The Constructal Law and Behaviorism
I’m surprised, as a matter of fact stunned, that modern behaviorism isn’t taking notice of the Constructal law as articulated by Adrian Bejan in his book “Design In Nature.” To me the implications of the Constructal law are overwhelming and yet no behaviorist or biologist is taking note. So about a month ago I had a discussion with Eric Brad about NDT relative to learning theory and I interjected the Constructal law into the discussion to see what kind of reaction it might engender. Eric did a quick survey of the Internet and noted the criticism of the theory that can be found on Wikipedia. I’m surprised he was satisfied by the vagueness of the critique but I didn’t pursue the matter at that point because I wanted to stay on topic, to wit: Natural Dog Training as behavioral snake oil. But now I return to the matter since at least I have Eric’s response to work with. Being well read and articulate I presume that Eric’s take would be what most behaviorists and biologists would come up with.
Eric Brad: “I do have some difficulty with your linking “emotion” to the “current” that Bejan describes in his work on constructal theory. I’ve done some reading today and even found a great TED talk that Bejan delivered on constructal theory. Very informative stuff but it differs in some significant ways from what you are characterizing here. For example, the Wikipedia page describes the major scientific criticisms of constructal theory as follows: “there is neither a mention of what these “currents” are nor an explicit definition of what “providing easier access” means. As a result, constructal theory is very versatile, but often unconvincing: depending on the choices made for the currents and the “access” to them, it can lead to extremely different results.” (Really? What different results are possible from Bejan’s explication of organ size and locomotion in terms of moving a physical body over long distances?)
“In defense of his theory (again from the Wikipedia page on Constructal Theory), Bejan has countered his critics by stating “the constructal law is not about what flows, but about the physics phenomenon of how any flow system acquires its evolving configuration (design) in time. The constructal law is not about optimality (max, min, opt)—it is the definition of “life” in physics terms, and of the time direction of the changes in flow configuration.” Taken in that context, it seems clear to me that what Bejan describes would be the tendency of any system with flowing matter to evolve toward a more efficient form over time. If one applies such a theory to nature (plants and animals), it seems to agree with Darwin’s theories as the more efficient species would be more likely to survive in order to breed.”
“In my reading thus far, I can find no reference by Adrian Bejan to this “other” current that you link to “emotion.” If one can accept your premise that such an “energy” does exist, then your writings make sense to me. I think it falls down for me because I do not accept as fact the existence of this unmeasurable “other” energy you choose to define as emotion.”
KB: Yes, Bejan doesn’t talk about emotion as a current. This is my interpretation of behavior, one I’ve been pursuing since the seventies. The point with the Constructal law however is its finding that whenever we find something persisting over time, then it is due to an order that has configured itself around a current. In other words, order and currents are inseparable. The question therefore is what is the current? Biology and behaviorism should be consumed with this question. Meanwhile biology and behaviorism says that order is the result of intention in service to an underlying genetic compulsion. Therefore if behaviorism and biology want to reconcile their theories with the Constructal law, the exchange and the perpetuation of genes must be the current.
A current is the movement of something, anything, it could be an object of mass, a liquid, a stream of electricity, an idea, and of course a gene. And this movement always follows a spreading branching architecture that vascularizes whatever area it is flowing through. It ultimately saturates that area so that flow is eased and access is enhanced, otherwise it cannot persist over time. All order evolves according to this architecture. This is inarguable. Genes can mutate all they want, but they must always configure their expression so that it renders a spreading tree-like architecture. Furthermore, they must always mutate so that the order they configure always meshes with every other order that is likewise configured around a current. But note that behaviorism and biology have been arguing for decades that the order which persists over time in the animal kingdom (such as canine social structure), is due to the mathematics of gene replication, that the proliferation of one gene as opposed to another results from a competition between genes that vary at random. Those traits that are most suited to a given climate or competitive pressure have an advantage and eventually occupy a greater share of the genome. This contravenes Constructal law because it explicitly states that enhancement of flow does not result from a competition. It results from objects of resistance becoming incorporated into the configuration.
These days behaviorism has modified the dominance theory to mean that a dog has an intention to maintain access to a resource, and again in service to the perpetuation of its genes. Learning theorists also argue that behavioral structure, i.e. the capacity to learn, the capacity of an individual to perpetuate itself over time, are due to high cognition, a problem solving capacity to conduct a cost/benefit analysis of various options. But the Constructal law demonstrates that this is faulty reasoning. What we perceive of as dominance is a vascular structure, it cannot possibly be about controlling access to a resource because by definition it has to be about enhancing the flow--FOR THE ENTIRE STRUCTURE. If a dog becomes fixated on a particular resource the only accurate statement that can be made is that the dog’s behavior is the structure sprouting a new “branch” in order to enhance flow, that’s it. Anything else is story telling. The only coherent explanation must be predicated on this flow logic rather than a logic of gene replication or reinforcements according to consequences.
The purpose of the structure is to saturate an “area;” be it a physical space or an expanse of time. It is a branching architecture that intimately entwines the individual with all other currents moving through other branching architectures. It is not due to a random process of genetic mutations and is not due to cognition because this spreading architecture predates by billions of years the existence of genes or cognition. This makes genes a suspect candidate as the current. Also modern behaviorism which derives its intellectual sustenance from Neo-Darwinism says that evolution has no direction. But this is also contradicted by Constructal law. As Bejan points out, were we to replay the evolutionary tape, we would get the same basic outcomes, from anatomy, locomotion, behavior and social structure and with the same basic mix of species variability that we have today because the evolution of all things is predicated on a current that is constantly improving its flow through a design predicated on a branching architecture. The evolutionary process is not random, it works according to principles of physics as opposed to an abstract mathematical rationale of gene replication. Gene replication is likewise a part of this Constructal process, but it is not driving it, it is subsumed by it. Furthermore, all flows confluence into the one overarching flow. Genes can’t possibly do this because unless genes are being culled, or transmitted through sexual recombination, then no genes are in transaction in the vast majority of interactions between individuals. And we must also remember that gene replication is not scalable since the genes of one species are not necessarily transferable to another species.
For this and other reasons reinforcement theory does not have a coherent definition of information. Saying that a reinforcement is that which increases the frequency of a given action, is not a flow statement. It is not scalable because it can’t answer why what is a reinforcement for a lion, isn’t a reinforcement for a gazelle. In a flow system, the fundamental unit of information is a principle of conductivity, a principle that also encodes for its improvement over time. It is not the standard 0 or 1 binary value, or the capacity to evaluate two equal alternatives, or a gene that is slightly more adaptive than another gene. The principle of conductivity as information is scalable. (For example the Constructal law shows the same principle is responsible for the locomotion of birds, animals, insects and fishes, as well as trucks, trains, and airplanes.) The prime unit is consonant with the most advanced complicated construct, like an intricate Lego structure built with a set of standardized blocks. Whereas behavioral/gene replication theory gets more and more complex as it must constantly paper over this internal inconsistency. (See "The Red Queen" or "Why Sex Is Fun" for convoluted theories as to the evolution of sexuality.)
Eric mentions efficiency as a means of reconciling Neo-Darwinian theory (on which learning theory draws its intellectual sustenance) with the Constructal law. But efficiency is not synonymous with flow and its enhancement. The story of evolution isn’t efficiency, the story of evolution is turning CHANGE into information, i.e. enhanced flow. For example American car companies in the seventies and eighties became ultra efficient by making interchangeable parts across model lines. This significantly lowered production costs but eventually what made an Oldsmobile different from a Buick, Pontiac or Chevy, became the badge on the hood and some glossy fender work. I can remember the uproar when it was discovered that Cadillacs were being powered by Chevy engines. When the consumer realized this, each brand lost emotional value and they turned to foreign cars because they had distinctive handling and styling characteristics which made them substantially different from other makes of cars. In the interest of efficiency, American car companies violated the principle of a branching architecture which itself is derived from a principle of motion, to wit: every action has an equal and opposite reaction. By being less efficient, but more polarized, they could have persisted longer over time. (Anyone seen a new Oldsmobile, Pontiac or Saturn lately?) This distinction between flow and efficiency further disqualifies genes as candidate for the current. Genes are “self” motivated, they can’t embrace the full architecture by which all of nature is interconnected.
Of course genes are part of evolution but their most likely role would be to lock in a particular flow configuration so that it can be passed through time. Genes are not the agency by which the configuration evolves. The configuration evolves IN THE MOMENT because the current flows IN THE MOMENT. And the only plausible candidate for the current around which the animal mind configures is emotion. The configuration emerges and improves by way of emotion becoming unresolved emotion (stress) and then resolved emotion: E->UE->RE, this is the full statement of the principle of emotional conductivity and paradoxically as it might first appear, E->UE->RE is the most basic unit of information. (Thus the car we buy resolves some layer of stress we carry.)
The emerging science of Epigenetics is beginning to illuminate the role that emotion plays over genetics. Epigenetics shows that the timing of a gene’s expression is more vital than the gene itself and that the biochemicals affiliated with stress inhibit the expression of genes. Emotion becoming unresolved by meeting with objects of resistance, and then resolved by importing these objects into the configuration, is how the energetic principles of nature shape and fine tune genetic expression so as to adapt to the environment. Social structures are statements of resolved emotion. Eventually genes reflect these shifts.
Finally Eric repeats the common behavioral refrain that emotion-as-energy is not measurable. But the truth is that science measures emotion-as-energy every day. Every brain scan, blood chemistry assay, hormonal surge obtained in lab experiments is measured and recorded but then it is misinterpreted as being in service to the mathematics of gene replication. Science misinterprets this data because it is reductionist and is not trying to construct a model. Whereas if we were to secure a dog to a post and stimulate him with a prey object and then measure every variable in its mind and body, the resulting data would constitute full energization. We would have an emotional baseline, we would have identified the main channel draining the emotional watershed that constitutes animal consciousness. And then if we were to do the same tests with all other aspects of its existence so that we could compare the values of these various emotional states, we will find a spreading architecture, with hard mathematical and biochemical correlates, configured around the main emotional current, the predator-->prey module as the chief conduit of emotion.
Also in regards to an “unmeasurable other current:” consider, that if someone talks to a psychiatrist about what they are feeling, describing their internal emotional states and how these impact their minds, the good doctor gives this direct testimony a lot of weight in the formulation of a remedial program. Direct testimony is part of the current scientific approach to understanding emotion. It is a clinical research tool (Kagan; “What Is Emotion?”). Likewise, if cultures around the world describe in great detail the energy-like impacts of emotional experience, we should also give this direct testimony great weight in the formulation of a theory of emotion. Every culture speaks of emotion as a current, as in for example the common expression: “emotional juices flowing.” This vast aggregate of experience edits out subjective impressionability. And then if the signature of an electric, magnetic and gravitational charge can also be detected in the behavior of animals, and if interpreting animals in light of the immediate-moment, which means emotion-as-energy, is the only means of analysis which has to date rendered a model, this is further corroboration of such evidence.
Dogs do not learn by reinforcement. A behavior is reinforced in terms of how it contributes or detracts from a spreading architecture. In other words, everything dogs learn is in terms of a social matrix, a construct that always has increased sociability as its norm, no matter the particular consequences of the dog's experience.
Behaviorism, your turn. What is the current?