Aside from my dissertation, this paper is the first to describe the use of dynamically-updated verbal descriptions by blind people to support complex spatial tasks in both real and virtual environments. Of note, performance is similar between groups, a result that supports my long-standing argument that challenges to blind navigation are related to lack of information access rather than lack of visual access to the environment. The paper also gives a good description of the verbal protocol underlying the dynamically-updated geometric descriptions used in a number of subsequent papers.