Cognition and Behavior The Brain-Cognitive Behavior Problem: A Retrospective
Cognition and Behavior The Brain-Cognitive Behavior Problem: A Retrospective
Background
In two thousand one, I was invited to write a review for a prominent journal. I thought that the best way to exploit this opportunity was to write an essay about my problems with ill-defined scientific terms and question whether the dominant framework in neuroscience is on the right track. My main argument was that many terms in neuroscience are inherited from folk psychology and are often used in two ambiguous ways: both as the thing-to-be-explained (explanandum) and the thing-that-explains (explanans; e.g., "we have memory because we remember," "we remember because we have memory"). These postulated terms are assumed to be entities with definable boundaries, and within this framework, the goal of neuroscience is to find homes and mechanisms for these terms in the brain with corresponding boundaries (I called this "the correlational approach"). I warned that a framework dictated by human-centric introspection might not be the right roadmap for neuroscience and argued that there should be another way of carving up the brain's "natural kinds."
A month later, I received the rejection letter: "Dear Gyuri, ... I hope you understand that for the sake of the journal we cannot publish your manuscript" (emphasis added). One reviewer was very enthusiastic while the other strongly dismissive. I took a deep breath, put the issue on the back burner, and went back to the lab. Perhaps the harsh reviewer was right. Although I recognized the problem correctly, I failed to provide the right strategy to solve it. Yet, the issues I exposed in the manuscript kept bugging me, and I have since written two books as attempts to clarify my views and offer alternative strategies to explore brain-behavior relationships. In the intervening two decades, perhaps thinking along similar lines, other investigators have also addressed the explanandum-explanans issues. By digging into the literature deeper, I also realized that many of "my ideas" had been considered already, often in great detail and depth, by numerous scientists and philosophers, although those ideas have not effectively penetrated mainstream neuroscience. After re-reading the text of the rejected manuscript eighteen years later, I found that many of the ideas presented there have become "contemporary." Hence, I did an experiment.
I resubmitted the original manuscript (nearly two decades later!), and the current editor kindly agreed to send it out for review. Remarkably, three of the four referees recommended publication with minor revisions, whereas the fourth recommended major revision. They correctly pointed out that pitting theory against methods cannot be the right solution. I wholeheartedly agree with the reviewers that science is not just the art of measuring the world and that cutting-edge technique is not all there is to neuroscience. We have to be careful that new methods will not simply reveal more and more about less and less. Observations need to be organized into coherent theories to allow further progress. Yet, when caveats are recognized in a dominant model, alternative solutions oftentimes emerge by exploring the mechanical details of the substrate which the model is supposed to explain. Novel methods provide new windows on the same problems and can facilitate alternative interpretations. I have devoted several pages to this issue in my recent book. I still believe that promoting large-scale recording back in two thousand one was a progressive idea, a method that has caught fire only recently. Yet, I support the view that technology cannot be the savior of the discipline. This latter sentiment is the key element of The Brain from Inside Out. Nearly all other comments and suggested revisions were based on papers published after two thousand one. Despite caveats and shortcomings of the original manuscript, it is interesting that just by virtue of time, or perhaps because of a shift in scientific perspective, the piece which was rejected eighteen years ago became publishable.
Obviously, I did not want to revise the old text because by such revision would negate its historical value. Below is the original text and reference list two thousand one. The point of my exercise is to illustrate how dominant frameworks shared and defended by community consensus can oppose or welcome outside-the-box views. One just has to calculate the right time for publishing a particular idea. Of course, I do not expect that all my colleagues will agree with my views. Yet, providing alternative views is what makes progress in science.
The Original Paper
The Original Paper
The title of the original paper was "The brain-Cognitive Behavior Problem: Rethinking the Correlational Approach" with the following summary:
Although the goal of neuroscience is to understand how the brain generates behavior, current practice in behavioral/cognitive neuroscience appears to work in the opposite direction. Subjectively defined terms are selected as independent variables and their dependent (brain) variables are searched for. These terms, inherited from philosophy rather than derived from objective investigation of the brain itself, are assumed to be "real" entities. In such a framework, the major problems have been clearly outlined by previous introspection and the sole role of neuroscience is to discover and explain their mechanisms. In today's practice, the reliability of brain-behavior correlations is judged by "inter-laboratory consensus" rather than by an independent assessment of the experimental error.
I suggest that behavioral-cognitive neuroscience should establish its own vocabulary, based on brain mechanisms, to formulate testable hypotheses. It should start with the brain and define descriptors of behavior that are free from philosophical connotations and can be communicated across laboratories, languages and cultures. A potential approach, using ensemble recording of neurons, is suggested to provide a quantified description of behavior and the associated experimental error.