of PCT.
Remember, phenomena phirst! PCT is used to model behavior only after
you've determined that the behavior under study actually involves
control. So you would not apply PCT to the behavior of, say, a mass on
a spring. The position of the mass is clearly not under control (not a
controlled variable or related to a controlled variable) as
demonstrated by the fact that disturbances to the position of the mass
are completely effective (and perfectly predicted by the open-loop
causal model of physics). You only apply a control model when you know
that the behavioral phenomenon involves control, which means that
there is, in fact, a variable (or many variables) under control, even
if you don't know precisely what that variable is (or what those
variables are).

Tests of PCT involve seeing how well the PCT model accounts for the
observed controlling in comparison to other models of the same
phenomenon. The fit of the PCT model will, indeed, be improved by
finding better definitions of the controlled variable (which is
equivalent to finding better definitions of the perceptual function in
the model). But there are other aspects of the PCT model that could
influence how well it fits the data. But the main goal is to compare
the PCT model of control to other models. We usually compare PCT to
the open-loop model of conventional psychology, which is kind of not
fair since the open-loop model is not a model of control. But we hope
to show, by demonstrating the overwhelming superiority of PCT model,
that the behavioral phenomena under study in psychology are control
phenomena that are best handled by a control model, like PCT.

I think the problem here is created by the word "falsify". I think it
would be difficult to come up with an experiment on control behavior
that produces results that "falsify" PCT in the sense that they are
completely incompatible with PCT; where, for example, the predictions
of the PCT model are as poor as what we would get by simply running a
random number generator and matching its output to the observed
behavior.

If by "falsify" you mean showing that the whole PCT model of control
is wrong, then I agree that PCT is not falsifiable in this sense. But,
then, neither is Newtonian physics. It turns out that we now think
that Newtonian physics is false, being replaced by Relativity physics.
But the fact is that Newtonian physics works just fine for velocities
well under that of light. So we still use Newtonian physics to build
bridges and fly satellites (though we do use Relativity to correct the
clocks on our satellites, which are affected by the speed of the
satellite).

I think eventually PCT will be "falsified" in the same way that
Newtonian physics has been falsified; the basic equations of control
will probably hold up but there may be experimental tests that produce
results that "falsify" assumptions about the way control systems are
organized (it might not be neatly hierarchical, as per the current
version of PCT) or about the way consciousness interacts with these
control systems.

PCT is a specific model of how the controlling done by living systems
works. As alluded to above, there are other models of control -- such
as the manual control model developed by McRuer & Jex . Some of the
strongest experimental tests of PCT involve seeing how well PCT
compares to these alternative models of control. A good example of
such a test can be found in Chapter 5 of Powers' LCS III: The Fact of
Control. I highly recommend it!

This chapter shows that a simple two level PCT model can account for
the the apparent "adaptation" (change in control characteristics) that
occurs when there are changes in the environmental feedback connection
from output to input while controlling. Bill shows that the observed
data can be accounted for with a simple "non-adaptive" PCT model just
as well as (or better than) the more complex adaptive McRuer-Jex
model. So by Occam's razor PCT gives a better account of the data than
an alternative control model.

So there is a great deal more to doing PCT science than just
determining the best definition of the controlled variable. Correctly
defining the controlled variable is certainly a fundamental part of
PCT science, but it plays a more important role in the beginning of
research on a particular kind of behavior, when we are just trying to
get a handle on what the behaving system is "doing" (what it is
controlling), than when the research is more mature (as in the
tracking studies) where the goal is to find out more about _how_ these
variables are being kept under control.

Best

Rick

···

--
Richard S. Marken PhD
rsmarken@gmail.com
www.mindreadings.com

[Bruce_Gregory6]

[From Bruce Gregory (2010.04.18.1848 EDT)]

[Martin Taylor 2010.04.18.11.00]

Second question, specifically to Bruce G. Did my essay on “why” that started the thread under its current subject line answer your comment about PCT not addressing why we do what we do?

Yes, indeed. And your post on “Is PCT falsifiable” was extremely helpful in clarifying my own view. Thanks.

Bruce