Western Association for Biofeedback and Neuroscience — Spring 2017 21 that Sterman might be pre-empting my talking points. I need not have worried. His reference to observation hung in the air without elaboration. His comment did, however, hint at the essence of the di- vide, which relates to the context in which observation takes place. The state of affairs can be dissected nicely in terms of the left-brain/right- brain dichotomy as well as in terms of bottom-up versus top-down interaction with the empirical realm. Observation, as we know, is not a neutral process. In the words of Albert Einstein, “It is the the- ory that tells us what we may observe.” If the observational process relates to an experimental design, then we have a top- down process where observation is neces- sarily constrained to the variables being tracked. If the observations being made are goal-directed, then we don’t see the man in the gorilla suit walking through the scene. This is nicely illustrated in Sterman’s early work, where the neurofeedback was done with UCLA graduate students. Their most prominent report was of a subsidence of anxiety and depression, and this was duly registered in the reports to the contract monitor. When we later claimed to have used SMR/beta training beneficially in application to both anxiety and depression, Sterman was completely non-plussed. “But you do temperature training for anxiety.” With regard to de- pression, he did not even dignify the topic with a response. He was stuck in compart- mentalized, paradigm-bound thinking. His own earlier findings had not regis- tered in memory because at the time they had not been the explicit objects of study. The alternative empirical approach— truly observational science—was modeled for us by Nikolaas Tinbergen and Konrad Lorenz in the field of ethology. Confront- ed with the complexity of animal behav- ior, it was incumbent on the scientists not to impose their own pre-conceived notions on their investigations, lest they miss important features. The models emerged later out of their experiences, and both were recognized with the Nobel Prize for their collective work. This is a good example of bottom-up observational science. Other examples are William James, Ivan Pavlov, Charles Darwin, and Stephen Jay Gould. Pavlov did not ‘discover’ classical conditioning in his formal experiments. He had observed in the course of his studies on gastric se- cretions that his dogs were already salivat- ing when the food dishes were delivered. An initially random observation became systematic and eventually culminated in formal investigations, leading finally to established science. The process went in natural progres- sion from bottom-up observation to top- down experimental proof. In the words of Elmer Green: “First the findings; then the science.” It has to be that way whenever we are confronted with novelty, and there is no better example of this than Sterman’s own landmark experiment. Understand- ing followed the experiment; it did not precede it. Indeed, the following morning Sterman came up to me to acknowledge that the bottom-up process comes first. It’s just that those early ventures of inqui- ry don’t usually make it into the journals. In our own field, the initial authoriza- tion to work with human brains in this manner had to be furnished by formal research. But as soon as this method tran-