An abstract noun, indirect-ness essentializes what it denotes. It invites us to think of indirectness as an underlying “property” or “quality” or “feature” of discourse, as if it could be studied without reference to communicative norms and ideals. If, instead, indirectness should be understood in relation to ideals of, say, directness, are such ideals universal, or are they historically contingent, group-relative norms, perhaps instances of what has been called “language ideology”?2 Beyond the familiar preoccupation with what is universal and what isn’t, we face the more elementary question of whether indirectness names a singular object of study at all.
Indeed, quite a few things have been called indirect. From this diversity let me highlight two distinct senses (with the proviso that this distinction is intended only to expose variation in the literature; it is not meant typologically, since the facets of indirectness distinguished below often overlap in practice).
“Indirectness,” in the first and familiar sense, has been a shorthand for talking about the exploitation of pragmatic ambiguity in discourse, where “utterances … convey something more or different from their literal meaning” (Blum-Kulka 1987: 141). If at one end of the spectrum lie what John Austin (1962), the chief architect of speech-act theory, called “explicit performatives,” like I bet you sixpence it will rain tomorrow, which seem to denote what they try to do, at the other ...