Most of the research on traffic behaviour has focused on the individual road user, despite the fact that road users rarely encounter traffic situations in which they are not confronted with other road users. Most models of the driving task represent the driving task from the perspective o f one individual road user. For example, in an overview article on driver behaviour models, Michon (1985) discusses models that primarily take only one active road user into account. Until now, models o f driving behaviour focus on the interaction between road users and their impact on each other. Take, for example, an intersection with two road users approaching each other from different directions. In this case, direct verbal communication with each other is not possible. Despite this, situations like these rarely develop into an accident. Michon (1985) points out that the interactions between road users are mediated by distinctive and frequently subtle cues. In the past, attempts to describe the relations between such cues and the road user’s behaviour have been made, which have often taken the form of ethological models (Bliersbach & Dellen, 1980; Shor, 1964; van der Molen, 1983). In these models an attempt is made to determine specific behaviour, which elicits stimulus configurations. Unfortunately, as the focus is on the observable interaction situation, these models have not been able to describe the information processing o f road users in interaction situations, which can be considered to drive the overt behaviour (e.g. paying attention to stimuli, interpreting them and deciding how to respond). It is remarkable that, with so many interactions in traffic, so few accidents actually occur, considering the limited communication possibilities between road users. Besides these communication limitations, other aspects contribute to the complexity o f interactions in traffic. As Chauvin and Saad (2000) point out, the amount of experience of road users may differ, as may their goals, knowledge and strategies. They mention a number of measures in driving situations that support and organise interactions. First of all, the infrastructure and formal rules play an important role in organising the way in which interactions will develop. Also, means of communication such as the use of indicator, headlights and horn are mentioned by Chauvin and Saad. W e could also add changes in approach-speed, gestures and eye contact to the list of means of communication. M ost accidents are currently attributed to the behaviour of the road users involved, which is often seen as deviating from some normative behaviour (Brookhuis, de W aard & Janssen, 2001). From this point of view, it is often assumed that the accident could only have been prevented by this same road user not exhibiting this behaviour. An aspect which is often neglected is that the accident could also have been prevented by the other road users involved performing some kind of behaviour which compensates for the behaviour of the first road user. As “deviating behaviour” occurs much more frequently than accidents, this could imply that this compensation mechanism must be quite robust. One should ask oneself when a particular action should actually be considered a “deviating behaviour”, as these are often compensated for so smoothly that they are hard to recognise as deviating. Also, there is no easily definable boundary of “normal” or “correct” behaviour which can be defined if we take into account that normative traffic rules are often open to interpretation and even require that interpretation in order to make them applicable to the diversity of situations met in practice. Un till now, a detailed understanding of this compensatory mechanism has not yet been achieved.
Abstract