📄 Teaching English Pragmatics by Taking a Sociocultural Stance

Summary

The paper asks whether instruction grounded in Vygotsky's sociocultural theory helps Iranian EFL learners use the speech act of criticizing appropriately. It targets two gaps in instructed-pragmatics research. Most studies compare only implicit and explicit teaching, and few check whether gains last. Two intact classes took part, which the Persian abstract describes as 25-learner groups: one experimental and one control. The experimental class worked through interactive, cooperative tasks. Learners read a peer's work and voiced criticism of it orally, while the teacher gave individually tuned, mediated feedback (scaffolding within the learner's ZPD). Both groups took a pre-test, an immediate post-test and a delayed post-test, each consisting of a Discourse Completion Test and a Role-play test. Scores were compared with paired and independent t-tests. The experimental group improved significantly and outperformed the control group on both short-term and long-term measures. Follow-up interviews showed learners felt less stress and more motivation and fun, and found the instructions clear. The authors argue that sociocultural methods bring both learning and emotional benefits in EFL settings like Iran, where learners rarely hear authentic English.

Key points

• Target pragmatic feature: the speech act of criticizing. The analysis attends to external modifiers, internal modifiers and hedges.
• Design: experimental vs. control class, with a pre-test, an immediate post-test and a delayed post-test.
• Instruments: a written Discourse Completion Test (DCT) and an oral Role-play test (RP), followed by interviews.
• Statistics: paired and independent t-tests. The experimental group was significantly better and kept its gains at the delayed test.
• Theoretical basis: ZPD, scaffolding, collaborative dialogue and interaction. The paper contrasts these with Schmidt's noticing and Sharwood Smith's awareness-raising, which underlie the usual implicit/explicit dichotomy.

Interesting ideas

• I1. Pairing an immediate test with a delayed one separates gains that last from gains that fade, which a single post-test cannot do.
• I2. Measuring the same skill with both a written instrument (DCT) and an interactive one (RP) matters, because a method can look effective under one format and weaker under another.
• I3. Criticism is analyzed as a core act plus softening devices (external modifiers, internal modifiers, hedges) rather than as a single label.
• I4. In this paper, scaffolding means help that is fitted to the individual and moves from implicit to explicit only as far as needed.
• I5. Having learners critique a peer's work turns production and evaluation into a single interactive exercise.

For Mova

How these ideas could serve Mova, our small open language model — its files, stages and tests:

• I1 (learn): After each new self-induced repair rule or retraining round, rerun the affected benchmarks a few training cycles later to confirm the improvement has not quietly decayed.
• I2 (answer): Test speech-act and implicature reading on several task formats, such as isolated sentences vs. dialogue turns in tales, so a module is not credited for exploiting one format.
• I3 (frames): In crate prag, represent a criticism or other face-threatening act as a head act with attached mitigator slots, so the explanation can show how strongly and how politely something was criticized.
• I4 (learn): LLM teacher tiers could follow this pattern, giving the student parser a minimal hint first and the full correct analysis only if it still fails, which shows where its competence actually ends.
• I5 (learn): In self-play, one Mova instance could produce a parse or answer and another critique it, with disagreements becoming candidates for silver data or repair rules.