Reviewed by Scott Kiesling, University of Pittsburgh

This volume presents a study of epistemic markers used in a subset of the Santa Barbara corpus of spoken American English. Elise Kärkkäinen’s goals are to review the literature of stance taking in ‘naturally-occurring’ (4) conversations and to ‘highlight the essentially interactive nature of stance-taking’ (5) in American English. The first goal, the topic of Ch. 2, is nicely achieved. Although K could have cast her net wider into the epistemicity literature, she very wisely limited her review to studies relevant to interaction.

Indeed, one tension that K fights is the reification of stance as a preexisting category rather than a category that arises in conversation. After discussing the intonational unit (IU) in Ch. 3, K presents her quantitative analyses in Ch. 4, obtained through a valuable and carefully-constructed study (although based on a limited geographic range—west of the Rockies—and a small group of fifteen speakers). She finds that stance is usually indicated preceding the assertion it marks, and that ‘stance-taking can be viewed as highly regular and routinized’ (35). The latter claim applies to the types of stance found in the data, although the data do not conform to K’s interpretation: while I think is used more than any other marker, the analysis of ‘highly regular and routinized’ (35) is highly relative. K shows that I think is used for nine percent of the total tokens, while there are at least seventy-seven token types in the 503 tokens in her data (37). I do not interpret this as highly routinized but rather as extremely varied. In any case, K needs to define routinized and compare her findings to those of another language to show that this is a ‘limited set of items’ (37). The routinization of stance placement is more convincing: the majority of instances are in IU-internal position. However, it is not clear that this is an interactional constraint or one imposed by English syntax. (It might be argued either way, although K plays down the possible syntactic constraints.)

Ch. 5 moves to a study of how stance-taking is an interactive activity: ‘displaying stance is thus engendered by what happens […] in prior discourse’ (105). This is a welcome and important step that shows both the link between interpersonal and epistemic stance and that stance is not only (even rarely) limited to the expression of the state of knowledge in the speaker’s head. The analyses are nicely done, and K manages to make some interesting interactional generalizations about how I think is used by speakers in interaction. Ch. 6 summarizes K’s study and argues for more work in interactional linguistics.

This is a valuable work that anyone studying stance or epistemicity should examine, for both its method and findings, especially those on the interactional nature and interpersonal function of stance-taking.