Daniel C. O'Connell

Turn-Taking: A Critical Analysis of the Research Tradition

In the following we present a radical critique of the assumptions, concepts, methods, statistics and interpretation of data, and theories that have characterized the recent research tradition concerned with turn-taking. The principal representative of this tradition is the “simplest systematics” of Sacks, Schegloff, and Jefferson (1974). Attempts to describe the generalizable properties of turn-taking have quite inappropriately and unsuccessfully been limited for the most part to formal approaches that have deliberately excluded considerations of conversational content and purpose. We start instead from the assumption that the ultimate criterion for the success of a conversation is not “the smooth interchange of speaking turns” (Cutler & Pearson, 1986, p. 139) or any other prescriptive ideal, but the fulfillment of the purposes entertained by two or more interlocutors. Our approach is that of a psychology of language use based on Bühler (1927; 1934/1982) and Rommetveit (1974). The emphasis is deliberately placed on social aspects of language as means of communication.

Transcription and the Issue of Standardization

For millenia, the human family has relied on transcription of spoken discourse into permanent written records as an indispensable tool of culture, history, and science. Currently, various transcription systems are used by linguists, psycholinguists, sociolinguists, and ethnographers to encode verbal, prosodic, paralinguistic, and extralinguistic features of spoken discourse for scientific analysis. The recent history of such systems, some problems involved in their scientific application, and arguments against the mandatory standardization of such systems are presented. Guidelines for the construction and use of transcription systems are formulated.

HOW DO TRANSCRIBERS DEAL WITH AUDIO RECORDINGS OF SPOKEN DISCOURSE

Four undergraduate volunteers, two women and two men, transcribed an audiotaped interview of former United States President Ronald Reagan by Dan Rather. Their first transcript was made from a single complete playing, with only pausing, but no repetition or replay allowed. Thereafter, two of the transcribers made corrections by an “on-line” method—complete playings as often as they wished. The two others were allowed an “off-line” method—unlimited playback of any portions. None produced a verbatim transcription, but all preserved semantic content quite well. Still, deletions were numerous, particularly of discourse markers and hesitation phenomena, both of which characterize spoken, not written discourse. Significantly more deletions in the on-line than in the off-line condition indicated the difficulty of audiotape processing without off-line replay. These results are discussed in light of recent theories of speech processing.

Interplay of Literacy and Orality in Inaugural Rhetoric

Inaugural addresses require a complex interplay of literacy and orality. They are written to be performed, but oral performance is not explicable in terms of textual analyses alone. Texts of the 55 inaugurals of the 42 U.S. presidents and audio recordings of the 16 dating from F. D. Roosevelt were the corpus for this study. The general hypothesis was that changes in media technology and in Presidential governance have moved both text and performance of inaugurals in the twentieth century in the direction of “conversational style.” Textual response measures were frequency-of-occurrence ratios of words (per paragraph, sentence, punctuation, and discourse marker), of syllables (per paragraph, sentence, word, punctuation, and discourse marker), and of first-person pronominal forms. Performance response measures were speech and articulation rates, percentage of pause time, pause duration, and phrase length. Use of contractions was also analyzed. Textual analyses showed a shortening of units and a shift from singular to plural first-person pronominal forms in the course of 200 years. Performance of the inaugurals over the past 60 years showed no diachronic changes, but was dramatically slower than that of other speech genres. Use of contractions was limited to three recent inaugurals. Various published texts of Reagans first inaugural and Bushs and Clintons inaugurals were compared with one another and with the audio recordings and were found to differ from one another in text, punctuation, and format, and from the audio recordings in text. The notion of conversational style is critically discussed, particularly in terms of the boundaries imposed upon it by the norms of both literacy and orality.

Orality and literacy in public discourse: An interview of Hannah Arendt☆

Abstract Orality and literacy were investigated in a 1964 German television interview of Hannah Arendt by Gunter Gaus and publications thereof. The interview itself was awarded the Adolf Grimme Prize (1964); and yet, the publications were extensively edited. This paradox raises the question: what was changed in this transition from successful oral performance to readable literate prose? Information regarding temporal organization was inevitably lost in the publications. Back channel signals, hesitations, contractions and elisions, and paralinguistic phenomena (e.g., laughter) were virtually eliminated. Turn-taking was regularized, and many discourse markers were removed. Comparisons with another prize-winning interview (Princess Diana with Martin Bashir; BBC, 1995) revealed optional, rather than prescriptive ways of being successful: although spoken syllables in both interviews were allocated to interviewers and interviewees in a 25% to 75% ratio, turns of both interviewer and interviewee were much longer in the Arendt than in the Princess Diana interview; Arendt spoke more slowly and used more hesitations than any other speaker; although both interviewers remained atypically off-camera, their styles were different. Gaus used frequent back-channeling, and Bashir none at all; Bashir articulated more slowly and less hesitantly, interrupted less frequently, and his pauses before taking his turns were twice as long as Gauss. Characteristics of orality and literacy in excellent public discourse are discussed.

Race and Gender in Current American Politics: A Discourse-Analytic Perspective

Male and female, black and white political interviewees (M. Albright, B. Clinton, H. Clinton, B. Obama, C. Powell, and C. Rice) of Larry King on CNN TV are used to ascertain whether ethnicity and gender affect the way politicians actually speak. Qualitative comparisons are made of Obama’s hesitations and rate with and without a threatening context. A number of normalized response measures are evaluated quantitatively: percentage of syllables spoken by each interviewee, and use of interjections, interruptions, self-referent I, non-standard English, y’ know, and syllables of laughter. Senator Obama and Secretary of State Rice become the focus of the comparative evidence that both ethnicity and gender do indeed influence the speaking of politicians.

Maya Angelou's Inaugural Poem

Maya Angelous performance of her poem On the Pulse of Morning at the inauguration of William Jefferson Clinton as 42nd President of the United States on January 20, 1993, was compared with other poetry and prose performances. Measurements of her pause frequency, pause location, pause duration, phrase length, speech rate, articulation rate, and percentage of pause time all uniquely characterized her performance. Printed versions of Angelous inaugural poem were also analyzed. Inconsistencies in spelling, punctuation, spacing, and line and stanza breaks, along with additions, omissions, and sequence changes of words and phrases, were found. The poem performed possesses a richness unpredictable from either the extant literature on poetry readings or from Angelous own printed pages.

Reading Aloud from Logographic and Alphabetic Texts: Comparisons Between Chinese and German

The following study investigated the effect of writing systems (logographic vs. alphabetic) on the temporal organization of reading aloud. More specifically, we wished to test Perfetti and Zhangs (1995) assertion that “the reader of an alphabetic system can do better at recovering the phonological form, less well at recovering the semantic category of the word” (p. 186f.). Native speakers of Chinese and German read a semantically identical passage from texts written in their own language. The Chinese version consisted of 132 characters (132 syllables), the German of 80 words (also 132 syllables). In accord with Perfetti and Zhangs position, Chinese readers articulated significantly more slowly; they also used significantly more pauses. Moreover, German readers used a set pattern of pause positions, whereas Chinese did not. Logographic and alphabetic determinants of reading aloud are discussed.