Time scale of reorganization

When reorganization is considered to be the active process in MoL, this observation (B:CP p. 185 bottom) is relevant:

Since this is the most generalized control system so far considered, it will also operate on the slowest time scale of all—a point to keep in mind as we consider how this system reacts to various events. To the reorganizing system, the disturbances associated with a single trial in an experiment may be as the blink of an eye—barely noticeable.

Hi Bruce,

It’s always interesting to witness what happens in reorganization processes in the course of a therapy session. The speed of the process differs a lot over the course of the session, depending on what level changes are happening. The higher up the hierarchy, the longer the time scale. But at the same time, lower in the hierarchy, many more control systems are involved. Reorganization involving many sensations, as what we showed at the IAPCT presentation, is intense and involves many rapid changes of sensations and configurations. Reorganization at the higher levels, for instance at the principle level, is slower. And change in your system concept level might take months or years.

I think reorganization in practice usually involves many levels and many control systems at once. If you change your higher level references, this will also mean that many lower level control systems change their references. And if you unblock lower level control systems (e.g. in case of dissociation), their involvement in the larger hierarchy will mean that other control systems will also start to shift and change.

One thing to keep in mind is that the reorganization stretches beyond the course of the session. This even happens in MOL training sessions, where participants return a few weeks later stating that a single practice talk they had started something that kept reorganizing days after the training.

Bill talks about the ‘Reorganization System’ that e.g. in (2009) “From reorganization to evolution and back”, and before that in (1995) “The origins of purpose: the first metasystem transitions”, and of course in (1973) B:CP, the 1963 joint articles, and earlier.

Following Ashby, the Reorganization System is described as controlling intrinsic variables. Its operation is not accessible to consciousness. Reorganization is trial and error until error goes to (near) zero.

There can also be conscious and deliberate trial and error until error goes to (near) zero. That’s what happens at the Planning level (so-called Program level), where alternative sequences are tried in imagination until a path is found from the currently perceived situation through to the desired situation.

I suspect that part or all of this process at the Planning level may proceed without awareness between sessions and need not be conscious at all.

If this is the case, we don’t have to invoke intrinsic error to understand the reorganization by which an MoL participant resolves internal conflict.

It does seem unreasonable to me to assume that internal conflict is always life-threatening. It seems more plausible that MoL may often involve reorganization at the Planning level, and that this can go on without awareness.

In a conflict, there are two desired situations, A and B. The currently controlled sequence terminating in A includes an intermediate situation (the controlled result of a sequence step along the way) in which there are disturbances to control of B or disturbances to perceptions within the currently controlled sequence terminating in B. Efforts to sort out (in imagination) a working sequence leading to A go on independently of efforts to sort out a sequence leading to B, working on now one, now on the other. Bringing them together permits the system calling for A and the system calling for B simultaneously to engage in a Planning process.

If internal conflict is distressing enough, the paleo-mammalian parts of the brain may well perceive a situation as life-threatening. Could that generate intrinsic error signals? Intrinsic values like body temperature, cortisol level, and blood pressure might be affected. But maybe such distress also affects memory through the hippocampus, hence affecting reference values and gain.

Hi Bruce, isn’t trial and error change in sequential order, reorganisation happening at that level in the hierarchy? Also, surely an intrinsic variable needs to be linked with survival but it doesn’t have to be physiological? Could it not be an unlearned sensory gradient of a predator’s smell, for example?

Is there any evidence for e.g. recognition of predator-aroma to be genetically determined? Not sure zoo conditions comport with that.

I think it would be almost impossible for olfactory reference values not to be largely predetermined. An animal doesn’t have time to learn to smell everything that is food.

More importantly to the animal, it doesn’t have time to learn the smells of what might be dangerous, whether predator or rotten food. To find what is food can safely take time, at least in mammals, but to detect what will kill you cannot wait.

Absolutely, I expect there will be many intrinsic sensory variables of this kind that are not physiological but are nonetheless necessary to control to survive…

Warren, you raised two good questions, each with its own line of discussion. Here’s the first, and I’ll respond to the second question with a separate post under this same topic.

Yes, that takes place at what I call the Planning level. My introspective investigation indicates to me that planning is a trial of different sequences of controlled perceptions, the control of one providing necessary preconditions to bring the next under control. It can involve the building of a new sequence, looking in memory and in the environment for preconditions for controlling the final outcome (necessary or possible prior CVs on the way to it) and looking in memory and in the environment for means of controlling those preconditions.

As I understand it, Sequences that are repeatedly successful are remembered as routines, ‘habits’, familiar routes: when perception x is controlled branch from that perceptual input function go to every higher-level perceptual input which has made use of it in the past, but this branch of the perceptual signal has no effect if the perception to which it is input is not currently being controlled. By Hebbian learning, this is an important structural form of memory (immanent in the branching structure of the hierarchy).

(I call it the Planning level because the notion of a Program level is an analogy to the Boolean logic of computer programming, which derives from and is dependent upon language).

Your question above sounds like you agree with what I was proposing: that the ‘reorganization’ experienced in MoL is (or may often be) of this sort.

To the extent that this may be so, the dilemmas of internal conflict may often be pairs of controlled sequences which intersect by each having the same controlled perception y as part of the sequence, except at different values. Or they may each be closed through the same perceived aspect(s) of the environment as means of control (the same atenfel), but the atenfel must be in a different state (happenstance or controlled by oneself or by others) depending on which sequence is to be controlled: for example, a door must be open if you are to control one sequence to work and closed if you are to control the other.

But the trial and error which responds to intrinsic error is not a trial of different sequences. According to Bill’s model of the reorganization system, reorganization of the control hierarchy in the nervous system is done by nerve cells making and breaking synapses, extending and retracting axons and dendrites.

This can occur at other levels. Learning to play a musical instrument involves reorganization all the way down to input and output intensities. That’s Hebbian learning, I think, and not instigated by intrinsic error–although the repeated exercise does cause changes in the intercellular chemical environment which may be similar to the chemical changes that I postulate are the intercellular expression of intrinsic error which causes reorganization.

Here’s the second question.

I was asking about relevant research, so I looked. Perceiving a predator’s smell has a decidedly physiological basis. You are right, there are indications that the input function that constructs the perception and the reference value for controlling it are genetically determined.

Much more than predator smell, there’s a rich domain of ‘infochemicals’ in the environment between multicellular organisms, precisely analogous to ‘infochemicals’ in the intercellular environment within those organisms.

Here’s some that bears out your specific point about predator odor:

An overview

Animals possess an inborn ability to recognize certain odors to avoid predators, seek food, and find mates. Innate odor preference is thought to be genetically hardwired. Here we report that acquisition of innate odor recognition requires spontaneous neural activity and is influenced by sensory experience during early postnatal development. Genetic silencing of mouse olfactory sensory neurons during the critical period has little impact on odor sensitivity, discrimination, and recognition later in life. However, it abolishes innate odor preference and alters the patterns of activation in brain centers. Exposure to innately recognized odors during the critical period abolishes the associated valence in adulthood in an odor-specific manner. The changes are associated with broadened projection of olfactory sensory neurons and expression of axon guidance molecules. Thus, a delicate balance of neural activity is needed during the critical period in establishing innate odor preference and convergent axon input is required to encode innate odor valence.

Another overview:

Descriptive taxonomies tend to verge into dormitive principle territory. They are a useful start if we don’t get stuck in a Procrustean bed. I see literature on ‘infochemicals’ of three kinds:

  • Pheromones are informative from an animal to a conspecific animal.
  • Allelochemicals are interspecific.

Among allelochemicals

  • An allomone is beneficial to the animal that produces it.
  • A kairomone is beneficial to the other animal.
  • A synomone is beneficial to both.

Predatators use keromones to locate prey.

They have been described as pheromones and allomones that have evolutionarily backfired and therefore may not represent a distinct class of chemical signals themselves. Kairomones may be hormones, pheromones, or allomones that are normally used by one organism but exploited by an illegitimate receiver.

Examples in Marc J. Klowden, in Physiological Systems in Insects, quoted here.

Thanks for looking! I’ll hope to cite those!

I think we might be talking at cross purposes here. It’s certainly helpful to think more about how the program and sequence level might work. My point was more about the fact that sometimes to generate a truly novel sequence or program, surely classic reorganisation is needed ?

And not only between multicellular organisms, according to articles in March and June 2019 Nature, both by Elie Dolgin (I have paper tear-outs from the paper issues, but they should be easy to find in the Nature archives). The first is titled “The secret conversations inside cells”, the June one “The secret social life of viruses”.

I don’t know what this has to do with the “Time scale(s?) of reorganization”, but I thought it might be worth a mention. Maybe reorganization in a virus is represented by its mutation rate.