Turn-taking


Turn-taking is a type of organization in conversation and discourse where participants speak one at a time in alternating turns. In practice, it involves processes for constructing contributions, responding to previous comments, and transitioning to a different speaker, using a variety of linguistic and non-linguistic cues.
While the structure is generally universal, that is, overlapping talk is generally avoided and silence between turns is minimized, turn-taking conventions vary by culture and community. Conventions vary in many ways, such as how turns are distributed, how transitions are signaled, or how long the average gap is between turns.
In many contexts, conversation turns are a valuable means to participate in social life and have been subject to competition. It is often thought that turn-taking strategies differ by gender; consequently, turn-taking has been a topic of intense examination in gender studies. While early studies supported gendered stereotypes, such as men interrupting more than women and women talking more than men, recent research has found mixed evidence of gender-specific conversational strategies, and few overarching patterns have emerged.

Organization

In conversation analysis, turn-taking organization describes the sets of practices speakers use to construct and allocate turns. The organization of turn-taking was first explored as a part of conversation analysis by Harvey Sacks with Emanuel Schegloff and Gail Jefferson in the late 1960s/early 1970s, and their model is still generally accepted in the field.
Turn-taking structure within a conversation has three components:
  • The turn-taking component contains the main content of the utterance and is built from various unit types. The end of a TCU is a point where the turn may end and a new speaker may begin, known as a transition relevance place or TRP.
  • The turn allocation component comprises techniques that select the next speaker. There are two types of techniques: those where the current speaker selects the next speaker, and those where the next speaker selects themself.
  • Rules govern turn construction and give options to designate the next turn-taker in such a way as to minimize gaps and overlap. Once a TRP is reached, the following rules are applied in order:
  1. The current speaker selects the next speaker and transfers the turn to them; or
  2. One of the non-speakers self-selects, with the first person to speak claiming the next turn; or
  3. No one self-selects, and the current speaker continues until the next TRP or the conversation ends
This order of steps serves to maintain two important elements of conversation: one person speaking at a time and minimized space between when one person stops talking and another begins. Because the system is not optimized for fairness or efficiency, and because turn-taking is not reliant on a set number or type of participants, there are many variations in how turn-taking occurs.

Timing

Another cue associated with turn-taking is that of timing. Within turn-taking, timing may cue the hearer to know that they have a turn to speak or make an utterance. Due to the very nature of turn-taking and that it is dependent on the context, timing varies within a turn and may be subjective within the conversation. Vocal patterns, such as pitch, specific to the individual also cue the hearer to know how the timing will play out in turn-taking.
Deborah Tannen also shows timing differences in relation to turn-taking. For a particular study, she used a recording of a conversation between a group of her friends at dinner. The group included men and women from across the United States of mixed ethnicities. She concluded that while the amount of space left between speakers may differ, it differs most dramatically between people from different regions. For instance, New Yorkers tend to overlap in conversation, while Californians tend to leave more space between turns and sentences.
Kobin H. Kendrick argues that rules and constraints that are established within a turn-taking system are done so to minimize the amount of time spent transitioning between turns. Not all transitions are minimal; Schlegloff found that transitions before turns that incorporate other-initiations of repair were found to be longer than other transitions.

Overlap

When more than one person is engaging in a conversation, there is potential for overlapping or interruption while both or many parties are speaking at the same time. Overlapping in turn-taking can be problematic for the people involved.
There are four types of overlap including terminal overlaps, continuers, conditional access to the turn, and chordal. Terminal overlaps occur when a speaker assumes the other speaker has or is about to finish their turn and begins to speak, thus creating overlap. Continuers are a way of the hearer acknowledging or understanding what the speaker is saying. As noted by Schegloff, such examples of the continuer's phrases are "mm hm" or "uh huh." Conditional access to the turn implies that the current speaker yields their turn or invites another speaker to interject in the conversation, usually as collaborative effort. Another example that Schegloff illustrates is a speaker invited another to speak out of turn when finding a word in a word search. Chordal consists of a non-serial occurrence of turns; meaning both speakers' turns are occurring at once, such as laughter. The above types of overlap are considered to be non-competitive overlap in conversation.
Schegloff suggested an overlap resolution device, which consists of three parts:
  • A set of resources that are used to compete for the turn space
  • A set of places where the resources are used
  • An interactional logic of the use of those resources at those places
Gail Jefferson proposed a categorization of overlaps in conversation with three types of overlap onsets: transitional overlap, recognitional overlap and progressional overlap.
  • Transitional overlap occurs when a speaker enters the conversation at the possible point of completion. This occurs frequently when speakers participate in the conversation enthusiastically and exchange speeches with continuity.
  • Recognitional overlap occurs when a speaker anticipates the possible remainder of an unfinished sentence, and attempts to finish it for the current speaker. In other words, the overlap arises because the current speaker tries to finish the sentence, when simultaneously the other speaker "thinks aloud" to reflect their understanding of the ongoing speech.
  • Progressional overlap occurs as a result of the speech dysfluency of the previous speaker when another speaker self-selects to continue with the ongoing utterance. An example would be when a speaker is retrieving an appropriate word to utter when other speakers make use of this gap to start their turn.
Sacks, one of the first to study conversation, found a correlation between keeping only one person speaking at a time and controlling the amount of silences between speakers. Although there is no limit or specific requirement for the number of speakers in a given conversation, the number of conversations will rise as the number of participants rise.
Overlaps can often be seen as problematic in terms of turn-taking, with the majority of research being between cooperative versus competitive overlap. One theory by Goldberg argues the dynamic relationship between overlap and power over the conversation by suggesting that two types of overlap are power interruptions and displays of rapport. During conversation, a listener has an obligation to support the speaker. An interruption impedes upon this obligation by infringing upon the wishes of the speaker. The difference between a power interruption or rapport is the degree to which the speakers' wishes are impeded upon. Rapport interruptions contribute to the conversation in that they ultimately cooperate and collaborate with the speaker in order to reach a mutual goal of understanding. Power interruptions are generally hostile and do not cooperate with the speaker. The goals of the power interruptor are both divergent from and regardless of the goals of the speaker. Power interruptions are further categorized into two types: process control interruptions and content control interruptions. Process control interruptions involve attempts to change the topic by utilizing questions and requests, and because they return control to the original speaker are generally seen as the less threatening of the two. Content control interruptions involve attempts to change the topic by utilizing assertions or statements that are unrelated to the current topic. Content control interruptions are viewed as problematic and threatening since they seize control of both the topic and attention away from the speaker.
However, while overlaps have the potential to be competitive, many overlaps are cooperative. Schegloff concludes that the majority of overlaps are non-problematic. Konakahara et al. explores cooperative overlap by observing 15 graduate students from 11 different lingua-cultural backgrounds in an ELF conversation, or an English-based conversation among individuals of multiple native languages. Two types of overlap were observed: overlaps that were continuers or assessments and did not substantially contribute to the conversation or demand attention away from the speaker, and overlaps that were questions or statements and moved the conversation forwards. The majority of overlap during the study consisted of continuers or assessments that were non-interruptive. Overlapping questions and their interactional environment were analyzed in particular. It was found that overlapping questions demonstrate the speaker's interest in the conversation and knowledge of the content, act as clarifiers, and progress the conversation. In response, speakers who are interrupted by overlapping questions continue on to clarify their meaning. This suggests that overlapping questions, while interruptive in the fact that they demand attention away from the speaker, are cooperative in nature in that they significantly contribute to achieving mutual understanding and communication.
While Goldberg's study primarily focuses on the distinctions and characteristics between power interrupters and displays of rapport, Konakahara et al. explores the ways in which overlap, in particular overlapping questions, can be collaborative and cooperative.