How do you pronounce biopic, synod, and Breughel? - and why? Do our cake and archaic sound the same? Where does the stress go in stalagmite? What's odd about the word epergne? As a finale, the author writes a letter to his 16-year-old self.
Date: Fri, 6 May 2005 15:44:58 -0500 From: Brian Reese Subject: True to Form: Rising and Falling Declaratives as Questions in English
AUTHOR: Gunlogson, Christine TITLE: True to Form SUBTITLE: Rising and Falling Declaratives as Questions in English SERIES: Outstanding Dissertations in Linguistics PUBLISHER: Routledge (Taylor and Francis) YEAR: 2003
Brian Reese, Department of Linguistics, University of Texas at Austin
"True to Form" by Christine Gunlogson is an investigation of the use of rising and falling declarative sentences as questions in English. The central claim is that the meaning and use of these sentences is best understood in terms of the interaction between their defining formal characteristics: sentence type and intonational contour. The primary goal is to provide a compositional analysis of these formal elements and, given such a formal analysis, to explain facts about the distribution across contexts of rising interrogatives, rising declaratives and falling declaratives used with the force of a question. The book should be of value to researchers interested in the semantics and pragmatics of questions and intonation.
Chapter 1 provides an overview of the issues introduced above, while simultaneously introducing the book's main methodological and formal assumptions. Throughout the dissertation, Gunlogson uses a minimal pair methodology, focusing on the use as questions of sentences like those in (1). (A question mark indicates a final rise, while a final fall is indicated with a period; I use Gunlogson's numbering throughout the review.)
(1) a. Is it raining? b. It's raining? c. It's raining.
There are two minimal pairs in (1): (1a-b) and (1b-c). Each of the examples in (1) has the same propositional content, but (1a-b) differ in sentence type (interrogative vs. declarative) and (1b-c) differ in intonational contour (rising vs. falling). Each minimal pair, then, isolates a distinct formal element: sentence type vs. intonational contour. Contexts where rising and falling declaratives pattern together to the exclusion of rising interrogatives provide insight into the semantic contribution of rising intonation. Likewise, contexts where rising interrogatives and rising declaratives pattern together to the exclusion of falling declaratives provide insight into the semantic contribution of sentence type. Gunlogson's thesis is that both of these formal elements contribute to the distributional restrictions on uses of declarative sentences as questions. The primary goal of the dissertation, then, is to provide a compositional semantic analysis of sentence type and intonation, with a secondary goal of explaining how the meaning of rising and falling declaratives relates to their interpretation as questions.
Gunlogson is careful to explicitly state her assumptions regarding the formal elements isolated by her minimal pair methodology, especially phonological assumptions relating to intonational contour. Final rises and falls are treated as elements of an intonational lexicon (Ladd 1980). Gunlogson characterizes a final rise as a non-falling pitch contour from the nuclear pitch accent to the end of the utterance ending at a point higher than the nuclear accent. In the Pierrehumbert and Hirschberg (1990) system this definition would encompass the following tunes: H* H H%, L* H H%, L* H L% and L* L H%. A final fall is characterized by a non-rising pitch contour from the nuclear pitch accent to the terminus ending at a point lower than the nuclear accent.
Chapter 1 closes with a brief discussion of the previous literature.
Chapter 2 lays the empirical groundwork for Gunlogson's formal analysis. Specifically, she presents data supporting the generalizations in (8) and (9).
(8) Declaratives express a bias that is absent with the use of interrogatives; they cannot be used as neutral questions.
(9) Rising declaratives, like interrogatives, fail to commit the speaker to their content.
She bases the generalization in (8) on the observation that interrogatives , but not rising or falling declaratives, are felicitous in contexts that require neutrality. Polar interrogatives can be used as exam questions, for example, as shown in (15).
(15) [as an exam question] a. Is the empty set a member of itself? b. #The empty set is a member of itself? c. #The empty set is a member of itself.
Gunlogson uses a similar kind of argument to support (9), finding contexts where rising declaratives pattern with rising interrogatives. She notes that rising declaratives and polar interrogatives are consistent with a wider range of speaker attitudes than falling declaratives, from routine acceptance to skepticism. Falling declaratives on the other hand only allow propositional attitudes consistent with commitment. Furthermore, rising declaratives and polar interrogatives are consistent with contexts where the knowledgeability of the addressee rather than the speaker is at issue.
Gunlogson concludes that declaratives express commitment to the propositional content by some discourse participant. This rules out declaratives as questions in contexts where neutrality is called for. It is the intonational component of a locution that determines which participant is committed. A rising declarative commits the addressee, while a falling declarative commits the speaker. Hence a falling declarative only allows speaker attitudes consistent with commitment.
Chapter 3 presents a formal, compositional analysis of these intuitions that builds on Stalnaker's (1978) notion of common ground. Whereas a classic "Stalnakerian" common ground is a set of propositions representing the mutual beliefs of the participants in a discourse, Gunlogson's common ground is a tuple of sets of propositions representing the public beliefs or discourse commitments of each participant. A classic common ground is recoverable by collecting those propositions common to each set in the tuple. Crucial to Gunlogson's analysis is the notion of "public belief." In a context where A has argued for "p" and B for "not p", neither "p" nor "not p" will be in the common ground, but "A believes that p" will be (among other propositions). So Gunlogson's common ground records the beliefs that each discourse participant has publicly committed themselves to, be it through linguistic or extra-linguistic means. Stalnaker's "context set", the set of possible worlds consistent with the mutual beliefs of the discourse participants, is mirrored in Gunlogson by "commitment sets", the set of worlds consistent with a participant's public beliefs. There is a commitment set associated with each discourse participant.
A number of auxiliary notions are defined which describe the status of a proposition p with respect to a context C, e.g., commitment, joint commitment, resolved, controversial. Most important are formal definitions of bias and neutrality. A context C is said to be biased toward a proposition p iff p is a commitment of at least one discourse participant and not-p isn't a commitment of any discourse participant. In other words, it is possible for p to become a joint commitment of the discourse participants without revising the context. C is neutral with respect to p iff no discourse participant is committed to either p or not-p.
With this extended notion of context in hand, Gunlogson proceeds to provide a semantics for sentence type and intonation in terms of their context change potential (CCP) (Heim 1982). A declarative sentence, for example, updates the commitment set of one of the discourse participants with the descriptive content of the declarative. Whether it is the speaker's or addressee's commitment set is determined by the intonational component: a final fall commits the speaker, whereas a final rise commits the addressee. A rising declarative, then, updates the addressee's commitment set with the descriptive content of the declarative. Interrogatives, on the other hand, do not result in an update of the targeted commitment set. If L is an utterance of a rising interrogative in a context C, C + L = C. So with respect to any context C, no declarative sentence is ever neutral and no interrogative is biasing. This result is consistent with the descriptive generalization in (8): declaratives always express commitment on the part of some discourse participant. The generalization in (9) is also covered, since a final rise always targets the commitment set of the addressee, not the speaker.
Chapter 3 closes with the definition of a number of auxiliary notions such as entailment, uninformativeness and vacuousness. Entailment is defined, as usual, in terms of inclusion. A locution L based on a sentence S is uninformative with respect to an individual commitment set iff S is entailed by that commitment set (and assuming that L's presuppositions are satisfied in the context). A locution is vacuous iff its descriptive content p is a joint commitment in the context C.
Chapter 4 provides a definition of the pragmatic category "polar question" in distributional terms. A locution L, Gunlogson claims, is interpretable as a polar question in a context C iff L is uninformative with respect to the context set of the addressee. Interrogatives are uninformative by definition, so any utterance of a polar interrogative qualifies as a polar question. Declaratives, on the other hand, express commitment either on the part of the speaker or the addressee and so count as polar questions only if the descriptive content p of the declarative is entailed by the addressee's commitment set, i.e., if the addressee is already publicly committed to p. Gunlogson refers to this constraint as the Contextual Bias Condition. She provides evidence that uninformativeness with respect to the addressee's public beliefs is critical, rather than what it might be reasonable for the speaker to assume about the beliefs of the addressee.
Section 4.2 addresses declarative questions in more detail. Gunlogson introduces uses of rising declaratives as questions that do not satisfy the uninformativeness condition. For example, she introduces data where the descriptive content p of the declarative is not a logical consequence of the addressee's public beliefs. In such cases she suggests the possibility of accommodating (Lewis 1979) the rising declarative as a question. She mentions a number of factors that may affect the addressee's ability to accommodate a declarative question, for example the relative knowledge of the discourse participants.
Section 4.3 notes the affinity between polar interrogatives and rising declaratives given the formal apparatus from Chapter 3 and the Contextual Bias Constraint, accounting for the naturalness of rising declaratives as questions. She then offers a revision of the definition of polar questions that is needed in order to capture the narrower distribution of falling declaratives as questions compared to rising declaratives. Falling declaratives as questions commit the speaker to the descriptive content p of the locution and so differ from interrogatives and rising declaratives in contexts where the Contextual Bias Condition is met. She suggests that in uses of falling declaratives as questions, it is understood that the speaker's commitment is contingent on the addressee's commitment.
Gunlogson notes that rising declaratives as questions are more marked than polar interrogatives and that falling declaratives as questions are more marked than rising declaratives, where markedness is understood in terms of distribution across contexts. This follows from the update semantics given Chapter 3 and the final, revised definition of polar question in Chapter 4. The chapter closes with a discussion of the discourse functions consistent with locutions satisfying the Contextual Bias Constraint.
Chapter 5 summarizes the empirical observations and formal analysis presented in the preceding chapters.
"True to Form" is an important contribution to the literature on the meaning and pragmatic effects of specific intonational contours. Gunlogson takes a difficult set of data and presents it in a way that allows clear empirical generalizations to emerge. This is due, no doubt, to the explicitness with which Gunlogson conducts her investigation, especially as regards the minimal pair methodology. The clarity of the discussion is helped by the fact that Gunlogson circumscribes the empirical domain of the investigation rather narrowly. This simplifies the discussion greatly, but consequently it is not clear how her analysis extends to other types of data, e.g., negative polar interrogatives as in (15d), which have a biasing effect despite their interrogative syntax (Ladd 1981, Buring and Gunlogson 2000, Romero and Han 2004).
(15) [as an exam question] a. Is the empty set a member of itself? b. #The empty set is a member of itself? c. #The empty set is a member of itself. d. #Isn't the empty set a member of itself?
Gunlogson acknowledges this narrow empirical focus and recognizes the existence of such examples.
A more substantial criticism relates to the requirement that the addressee be publicly committed to the descriptive content of a a rising declarative used as a question. (Safarova 2005 raises similar concerns.) The Contextual Bias Condition predicts that only affirmative answers to the question should be possible since the proposition corresponding to the positive answer must be a public belief of the addressee. Although this seems right for "echoic" uses of rising declaratives, it runs into problems with respect to the so-called "inferential" questions discussed in Chapter 4. A negative answer seems equally likely in these cases. (120') is a modified version of the dialogue in Gunlogson's (120).
(120') A: Mark and Helena are leaving for Japan this week. B: Oh... You talked to Helena? A: No. I saw her sister at the supermarket.
Gunlogson accounts for this kind of example by accommodating a conditional of the form q --> p in the addressee's commitment set where p is the descriptive content of the rising declarative and q is a public belief of the addressee on which the inference is based. The addressee, then, is effectively committed to p. Yet, a negative answer is felicitous in this context. These facts weaken the argument for Gunlogson's definition of a polar question, in particular the Contextual Bias Condition, which follows from her compositional semantic story.
Overall, however, "True to Form" is an important step in research on the relationship between intonation, meaning and illocutionary force and should serve as a model for its clarity and formal explicitness. In addition, it provides a formally precise and clearly articulated extension of the classic notion of common ground that should be useful to researchers working on a variety of topics in formal semantics and pragmatics, especially those interested in intonational meaning and discourse.
Buring, Daniel and Christine Gunlogson. 2000. Aren't positive and negative polar questions the same? Manuscript.
Heim, Irene. 1982. The semantics of definite and indefinite noun phrases. Ph.D. dissertation, University of Massachusetts, Amherst.
Ladd, D. Robert. 1980. The structure of intonational meaning: evidence from English. Bloomington: Indiana University Press.
Ladd, D. Robert. 1981. A first look at the semantics and pragmatics of negative questions and tag questions. Chicago Linguistic Society 17, pp. 164-171.
Lewis, David. 1979. Scorekeeping in a language game. Journal of Philosophical Logic 8: 339-359.
Pierrehumbert, Janet and Julia Hirschberg. 1990. The meaning of intonational contours in the interpretation of discourse. In Philip R. Cohen, Jerry L. Morgan, and Martha E. Pollack, eds., Intentions in communication. Cambridge, MA: MIT Press, pp. 271-311.
Romero, Maribel and Chung-hye Han. 2004. On negative yes/no questions. Linguistics and Philosophy 27: 609-658.
Safarova, Marie. 2005. The semantics of rising intonation in interrogatives and declaratives. In Emar Maier, Corien Bary and Janneke Huitink, eds., Proceedings of SuB9. www.ru.nl/ncs/sub9, pp. 355-369.
Stalnaker, Robert. 1978. Assertion. In Peter Cole, ed., Pragmatics: Syntax and Semantics, Volume 9. New York: Academic Press.
ABOUT THE REVIEWER:
ABOUT THE REVIEWER
Brian Reese is a PhD candidate in the Department of Linguistics at the University of Texas at Austin working under the supervision of Professor Nicholas Asher. His main research interests include formal semantics and pragmatics and computational linguistics. He is currently preparing a dissertation on bias in questions, addressing, in particular, the semantics and pragmatics of questions containing negative polarity items, negative polar questions and the role of intonation in promoting bias.