Claire F. Michaels

The relevance of action in perceiving affordances: Perception of catchableness of fly balls

The catchableness of a fly ball depends on whether the catcher can get to the ball in time; accurate judgments of catchableness must reflect both spatial and temporal aspects. Two experiments examined the perception of catchableness under conditions of restricted information pickup. Experiment 1 compared perceptual judgments with actual catching and revealed that stationary observers are poor perceivers of catchableness, as would be expected by the lack of information about running capabilities. In Experiment 2, participants saw the 1st part of ball trajectories before their vision was occluded. In 1 condition, they started to run (as if to catch the ball) before occlusion; in another, they remained stationary. Moving judgments were better than stationary judgments. This supports the idea that perceiving affordances that depend on kinematic, rather than merely geometric, body characteristics may require the relevant action to be performed.

Affordances: Four points of Debate

In this article, attention is given to 4 issues regarding affordances. The first relates to the ontological status of affordances: Affordances do not have to be perceived to exist. The second concerns whether affordances are necessarily related to (ones own) actions. The third considers the relation between affordances and effectivities. Finally, the nesting of affordances is addressed, as is whether some particular level of description should be considered privileged. Some issues seem settled; others require more theory and experimentation.

Information and action in punching a falling ball

Lee, Young, Reddish, Lough, and Clayton (1983) reported that the timing control of jumping and vertically punching a dropping ball exploits the inverse of the rate of change of optical expansion, τ (r). We raise a number of methodological and logical criticisms against their experiment and conclusions and attempt to rectify them by examining elbow joint angles only, in seated punchers, under both monocular and binocular conditions, with two ball sizes, dropped from two heights. Differences between the binocular and monocular cases suggest the exploitation of different information. We present several techniques to help determine the operative variable(s) controlling the action. The optical variable used to initiate and guide flexion appeared to be expansion velocity (looming), rather than τ (r); extension appeared to be under the control of different variables in the monocular and binocular cases. Simulations using single variables and single perceptuo-motor intervals were of mixed success.

Information, Perception, and Action: What Should Ecological Psychologists Learn From Milner and Goodale (1995)?

Two issues motivated the reassessment of a core concept of ecological psychology: the relation between perception and action. The first was Milner and Goodales (1995) claim, based on neurological and behavioral evidence, that vision for perception is distinct from vision for action. The second was the apparent involvement of lower order, nonspecific optical variables in the control of action as reported by Michaels, Zeinstra, and Oudejans (in press). Perception in the usual ecological sense of the detection of information is obviously needed for action; however, to the extent that perception is defined in terms of explicit knowledge or awareness of environmental properties, including animal-referential ones, a separation was deemed justified. Perception, so construed, together with the ventral stream, was suggested to be about telling rather than acting.

Higher Order and Lower Order Variables in the Visual Perception of Relative Pulling Force

In 7 experiments, undergraduates judged the force exerted by a videotaped standing puller, a computer-generated (stick-figure) puller, or a computer-generated inverted pendulum. Single and stepwise multiple regression analyses determined the kinematic variables exploited by the participants. Results show that (a) judgments correlated highly with force and improved with feedback; (b) judgments correlated more highly with lower order kinematic variables than with force itself; (c) participants differed in the kinematic variables exploited; (d) participants changed over blocks of trials in the variables exploited; (e) some participants used compound kinematic variables; (f) the variables exploited depended on the type of feedback; and (g) judgments to upright pullers, inverted pullers, and simple pendula showed the same qualitative patterns. Implications for theories of direct perception, directed perception, and heuristics are considered.

The role of feedback information for calibration and attunement in perceiving length by dynamic touch

Two processes have been hypothesized to underlie improvement in perception: attunement and calibration. These processes were examined in a dynamic touch paradigm in which participants were asked to report the lengths of unseen, wielded rods differing in length, diameter, and material. Two experiments addressed whether feedback informs about the need for reattunement and recalibration. Feedback indicating actual length induced both recalibration and reattunement. Recalibration did not occur when feedback indicated only whether 2 rods were of the same length or of different lengths. Such feedback, however, did induce reattunement. These results suggest that attunement and calibration are dissociable processes and that feedback informs which is needed. The observed change in variable use has implications also for research on what mechanical variables underlie length perception by dynamic touch.

Learning to visually perceive the relative mass of colliding balls in globally and locally constrained task ecologies

Novice observers differ from each other in the kinematic variables they use for the perception of kinetic properties, but they converge on more useful variables after practice with feedback. The colliding-balls paradigm was used to investigate how the convergence depends on the relations between the candidate variables and the to-be-perceived property, relative mass. Experiment 1 showed that observers do not change in the variables they use if the variables with which they start allow accurate performance. Experiment 2 showed that, at least for some observers, convergence can be facilitated by reducing the correlations between commonly used nonspecifying variables and relative mass but not by keeping those variables constant. Experiments 3a and 3b further demonstrated that observers learn not to rely on a particular nonspecifying variable if the correlation between that variable and relative mass is reduced.

Lateral interception I: operative optical variables, attunement and calibration

J. J. Gibson (1966, 1979) suggested that improvement in perception and action can be attributed in part to changes in which variable is attended to. Such reattunement has been demonstrated with observers making judgments in response to simulations. The present study sought attunement changes in the perception of real events and in visually guided action. In 3 experiments, adults judged the passing distance of or attempted to catch balls. Discrete measures and the predictions of a modified required velocity model (e.g., R. J. Bootsma, V. Fayt, F. T. J. M. Zaal, & M. Laurent, 1997) were used to reveal which variables were exploited. Participants differed from each other and, to some extent, changed in the optical variables used, in catching as well as judging. Nevertheless, the changes were much smaller than in previous simulation-judgment studies; calibration was also found to underlie the improvements in performance.

On the Apparent Paradox of Learning and Realism

This article first summarizes how the definition of perception as the detection of information follows from the assumption of realism (e.g., Shaw, Turvey, & Mace, 1982). The realist position appears to be inconsistent with the empirical finding that novice perceivers often use nonspecifying variables and converge on the use of information only after practice with feedback (e.g., Michaels & de Vries, 1998). We argue that the appearance of inconsistency is due to the application and evaluation of realist principles beyond the scale of phenomena to which they apply. If the relevant principles are considered at the appropriate scales, convergence on information and realism imply each other. We also argue that the possibility of convergence and the associated use of nonspecifying variables should always be considered in the analyses of experimental results, especially if the information-granting constraints prevailing in the experiment are different from those prevailing in natural ecologies.

Shedding Some Light on Catching in the Dark: Perceptual Mechanisms for Catching Fly Balls

To catch a lofted ball, a catcher must pick up information that guides locomotion to where the ball will land. The acceleration of tangent of the elevation angle of the ball (AT) has received empirical support as a possible source of this information. Little, however, has been said about how the information is detected. Do catchers fixate on a stationary point, or do they track the ball with their gaze? Experiment 1 revealed that catchers use eye and head movements to track the ball. This means that if AT is picked up retinally, it must be done by means of background motion. Alternatively, AT could be picked up by extraretinal mechanisms, such as the vestibular and proprioceptive systems. In Experiment 2, catchers reliably ran to intercept luminous fly balls in the dark, that is, in absence of a visual background, under both binocular and monocular viewing conditions. This indicates that the optical information is not detected by a retinal mechanism alone.