What counts as (non) cognitive? A comment on Rowe and Healy

As someone who finds that noncognitive factors, like motivation and energetic state, undoubtedly have an influence on cognitive performance, Rowe and Healy’s (2014) argument that we should take such factors more seriously strikes a definite chord. With luck, their article will help ensure mistakes can be rectified in the present and (hopefully) avoided in the future. Here, I suggest another reason why “noncognitive factors” should be taken seriously: namely, some noncognitive factors may not be as noncognitive as we assume. The burgeoning literature on “e-cognition” (embodied, embedded, enactive, and extended) is both interesting and potentially useful here (e.g., Clark 1997, 2008; Brooks 1999; Pfeifer and Bongard 2007; Rowlands 2010). Although important differences exist between these approaches, all agree that body and environment contribute to cognitive processes in a constitutive and not merely causal way: an organism’s cognitive system involves more than just its brain, encompassing other bodily structures and processes, as well as exploiting environmental structure. Although these ideas can sound rather bizarre, they are exactly what one should expect from a thrifty evolutionary process like natural selection (Clark 1997; Rowlands 2010). If there is reliably recurring structure in the environment, then, as Brooks (1999) has argued, why evolve expensive neural tissue to build an internal model? Why not let the environment itself guide you? Similarly, being made of the right kinds of materials, organized in the right kinds of ways, can obviate the need for neural control, enabling problems to be solved more cheaply through the exploitation of bodily structure. We are already familiar with the idea that animal-built structures are constitutive of physiological processes (Turner 2002). Why should cognitive processes be any different? There are many excellent examples one can offer here, from the manner in which phonotaxis in crickets relies heavily on the physical structure of a female cricket’s auditory system, to the arrangement of ommatidia in flies’ eyes, which automatically compensate for motion parallax, to the way in which the complex detouring behavior of Portia spiders relies as much on the physical structure of its “active” vibrating eyes, as its brain (see Barrett 2011 for a review). Of course, it is easy to dismiss such examples by pointing out these are all animals with very little brain to speak of. Bigger brained animals must surely employ more obviously “cognitive” mechanisms. Recent work on New Caledonian crows, however, suggests otherwise. Specifically, Troscianko et al. (2012) demonstrate that shape of the crows’ bills and the positioning of their eyes make a large contribution to the superior tool-using and problem-solving skills. New Caledonian crows have very high binocular overlap compared with other corvids, and their bills are also very straight, allowing them to maintain a more stable grip on tools, as well as look directly along the length of the tool as they use it. This undoubtedly offers an advantage over birds that are less able to guide their actions visually. Individual variation in performance within and between species could well be related to variation in body morphology, such as degree of bill straightness, or lesser or greater convergence of the eyes, in addition to other noncognitive factors. This, in turn, raises the issue of what counts as “cognitive” versus “noncognitive”: if animals have evolved certain bodily structures that contribute to their success at cognitive tasks, then embodied cognition theorists would argue that they are legitimate parts of the cognitive system (and extended cognition theorists would argue further that the tool itself constitutes part of the birds’ cognitive system, e.g., Maravita and Iriki 2004; Clark 2008). Findings like these thus seem to hold implications for Rowe and Healy’s (2014) argument that we should attempt to control for all possible sources of variation, ensuring that performance is attributed accurately to an animal’s brain power. If brain power alone is not the secret to success, however, how should we then proceed? One suggestion is that, in addition to attempting to control for individual variation under certain circumstances, we should also actively exploit it in others, examining how differences in body size and shape influence task performance. In some cases, what we assume to be noise may be an important part of the signal. L.B. is supported by the Natural Science and Engineering Research Council of Canada’s Discovery Grant and Canada Research Chair Programs.

Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.