Social Saturdays’ posts look at the social side of language. After all, language isn’t just a way to communicate, it is communication.
Conversation works under a basic rule: One person speaks a time.
The two deviations from this rule (more than one and fewer than one person speaking at a time) occur very rarely.
On average, less than 5% of conversation occurs in overlap (more than one person at a time) and the gaps between the current and the next speaker average something like a tenth of a second.
This is true for two people in conversation, but also for twenty.
How do we manage to have such little gaps with so rare an overlap?
It could be gaze. We look at someone at the end of our “turn” to indicates they can begin.
However. These gaps actually become smaller when a conversation is not face-to-face (on the phone, for instance), and the overlaps even more rare.
Intonation seems much more likely, and a lot of the time when we come to the end of a “turn” we finish it with a question, whether it’s “what do you think?” or “would would like a glass of wine?”.
Sacks, Schegloff and Jefferson, in one of the most pretentiously worded journal articles I have ever read, propose a rule system for selecting next speaker.
Firstly, the current speaker can select the next speaker. This is, without exception, the first half of what we call an adjacency pair (a question requiring an answer, an accusation requiring admission or denial, and so one).
For instance “John, would you like some more wine?” would select John to be the next speaker, and also dictate what he talks about, as he is expected to answer the question.
If the current speaker does not select the next speaker, then the next speaker can self-select. This is sometimes easy, as there is one person who has something relevant to say. Sometimes more than one person has something to say, in which case it’s an all-out race. The first person to speak gets the floor.
If no-one says anything at this point, the silence goes on for longer than one tenth of a second. Something in our brain, at this point, detects that the silence is slightly too long, and someone speaks. This could be the current speaker continuing and extending their turn, or again the first person to speak could self select.
This means that there are rarely overlaps and silences, because these rules (named, inventively, rule 1 (a), (b) and (c)) continue round until someone speaks.
Interruptions occur when either someone mistakes the end of someones turn. The (possible) end of a turn is called a Transitional Relevance Place (TRP), and if someone mistakes a TRP, they might begin their own turn.
S1: Would you like some more wine, [Jane?]
S2: [yes please]
Alternatively, two speakers could begin their next turn simultaneously. When this happens, one speaker generally backs down before they get to the end of their turn. If one doesn’t back down voluntarily, then the louder, s l o w e r, lo-o-onger vo-o-owled speaker will “win”, so to speak.
The obvious third occurrence is an outright interruption. In the case that this is mistaken, there are “repairs” (“Gosh, sorry”, “you go ahead”, etc.) that can come into play. If it’s deliberate, then someone’s either trying to be rude or it’s during an argument, so they will probably already be loud and slow and using elongated vowels.
These rules, along with intonation, serve to cut down the number of interruptions and overlaps.
If you manage to win the floor, however, what you do with it is a different matter…