Motivational theories of driver behaviour invoking risk homeostasis or behaviour adaptation have elicited considerable interest. However, their generality may be less than sometimes assumed. Specifically, they appear incomplete with respect to poor - visibility road casualties. Motivational theories imply a broadly constant rate of accidents, given steady motivational parameters, but accidents are consistently more frequent in poor visibility. An implicit assumption of motivational theories is that input to the driver is accurately and veridically processed in order to adequately compute risk. However, the driver's motion perception is likely to be very dependent on relative motion at the retina. In poor visibility the associated processing may be degraded - as reflected in poor thresholds for detecting object motion - or non-veridical - as reflected in simultaneous motion contrast. This paper illustrates the difficulty of developing global theories to explain "real-world" activities such as driving: theories addressing restricted aspects of road behaviour may be more fruitful.
Abstract