Skinner's Fundamental Insight and Fundamental Error

Behavioral selection was Skinner's key insight; radical behaviorism his error.

Posted Sep 02, 2016

B. F. Skinner is arguably psychology’s most influential academic, and is perhaps second only to Freud in terms of psychological scholars who have had an impact on society at large. And as with Freud, Skinner is both revered and reviled in various corners of modern psychology. From the vantage point of the unified theory, both Skinner and Freud had powerful insights and both made important errors. An earlier blog articulated how we can separate the Freudian baby from the Freudian bathwater; this blog engages in a similar analysis with Skinner’s work.

Source: Skinner/Wikicommons

Anyone who has any familiarity with Skinner will likely be aware that his work was centered in someway on rewards and punishments. If you want to get a kid to do his homework, give him a sticker or a lollypop or some praise. If you want to keep a rat from pressing a bar, follow it with an electric shock. Behavior, folks learn from Skinner, is all about rewards and punishments.

While there is certainly a grain of truth to this characterization, it glosses over the essence of Skinner’s key insight and, in many ways, trivializes it. This trivialization, combined with Skinner’s error (discussed below), prevents Skinner’s views from being deeply integrated with a modern understanding of human psychology. Skinner’s key insight is not so much about rewards and punishments per se, but about the process of how animal behavior evolves over the course of its lifetime. According to Skinner, animal behavior evolves (i.e., animals learn) in a way that directly parallels the way organisms evolve across the generations. Skinner called this idea behavioral selection, and if you understand it you will see it is much cooler than stickers and lollypops.

There are three components of behavioral selection. The first component is that that animals vary in the behaviors they “emit” (i.e., the process starts with some form of either random or innate behavioral output). The second point is that these varying behaviors have different consequences. And the third element is that those consequences play a determinative role in the frequency, intensity, and duration of behaviors emitted in the future.

Notice how this directly parallels Darwin’s theory for how organisms evolve over time. For Darwin, that there is variation in a group of organisms, some variants lead to greater rates of survival and reproduction, and the next generation of organisms inherits those characteristics, leading to greater adaptation over time. Here is a depiction of the modern evolutionary framework for Life.

Gregg Henriques
Source: Gregg Henriques

Although the psychologist E. L. Thorndike was the first to outline the idea animal behaviors were centrally influenced by the effects they had, it was Skinner who saw clearly how crucial the process of behavioral selection was to understanding virtually all complicated animal behaviors. Unfortunately, most introductory psychology texts teach about Skinner primarily in terms of behavior modification (i.e., stickers, lollypops and electric shocks). But the nature of behavioral selection is much more nuanced than this. To see how, consider the extent to which you are CONSTANTLY interacting with the environment and how closely and immediately your actions are tied to those consequences.

Here I am, for example, typing away at my computer. When I hit a key, a consequence is a letter appears on the screen. My typing behavior is directly tied to this consequence. What would happen if, when I punched the key, no letter would pop up? First, I would likely push harder on the letters, and also vary which keys I hit. If those behaviors did not change the outcome, my typing behavior would quickly extinguish, and I would initiate another pattern of behavioral investment, like searching to see if the keyboard had come unplugged. This analysis shows that the letters on the computer screen are reinforcing my typing behavior, and if that reinforcer stops occurring, my typing behavior quickly alters. It is in the fine-grained analyses of the relationship between actions and consequences are where you can see the magic of behavioral selection.

This example is salient because this actually has happened to me and, not surprisingly, it has directly shaped my behavior. When I am working on my email Outlook account at home, after about 7 minutes or so it freezes for about 45 seconds, where it does nothing, no matter what I do. I don’t know why it does it—it is an annoying glitch. But the consequences have had a striking impact on my behavioral output. After going through a wide variety of behavioral investments to fix this problem which failed, I now have used the stoppage as cue to engage in another behavioral investment pattern that is more productive for that interval. That is, as soon as the computer freezes, I now stop, get up, and either grab a cup of coffee or hit the restroom, and then come back and resume my work a minute or so later. The point here is that your “doing”—whatever it is that you are doing—is completely and intimately and immediately tied to the consequences that your doing is having. Unfortunately, Skinner’s paradigm is often taught in a very blunt way, with examples of how things like candy, money, or blame might shift someone’s behavior. In actuality, we are constantly being shaped by the processes of behavioral selection.

I think it is also important to note that Skinner was not as hostile to inner experiences as some think that he was. He regularly acknowledged the existence of private thoughts and feelings; he just conceptualized them as forms of behavior (what he called “covert behavior”) and thought the primary problem was that they were difficult to study. And if we allow this, we can allow the idea that animals can run simulations of behaviors internally and the internally anticipated consequences can shape behaviors. (Skinner's framework waffles on this point philosophically, but we can see that empirically there is clear evidence that animals do exactly this, see here). In this light we can see that, either overtly or covertly, we are constantly emitting a variety of (mental) behaviors, which in turn produce various (internal or external) consequences that either reinforce or extinguish future actions. This is a foundational insight into understanding animal/mental behavior.

If behavioral selection was Skinner’s fundamental insight, what was his fundamental error? His fundamental error was his radical behavioral philosophy. To understand his radical behaviorism, we need to recognize that there is a fundamental divide in the philosophy of underlying psychology between behaviorists on the one side and mentalists on the other. The fundamental difference is that for the behaviorists, “the mind”, to the extent it exists at all, is, at base, just another form of behavior. In contrast, mentalists argue we can talk about “the mind”, at least in some ways, as a cause of (overt) behavior. In other words, behaviorists think of the mind as something to be explained by things that cause behavior in general (i.e., associations and consequences in the environment), whereas mentalists (also called cognitivists) think we need to understand “the mind” so we can understand the hidden processes that cause observable action.

Skinner was, of course, firmly in the behaviorist camp. In August of 1990—the night before he died—Skinner completed an article for American Psychologist summing up his argument for why psychology could never be a successful science of the mind. Skinner’s anti-mentalistic perspective can be summarized as follows: First, in a manner directly paralleling the ToK System, he argued that human behavior was the product of three separate levels of variation and selection: 1) natural selection; 2) behavioral selection; and 3) verbal selection. He also corresponded each level to its own discipline: 1) biology; 2) psychology; and 3) anthropology/social sciences. Second, Skinner defined the mind as an unobservable cause of behavior, akin to a vitalistic life force that some used believe was needed to account for the behavior and complexity of living things. Third, Darwin’s theory of natural selection provides the framework for understanding how an environmental selection process can create biological complexity and in doing so it removed the need for vitalistic substance. Finally, Skinner concluded that in the same manner that natural selection removed the need for vitalism, behavioral selection removed the need for “mentalism.” In short, Skinner argued that if we are to ever become a real science like biology, we must give up our notion of unobservable, mentalistic substances or forces causing animal/human behavior.

To Skinner and his radical behavioral followers, this argument is straightforward, sound, and confers many scientific benefits. For example, it clearly defines the proper subject matter of psychology as the behavior of the animal as a whole. Second, it differentiates psychology from biology with the same basic logic that biology is differentiated from the material sciences. Third, it defines psychology as a science of behavior and removes the problematic concept of something nonbehavioral (i.e., nonphysical), causing something physical to behave.

All of these benefits are genuine; however, the argument is not entirely sound. In fact, there is a glaring problem. According to the ToK System, Mind is the same type of concept as Life. Both are emergent levels of complexity generated by feedback loops of variation, selection, and retention. Darwin’s theory of natural selection removed the need for the concept of vitalism, but it did not, of course, remove the need for the concept of Life. Indeed, the idea of Darwin being “anti-life” is absurd. Biology is crisply defined as the science of life, and the set of living behaviors are what biologists are attempting to describe, explain, and predict.

The problem deepens for Skinner’s philosophy when we extend the analysis further. As Darwin was fully aware, his theory needed a mechanism of inheritance. Indeed, biologists later on found exactly what Darwin needed to account for how traits were retained across the generations: they were retained via genes. Genes can be thought of as the informational codes that are stored in an organism’s DNA. And biology took a giant leap forward when biologists realized that natural selection operated on genetic combinations over time.

So if a much fuller understanding of biological evolution emerged when Darwin’s model of selection was combined with the idea that it operated on (genetic) information processor, we can continue with the parallel, and ask, what, exactly is the way in which the impact of consequences are retained in an animals lifetime? The answer is provided by the basic insight of cognitive neuroscience: memory, attention and learning happen via neuro-information processing! In this light then, neuro-information processing framework is the key missing ingredient in Skinner’s account of how animal behavior evolves.

The unified theory, via the ToK System and the philosophy of mental behaviorism, provides us a clear framework for how we can absorb Skinner’s key insight and correct his fundamental error. Skinner’s theory of behavioral selection, when merged an information processing view of the nervous system, provides the causal explanatory framework for the emergence of Mind.

Gregg Henriques
Source: Gregg Henriques