Speaking of Crime: The Language of Criminal Justice. By Lawrence M. Solan and Peter M. Tiersma. Chicago: University of Chicago Press, 2005. Pp. 264. $55.00 cloth; $22.00 paper. Reviewed by Katherine Beckett, University of Washington In Speaking of Crime: The Language of Criminal Justice, Solan and Tiersma offer a simple yet powerful thesis: many legal decisions hinge on the interpretation of words, but the criminal justice system adheres to deeply entrenched but often incorrect or problematic assumptions about language. These errors are not randomly distributed, but systematically favor the state. As Solan and Tiersma put it, [t]he legal system is willing to ask serious questions about how understandable language is so long as the answers do not threaten important legal institutions (p. 27). The central objective of this book is to bring linguistic education and enlightenment to the legal community. In this goal the authors succeed admirably: Speaking of Crime provides a clear, concise, and lively analysis of how legal decisions and processes are informed by questionable assumptions about language. The analysis is replete with fascinating examples taken from recent and important legal rulings. Solan and Tiersma begin with a brief primer in linguistics that emphasizes two things. First, the meaning of words is determined in part by the context in which utterances occur. Indirect requests for access to one's glove box, for example, are likely to be interpreted and responded to differently if the comes from a car wash operative than from a state patrol officer. second, utterances convey meaning, but they may also be performative acts aimed at achieving particular ends. Many legal rulings ignore these insights. Subsequent chapters provide numerous examples of legal miscomprehension of the nature of language. Police efforts to obtain consent for searches are a particular concern. The authors begin by pointing out that most searches do not involve a warrant, and that many people (some of whom are guilty) give consent even when it is quite likely that doing so will lead to the discovery of incriminating evidence. Solan and Tiersma argue that this counterintuitive pattern reflects, at least in part, the widespread assumption that police requests for consent to search are in fact commands. This inference is supported by linguistic studies indicating that people routinely make requests and demands indirectly. Requests and demands are closely related speech acts: both are an attempt to impel someone to do something. Yet commands are often expressed as requests. Saying Would you mind cleaning up your room now? rather than Clean your room now allows the recipient of the command to save face and may be perceived as more polite. Nonetheless, if the former request is made by a parent to a small child, it may be appropriately understood as a command. An apparent might be a request, but it might be a demand, depending, in part, on the context in which the is made. Police requests for consent often occur after law enforcement officers have pulled someone over and ordered them out of the car. …