Anticipatory systems and Time...

[Judith Rosen <***>]

Hi Folks,

I've been engaged in a few discussions off list that have developed into something that I think should also become a discussion here. I know that several list members are interested in artificial life and quite a few are interested in applications from Rosennean theory for human problems and human systems, still others are interested in problems of entailment and mysteries of evolution, and a fourth category of interest has to do with the human mind, or psyche, and issues pertaining to that, including the interfaces between mind and computer, the limits of computation, AI... All of these are potentially impacted if the set of ideas articulated in Robert Rosen's book; "Anticipatory Systems, Mathematical, Philosophical, and Methodological Foundations" are on the mark. In fact, if he's correct in his interpretation of this aspect of biological phenomena, then most of the unanswered questions about the aforementioned will likely remain unanswered unless they are approached with an understanding of what anticipatory systems are like and why. From my own, more or less "common-sense" perspective, particularly as an avid gardener, I don't see how certain behaviors in plants can be explained any other way. My feeling is; if plants are exhibiting anticipatory behavior, it has to be a built-in capability because there is no way to fob off their behavior on a thought process. If it's a built-in capability that's peculiar to all life forms, then it behooves us (as a life form, if nothing else) to figure out how WE work. Frankly, I don't see how humanity is going to solve our own problems, most of our own creation, if we don't really have a clue as to the causal basis of the side effects we always generate.

In that book, my father made several seminal observations about both science and about the nature of the universe. I will stick mainly to his points about the universe, in this discussion (unless others have a desire to branch out farther). Among the well-known ones (like his assertion that causality is a matter of relational interactivity as opposed to material particles) were his observation that the existence of the capacity for a modeling relation between two or more systems gives us a glimpse into one of the fundamental pieces of information about entailment in this universe. He viewed this as something so basic about the universe that he called it part of "Natural Law". What it means is that relational entailment patterns (and, by consequence; Causality) are both consistent and "exportable"-- Because of this, it is ultimately possible for a model to be constructed using the transplanted entailment pattern of a natural system, such that the model can accurately recreate (and therefore, accurately predict) the causal behavior of the system it models. All learning is predicated on this aspect of the universe and, indeed, science would be impossible without it. Because we can learn, and make models which accurately predict causal behavior of corresponding systems, this much must be true.

He further observed that one natural system can be similar in its entailment patterns to another. And, to the degree they are similar, they can both be modeled in some sense by the same model constructed for one of them. Because of this relational property, it can also be said that one of the natural systems can be used as a model for the other. We do this all the time in science, particularly in medical science, where we use one natural system as a surrogate (a model) for another. We interact with the entailment patterns of one (a chimpanzee, or a lab rat) in order to learn about the entailment patterns of the other (humanity), because by interacting with the surrogate, we provoke causal behavior. In other words, we induce the system to manifest the behaviors which are/were always entailed by it's organization. That way, among other things, we are able to tell the difference between the entailment patterns of a sleeping dog, say, and those of a marble statue of one... or of a compu-pet. But, of course, this kind of surrogacy is only reliably accurate up to a point. According to his observations in biology, RR said this ultimate limit will be true of all models of complex systems. The fact that makes the case for me is to realize that even if we were to use one of a set of identical twins as a model of the other, we will still run into divergence between their behaviors. It's guaranteed. That's the nature of complexity. And it can be further predicted in modeling complex systems that the discrepancy between the behavior as predicted by the model and the actual behavior of the system being modeled will widen as a function of time. We've all see this in action, via the weather forecast for later in the week.

My father studied all aspects of the modeling relation because it is so important for science and because it also turns out to be a natural aspect of the universe that living organization exploits. The fact that entailment patterns are both consistent and exportable is an aspect of the evolutionary context of all living organisms and therefore I guess it shouldn't really surprise us that it is an aspect that has been incorporated into living system organization, itself. Perhaps this is what makes a living system alive? (I would be asking my father if he were here.) All we can say for sure is that it is concurrent with life and both are properties which are representative of a particular type of complex organization. The inclusion of entailment patterns of our environment into our own organization also has multiple temporal aspects to it. Again, multiple temporal aspects are a natural aspect of the evolutionary environment, so it stands to reason that these could become part of newly developing system organization as well. In any case, the interactions between real environment-through-time with the temporally different behavior of the encoded information about the environment (fast-forward? rate-scaled? somehow ahead of "real time") is what gives organisms their characteristic behavior patterns of preparing for changes before they happen such that by the time they do happen, the organism will be in an appropriate condition to maintain optimality (health) and system stability (survival).

The aspect of organism behavior that is the clue to the model-based nature of these behaviors is the characteristic errors that can be produced when the model and the system being modeled are not close enough in entailment.  Because there is no thought process whatsoever involved in generating these behavior patterns, it's "mindless" and can be easily interfered with or thwarted. If we take a plant out of its native environment, it will still be acting on the predictions of its models-- but those models may no longer be appropriate, as in the case of a tropical plant being grown in my garden in Western New York. If I transplanted the tropical plant from the Amazon basin to a similar climate in Asia or Africa, it may thrive and spread all over the new environment within a decade as many invasive species of plant have done. In my garden it's killed by the first freeze. It totally ignores all the daylength cues that are inducing the native plants to begin battening down the hatches and it keeps on blooming right up until the temperature dips below 32 degrees F. which kills it outright. Fascinating! That's the reactive paradigm in action. If all living organisms were as science currently describes them, they would all suffer the same fate. But the native plants have been preparing for weeks, and we still have not had a frost here yet. They are not preparing because it's "too cold for growth" or photosynthesis-- the tropical plants are blooming up a storm right now. So what's going on? I think my father had it right.

Further proof that living systems incorporate aspects of their evolutionary environment into their own organization can be seen in the fact that plants which evolved in a climate like Western NY State actually require a cold phase before they will enter a bloom phase. Not only do they survive harsh winters, they actually depend on them in order to reproduce. If the cold phase doesn't happen, the bloom phase won't happen. This is easily tested and anyone who gardens has seen proof of it firsthand. This can be very frustrating for gardeners who want to grow tulips  on the Gulf coast; they have to "pre-chill" their  bulbs in a refrigerator for a specific length of time before planting them or else there will be no bloom. In other words, they have to change their environment to be in congruence with the plants evolutionary models in order to trigger the desired behavior to manifest itself. The plant trade does this all the time with bloom cycles and the way it does so is by mimicking each species environmental triggers. 

So, how does experiencing a frigid winter trigger a bloom cycle? There is no way to explain this behavior via a reactive paradigm. Only by viewing the evolutionary environment's "normal" cyclical behavior can the causal mechanism be seen. The mechanism has several temporal aspects to it, such as sequence and length of cold-phase. The connection is, of course; SPRING. Spring doesn't happen unless winter happens, in a tulip's native habitat.

Once I began looking for aspects of environmental context that are visibly incorporated into living organism "design" I began to see evidence all over. People make these logical connections all the time, even scientists ("...the wolf is a carnivore; we can see this in the shape of the teeth and the design of the digestive system...") but don't follow the logic where it leads. This is one reason why evolution is so "murky"-- there is no factoring in of the entailments of how selection would act on anticipatory systems-- it's all been done with an underlying assumption of purely reactive systems. The reactive paradigm is the model that has been used and the entailments are off; science can't explain how the heck we got here from "there" and I think it's because the model doesn't predict this.

Well, I think it's time for coffee. Any comments from the group? I can never tell whether I'm being clear or not, and if I'm being clear; how clear?

Judith 


Web address: http://www.rosen-enterprises.com
BioTheory: An electronic journal of general science based on the Relational (Rosennean) Complexity Paradigm