Beyond the general themes discussed in IV.A, there are various ways in which Simon’s own theorizing seems to encourage an embodied perspective, as this subsection demonstrates.
IV.B.1. Stress, duress, and taxing environments
Notice the way in which Simon and Newell talk about the discovery of parameters: “Only when the environment stresses its capacities along some dimension…do we discover what those capabilities and limits are, and are we able to measure some of their parameters (Simon, 1969, Ch. 1 and 2).” (1971, 149) And, unsurprisingly, their flagging of Simon 1969 (which I’ve been citing in its most recent edition, 1996) is spot on. When a system is taxed, Simon claims, or makes a mistake, the properties of the system’s material substrate, which would not otherwise show themselves, become behaviorally manifest (1996, 12, 58, 83).
Interestingly, contemporary embodiment theorists often emphasize that cognition is time pressured and evolution kludge and that this affects cognition through and through. What embodiment theorists add to Simon’s perspective, then – which addition Simon can naturally accommodate – is that the limiting properties of the inner system virtually always show through. Because thought is commonly time pressured – or in some other way pressured by context, including internal context – the human cognitive system is almost always being taxed, and thus significant and contingent aspects of the material substrate of cognition are continually on display. A constant stream of external stimuli together with the ongoing fluctuations in internal context (neuromodulators, hormone levels, neural correlates of emotional states and moods) create a shifting cognitive context, a series of “perturbing influences” on cognitive processing; and under these conditions, the fine-grained physical properties of our brains – and bodies and external environments, perhaps – reveal themselves in all sorts of ways relevant to our understanding of human performance and intelligent behavior.
IV.B.2. Beyond duress
I’ve tried to extend the scope of Simon’s category of a system under duress, so that it covers a wide range of cases in which embodied contributions to cognition reveal themselves. Circumstances involving what is more clearly conceived as breakdown also reveal fascinating aspects of the human cognition, in ways that Simon would have no reason to resist. Over the decades, cognitive scientists have learned no small amount about the human cognitive system – its architecture and component mechanisms – from the study of subjects suffering from autism spectrum disorders, Capgras syndrome, hemispatial neglect, prosopagnosia, Balint’s syndrome, and many more disorders.
Experimental manipulations, too, have led to the discovery of a variety of parameters (and their values) as well as systemic properties that emerge from interaction among sometimes quirkily functioning component parts, all of which Simon can gladly take on board. Consider, for instance, Pylyshyn’s proposal that there are four FINST pointers (Pylyshyn 2000), or studies showing that tactile discrimination becomes more sensitive when experimenters activate a visual representation of the area being touched (Serino and Haggard 2010), or that one’s report of the timing of one’s conscious intention to press a button shifts when pre-SMC is subject to TMS after one has pressed the button (Lau, Rogers, and Passingham 2007). Presumably, the material basis of human cognition, including materials that have more or less to do with specifically sensory or motor contributions to cognition, determine the presence in humans of such cognitively relevant architectural elements and their specific causal profiles.
In his attempt to understand the complexities of human problem solving, Simon himself identifies numerous contingent architectural elements of the human cognitive system and phenomena that would seem to arise contingently from the interaction of these elements, and many of these seem amenable to an embodied treatment. Among Simon’s most enduring contributions to cognitive scientific theorizing are his notions of bounded cognition and the associated idea of satisficing. Humans have limited cognitive resources, limited long- and short-term memory capacities, for example. But, by satisficing (looking for a satisfactory, or “good enough,” solution), we compensate for the limited (that is, bounded) nature of our cognitive systems and thereby solve (well enough) problems we couldn’t easily (or even possibly!) solve optimally, by, for example, exhaustive search through, and evaluation of, all possible options. Most of us don’t undertake the project of finding the very best wine to go with our dinner; instead, we find one that’s good enough. We achieve this partly by employing search heuristics (we buy a wine we’ve heard of, that seems priced about right, and that has a high Wine-Spectator score posted beneath it).
At this point, one might reasonably wonder whether Simon wasn’t onto one of the most substantive insights of the embodiment-based literature. Think of how deeply nonoptimizing is the human use of heuristics and biases. Cognitive scientists with time on their hands and lots of computing power can explain to us why the shortcuts we use represent reasonable approaches to time- and resource-pressured problem solving. But, we, in situ, can’t do such calculations; we would lose the advantage of employing heuristics to satisfice if we were to attempt calculate the costs and benefits of using those heuristics in a given case (McClamrock 1995). Instead their use must come naturally, in some sense. And, the only plausible ways for them to come naturally are the ones emphasized by embodiment theorists: we solve problems by relying automatically on shortcuts built into the neural system, into our bodily structure, and into our environmental tools, and by learning to exploit ways in which the fine-grained details of our bodily and neural structures interact with fine-grained aspects of our environments.