Prosody and Humor
Thomas Flamson, Gregory A. Bryant, and H. Clark Barrett 1
tải về 1.21 Mb. Chế độ xem pdf
|
21-tran thanh ai
| 62 Thomas Flamson, Gregory A. Bryant, and H. Clark Barrett
1. Introduction In most contexts of human social interaction people often behave in ways that can be broadly described as humorous. We tell jokes, make faces, and laugh with one another for an infinite number of reasons, many of them related to our folk under- standing of what it means to be funny. But the incredible variety of circumstances in which this kind of behavior occurs makes the scientific study of humor theoret- ically and methodologically difficult. Recently, however, researchers have begun analyzing specific aspects of humor in an attempt to characterize this pervasive and complicated human behavior. In this book, a particular emphasis is placed on the underlying prosodic patterns associated with humorous speech — that is, what are the tonal and rhythmic properties of speech typically associated with verbal joking? How do people mark their attempts at being funny, and do they gener- ally even mark them at all? Here we report an analysis of spontaneous humorous speech in a rural Brazilian collective farm, and we aim to show that, in fact, people often do not signal explicitly that they are being funny. We present this data in the context of a recently developed encryption theory of humor (Flamson and Barrett 2008). Rather than marking humorous utterances using explicit means such as speech prosody, speakers often rely on situational context and common knowl- edge in being understood. As described below, one likely function of humor is not merely to entertain an audience, but rather to signal encrypted knowledge and to communicate useful information in the context of recognizing and acknowledg- ing social alliances. In this capacity, explicit marking should often be avoided, and in our sample, this is precisely what we found. Most work on the pragmatics of conversational humor has focused primar- ily on the level of interaction between interlocutors (e.g., Kotthoff 2006a), or on performative aspects such as timing and framing (e.g., Norrick 2003), rather than on specific acoustic features. But the existing research on prosody and humor has uncovered some interesting results. For example, a study of conversational humor among brain-damaged patients found a significant correlation between stroke patients’ ability to decode prosody and their appreciation of humor (Heath and Blonder 2005) suggesting that nonverbal vocal signals can be important for iden- tifying humor. Other recent work has examined the specific acoustic features of joke telling, as one prevalent type of humor common in ordinary conversation (Pickering et al. 2009). In this study, undergraduate students told jokes and the punch lines were analyzed for distinctive pitch, loudness, and speech rate features. Contrary to suggestions in the literature (e.g., Norrick 2001), this work found that while punch lines did have lowered pitch and slowed speech rate relative to earlier portions of the jokes, these effects were not considered to be above and beyond what one would expect given the final position in a spoken monologue Prosody in spontaneous humor: Evidence for encryption 63 independent of humor. That is, speakers did not seem to provide distinctive pro- sodic signals that they were in fact ending a joke with a punch line. This result is especially interesting given the semi-scripted nature of the productions. Even with a stereotypical expectation of joke-telling behavior, people failed to produce highly stereotyped utterances to mark their humor. These results, albeit limited, suggest that one-to-one mappings between speech characteristics and humor are not common in everyday talk. Research examining how people signal ironic intent in spontaneous conversation also shows that rather than using some stereotyped form to communicate irony (often a form of hu- morous speech), speakers rather tend to contrast prosodic features quite variably (Attardo 2001; Bryant and Fox Tree 2005; Bryant 2010, 2011). Vocal strategies for communicating subtle meanings likely depend on many complex social factors, and the acoustic form speakers choose to adopt will vary according to the prag- matic function in a given communicative context. From this perspective, speech acts are strategic signals in a context of cooperation and conflict (Pinker, Nowak, and Lee 2008). The encryption theory of humor (Flamson and Barrett 2008) maintains that humor is produced by encrypting multiple implicatures within an utterance that can only be understood by audience members who share relevant background information (including attitudes, beliefs, and preferences, in addition to propo- sitional knowledge). When an audience member has access to this background information, he or she is able to “decrypt” the hidden implicature(s) — that is, they “get” the joke. This hypothesis builds on a relevance theoretic model of inferential communication (Sperber and Wilson 1995), and predicts that the explicit marking of humor will depend largely on context and audience composition (see also Clark and Schaefer 1987; Yus 2003). In contexts where people are particularly motivated to appeal to as many audience members as possible, they would be expected to mark their humor production in a variety of ways to facilitate the detection of encrypted content, and thereby enhance the breadth of audience members that would be able to detect their similarity with the speaker. Conversely, in contexts where speakers are motivated to solicit pre-existing ties, they would not be ex- pected to mark humor, highlighting the depth of similarity with those audience members who manage to detect the encryption. We examined spontaneous humor production in a community setting amongst individuals with long standing social ties. Based on the encryption the- ory of humor, we expected that there would be relatively little explicit marking of jokes in this specific context. The goals of most speakers revolved around official business of a cooperative work organization, and in this capacity, most speakers were highlighting depth of similarity with close social allies. When encrypting information, speakers should not provide disambiguating information to those 64 Thomas Flamson, Gregory A. Bryant, and H. Clark Barrett without an ability to decrypt the information, and thus, laughter can honestly reveal possession of relevant implied information (i.e., a key). As described be- low, we explored the acoustic features of speech containing humor and looked for measurable changes in speech prosody (i.e., prosodic contrasts), and we expected that the amount and nature of prosodic contrasts between punch lines and set-up information would be comparable to contrasts between set-up speech and speech immediately preceding it (baseline). tải về 1.21 Mb. Chia sẻ với bạn bè của bạn:
|