A blog about problems in the field of psychology and attempts to fix them.

Sunday, December 18, 2011

"But what about the brain?"

I received an email inquiry a few days ago from Eric Haaland, who has studied with John Shook. We met during the neuropragmatism conference in DC last summer, and he is hoping to be a kinesiology grad student next semester with Tom Stoffregen at the University of Minnesota. He gave me permission to post his email, lightly edited, to the blog along with a reply. He said....


I have been reading as much Holt and Skinner as I can find recently as well, and I knew that you were one I could get useful information from.  I know that they both insisted that the [mind] is not 'internal' to the organism, that there is no 'internal' - there is only organism as a process over time.  But I'm failing to put their interpretations of education into descriptive terms.  As animals, we are obviously learning beings, beings that have an innate understanding of our sensorimotor repertoire and how to manipulate the world around us to achieve goals (i.e. affordance perception); but this still seems to fall into the neural network, brain-activation paradigm, which I don't think is the case.
Sure, our nervous system is a decoupled regulatory system in that is not directly involved with conversion of the environment into metabolically-active content, but it is not
entirely decoupled.  It's derived from the ectoderm of the developing embryo for godsakes (i.e. the part of the embryo that forms the skin - the part that is in direct contact with the world).  The nervous system is not decoupled from the body-world interface; nor is the nervous system solely responsible for the organism's understanding of the being in the world.  But I still cannot seem to understand how behaviorism completely describes this concept.

I know this is a lot of content in a relatively short email and I don't expect an immediate reply.  However, if you would be willing to talk with me about some of this stuff just so I can understand how you interpret it, I would be extraordinarily grateful...

Best wishes,

Eric Haaland


Well, this is one of the big, big questions in psychology right now. I have lots of ways of trying to answer this question, but none I am fully satisfied with.

The easiest answer is to simply say that while, in humans, the brain clearly plays a crucial role in mental processes, it is not the whole story, and we need to develop better ways of articulating what, exactly, the role of the brain is.

Why isn't than fully satisfying?
  • First, I think that even saying that the brain plays a 'crucial role' in mental processes is misleading: The brain is only as crucial as every other part of the body that is involved... and if you are Andy Clark, the brain is also only as crucial as every aspect of the environment that is involved in the mental process. So, that is either completely crucial or not, depending on how you want to look at it.
  • Second, the throw away "in humans" line bypasses an important point: If we are interested in minds in general, then we are interested in at least some species without brains, and at least some species in which brains and brain-like structures are not nearly as important as they are in humans. In fact, we are likely interested in many species that meet those criterion, and certainly if we are just counting living organisms that psychologists might study to elucidate fundamental principles of the mental processes, then us big-brained humans are in the minority.
  • Third, the answer does nothing to make clear that there is a levels-of-analysis problem inherent in the question. If you met a reproductive biologist who was not particularly interested in the nuances of hormonal chemistry, that would seem reasonable, even though we all know that chemical interactions involving hormones are important to reproductive processes. Why would that seem reasonable? Because we understand that biology presents questions that can be answered without knowledge of chemistry. Of course, I am well aware that there is a board discipline of bio-chemistry, populated by scientists who do really great research. To the best of my knowledge, however, both biologists and chemists recognize that biochemistry is a peripheral topic; neither discipline confuses the periphery for the core. Taking that metaphor full circle: I am well aware that there is a field of psychobiology, and that brain research can, and should, play a crucial role in that discipline. But we have to remember that psychobiology is a peripheral discipline, at the border between psychology and biology - it is no way the core of the discipline, and the core issues of psychology can be studied without any knowledge of brain functioning.
Still, all of that doesn't do service to the nuance in Mr. Haaland's question. To try to do at least a little more, I'll focus on the questions relative to Skinner. Skinner's brand of behaviorism is the most popular at the moment, and has some major virtues. Skinner's approach emphasizes two things: The importance of operant conditioning in development*, and the ability of discriminative stimuli to control behavior**. I'll admit it from the start, both these things are overemphasized, occasionally to ridiculous levels... but they do explain much more about behavior than anyone would have believed 100 years ago, and more than most psychologists today appreciate.

Thus, the two cores of Skinnerian behaviorism are a pretty good theory of learning and a pretty good theory of organism-environment interaction. Nothing about either theory requires knowledge about brain-processes, and the entire science can be done at the level of organism-environment interaction. Certainly one could drop down a level of analysis and start talking about how the brain of a given organism changes over the course of learning, in the presence of given stimuli, or, most interestingly, how neural response to a given stimuli changes over the course of exposure to behavioral contingencies reliably present along with that stimuli. But one could just as easily drop down to a physiological level of analysis with an interest in hormonal response, immune response, digestive response, circulatory response, etc. This in not just hand-waving; I have seen very good talks about all the named topics (by researchers who are generally not Skinnerians) at ISDP, and much can be read about in Developmental Psychobiology as well as other venues. For example, there is some insanely good stuff coming out of the Brain-gut initiative at Columbia.

Now, Skinnerian behaviorism does not do a very good job at articulating the cyclical nature of perception-action processes, nor does it do the best job at evaluating what, exactly, organisms are responding to in their environments. On those topics, the ecological approach really shines. Holt is, in some important ways, the fountain-head for both approaches, so he blends together the strengths of both approaches. Alas, he did not have the benefit of the decades of sophisticated research the different traditions have generated, so he has some serious weaknesses as well. But that is a much longer discussion, and hopefully I have done something towards answering your question. There is even a chance, if we are lucky, that some of the readers here will throw additional insights into the mix.


* I suppose that instead of "operant conditioning" I could say that Skinner emphasizes the role of behavioral contingencies in development. For our purposes here, those are interchangeable.

** A discriminative stimuli is a stimuli that reliably indicates that a certain behavioral contingency is in effect. The classic example is a light, where pressing a lever will result in food when the light is on, but not when the light is on - e.g. receiving food is contingent upon lever pressing, but only when the light is on.


  1. One of the biggest problems that I have with traditional neurobiological approaches to psychology is that they do not offer any evidence for their base assumptions.

    Sure, you can have a number of rats and you can put them into differing experimental conditions and observe their behavioral patterns over time. Furthermore, you can say with certainty that certain neural structures in one group take up more volume in the brain relative to other groups. You might even be able to say that there was an upregulation in a certain protein in a particular region. But none of this is based upon the here-and-now, the lived experience of the animal.

    The here-and-now, the felt experience, the embodied self is a real aspect of our lives as animals. There are states of experience that can be triggered via drug administration or meditational practice, but to simply demand that those changes are due to functional states of the material brain is unwarranted.

    Learning occurs as a psychological process. Physical modification of the structure of a plastic neural network still takes time and coordination. The neurons within the brain are not snapping around in 'space', making new synaptic connections with other cells. This would require a system-wide calibration of where exactly each thought, opinion, or other psychological content is located in relation to all others. Furthermore, the 'mind' would be in charge of directing the cells around in this manner; sort of the organism's God who keeps all the planets orbiting correctly in relation to all the other things in the universe.

    This is just silly, but it's the only extrapolation from the traditional cog science foundation that I can think of. The psychological happens at a rate of speed that cannot be physically and functionally dependent on the underlying interconnectivity of neurons or even neural assemblages.

    In Kandel's 'Principles of Neuroscience', he discusses his experiments with slugs where he mapped out a reflex circuit. All these reflex circuits working together in the same organism are, as far as I can interpret it, what Skinner thought created the whole organic system. But if the physical connections between neurons are the things responsible for embodied experience of life, then the circuits must be able to coordinatedly make immediate shifts in connectivity pattern when things like insight occur. This seems to be problematic to me.

    I feel these questions are among the most important that can be asked because they may lead to a radical reassessment of what it means to be life, what it means to exist and learn. Without a narrated, synthesized explanation, all we have are a bunch of little data points that float ambiguously in academic journals.

    Thanks for taking my question head on!

  2. Well, one fundamental problem with the way most people do cognitive psych, or cog-neuro, is the assumption that mental terms refer to things happening in the head. Certainly there is stuff happening in the head, and it is very important, and worthy of study. However, most of the things we are interested in as psychologists are probably better understood at the level of behavioral interactions with the environment.

    We can get to that point just with the arguments made by the methodological behaviorists. While the advent of modern neuro-imaging meliorates those concerns somewhat, we still label brain parts based on behavioral observations, and some insidious problems arise when people forget that (e.g. when they forget that the "pleasure center" of the brain is merely a place that lit up under very particular experimental circumstances, associated with very particular behavioral responses).

    The more aggressive behaviorist position is to assert that the things we are interested in are whatever we observe when we see those things, i.e., the organism-environment interactions occurring at the behavioral level. This approach actually derives from Pragmatism, through Peirce and William James (and not from logical positivism).* From this point of view, there are crucial issues to be solved at the intersection of psychology and neuroscience, but the basic questions about psychology will all be answered with discussions of behavior. As Skinner would occasionally explain it: Psychological study constrains what interesting things neuroscientists can discover, but not the other way around.

    Of course, there is a little cheat there with the term "interesting" (interesting to whom?), and certainly at least a few discoveries from neuroscience have lead to interesting psychological discoveries... but the vast majority of insight is going in the direction Skinner asserted it would.

    By the way... one reason that psychologists have been able to maintain the "thing in the head" approach for so long is that we have a very simple view of how behavior is controlled. This is why all that complexity stuff, developmental systems stuff, dynamic systems stuff, and the ecological psychology stuff about ambient information and perception-action linkages, is all so important. A lot of the conceptual struggling in these circles is about finding a better way to talk about behavioral control, in a dynamic environment, and especially to find better ways of talking about the role of the brain in that larger system. I've been digging around quite a bit in the last two months for another project, and there is a surprising amount of convergence happening amongst fringe groups in psychology on the importance of these insights and on how to put them together. Exciting times.


    *And if you don't want to take my word for that, see an excellent analysis in Larry Smith's Behaviorism and Logical Positivism: A Reassessment of the Alliance