"Brown's book should be of interest to many computational linguists, especially those whose work is in topics related to anaphora, referring expressions, discourse structure, or speech understanding. In particular, it will interest those whose research is in the area of preventing or repairing miscommunication, a topic of current interest in the field. Brown's experimental tasks provide a straightforward method for reliably generating examples of how people address this issue....Language researchers considering alternatives for experimental design will also find the book useful. To its credit, the book is clearly written and well organized. In particular, the discussion of the design and motivation for the two tasks is clear and thought-provoking." Susan McRoy, Computational Linguistics