12. Studio Work

12.1 Practical exercises

Exercise 1 — The domain audit. Annotate one drafted exchange per §5.5, with the attributions stripped. Produce a table of turns against domains. Write 200 words on what the distribution reveals, paying particular attention to any empty column.

Exercise 2 — Mechanism substitution. Take the most on-the-nose turn in a draft and rewrite it four times, once each as an evasion, a changed subject, a too-precise reply and a delayed agreement (§6.4). Do not judge which is best until all four exist. Write 300 words on what each implies about the speaker.

Exercise 3 — The transcript test, run properly. Strip every tag and action beat from a two-page exchange. Give the bare lines to a reader who does not know the book, and ask them to mark where the speaker changes and to describe each speaker in two adjectives. Record the answers verbatim without correcting them.

Exercise 4 — The preference-organisation drill. Write an exchange in which one character refuses another, and produce the refusal in the full dispreferred format described at §3.2: delay, hesitation, token agreement, account, mitigation, and a refusal that is never stated outright. Then write the same refusal as a flat no. Compare what each does to the relational domain.

Exercise 5 — The maxim drill. Write one short exchange four times. In each version, one speaker flouts a different Gricean maxim at the same point in the sequence. Identify the implicature a reader computes in each case, and write it out in a phrase.

Exercise 6 — Report and quotation. Take a drafted exchange of two pages and rewrite it in Austen’s proportions (§3.5, §9.1): direct speech for the two turns that must be experienced, indirect report for everything else. Compare the length and the effect.

Exercise 7 — The silent third. Rewrite a two-hander with a third character present throughout who never speaks and is mentioned twice. Change nothing else. Write 200 words on what happens to the exchange.

Exercise 8 — The gap-closing move. Locate the moment in your draft where a character says outright what they mean for the first time. If there is none, write it. If there are three, decide which one the book has actually earned, and convert the other two.

Exercise 9 — Dialect without spelling. Take a passage of speech you have rendered phonetically and rewrite it so that syntax, idiom, lexicon and forms of address carry the same register, with no respelling. Then write 300 words on what has been lost and whether the loss is real (§8.9).

12.2 Seminar questions

1. Is subtext a technique or a period style? §6.6 argues that the two-channel ideal is a datable Anglo-American preference. Test the claim against Dostoevsky, Wilde and Compton-Burnett. Does the chapter’s model describe dialogue, or a fashion in dialogue?

2. Five domains or six? Argue for or against admitting the silence domain (§4.6) as a separate category, given the objection stated there — that silence is a means rather than an end, and that its value shifts are always relational or thematic on inspection. Use Wuthering Heights or Their Eyes Were Watching God as evidence.

3. Does the Grand Inquisitor break the model? §9.3 argues that a set speech is an exchange when the frame around it is contested. Is that a genuine defence of the model, or a redescription that makes it unfalsifiable? What would a passage look like that the model could not accommodate?

4. Phonetic dialect: whose rule, applied to whom? §8.9 argues that the prohibition has been applied unevenly by race and class. Read Hurston alongside a nineteenth-century English novel’s rendering of working-class speech. Is the difference one of technique, of position, or of the critic’s ear?

5. Jim’s speech in Huckleberry Finn. Critics have disagreed for a century about whether Twain’s rendering of Jim is a sympathetic representation, a minstrel inheritance, or both at once, and the disagreement is live rather than settled. Set out the strongest version of each position before deciding.

6. Is James’s dialogue a failure of test 7? In late James the principals share a register so completely that stripping attribution makes attribution difficult (§9.7). Is this a defect the chapter’s protocol correctly identifies, or a representation of a milieu that the protocol is too crude to recognise?

12.3 Assessment rubric

The following rubric is for a submitted dialogue exchange (1,000–1,500 words) with a 750-word critical commentary.