Abstract
In Cognitive Code: Post-Anthropocentric Intelligence and the Infrastructural Brain, Johannes Bruder provides an outline of the new epistemologies of neurosciences, which lie at the intersection of brain imaging, artificial intelligence, and data sciences. Mostly based on his ethnographic fieldwork in brain imaging institutes in the United Kingdom and Switzerland from 2009 through 2013 as well as analyses of complementary texts such as scientific papers, news articles, or science blogs, Bruder is able to paint an updated picture of the current state of brain sciences. In doing so, he also demonstrates why long-held criticisms and accusations regarding neuroscientific analysis (or more specifically brain mapping) as contemporary phrenology no longer apply. This updated picture is much needed, and Bruder’s illustration of a new epistemology of the brain is masterful—albeit at the same time somewhat overly optimistic.
The main thesis of Cognitive Code is that epistemologies of brain imaging are rapidly changing from analyzing the physical brain (its functions and anatomy) to simulations of non-physical data or “brains in silico” (p. 4) through cloud computing networks. Bruder calls this voluntary or purposeful process of integrating engineering and artificial intelligence analytical methods into the neurosciences as the “infrastructuralization of the brain” or “the action of obscuring physical structures” (p. 10). At the core of this infrastructuralization is the emerging informational division of labor between the experimenters (or the “real neuroscientists” [p. 27]) and the methodologists (or the “data monkeys” [p. 27]). Bruder portrays the disparate perceptions and reflections between these almost different sects of neurosciences with anecdotes and excerpts from his interviews. Methodologists come from disciplines like physics, engineering, or computer sciences and view brain imaging as an opportunity to apply their complex analytical refinement strategies into yet another avenue. The experimenters are usually the principal investigators (project leaders), scanning participants to reveal the neural substrates of cognition and behavior. According to Bruder, the informational labor created by the methodologists is powerful and challenges notions of the brain as a biological entity as well as traditional doctrines of cognitive neurosciences that the “real neuroscientists” hold dear, such as the functional or anatomical correlates of cognitive functions or malfunctions.
Throughout the book, Bruder describes the most important developments and sensational recent controversies in the neurosciences. First, Bruder focuses on the infamous story of the dead Atlantic salmon that showed brain activity during a social psychological mentalizing task (!) in 2009. The salmon itself was of course not some miracle of science but was rather a cautionary tale of what exactly can go horribly wrong if appropriate statistical correction methods are not applied in functional Magnetic Resonance Imaging analysis. However, the salmon controversy blew up exponentially with critics blaming social and cognitive neuroscientists for producing non-replicable false positives or so-called “voodoo correlations” in their data and social neuroscientists trying to defend their positions. Bruder argues that this controversy caused brain imaging methodologists to reimagine their statistical analyses.
In tandem with further advancements from the study of “false negatives” or brain activation previously ignored as being noise in the data, a much more fuzzy and dynamic account of the brain emerged. These previously ignored brain activities were referred to as the “default mode network,” which serves as the default or baseline activity in the brain. While most prior research focused on what the brain does while actively involved in experimental tasks, this new paradigm of research shifted its attention to what the brain does not do—or in other words what goes on in the brain when the participant does not do anything or simply during a “resting state.” According to Bruder, the idea of a default mode that is plastic and encompasses a set of wide, indeterminate networks of activation ultimately gave way to the conception of the brain as an “assemblage of information highways, switches, and hubs” (p. 89) or the “cognitive infrastructure” (p. 97).
Bruder further buttresses his main thesis by pointing out how modeling neural network traffic using graph theory (the study of networks) and cloud computing (analyzing and storing data on the internet rather than actual physical computer hardware) is contributing to the “waning importance of cerebral geographies” (p. 97). He uses the analogy of “the brain-as-cloud” (p. 106), in which he argues that brain mapping will likely face the same fate that the somewhat obscure attempts to classify and catalog the clouds in the sky into an International Cloud Atlas (issued by the World Meteorological Organization between 1896 and 1987) faced. Hidden is a double entendre here where the brain is cloud both because it is modeled in the non-physical spheres of the Cloud and because of the blurred visions of output image data resembling clouds. In either case, Bruder is detailing the declining significance of a tangible, physical brain with visible boundaries within.
While I mostly agree with Bruder’s characterization of the new labor force of brain imaging and cognitive neurosciences, I could not help but wonder about the hierarchies within this informational division of labor and how they might be contributing to these new epistemologies. For example, in most brain imaging centers such as those studied by Bruder, the so-called “data monkeys” are primarily transient post-doctoral researchers that often lack the job security, income, or status that come with principal investigator positions. Bruder alludes to the remarkable concept of “cognitive capitalism” on page 122, but this concept is dropped rather quickly without being discussed or connected to the book’s themes. Much of the mystery about Bruder’s interviewees is due to a lack of specification of his methodologies. How the research sites or the texts analyzed were chosen, how many interviews or literary texts were included in the analyses, how the data were analyzed—all were left to the imagination of the reader. It’s a rather unfortunate weakness—a book holding a magnifying glass to the epistemologies of a branch of science while at the same time completely clouding (pun intended) its own.
In his closing chapters, Bruder reemphasizes “an increasingly complex, restlessly active, and infinitely plastic brain” (p. 126) and uses autism spectrum disorder as an example of how scientists are no longer locating this condition within the localized centers of the brain. While contemporary neuroscience indeed shows that the brain is complex and highly adaptive, the complete picture is unfortunately not provided here. For example, autism spectrum disorder is actually a group of developmental disabilities with a wide range of symptoms and causes (and not one disorder); so it is no surprise to anyone for it to be related to a dispersed set of networks in the brain. Furthermore, many other diseases and disorders—such as Parkinson’s Disease, Huntington’s Chorea, and different kinds of aphasias (loss of comprehension and expression of speech)—are still located in certain brain regions and sometimes even in certain cells of the brain. So, while Bruder’s pursuit in delineating the contemporary contours of the new neurosciences is deft and exciting, the proposition that a productive uncertainty brought by these new epistemologies will challenge and replace traditional ontologies still seems overly optimistic.