Notes for Andy Clark “An embodied cognitive science?”
Key concepts: adaptive coupling, affordance, foveation, wideware.
Related theorists: J.J. Gibson, N. Katherine Hayles.
Example of Bluefin tuna actively exploiting local environment.
(345) The physical system whose functioning explains the prodigious swimming capacities of the Bluefin tuna is thus the fish-as-embedded-in, and as actively exploiting its local environment.
Robot controller ought to exploit intrinsic dynamics of the environment like the Bluefin tuna.
(345) To understand how the robot's “brain” controls the robot's motions, a shift towards an embodied perspective is required. The controller must learn to exploit the rich intrinsic dynamics of the system.
Task of robotic vision must be to efficiency in service of real-world, real-time action rather than building rich inner model.
The key insight here is that the task of vision is not
build rich inner models of a surrounding 3-D reality, but rather to
use visual information efficiently and cheaply in the service of
real-world, real-time action.
(346) Biological vision thus gears its computational activity closely and sparingly to the task at hand, making the most efficient use of the persisting external scene.
Action and affordance
Gibson affordance from ecological psychology challenges traditional ideas about perception and action assuming rich internal representations with adaptively potent equilibrium couping agent and world.
(346) Related insights stem from the
work of J.J. Gibson and the ecological psychology movement.
This approach stresses bodily movement, ecological context and the
action-relevant information available in the perceptual array. A
central organizing construct is the concept of an 'affordance'.
(346) Such co-ordination dynamics constitute something of a challenge to traditional ideas about perception and action: they replace the notion of rich internal representations and computations, with the notion of less expensive strategies whose task is not first to represent the world and then reason on the basis of representation, but instead to maintain a kind of adaptively potent equilibrium that couples the agent and the world together.
Beyond adaptive coupling?
Challenge of embodied approach dealing with off-line reason, cognition, imagination when there is no active, adaptive coupling.
(347) can the embodied, embedded
approach contribute to our understanding of so-called
(347) The mark of the cognitive, then, is the capacity to engage in something like off-line reason – reasoning in the absence of that which our thoughts concern.
(348) But (and this is the mild dilemma) it does seem that the more decoupled and abstract the target contents become, either the less applicable the sensorimotor simulation strategy is, or the less clearly it can then be differentiated from the more traditional approaches it seeks to displace.
Simple versus radical embodiment
Claims of radical embodiment theorists: new tools and methods necessary, reject traditional notions of representation and computation, decomposition into functional internal subsystems is misleading.
The source of much recent excitement, however, are the striking
claims involving 'radical embodiment'. Such claims can be found in
the work by Tim Van Gelder, Thelen and Smith, Kelso, Varela et
Turvey and Carello and others.
(349) [they claim] (I) that understanding the complex interplay of brain, body and world requires new analytic tools and methods, such as those of dynamical systems theory. (II) that traditional notions of internal representation and computation are inadequate and unnecessary. (III) that the typical decomposition of the cognitive system into a variety of inner neural or functional subsystems is often misleading, and blinds us to the possibility of alternative, and more explanatory, decompositions that cut across the traditional brain-body-world divisions.
Foveation as example of physical analog to computer-science pointer data structure advantageously playing dual roles, leading to point that external environment (artifacts, texts, media, institutions) plays role in cognition as wideware.
An example – which also demonstrates how a single research program
can combine elements of simple and radical embodiment – is Ballard
use of a notion of 'deictic pointers'. A pointer, in artificial
intelligence, is an inner state, which can act both as an object of
computation and as a 'key' for retrieving additional data-structures
or information. Deictic pointers, as Ballard et
describe them, are physical actions – such as foveating a certain
location in visual space – that play a similar kind of functional
(349) The external environment, actively structured by us, becomes a source of cognition-enhancing 'wideware' – external items (devices, media, notations) that scaffold and complement (but usually do not replicate) biological modes of computation and processing, creating extended cognitive systems whose computational profiles are quite different from those of the isolate brain.
Expanding cognition by imbricating new, fine motor control behaviors same point Hayles makes about electronic literature.
(349) In short, the world of artifacts, texts, media, and even cultural practices and institutions, might be for us what the actively created whorls and vortices are for the Bluefin tuna. Human brains, raised in this era of cultural tools might develop strategies for advanced problem solving that 'factor in' these external resources as profoundly and deeply as the bodily motions of the tuna factor in and maximally exploit the reliable properties of the surrounding water.
Radical embodiment must deal with challenge of representation-hungry problems and environmentally-decoupled reason.
(350) The major challenge for the vision of 'radical embodiment' described here lies with the class of 'representation-hungry' problems and phenomena of off-line, abstract, and environmentally-decoupled reason.
Clark, Andy. 1999. “An embodied cognitive science?” Trends in Cognitive Sciences 3:9, September 1999, pp. 345-51.
Clark, Andy. “An embodied cognitive science?” Trends in Cognitive Sciences 3:9 (September 1999): 345-51. Print.