A Constructionist Cognitive Model of Presence

This paper proposes a cognitive model of presence which emphasizes the interaction of top-down and bottom-up processes, and the subsequent creation of temporary representations of an environment in working memory. These temporary representations (which are termed constructions) are formations of meaning about the environment, which allow a subject to make inferences and interact in that environment. This paper argues that this constructionist position follows from the existence of hypothesis selection models of presence such as proposed by Slater & Steed. Finally, the paper contrasts this model to that of Wirth et al and discusses some of the implications of this model for presence in unfamiliar or unrealistic environments. 1. Cognitive models of presence With the recent move towards a comprehensive theory of presence, cognitive science has surfaced as an important basis for understanding presence. For example [1], in discussing breaks in presence, propose a filtering, information-processing model in which one of two information streams (real environment/virtual environment) is selected for processing. Many other examples of the application of cognitive theory in presence exist for instance [2] and [3]. Although it is generally accepted that presence is a complex, multifactor construct, cognition seems to be significant enough a factor that an understanding of the cognition of virtual environments (and indeed real environments) can lead to invaluable insights into fundamental aspects of presence. Therefore, although this paper presents an exclusively cognitive model of presence, it should be understood as an attempt to understand the role of cognition in presence, rather than an argument that a complete understanding of presence can come from examining cognition alone. 2. Constructionism in cognition Constructionism involves the interaction of topdown and bottom-up processes which create a mental construct reflecting some external or imagined situation in such a way that inferences and decisions can be made in relation to that situation [4] [5]. Furthermore, constructionism is generally associated with a dynamic state of cognition in which bottom-up processing activates particular top-down processes which in turn lead to a bias in bottom-up processing [6]. The ‘rabbit/duck’ switching illusion discussed in [1] is an example of two constructions arising from the same bottom-up data, but mediated by a different top-down bias. Constructions are capable of adapting to changes in the external world, but only to a degree; if the products of bottom-up processes lead to a significant mismatch with the construction, a reconstruction will occur so as to better represent the stimulus situation [7]. That constructionism is useful in presence research has already been argued by [8], and indeed, constructionism already exists in the literature in various forms. [9] for instance, argues for the interaction of internal and external factors as a determinant of presence. More recently, [10] describes a concept (perception and action in 3d space) in which a continual re-consideration of the environment in terms of the user’s cognition, perception and emotions occurs. Also, [11] proposes a model in which presence is an active process in which sensory and conceptual data are used to construct mental models of the environment (see section 7 below for a discussion of this model). 3. The need for constructionism: The virtual