Understanding the Brain as an Endogenously Active Mechanism

Understanding the Brain as an Endogenously Active Mechanism William Bechtel (bill@mechanism.ucsd.edu) Department of Philosophy, University of California, San Diego La Jolla, CA 92093-0119 USA Adele Abrahamsen (aabrahamsen@ucsd.edu) Center for Research in Language, University of California, San Diego La Jolla, CA 92093 USA The investigation of endogenous activity, though less in- fluential, has historical roots nearly as deep as those of the reactive approach. It was promoted by Thomas Graham Brown (1914), for example, who studied decerebrate and deafferented cats in Sherrington’s laboratory at Liverpool from 1910 to 1913. He found that the isolated spinal cord, even when not receiving inputs, generates patterns of activ- ity comparable to those exhibited during motor behavior elicited by stimuli. Brown’s emphasis on endogenous activ- ity initially was largely ignored (for discussion, see Stuart & Hultborn, 2008) but was revived several decades later when biologists recognized a class of neural circuits—central pattern generators—whose self-sustaining patterns of activ- ity generated rhythmic motor behavior even in the absence of sensory input. After Wilson and Wyman (1965) pio- neered this construct in their account of locust flight, others identified central pattern generators in the brain stem and spinal cord for walking, swimming, respiration, circulation, and other behaviors for which oscillatory control was cru- cial (Grillner, 2003). Endogenous activity has received far less attention from those studying sensory processing and central cognition rather than motor control, despite indica- tions of endogenous oscillatory activity in cerebral cortex using techniques ranging from single cell recording to EEG and fMRI. In the next section we describe highlights from this research and in the subsequent section briefly explore the implications for reconstruing how we understand cogni- tive activity. Most important, if the conception of the brain as endogenously active is taken seriously, it profoundly challenges the reactive perspective that has dominated much of cognitive science as well as neuroscience: stimuli or tasks must be regarded not as initiating activity in an inactive sys- tem, but rather as perturbing endogenous dynamic behavior. The slow pace at which these fields are achieving a change of perspective is unsurprising considering the his- tory of other sciences. Although Max Planck was exaggerat- ing when he said “A new scientific truth does not triumph by convincing its opponents . . . but rather because its oppo- nents eventually die . . .,” the considerable costs and uncer- tain benefits of change make it a tough sell. Uneven accep- tance of Einstein’s revolutionary proposals is a familiar ex- ample. Less remarked upon is the delayed impact of changes in the sciences on philosophy of science. For exam- ple, this young field (which did not even have a journal until 1934) did not exhibit acute concern with the epistemological foundations of science until it was confronted with Ein- Abstract Although a reactive framework has long been dominant in cognitive science and neuroscience, an alternative framework emphasizing dynamics and endogenous activity has recently gained prominence. We review some of the evidence for en- dogenous activity and consider the implications not only for understanding cognition but also for accounts of explanation offered by philosophers of science. Our recent characteriza- tion of dynamic mechanistic explanation emphasizes the co- ordination of accounts of mechanisms that identify parts and operations with computational models of their activity. These can, and should, be extended to incorporate attention to mechanisms that are not only active, but endogenously active. Keywords: philosophy of science; mechanistic explanation; dynamics; endogenous brain activity, resting state fMRI, brain default network Introduction Observe a living organism, from a bacterium to a fellow human being, and you see an endogenously active system. Introspect and you will observe, as did William James, a continual flow of thoughts. If pressed, most cognitive scien- tists will acknowledge that neural systems—from individual neurons to the brain as a whole—exhibit endogenous activ- ity. That is, some of the activity is internally (Greek endo) produced (German gennan); the causes and control of this activity is inside the system rather than reactive to inputs from outside the system. But cognitive scientists tend to disregard this when designing studies. Those in psychology present discrete stimuli in structured tasks designed to per- mit statistical analysis of the behavioral effects of independ- ent variables. Those in neuroscience, following the tradition of Charles Scott Sherrington (1923), commonly treat the brain as a reactive system in which sensory inputs initiate neural processing that results ultimately in motor responses. They may stimulate specific neurons or provide sensory inputs with specific properties so that recorded neural activ- ity can be analyzed in terms of responses to inputs. In both fields, variations in activity that cannot be associated with an input are treated as random fluctuations (noise). There is no doubt that this reactive framework in psychology and neuroscience has been enormously productive in identifying the parts, operations, and organization of the mechanisms responsible for cognition. It soon reaches its limits, though, in seeking accounts of the orchestrated functioning of those components: their dynamics and coordination in real time.