Eroding the Boundaries of Cognition: Implications of Embodiment

Abstract of article to be found here. Also check out Mog Stapleton’s paper.

To accept that cognition is embodied is to question many of the beliefs traditionally held by cognitive scientists. One key question regards the localization of cognitive faculties. Here we argue that for cognition to be embodied and sometimes embedded, means that the cognitive faculty cannot be localized in a brain area alone. We review recent research on neural reuse, the 1/f structure of human activity, tool use, group cognition, and social coordination dynamics that we believe demonstrates how the boundary between the different areas of the brain, the brain and body, and the body and environment is not only blurred but indeterminate. In turn, we propose that cognition is supported by a nested structure of task-specific synergies, which are softly assembled from a variety of neural, bodily, and environmental components (including other individuals), and exhibit interaction dominant dynamics.

From long before psychology became a separate scientific discipline, it has seemed useful to think of the mind as a collection of separate faculties. For example, in the late 18th century, Thomas Reid separated out judgment, reason, memory, and conception, among other intellectual faculties, to go alongside a suite of other faculties, including moral faculties and the will. Later, most famously in the hands of Franz Gall in the 19th century, each of these faculties was associated with a specific brain area, as can be seen in the phrenological diagrams produced at the time. In the intervening 200 or so years, these ideas have gone in and out of fashion several times. Both are decidedly in fashion at the moment. In cognitive science, it has long been fashionable to take cognitive abilities to be separable from one another and to take experiments to be about, for example, attention, but not also about conception or memory. This is the case because most cognitive psychologists take the cognitive faculties to be separable modules (see, e.g., Fodor, 1983; Uttal, 2003). Evolutionary psychologists take this even farther and posit “massive modularity” (Cosmides & Tooby, 1992), according to which there are far, far more faculties than Reid ever dreamed of, each with its own evolutionary history. Advances in neuroimaging have led cognitive neuroscientists to identify the brain areas that are especially active in the exercise of these faculties.

This is the scientific state of the art in which recent research on embodiment in cognitive science is set. The cornerstone of that research is the notion that behavior, bodily structure, and environmental resources are far more deeply implicated in an adequate explanation of cognition than the outline above would suggest. The variety of research investigating this very general claim includes such things as work on the role that neural resources involved in motor-control might play in supporting higher order cognitive processes like language understanding (Glenberg & Kaschak, 2002; Pulvermüller, 2005); investigations into the influence that metaphorical mappings from elements of our embodied experience like moving around or standing upright might have on the way we think about abstract things like planning or morality (Lakoff & Johnson, 1999); experiments documenting the influence of bodily sensations like weight and warmth (Ackerman, Nocera, & Bargh, 2010; Williams & Bargh, 2008) or bodily actions like extending one’s finger (Chandler & Schwarz, 2009) on cognitive processes like interpersonal judgment; considerations of the circumstances under which the manipulations of external resources must be considered to be part (and not just a causal side effect) of cognitive processing (Clark & Chalmers, 1998); and demonstrations of the influence of coordination dynamics on perception, action, and cognition (Chemero, 2009; Kelso, 1995).

We will discuss many of these research programs below, but our aim in this paper is not simply to catalog the array of interesting findings implicating body and environment as part and parcel of cognitive processing. Instead, we will argue that embodiment research questions some of the most deeply held beliefs in the cognitive sciences. First, it makes the localization of the cognitive faculties in specific brain areas problematic. It is not hard to see why this is the case: If the exercise of cognitive faculties happens in the body and, sometimes, the local environment, along with portions of the brain, one cannot localize the cognitive faculty in a brain area alone. Perhaps more surprisingly, recent research on embodiment also makes the identification of separate cognitive faculties problematic. Or, to put the matter differently, we will argue that one effect of research on embodiment is to undermine the architectonic principle of this special issue. One cannot isolate “cognitive” psychology from the study of perception, action, and social interaction. This essay, then, will not be about research on embodiment as pursued within cognitive psychology but about the nature of the cognitive system in light of research on embodiment.

To argue that research on embodiment has these effects, we will rely on four related concepts that might be unfamiliar to some readers: soft assembly, interaction dominance, synergy, and 1/f scaling. For convenience, we define these concepts here.

Certain systems, such as an automobile or a laptop computer, are composed of a series of parts, each of which has a particular role that it fulfills. Other systems, such as flocks of birds, are more fluidly put together. In the latter case, it doesn’t matter which particular birds are part of the flock—any old bird will do—and each bird is capable of taking up each position in the flock. Indeed, during flight each bird will take up multiple positions in the flock. The flock is softly assembled, in that it is composed of a temporary coalition of entities, engaged in collaborative task. Some softly assembled systems exhibit interaction-dominant dynamics, as opposed to component-dominant dynamics. In component-dominant dynamics, behavior is the product of a rigidly delineated architecture of modules, each with predetermined functions; in interaction-dominant dynamics, on the other hand, coordinated processes alter one another’s dynamics and it is difficult, and sometimes impossible, to assign particular roles to particular components. Sometimes softly assembled systems exhibiting interaction-dominant dynamics are called synergies. A synergy is a functional grouping of structural elements (molecules, genes, neurons, muscles, limbs, individuals, etc.) that are temporarily constrained to act as a single coherent unit (Kelso, 2009).

Work this decade has shown that 1/f scaling (a.k.a., 1/f noise or pink noise or long memory) is ubiquitous in smooth cognitive activity. 1/f scaling is temporal long-range dependencies in the fluctuations of a repeatedly measured behavior or activity. Analogous to spatial fractals, 1/f scaling denotes a fractal or self-similar structure in the fluctuations that occur over time (within a time-series of measurements). That is, higher frequency, lower amplitude fluctuations are nested within lower frequency, higher amplitude fluctuations as one moves from finer to courser grains of analysis (see, e.g., Holden, 2005; Kello & van Orden, 2009 for a more detailed description). 1/f scaling indicates that the connections among the cognitive system’s components are highly nonlinear (Ding, Chen, & Kelso, 2002; Holden, Van Orden, & Turvey, 2009; Kello et al., 2010; Riley & Turvey, 2002; Van Orden, Holden, & Turvey, 2003; van Orden, Holden, & Turvey, 2005). This nonlinearity indicates that cognitive systems are not modular. When systems are nonlinear, operations are not easily localizable to a relatively small spatial or temporal region of the system but rather are distributed throughout the system. Therefore, the “parts” of the system cannot be treated as truly structurally or functionally separate for the purpose of localization. The systems are synergies.

In the remainder of this paper, we will present a series of cases in which embodied cognitive science (ECS) provides evidence that the systems responsible for cognition are synergies, softly assembled systems which exhibit interaction-dominant dynamics. This, we will argue, plays havoc with attempts to localize cognitive faculties in circumscribed locations and, even, to understand the cognitive faculties as separate from one another.