The pitch tracks in (2a) and (2b) differ in two ways. First, the wh-word direct object in (2b) has its pitch boosted; compare the lower pitch of the direct object in the statement in (2a). Second, there is a domain, starting with the wh-phrase and ending with the wh-complementizer (shaded in the pitch track in (2b), which is characterized by pitch compression: the peaks in this domain (circled) are lower than they would normally be. Japanese wh-questions, then, involve a prosodic domain of some type which starts with the wh-phrase and ends with the complementizer. The proposal being defended here will be that all languages are attempting to do this; every language tries to create a prosodic structure for whquestions in which the wh-phrase and the corresponding complementizer are separated by as few prosodic boundaries as possible. How languages achieve this varies from language to language, depending on where the complementizer is and on what the basic rules for prosody are.Schematically, then, the proposal is this. Suppose we have an expression in which a whphrase and its corresponding complementizer are separted by prosodic boundaries, as in (3):