An interlocutor is the person one speaks with: the partner in any conversation or dialogue. The word joins Latin inter- ("between") and loquī ("to speak"), and in linguistics it stands alongside addressee, hearer, and conversation partner as the name for whoever receives and answers talk. Because meaning in conversation is built jointly, the identity and behaviour of the interlocutor shape the language each speaker produces.
Two settings give the term particular weight in language teaching: research on how interaction drives acquisition, and the speaking examination, where a designated interlocutor runs the test.
Interlocutor characteristics have been a central variable in research on how conversation supports second-language learning since the early 1980s. Whether the interlocutor is a native or non-native speaker, a peer or a more proficient partner, familiar or unfamiliar, and how proficient they are all affect the negotiation of meaning, the modified input a learner receives, and the output they are pushed to produce. Interactionist accounts treat the interlocutor not as a passive audience but as a co-participant whose clarification requests, recasts, and feedback create the conditions for learning. Speakers also align without intending to, converging in vocabulary, structure, and pronunciation, the accommodation that lets dialogue cohere.
For task design the interlocutor is a live variable: pairing a learner with different partners changes both task difficulty and the language a task elicits, independently of the task itself.
In a speaking test the interlocutor is the examiner who conducts the face-to-face interaction with the candidate. Because spoken language is co-constructed, the examiner's behaviour directly shapes the candidate's output, which makes the role a question of test validity and reliability. To hold that influence steady, examiners work from an interlocutor frame, a scripted set of questions, prompts, and timings that gives every candidate essentially the same test event.
Systems handle the role in two ways. Some, such as Cambridge C1 Advanced, split it: an interlocutor manages the interaction while a separate assessor observes silently and scores, which eases the load on each examiner and can sharpen scoring reliability. Others, including IELTS, combine interlocutor and assessor in one person and rely on standardisation training for consistency. On the IELTS Speaking test, O'Sullivan and Lu (2006) found that minor examiner deviations from the frame had little effect on candidate language in Parts 1 and 2, with more variation in the Part 3 discussion.