The Gold Standard of Anthropology | Sequential Position, Preference Organisation and Action Formation

1. Begin with the sequence, not the sentence

Back to contents

Start with enough prior and subsequent talk to see what problem the turn addresses. A first pair part such as a question, invitation, offer or complaint makes a range of responses conditionally relevant. The next turn displays how the recipient understood that action, although it can also resist or transform it. Avoid assigning an action from grammar alone. Show how position, wording, prosody, embodiment and the developing activity converge on the proposed analysis.

2. Distinguish action formation from action ascription

Back to contents

Action formation concerns how a speaker designs a turn so it can be recognised as doing something. Action ascription concerns how a recipient treats that turn in responding. The two may diverge. A speaker can design a remark as advice while the recipient answers it as criticism; later repair may reveal the mismatch. Analysts should not declare the speaker’s hidden intention. They should trace the public evidence through responses, reformulations and subsequent consequences.

3. Use adjacency pairs without forcing tidy pairs

Back to contents

Many actions are organised in paired relations: greeting–greeting, question–answer, invitation–acceptance or refusal. The first turn constrains what counts as a relevant next action, but insert expansions, delays, repairs and sequence closings can intervene. A response can fulfil several actions at once. Map the base sequence and its expansions, then explain how each addition manages information, entitlement, trouble or alignment. Do not cut the extract at the first apparent answer.

4. Read preference as structural organisation

Back to contents

In conversation analysis, preference does not mean what a person personally likes. It describes recurrent asymmetries in how alternative actions are designed. Acceptances may be prompt and compact, while refusals often arrive after delay, explanation or qualification. That pattern helps participants anticipate delicate outcomes. It is neither a universal politeness law nor proof of reluctance in every case; institutional rules, relationships and the specific action can reorganise what is treated as preferred.

5. Track projection, progressivity and response space

Back to contents

Turn beginnings, grammar, gaze and prosody can project what kind of action is coming and when completion is possible. Recipients use that projection to time responses. Delays, cut-offs and prefaces can suspend progress or reshape the response space. A question such as “Are you free Friday?” may project an invitation without delivering it yet. Analyse what participants can know at each moment rather than reading the completed transcript backward as if the outcome were obvious from the start.

6. Worked example: an invitation that becomes advice

Back to contents

In an invented student-project meeting, Lina asks, “Are you free after class?” Omar waits, then says, “I have practice.” Lina replies, “Okay—I was going to ask you to rehearse, but maybe send me your opening slide tonight.” The first question projects a possible invitation while withholding its full purpose. Omar treats it as a request requiring an account for unavailability. Lina accepts the refusal and transforms the activity into advice or a smaller request. The analysis follows that sequential change; it does not claim that “I have practice” directly answered a literal availability question only.

7. Build collections around a practice and its alternatives

Back to contents

Create a collection using explicit sequential criteria: the initiating action, position, design features, response and outcome. Retain cases where the projected action never arrives, the recipient responds differently or the sequence is abandoned. Compare alternative designs in comparable environments. Frequency can show distribution, but close analysis must establish what each case is doing. A collection built from a keyword alone will mix actions that merely share vocabulary.

8. Changed-context transfer check: a classroom correction

Back to contents

Move from peer planning to a teacher asking, “Is that the denominator?” after a student gives an answer. The turn has interrogative form, but its position and epistemic asymmetry can make correction or a prompt for self-repair relevant. The student may revise the answer rather than supply new information. The same sequential method transfers, while institutional rights and pedagogic goals alter the response space. The analyst should show how participants orient to those roles instead of assuming every teacher question is instructional.

9. Know the limits and common misconceptions

Back to contents

A next turn is strong evidence of understanding, not a perfect window into cognition. Recipients can strategically misread, joke or answer only part of an action. Preferred design does not equal moral approval, and dispreferred design does not prove dislike. Adjacency pairs are organising resources, not rigid scripts. Transcripts can hide gaze, timing and overlapping conduct. Claims about culture or institution require collections and contextual evidence beyond one elegant extract.

10. Audit recordings, transcripts and analytic alternatives

Back to contents

Retain the source recording, consent and access conditions, transcript conventions, timing, prosody, embodied conduct and enough surrounding activity to inspect the sequence. Record how candidate cases were found and why near-matches were excluded. Test rival action ascriptions against recipient uptake and later repair. Mark inaudible or ambiguous material rather than completing it silently, and preserve deviant cases that set limits on the claimed organisation.

11. Frequently asked questions

Back to contents

Is preference about personal choice? No; it concerns structural asymmetry in action design. Does a question always seek information? No; position and epistemic relations can make checking, challenge or correction relevant. Is the next turn definitive? It is central evidence, but later turns may revise the understanding. Can silence be a response? Timing can become meaningful, but the context must show how participants treat it. Are all sequences pairs? No; expansions and larger courses of action are common.

12. The gold standard makes action recognition inspectable

Back to contents

Strong sequence analysis connects turn design to the environment that gives it force. It distinguishes formation from ascription, treats preference as interactional organisation and follows how projection, delay, response and repair reshape an action. The result does not translate sentences into fixed labels. It shows, moment by moment, how participants build enough shared understanding to move an activity forward.