On reconstruction of task context after interruption

Theoretical accounts of task resumption after interruption have almost exclusively argued for resumption as a primarily memory-based process. In contrast, for many task domains, resumption can more accurately be represented in terms of a process of reconstruction-perceptual re-encoding of the information necessary to perform the task. This paper discusses a theoretical, computational framework in which one can represent these reconstruction processes and account for aspects of performance, such as measures of resumption lag. The paper also describes computational models of two sample task domains that illustrate the sometimes complex relationship between reconstruction and more general human cognitive, perceptual, and motor processes.