A teacher can explain the dictionary pronunciation of every word in a classroom instruction and still misdescribe what they actually said. Fluent speech contains choices of prominence, reduction, linking and pausing that spelling alone cannot supply.
For the refreshed Trinity DipTESOL Unit 3 introduced from April 2026, analyse your own teacher talk as speech evidence. Establish the actual utterance and context, use a clear transcription system, justify the features you identify and explain their classroom significance. Do not replace a recording with an idealised reading of the written sentence.
This guide uses Trinity's April 2026 specification, online edition 1.5 dated 23 July 2026, checked on 24 October 2026. The utterances and teaching cases below are original illustrations, not recordings, candidate submissions or official assessment samples. Confirm with your provider whether the refreshed or earlier assessment structure applies to your programme.
Find your next learning step
Prepare for the updated DipTESOL phonology task by separating recorded speech, transcription and classroom interpretation, with original diagnostic cases.
ROUTE 1 · CHAPTERS 1–3
Locate the speech evidence
Identify the refreshed task, transcription layer and classroom context.
ROUTE 2 · CHAPTERS 4–6
Analyse the listening demand
Investigate reduction, prominence and word boundaries.
ROUTE 3 · CHAPTERS 7–9
Work the teaching cases
Choose evidence-based focus points and examine two diagnostic cases.
ROUTE 4 · CHAPTERS 10–12
Explain and prepare
Connect speech models, follow-up questions and authentic preparation.
ROUTE 5 · CHAPTERS 13–15
Test the interpretation
Try a fresh context and consult the applicable assessment sources.
Full chapter index · Independent learning check · Education parent guide
Full chapter index
Locate the speech evidence · Chapters 1–3
Analyse the listening demand · Chapters 4–6
Work the teaching cases · Chapters 7–9
Explain and prepare · Chapters 10–12
Test the interpretation · Chapters 13–15
CHAPTER 1 OF 15 · Locate the speech evidence
1. Identify the updated Unit 3 before preparing
Trinity's current DipTESOL page explains the refreshed assessment and its transition arrangements. The April 2026 specification identifies Unit 3 as Phonology: Classroom applications and includes a short transcription of the candidate's own teacher talk.
This matters because older descriptions can concern a different transcription task or assessment arrangement. A preparation plan should name the version that applies to the candidate, rather than combine attractive details from several syllabuses.
The refreshed unit uses a three-task, 30-minute online interview: presentation, transcription with analysis and discussion. The transcription criteria include segmental and suprasegmental accuracy, contextual information, the speech model and type, the transcription system and four focus points.
Those official features define the task's frame. They do not turn this article's examples into a prescribed set of four features or an approved transcript. Follow the provider's current instructions for the actual selection, format and submission.
Begin by writing a precise preparation question: “What did I say in this classroom moment, how am I representing it, and why does the analysis matter for these learners?” That question keeps the qualification connected to teaching.
The existing CertTESOL learner-profile guide concerns a different qualification and Unit 3 task. DipTESOL's own-teacher-talk analysis should not be confused with diagnosing an individual learner for a CertTESOL profile.
If your question concerns professional status in England, the assessment-only QTS guide to experience and readiness addresses a separate decision. Keep its experience and standards requirements distinct from the DipTESOL phonology analysis developed here.
CHAPTER 2 OF 15 · Locate the speech evidence
2. Separate spelling, speech and transcription
Spelling represents words through an orthographic system. Speech is the sound event. A transcription is an analytical representation of selected properties of that event. These three layers can correspond without being identical.
Consider the original illustrative sentence “Can you move the blue card?” The written word can does not tell us whether the speaker used a strong vowel, a reduced vowel, particular prominence or a pause after the word. Those decisions require speech evidence.
A possible reduced form of can in an appropriate context is /kən/, while a strong form can be represented as /kæn/ in a suitable broad transcription model. These are illustrative possibilities. No recording has been supplied here, so neither is claimed as an observed token.
A teacher who begins with the spelling and inserts dictionary forms may produce a plausible pronunciation exercise. They have not necessarily transcribed their actual teacher talk. The direction of reasoning has changed from evidence to expectation.
For the real task, listen to the chosen speech before deciding which features belong in the account. A written transcript can help locate the words, but it should not settle phonological decisions that the audio contradicts.
Ask what the transcription system is intended to capture. A broad phonemic representation and a more detailed phonetic account make different analytical commitments. Use the system agreed for the task and explain it consistently, rather than mixing symbols and levels of detail without a purpose.
The Sengkang guide to observed, inferred and unknown offers a useful general discipline: distinguish the sound you can identify, the interpretation you propose and the details the evidence does not settle.
Distinguish the sound segment from the pattern around it
Segmental analysis concerns individual speech sounds, such as a vowel or consonant. Suprasegmental analysis concerns properties such as stress, prominence and intonation that organise material beyond an isolated segment. The categories help identify different questions about the same stretch of speech.
A candidate might identify the vowel quality of a particular token of can. That does not automatically describe its prominence in the phrase. Conversely, noting that a word stands out does not establish every detail of its vowel or consonants. A coherent account coordinates the observations without treating one label as a substitute for the other.
Use a short original comparison in practice. Ask what evidence would be needed to decide whether a vowel was reduced, then ask what evidence would support a claim that the word carried corrective focus. Some acoustic and contextual cues can interact, but the analytical claims are still distinguishable.
This distinction can improve the classroom implication. A learner might recognise the word's segments yet miss the contrast signalled through prominence. Another might recognise the intended contrast while failing to identify a reduced grammatical word. Those are different possible learning needs.
The candidate should not decide which need exists merely from the feature they find most interesting. Start with the learner's task and the response evidence, then choose an investigation that could separate the explanations.
CHAPTER 3 OF 15 · Locate the speech evidence
3. Establish classroom context before naming features
An utterance becomes more interpretable when the classroom purpose is known. Was the teacher giving a first instruction, repeating it after confusion, correcting a misunderstanding or modelling language for practice? The same words can serve different jobs.
In an invented first context, a class has several coloured cards and needs to know which one to move. In a second, everyone has already selected the blue card but is moving it instead of placing it beside another object. The potential information focus differs.
Do not infer an exact intonation contour from those written contexts alone. They give reasons to investigate prominence, not permission to invent what the teacher sounded like. The recording must support the account of the actual token.
A useful contextual note identifies the learners' task, the point in the lesson and what was already established. It need not become a long biography of the class. Include details that change the interpretation of the utterance or its teaching relevance.
For example, knowing that the teacher is contrasting blue with green may make prominence on blue relevant. Knowing the brand of classroom furniture probably does not. A precise selection of context keeps the analysis focused.
Context also helps distinguish deliberate modelling from ordinary instruction. A carefully articulated pronunciation model may legitimately differ from the teacher's less monitored language during classroom management. Describe the speech type rather than treating every difference as inconsistency.
The educational aim is to explain a relationship between sound, communicative purpose and learner access. A list of phonological labels without that relationship can sound knowledgeable while revealing little about teaching.
CHAPTER 4 OF 15 · Analyse the listening demand
4. Hear reduction without treating it as careless speech
Weak forms and other connected-speech features can make a familiar written word less transparent to a listener. A learner who knows the word in isolation may fail to recognise it when it is less prominent inside an utterance.
The British Council's connected-speech explanation provides primary teaching background for changes within and across word boundaries. Its discussion of weak forms helps connect such forms with listening.
Return to our invented instruction. A student might recognise can when it is said alone but miss a reduced token in the full question. That is one possible diagnosis, not a conclusion about an actual learner. A targeted listening comparison would be needed.
The teacher's next action could be a short contrast between a full form and a contextually appropriate reduced form, followed by recognition inside a new utterance. The aim is to connect known vocabulary with a speech realisation, rather than demand imitation of one accent as the only legitimate English.
Do not label every unstressed syllable as absent. A vowel may be reduced while the word remains present. Likewise, failing to hear a segment clearly on one replay does not prove that the speaker deleted it.
For transcription, slow the analysis while keeping the event intact. Listen to the surrounding phrase and compare repeated hearings. If a detail remains uncertain, seek appropriate feedback and represent the uncertainty honestly instead of forcing an expected form into the transcript.
A good teaching implication is specific: learners may need help recognising this kind of reduced form in relevant listening. It is weaker to claim that every pronunciation difficulty in the class is caused by reduction.
CHAPTER 5 OF 15 · Analyse the listening demand
5. Examine prominence as a communicative choice
Prominence concerns what stands out in an utterance. It interacts with information, contrast and the speaker's communicative intention. It should not be reduced to an instruction to stress every word that looks important on paper.
Use two original written scenarios. In the first, the teacher contrasts the blue card with the green one. In the second, the colour is already settled and the teacher contrasts putting the card beside a book with putting it under the book.
These contexts suggest different analytical questions. Does the speech make the relevant contrast salient? Which word is prominent in the actual recording? How might the prominence support or complicate the instruction?
The written scenarios do not determine the audio. A teacher might emphasise unexpectedly, use several prominent words or rely on gesture. The candidate should analyse the recorded event rather than improve it silently in the transcription.
Now consider an invented listening activity. Learners hear two teacher-created versions that differ in focus and choose which contrast each version answers. This can test whether they notice the communicative effect instead of merely copying a stress mark.
For classroom evaluation, observe what learners do with the distinction. Accurate repetition may show that they can imitate a model. Selecting the intended object after a fresh instruction gives different evidence about comprehension.
Avoid claiming that one prominence pattern always has one meaning in every context. Interpretation depends on discourse and can interact with other cues. A careful analysis explains the likely function in the actual setting and identifies what would support a stronger claim.
That is the professional move the unit can develop: connect a phonological observation with a bounded account of communication and a teaching decision that another person can question.
CHAPTER 6 OF 15 · Analyse the listening demand
6. Distinguish a boundary problem from a vocabulary problem
A learner can know the component words and still struggle to locate them in a continuous stream. Word boundaries are visible in writing but are not delivered as evenly spaced blocks in ordinary speech.
Suppose a fictional adult learner correctly identifies move, the and blue separately, yet cannot segment a short instruction containing them. The teacher should investigate the connected sequence rather than immediately reteach the vocabulary meanings.
A short task might ask the learner to match a heard phrase with two possible written sequences. Another might ask them to indicate where a known word occurs. The evidence should reveal whether the difficulty concerns recognition in context rather than only isolated word knowledge.
The teacher's own talk can supply a useful diagnostic starting point. How are adjacent words joined? Where are pauses actually placed? Which syllables carry prominence? Are repeated versions noticeably different in ways relevant to the learners' task?
Do not assume that every difference from citation pronunciation is a problem. Connected speech is ordinary language behaviour. The teaching question is how to help learners recognise relevant patterns and communicate effectively in their intended contexts.
Also avoid attributing a difficulty solely to a learner's first language without evidence. Individual experience, familiarity with the task, vocabulary, attention and the particular speech sample can matter. An observed error deserves investigation rather than a national-language stereotype.
The Bukit Timah guide to diagnosing the first unstable link offers a useful general teaching principle. Here the candidate should distinguish lexical knowledge, speech recognition and task understanding before choosing the repair.
CHAPTER 7 OF 15 · Work the teaching cases
7. Choose focus points that the actual speech supports
The refreshed specification requires four focus points as part of the task-specific information. It does not make every illustrative category in this article compulsory. Choose and justify points in accordance with the actual recording and the provider's instructions.
An effective focus point has a visible route back to evidence. Identify the relevant stretch, state what is represented and explain why it matters. A broad heading such as “intonation” is less informative than a clearly bounded question about a particular utterance.
Suppose an original practice extract contains an audible pause before a repeated instruction. The candidate might investigate its relationship to reformulation or learner attention. That would require the actual context and speech, rather than a claim that every pause improves comprehension.
Another possible practice extract contains a contrast between a strong and reduced form. The useful analysis identifies the tokens and conditions. It does not add an unrelated list of weak forms simply because the topic is familiar.
| Layer of an original practice analysis | Question to ask | Common overreach |
|---|---|---|
| Speech evidence | What can be identified in this stretch? | Transcribing an expected dictionary form |
| Representation | What does the chosen system show? | Mixing incompatible levels of detail |
| Interpretation | What might the feature do here? | Assigning a universal meaning |
| Teaching implication | What learner task would test its relevance? | Claiming learning from a polished explanation alone |
The table is an original analytical aid. It is not Trinity's scoring grid or a substitute for the official task instructions.
A focus point should earn its place by revealing something about teacher language and classroom work. Four decorative labels can satisfy the appearance of coverage while leaving the candidate unable to answer a follow-up question about the actual speech.
CHAPTER 8 OF 15 · Work the teaching cases
8. Work a case in which the transcript follows spelling
Imagine fictional teacher Nadia selecting a short instruction from her own class. In a practice analysis, she writes strong dictionary forms for every word because those are the forms she knows how to transcribe confidently.
On replay, her tutor notices that Nadia's notation does not account for a less prominent form in the actual extract. Nadia initially says that the transcript is “clearer” than the recording. That response reveals the task confusion: she has improved a model rather than represented the evidence.
The repair starts by separating two legitimate products. A teaching model can be deliberately articulated for a purpose. A transcription of a chosen speech event must describe that event under its stated system. Neither should silently pretend to be the other.
Nadia returns to one uncertain word in its surrounding phrase. She listens, identifies the relevant difference and discusses the appropriate representation. Then she checks another stretch without the tutor naming the feature first.
This small independent check matters. If Nadia changes only the token the tutor identified, she may have corrected an answer without changing the method. A second stretch tests whether she now listens before selecting a familiar symbol.
The teaching implication is also bounded. Discovering a reduced form in her own instruction suggests a useful listening focus to investigate. It does not prove that the class misunderstood the instruction because of that feature.
Nadia would need evidence from learner responses or a suitable follow-up task to support that stronger claim. The analysis becomes more professional when it keeps the observation, possible effect and next inquiry distinguishable.
CHAPTER 9 OF 15 · Work the teaching cases
9. Work a case in which successful repetition is overinterpreted
Fictional teacher Leo introduces a short connected-speech listening activity. Learners repeat a model accurately after several demonstrations. He writes in his practice evaluation that they can now recognise the form in ordinary listening.
That conclusion is broader than the evidence. The observed task involved immediate repetition of a known model. It did not yet test recognition of a fresh utterance or understanding in a different speaking context.
A better next task changes the surface while preserving the relevant listening demand. Use a new short instruction with familiar vocabulary and ask learners to identify the intended action. Keep the language manageable so that the result can be interpreted.
If learners succeed, Leo has additional evidence of recognition under that new condition. If they struggle, he should inspect whether the difficulty concerns the reduced form, vocabulary, prominence, task instructions or another feature. One failed transfer does not settle the cause automatically.
The CertTESOL Unknown Language journal guide examines the difference between feeling supported, following a task and using language independently. It is a different assessed unit, but that distinction helps prevent a pronunciation activity from being judged only by enthusiasm or fluent imitation.
For DipTESOL, Leo should explain his phonological rationale and the evidence behind his evaluation. A pleasing classroom moment can be worth retaining without being promoted into proof of lasting learning.
This is a constructive form of restraint. It does not require a teacher to distrust every success. It asks them to name what the task actually checked and choose a reasonable next task for the capability that remains untested.
CHAPTER 10 OF 15 · Explain and prepare
10. Keep the teacher's speech and the learner's target distinguishable
Analysing your own teacher talk does not mean presenting your personal pronunciation as the only acceptable destination for every learner. The speech sample is evidence about a particular classroom event and model.
A candidate should identify the relevant speech model and type clearly. Their own accent, the learners' intended contexts and the materials used in the course can interact. The analysis should not hide that variety under a vague claim about “correct English”.
In an original class, learners need to understand several speakers during workplace conversations. A teacher's model can provide a manageable starting point, but successful recognition of that one voice does not establish readiness across all speakers.
A changed-context task might introduce another suitable speaker or a different familiar phrase. The teaching question is which recognition skills transfer and which require more exposure. It is not a demand to erase the learner's identity or force every voice into the same pattern.
The candidate can also distinguish receptive and productive goals. Recognising a reduced form in listening and choosing to use a comparable form in speaking are related but different capabilities. A lesson should not claim one merely because it practised the other.
This distinction keeps the classroom implication realistic. A feature observed in teacher talk may be useful for receptive awareness even if the immediate speaking objective is a different matter. State the purpose before judging the activity.
The Education parent guide places knowledge beside adaptation. Here adaptation means selecting a phonological focus that serves the actual learners and communication, while remaining accurate about the evidence and model.
CHAPTER 11 OF 15 · Explain and prepare
11. Prepare to explain the analysis under questioning
A written transcript can conceal a weak decision until someone asks why a particular symbol or interpretation was chosen. Preparation should therefore include explanation, alternatives and evidence, rather than only visual polishing.
Choose one uncertain feature from the actual practice work. State what you heard, how the system represents it and what alternative you considered. Then identify what in the speech or context supports your choice.
A tutor can change one condition in a practice discussion. What if the same words were used for correction rather than a first instruction? What if the phrase were a deliberate pronunciation model? Which parts of the analysis would stay and which would need reconsideration?
The candidate should not revise an observed transcription merely because a hypothetical context changes. The hypothetical exercise tests interpretation. The original recording remains the source for what was actually said.
This is a useful distinction during spontaneous questioning. A candidate can say that a feature might serve a different function in another context while preserving the evidence-based account of their own extract.
Use technical terminology to make the explanation more precise. A term should identify a distinction that matters. If replacing it with a simple description reveals that no actual feature has been identified, return to the evidence.
The Reading Evidence and Answer Scope guide provides a parallel discipline for answering questions: select the relevant evidence and stop the claim where that evidence stops.
Do not rehearse a fixed speech that can only proceed when every question arrives in the expected order. A flexible explanation should survive a request to justify one small decision in the middle.
Audit a disagreement without silently changing the model
Imagine Nadia and a peer disagree about one part of a practice transcript. Before voting for the more familiar symbol, check whether they are representing the same extract, using the same transcription system and analysing the same speech model.
One person may be describing a broad contrast while another tries to capture a fine phonetic detail. The difference in notation can then reflect different representational purposes. Alternatively, they may genuinely disagree about what is audible. Those possibilities require different responses.
If the system is inconsistent, agree the appropriate framework with the tutor and revise the account consistently. If the listening remains uncertain, return to the relevant stretch and seek informed feedback. Do not add precision that the evidence cannot sustain.
A candidate can learn from such a disagreement without presenting uncertainty as incompetence. The professional capability is to identify what the disagreement is about and resolve it through the task's evidence and conventions. Confidently defending an unrelated dictionary form does not do that work.
After revision, ask Nadia to explain a second feature independently. That checks whether she understands the analytical distinction or has only accepted a peer's correction. The same principle applies to classroom learning: a repaired answer becomes more useful when the learner can apply the reason elsewhere.
CHAPTER 12 OF 15 · Explain and prepare
12. Build an authentic preparation routine
Begin with the correct assessment version and provider instructions. Establish what material is required, how it is to be represented and how the candidate's own teacher talk is selected. Do not assume that an older sample task governs the refreshed route.
For the analytical work, separate listening, transcription and interpretation into manageable passes. The separation is a teaching routine, not a claim that Trinity mandates a particular number of passes or a particular software tool.
On one pass, establish the words and relevant context. On another, examine selected phonological details under the chosen system. Then consider the classroom implication. Revisit the audio when an interpretation tempts you to alter the evidence.
Keep the recording or authorised reference stable while revising the analysis. If the underlying extract changes, check the transcript and commentary again. A well-written discussion of an earlier extract can become inaccurate when attached to a different version.
Use the provider's arrangements for recording, access and handling classroom material. Do not expose learners' identifiable speech or information merely to make a practice discussion more vivid. An educational guide does not grant permission to record or share a class.
A peer can help by asking a question about an actual decision, rather than rewriting the candidate's account. The candidate must be able to explain their own speech analysis and professional judgement.
The learner benefit of this routine is concrete: teacher language becomes something that can be examined, adapted and evaluated. It moves from an invisible habit to a set of choices whose effects can be investigated with care.
CHAPTER 13 OF 15 · Test the interpretation
13. Try an independent teacher-talk analysis
Here is an original written practice scenario. A teacher first says, “Put the blue card beside the book.” After a learner selects the green card, the teacher repeats the same words. No recording is supplied.
What can be concluded? The second utterance occurs in a corrective context. The colour contrast is potentially relevant. What cannot be concluded from the written scenario alone? The exact vowel qualities, prominence pattern, pauses and intonation of either spoken token.
A strong first answer therefore identifies a listening question: did the teacher's second version make the relevant colour distinction salient, and how? It does not invent a completed transcription from the spelling.
Now suppose an actual authorised practice recording is available and a clear prominence difference can be identified. The candidate can describe that observed difference under the chosen system and discuss its plausible role in the correction.
Learner understanding remains another question. If the learner then moves the blue card, the action is relevant evidence in that moment. It does not establish that prominence alone caused the correction; gesture, repetition and other cues may also have contributed.
Change the scenario. The learner already has the blue card but places it under the book. The likely corrective issue now concerns the spatial relationship. A candidate should revise the interpretive question rather than keep the earlier colour explanation unchanged.
For a final variation, imagine the utterance is deliberately slow language modelling. The speech type has changed. Compare it with ordinary teacher instruction only after identifying that difference. Accurate analysis begins by asking what kind of event the sample represents.
The independent skill is disciplined coordination: identify the context, inspect the speech, choose a representation, bound the interpretation and propose a classroom check. Each decision should remain traceable to its own evidence.
CHAPTER 14 OF 15 · Test the interpretation
14. Questions teachers preparing for Unit 3 ask
Does this guide apply to every DipTESOL candidate in 2026?
No. Trinity identifies transition arrangements. Confirm with your provider whether your programme uses the earlier or refreshed assessment structure.
Can I transcribe the written sentence from dictionary forms?
That can produce a pronunciation model, but it does not necessarily represent the actual teacher-talk token. Analyse the speech evidence required by the task.
Must all four focus points be the categories used here?
No. These are original teaching examples. Follow the current task and provider guidance, choosing points that the actual extract supports.
Is a reduced form automatically a pronunciation error?
No. Reduction can be an ordinary feature of connected speech. Its relevance depends on the model, context and communicative purpose.
Does accurate repetition prove listening transfer?
It shows performance on that repetition task. Test recognition or use in a fresh condition before making a broader claim.
Can I infer intonation from punctuation?
Punctuation does not provide a recording's exact intonation. Use the speech event for the actual analysis and treat written scenarios as prompts for questions.
Should my learners reproduce my accent exactly?
The teacher's speech model and the learners' communicative needs should be made explicit. Analysing a model does not establish one voice as the only acceptable English.
What if I cannot hear a feature confidently?
Recheck the relevant stretch in context, use appropriate feedback and avoid inserting an expected feature merely to make the transcript appear complete.
Is this a model submission?
No. The utterances, cases and routines are invented educational material. The assessed analysis must accurately concern the candidate's own required evidence.
What makes a useful practice question?
Ask why a particular representation or interpretation is justified, then change one condition. The explanation should remain connected to the actual evidence rather than depend on a memorised order of questions.
CHAPTER 15 OF 15 · Test the interpretation
15. Sources and the next classroom investigation
The qualification sources checked on 24 October 2026 are Trinity's DipTESOL page and transition information and the April 2026 Qualification Specification and Assessment Information, online edition 1.5 dated 23 July 2026. Use the applicable Unit 3 requirements with the provider's instructions.
The British Council's Connected speech teaching-knowledge entry and Connected speech: part 1 provide background for the relationship between speech features and listening. They are teaching resources, not Trinity's assessment rubric.
All written utterances, fictional teachers, tables and diagnostic activities here are original. No audio was analysed for these cases, no classroom outcome was measured and no exact pronunciation is attributed to a real speaker. The small IPA examples illustrate possibilities under a stated broad representation.
For the next preparation session, choose one feature in the required teacher-talk evidence that you can justify carefully. Explain what was observed, how it is represented and what classroom question follows. Then identify the new learner task that would test the proposed implication.
That sequence keeps professional curiosity attached to evidence. The result is more than a technically decorated transcript: it is a teacher who can examine their own language, explain a decision and investigate whether an adaptation helps the learners in front of them.