Louise Barrett, beyond the brain: how body and environment shape animal and human minds
Beyond the brain: how body and environment shape animal and human minds is an eye-opening and thought-provoking book that sets out a much-needed contribution to the study of the relationship between animals, cognition, and the environment. The volume provides remarkable new insights into how to understand animal (including human) behavior, raises interesting questions about the role of environmental affordances in the emergence of complex cognitive processes and provides the reader with a refreshing break from the wearisome excess of brain-centric literature that still pervades much of the debate surrounding evolutionary psychology. In embracing the theoretical framework endorsed by proponents of embodied cognition, Barrett adopts an ecological approach to psychology that aims at challenging any attempt to describe complex thinking and flexible behavior as mere by-products of internal cognitive activity. Ecological approaches in psychology typically provide a powerful and coherent alternative to standard information-processing paradigms [see Marr 1982 and Poggio 1981 just to name a few] that postulate the existence of distinct and hierarchical theoretical levels to understand complex information-processing tasks such as visual perception (Rowlands 2006, p.8). Ecological approaches resist these computational understandings of visual perception and rather explain perceiving as “an achievement of the individual, not an appearance in the theater of his consciousness; as a keeping-in-touch- with the world and experiencing of things rather than a having of experiences; as an awareness of instead of just awareness. Perceiving, they continue, may be awareness of something in the environment or something in the observer or both at once, but there is no content of awareness independent of that of which one is aware” (Gibson 1979, p.239). On these accounts, visual perception is not therefore the mere reception of stimuli from the environment followed by the construction of internal representations, but rather an active sampling of the optic array, driven by environmental affordances,1 that allows an animal to discover the information present in the world (p.106).
In endorsing this “Gibsonian” perspective, Barrett successfully merges and deeply interweaves different disciplines (ranging from ethology and ecological psychology to artificial intelligence, robotics, and philosophy) to create an interdisciplinary space for the study of the evolution of cognition that ultimately calls into question any orthodox, brain-bounded, purely intracranialist understanding of the nature of the human mind. Barrett’s criticism is essentially directed at those views that deny any possibility of body–brain integration. Such views, she notices, find both their communal ancestor and their theoretical justification in the conceptual palette provided by Cartesianism.2 Cartesianism is the idea, affecting much of modern psychology, that the mind is separated from the corporal body and entirely in control of it. In this book, Barrett powerfully questions Cartesianism by presenting a compelling bulk of empirical evidence that shows how heavily animals and humans rely on their bodies and milieus to behave intelligently. The main scope of the volume is therefore to demonstrate that: (1) the brain isn’t the sole locus of intelligence and (2) that there isn’t, properly speaking, such a thing as a “naked brain”—an isolated mental substance that dominates the physical from above. A related goal of the book is also to prove that our knowledge cannot be entirely derived, through reason and thinking, from a set of constrained innate ideas but that it rather emerges in the interplay between brain, body and the wider world. To the extent that the book unveils these erroneous conceptions of the mental, it also encourages us to drop our anthropomorphic stance by revealing that intelligent behavior is always profoundly influenced and directly governed by physical and socio-cultural circumstances in a way that has lead both Paul Maclean (1949) and Blessing (1997) to talk of a “visceral brain.” For this reason, Barrett’s book can be taken as a refreshing antidote for anyone who is troubled by old-fashioned tendencies that attempt to describe the human brain as a Swiss army knife (Cosmides and Tooby 1992); a collection of highly modular, domain-specific computational tools (Fodor 1975, 2008; Pinker 2003; Byrne and Bates 2006). In line with a long-established tradition of writings in phenomenology [see Heidegger 1927 trans. 1962 and Merleau-Ponty 1962 for instance] and in agreement with more recent works in the cognitive sciences [such as Clark 1997, 2008; Menary 2007; Rowlands 2010; Sutton 2010;Wheeler (forthcoming)] that describe the brain as an integral and inseparable part of the complex loop that accompanies the formation of mental states; Beyond the Brain therefore contributes to further illustrate the importance of the Umvelt—the wider world as it is experienced by a particular organism (von Uexküll 1957)—for the emergence of complex cognitive strategies and successful adaptations.
The volume is organized in 11 chapters centering, as we have seen above, around one main theme: the Gibsonian idea that environmental affordances are crucial for the emergence of our cognitive functions. In the course of this review, I have already presented standard ecological approaches and quickly explained how they follow Gibson in taking the environment to be the richest source of information we can directly access through perception–action cycles. In what remains of this review I will now try to provide the reader with a full sense of Barrett’s project by condensing, in a page or two, the most important contents of the volume.
In chapters 1 and 2, Barrett mounts a sustained attack on chauvinistic forms of anthropomorphism. Anthropomorphism is basically the attribution of human characteristic to gods, animals, weather (in this case is called “pathetic fallacy”) or inanimate objects.3 Through a series of compelling empirical examples (involving facial recognition in babies [Kanwisher et al. 1997; Gauthier et al. 2000] and social intelligence in apes [Brothers et al. 1992; Perrett 1999; Barton 2006]), Barrett invites us to renounce our anthropomorphic perspective by demonstrating how, in investigating cognition, a strictly human-centered perspective is not only undesirable but also wrong both theoretically and empirically. This kind of chauvinistic anthropomorphism should be abandoned, Barrett argues, because it “invites us to ask scientific questions that simply reflect our own concerns” (p.3) and therefore encourages us to think that “other creatures are interesting only to the extent that they resemble us” (p.5). In essence, in the first chapters of her book, Barrett attempts to resist the idea that the whole of evolution was crafted to produce us as the most supreme creature and rather tries to establish the claim that anthropocentrism is a hazardous bias that considerably endangers scientific inquiry.
Chapters 3–5 are principally devoted to explore and appreciate how much a small brain can achieve. The goal of these chapters is therefore to prevent another bias, what Barrett labels as the “big head bias”—the idea that a big brain can allow us for an assumption of superiority over creatures that possess vastly smaller and less developed brains than ours. Barrett calls into question this idea and effectively demonstrates that there is no connection between the complexity of an observed behavior, and the complexity of the mechanism that produced it. As evidence for this claim, she adduces two interesting examples: (1) the “fertile turtles” and (2) the behavior of Portia spiders, which she discusses in chapters 3 and 4, respectively. In the early 1950s, in order to test the idea that complexity of brain functions doesn’t necessarily arise from the number of neurons but rather emerges from the richness of their interconnections (Holland 2003); neurophysiologist and robotics pioneer Grey Walter built two motorized robots (Elsie and Elmer) that looked like tortoises. These tortoises had only two “brain cells,” which could produce one reflex response each; one for the light, the other for touch. The sensor for detecting the light was a photoelectric cell mounted on the front of the steering column of the robot; the touch sensor was instead an electrical contact that was turned off whenever the shell of the tortoise encountered an obstacle (p.43). These tortoises could only work in two modalities: (1) seeking light (2) moving towards it. Despite their relatively simple architectures, these robots, once freed in the environment, were able to explore the surroundings by slowly moving in a series of arches, detecting and hurrying towards sources of light and avoiding or pushing obstacles around. They were also reported, when placed in front of a mirror, to flicker, twitter, and jig like clumsy Narcissuses (Grey Walter 1951, 1953) in a sort of mating dance that led many to attribute them what would have normally been accepted, if observed in living animals, as evidence of self-awareness. Without any kind of internal, programmed knowledge or computer-like structure, these tortoises were able to exhibit remarkably complex behaviors and these “emerged” as a direct result of their interaction with the environment. Barrett uses the tortoises’ example to show that sometimes the structure of the environment and the physical shape of the animal’s body can play a highly active role in forging and molding adaptive strategies.4 Analogous results are also observed in the animal kingdom. In chapter 4, Barrett takes us into a fascinating journey in the world of spiders and invites us to focus on the behavior of the Portia. Portia spiders are spiders that have brains with the size of a pin, yet this doesn’t impede them to exhibit one of the most complex and flexible hunting behavior displayed in the animal kingdom. Portia spiders in fact, while chasing their prey, successfully engage in a series of intricate stalking strategies (including aggressive mimicry, detouring to break visual contact, and smoked-screen behavior) that have led many scientists to accept the idea that they behave intelligently, being capable of learning on the fly and of deploying complex hunting tactics (Wilcox and Jackson 2002; Ross 2007). Again, Barrett uses the case of Portia spiders to further demonstrate that articulated behavior isn’t necessarily tied to big brains but may indeed be triggered by relatively simple internal mechanisms responding to complex environments.
Chapters 6–9 are probably the most crucial in the book, as they are used to introduce, critically assess, and ultimately endorse the case for the ecological approach in psychology that Barrett wants to establish. Though crucial, this isn’t however the only key theme that span across these four chapters. Additional topics worth a mention are those that relate to other important aspects of embodied approaches to cognition such as the critique of classical computationalism conducted through the analysis of the so-called Watt governor metaphor (Van Gelder 1995) and through the idea of soft-assembly (Thelen and Smith 1994). The Watt governor metaphor is a metaphor used by Barrett to attack the “brain as a computer analogy”—the idea that our brains work in the same way as digital computers work by means of symbolic manipulations carried out by dedicated mechanisms or pre-specified modules. The Watt governor was introduced by the Australian philosopher Tim van Gelder (1995) as an alternative metaphor better able to capture the dynamic interactions between animals and the surrounding world. Cognition, in his view, is described as a dynamical process arising when, “inputs, internal processes and outputs are tightly coupled like the spindle, angle arm, and throttle valve” in the centrifugal governor (p.129–130). Barrett takes dynamical approaches to be in competition with computational accounts of brain function but this opposition may well be more apparent than real. Clark (2008) and Wheeler (2005) in fact both talk of dynamical computationalism. Dynamical computationalism, they argue, is a view that merges dynamical systems theory with old-fashioned computationalism and asserts that in order to provide with the right kinds of temporal coordination needed to trigger the temporally rich adaptive behaviors typical of forms of higher cognition, the brain needs to span the body–world boundaries creating crossing architectures that encompass these two realms.
The other important point that Barrett tries to make in these central chapters of her book is related with the idea that embodied systems are preferable to classical computational systems because they can express more variability. In order to illustrate this point, Barrett introduces the idea of soft assembly. Softly assembled systems are provisionally assembled units working in the service of successful phenotypic transmission. They are made up via temporary bio-ecological operations of assemblage that occur between neural and non-neural resources and are characterized by operations that deal with specific problems arising in the course of one’s experience. Softly assembled systems are richly variable systems that act as pacemakers and rhythm-setters (Wheeler 2005) to couple an animal with its own world. These systems are systems in which there is no central controller and no planning and according to Thelen and Smith (1994), they should be taken as the defining feature of a dynamical account of development. Barrett deploys the idea of soft assembly in the context of human babies (chapter 10) to show that (1) “our knowledge is tailored to the specific environment we encounter” (p. 176) and (2) to demonstrate that the environment is an “equal partner” in producing cognitive behavior (Clark 1997). The book ends (chapter 11) with a final chapter in which Barrett discusses the idea of the body schema and the incorporation of tools. In recalling a series of studies involving the plasticity of the body schema (Iriki et al. 1996; Maravita and Iriki 2004), Barrett focuses on the negotiability of our own bodies and tries to demonstrate that organism are “always mutually entangled with their environments and do not stand apart from them as some kind of hermetically sealed, disembodied cognisers” (p.200). Although I am very sympathetic with this line of argument [see Kiverstein and Farina (in press)], in this book, Barrett fails to address and rule out alternative and competing approaches (such as the embedded view) that invite us to distinguish the “fairly acceptable” claim that incorporation yields new agent-world circuits, entangled and re-engineered bio-mechanical architectures, from the more radical assumption that buys the inference to extended cognition and therefore envisages literal cognitive extension. In particular, Rupert (2009, chapter 8) grants the existence of the former but gives an embedded account of cases of incorporation on the grounds of his notion of brain-centered realizers of the body schema. Barrett surprisingly doesn’t discuss this objection.
To conclude, the impressive sheer of wealth and breadth of information presented in this book combined with a charming and engaging prose makes this volume very enjoyable to read. Beyond the Brain is indeed an amusing and entertaining read, but one with an extraordinary analytical rigor and eloquence of argument. Very accessible, enticing, and lucidly written, it can be enjoyed both by professional academics and laypeople. Readers—be they novice or seasoned—will certainly find the volume uplifting and inspirational, Barrett’s style brisk and delightful and her intellectual playfulness quite solacing. An intriguing, illuminating, very ambitious, and stylish book that will zealously and passionately set the stage for the next round of academic speculations by becoming an essential source of discussion for some time to come. A must read for the next generations of cognitive scientists and for all those who are interested in the study of comparative cognition.
Affordances are the terms in which perceptual information is made available.
John Sutton has written a fair bit about how Descartes did allow for this kind of body–brain integration. In Sutton’s view; modern forms of Cartesianism must be intended as deviations from the Magister’s Dictum because they have failed to keep abreast of the spirit of his philosophy. So, Cartesianism would interestingly be not true to the spirit of Descartes! For more specific details, see Sutton (1998) and Gaukroger et al. (2000). Thanks to Julian Kiverstein for the hint.
This seems to provide nice empirical evidence for what Mark Rowlands (1999) has called the “barking dog principle.” “If it is necessary for an organism to be able to perform a given adaptive task T, then it is differentially selectively disadvantageous for that organism to develop internal mechanisms sufficient for the performance of T when it is possible for the organism to perform T by way of a combination of internal mechanisms and manipulation of the external environment” (Rowlands 1999), p.80]. This principle is very similar to the 007 principle formulated by Clark in 1989. It is important to note that the 007 principle has been consistently used, in the literature directed at the extended mind thesis, to endorse the case of cognitive extension. For an interesting analysis of the relationships between the two, see Shapiro 2010.
I am deeply indebted to Julian Kiverstein and John Sutton for their continuous guidance and invaluable support throughout my doctoral studies. My warmest gratitude also goes to Andy Clark and Richard Menary for their stimulating feedback on earlier drafts of this review. Thanks also to Joel Krueger for the precious assistance through the editorial process. Last but not least, I would like to express my appreciation to the ARC Centre of Excellence in Cognition and its Disorders (CCD) and to Macquarie University for generously financing my research. Needless to say, any remaining errors are mine and mine alone.