Adjacency Pairs

Among the first things that conversation analysts noticed when they started looking carefully at conversations was ordered pairs of expressions, like questions and greetings (Schegloff and Sacks 1973). Once the first part of a pair occurs, the second part is expected.

Sacks (1992:3) noticed that workers at a psychiatric hospital’s emergency telephone line greeted callers by saying something like, ‘‘Hello. This is Mr. Smith. May I help you?’’ Most of the time, the response was ‘‘Hello, this is Mr. Brown,’’ but on one occasion, the caller responded, ‘‘I can’t hear you.’’ When the worker repeated his greeting, ‘‘This is Mr. Smith,” with an emphasis on Smith, the caller responded ‘‘Smith.’’ In this case, the rule for an adjacency pair was not really being broken. It was being negotiated by both parties, on the fly, during the conversation. Mr. Smith, the suicide prevention worker, was trying to get the caller to give his name, and the caller was trying not to give his name.

If conversations are dynamic things, with constant negotiation, then, to understand the rules of conversations, you can’t interview people about the rules of, say, greetings. You have to study real conversations in which greetings occur.

Here is an example of an adjacency pair—a question-answer pair—that’s broken up between its parts.

  • 1. A: Is it good? The Szechuan pork?
  • 2. B: Y’like spicy, uh—
  • 3. A: [Yeah
  • 4. B: It’s kinda, /know, hot.
  • 5. A: Great
  • 6. B: Yeah, me too.

There is a lot going on here. A asks B if a particular dish is good. This is the first part of a question pair, which calls for an answer. B responds with another question, which creates a new expectation for an answer. A anticipates this and answers ‘‘yeah’’ (line 3) before B finishes her thought (which presumably was “y’like spicy, uh, food?’’).

B doesn’t want to give up her turn just yet, so she completes the thought, in line 4. This gives A the opportunity to reaffirm that she does, in fact, like spicy food (that her ‘‘yeah’’ in line 3 was not just an acknowledgment of B’s question) before she answers the first one. A reaffirms her ‘‘yeah’’ response with ‘‘great’’ in line 5. The segment ends with B agreeing that she, too, likes spicy food.

Looking at this segment, it seems like the first adjacency pair, the question about whether the Szechuan pork is good, has never been answered. But it has. The answer is implied in the inserted question in line 2 and the retesting in line 4, all of which is a shorthand for: ‘‘Yes, if you like spicy food, then the Szechuan pork is good; otherwise it isn’t’’ (Further Reading: conversation analysis).

If this seems complicated, it’s because it is. It’s very, very complicated. Yet, every native speaker of English understands every piece of this analysis because: (1) The segment of conversation is rule based; (2) We understand the rules; and (c) We understand that the outcome is the result of cooperation by both speakers, A and B, as they attend to each other’s words, draw on cultural knowledge that they share, and build a conversation dynamically. In this case, two people, in a conversation among equals, have worked together to build a particular outcome.

And if you think this is complicated, just add more speakers. With two people in a conversation, A and B, there are two pairs of people—AB (as seen from A’s perspective) and BA (as seen from B’s perspective)—figuring out each other’s motives and behaviors and negotiating their way through the interaction. With three people in a conversation (A, B, and C), there are six pairs (AB, BA, AC, CA, BC, CB) doing the negotiating. With six people, there are 30 such pairs. Deborah Tannen (2005) analyzed 2 hours and 40 minutes of conversation among six friends at a Thanksgiving dinner. The analysis alone is a 170- page book, and that doesn’t include the hundreds of pages of transcripts (box 18.4).

BOX 18.4


Some conversations require special rules so that they don't break up, like dinner parties do, into lots of smaller conversations. Pilots and air traffic controllers around the world wait for each other to finish a thought before taking the next turn. Juries have special rules for turn taking, with a leader (or foreperson) who ensures that the cultural rules for these formal conversations (written down nowhere, but widely understood) are followed. Among other things, the rules require that no one gets completely drowned and sidelined, unless they deserve to be (by first breaking the cultural rules themselves and trying to dominate). (See Manzo [1996] for a detailed analysis of turn-taking by a jury.)

We turn next to methods that involve coding a corpus of text for themes and analyzing the themes for patterns.


Native ethnography. Firbank (2008); Fournillier (2007); Narayan (1993); Rudolph (1997); Sharma (2006).

Interpretive analysis: Garot (2004); Mann (2007); Rabelo and Souza (2003); Segre (1988); Yakali- Camoglu (2007)

Narrative analysis: Bridger and Maines (1998); Griffin (1993); Ozyildirim (2009); Riessman (1993). Performance and ethnopoetics: Blommaert (2006); Fabian (1990); Pratt and Wieder (1993); Webster (2008).

Transcription: Atkinson and Heritage (1984); Psathas (1979); Psathas and Anderson (1990). Conversation analysis: Drew and Heritage (2006); Markee (2000); Psathas (1995); Sacks et al. (1974); ten Have (1999).

< Prev   CONTENTS   Source   Next >