ABSTRACT
This chapter examines the difficulties involved in understanding recorded conversations, especially surreptitiously recorded ones, for example those made when a police agent wears a wire. Such recordings can be hard to understand, because of background noise, overtalking, and the fragmented nature of conversation. These problems can give rise to misperceptions due to listener expectations or bias. Areas of linguistics that can help interpret such hard-to-understand recordings include phonetics (how speech sounds are produced, and the structure of the sound waves of speech), phonology (organized patterning of speech sounds), and dialectology (regional and social differences in speech). These fields can help make clear what words were said. What speakers intended to communicate can be addressed by discourse analysis and pragmatics (how what is communicated can differ from what a person literally says). This chapter includes a summary of the 1930s Lindbergh child kidnap case as an introduction to the problem of speaker identification; an introduction to sound spectrograms and their possible use for speaker identification; an account of William Labov’s use of phonological dialectology to prove the innocence of a suspect from Boston in a telephoned bomb threat case in which the actual caller, Labov showed, was a New Yorker; and reports of work by Roger Shuy and Ellen Prince on interpreting what a speaker said and meant.
