“The”, “Three” and “Tree”: English Pronunciation, /ð/, /θ/ and /t/ for Clear Speech

Teaching Guide: Hear, Place, Produce and Transfer /ð/, /θ/ and /t/

Pronunciation improves fastest when the learner can separate four jobs: hear the contrast, understand how the sound is made, produce it in a controlled word, and then keep the distinction while speaking naturally. Students often fail because they jump from explanation to full conversation in one step. A learner may know that the tongue belongs near the teeth and still substitute /t/ under pressure. Another may pronounce a single word correctly but fail to hear the difference when someone else speaks quickly.

The durable goal is intelligibility and control, not accent erasure. English has many legitimate accents. A pronunciation lesson should target a contrast when the learner’s current production repeatedly changes meaning, causes misunderstanding or prevents confident reading aloud. The examples below are original teaching material and can be adjusted for age, language background and school needs.

The Three Sounds Use Different Articulatory Systems

/t/ is a voiceless alveolar stop. The tongue contacts the ridge just behind the upper teeth, air pressure builds briefly, and the release produces the consonant. /θ/ is a voiceless dental fricative. Air continues through a narrow gap near the front teeth. /ð/ is the voiced partner of /θ/: the tongue position is similar, but the vocal folds vibrate. The learner should feel the difference between stopping air and allowing friction, and between voiced and unvoiced production.

Use the throat-touch test for voicing. Place fingers lightly on the throat and alternate /θ/ and /ð/. The second should produce noticeable vibration. Then alternate /t/ and /θ/. Both are voiceless, so the useful contrast is not vibration but airflow and tongue placement. This prevents the student from using one vague instruction for three different sounds.

Listening Comes Before Reliable Production

If a learner cannot consistently hear the contrast, production practice can become guesswork. Begin with discrimination: teacher says thin or tin; student points to the matching card. Use short sets and vary speaker, position and pace. Then ask the student to repeat only after identifying the word. Listening errors and production errors should be recorded separately because they need different repairs.

Do not use only written words during listening practice. Seeing the spelling can give away the answer and hide an auditory weakness. Alternate listening-only tasks with print-supported tasks. A learner who succeeds only when the word is visible still needs more auditory discrimination.

Minimal Pairs Reveal Whether Meaning Changes

Useful /θ/ versus /t/ pairs include thin–tin, thank–tank and, for some accents and learners, carefully selected contrasts where the rest of the word remains stable. The purpose is not to collect dozens of pairs. It is to show that one sound can change lexical identity. When the student hears and produces the distinction reliably, move into longer phrases.

For /ð/, exact minimal pairs with /t/ are less common in everyday vocabulary, so use contrastive phrases: then versus ten, those versus toes where the student’s accent makes the comparison useful, and sentence-level examples such as “Those three trees” or “Then Tina thanked them.” The aim is controlled contrast, not artificial wordplay for its own sake.

Spelling Does Not Map Perfectly to Sound

The letters th can represent /θ/ as in three, thin, thought, or /ð/ as in the, this, those, mother. Students should therefore avoid assuming that one spelling has one pronunciation. Build word families around sound as well as spelling. Function words such as the, this, that, these, those, then, they frequently use /ð/, while many content words such as think, three, theory, thumb use /θ/.

There are historical and lexical exceptions, so the practical teaching rule is frequency plus listening. Students should encounter the word in speech, record its sound and use it in phrases. A pronunciation dictionary or reliable audio source can settle uncertain items better than an invented universal spelling rule.

Worked Drill 1: From Sound to Word

Teacher says ten items in random order: thin, tin, three, tree, then, ten, thought, taught, those, toes. Student marks which target sound is heard. Replay only the missed items, then ask for the student’s own production. Record whether the error was perception or articulation. A learner who heard three correctly but produced tree needs production repair; a learner who heard both as the same word needs listening contrast first.

Repeat several days later without showing the original list. Delayed discrimination is stronger evidence than immediate repetition. Add a new speaker or recording when possible. Pronunciation learning becomes robust when the distinction survives voice, speed and context.

Worked Drill 2: Phrase-Level Control

Use short phrases that place the sound next to different vowels and consonants: three things, thin thread, thank them, those three, this tree, the third task. Say each slowly once, then at a natural pace. The student should keep the target consonants distinct without inserting an exaggerated pause between words.

If the learner loses the sound only at natural speed, the problem is transfer rather than knowledge. Reduce the phrase length, rebuild the sound, then increase speed gradually. Do not return immediately to isolated words unless the articulation itself has broken.

Connected Speech Changes Timing but Not the Need for Intelligibility

Natural speech is not a string of dictionary pronunciations separated by equal pauses. Sounds influence one another, function words can become less prominent, and rhythm changes with emphasis. Learners should therefore practise target sounds in meaningful sentences after word-level control is stable. “Those three students thought about the answer” creates multiple target opportunities inside normal syntax.

The goal is not to pronounce every /th/ with theatrical force. Excessive protrusion can sound unnatural and slow speech. Once placement is reliable, reduce the movement to the smallest gesture that still produces clear friction. Pronunciation should become efficient enough to survive ordinary conversation.

Reading Aloud Should Preserve Meaning

Pronunciation practice can damage reading fluency if the student focuses so intensely on one sound that sentence meaning disappears. Mark only a small number of target words in a passage. First read for meaning. Then practise the target words. Finally reread the sentence with appropriate phrasing and stress. The sound should support communication rather than become the entire performance.

Ask the listener whether any word changed identity or became difficult to understand. This keeps intelligibility as the measure. Perfect imitation of a teacher’s accent is unnecessary if the student’s speech remains clear.

A Common Singapore and Multilingual Learner Issue: Substitution

Different first-language sound systems make different substitutions more likely. Some learners may replace /θ/ with /t/, /s/ or /f/, and /ð/ with /d/, /z/ or /v/. These substitutions are not signs of laziness. They reflect the learner mapping English onto an existing sound inventory. The repair begins by making the new articulatory contrast available, then building enough repetition that the new category can be retrieved quickly.

Do not assume every speaker from one language background will make the same error. Diagnose the individual production. One learner may need tongue placement; another already produces the sound in isolation but loses it in clusters. A third may speak intelligibly and need no intensive correction at all.

Self-Monitoring Without Becoming Self-Conscious

Students can record a short sentence, listen once for meaning and once for the target sound, then attempt a second version. Limit the number of targets. If every syllable is analysed, fluency and confidence can collapse. One or two recurring contrasts per practice session are enough.

The learner should also know when to stop correcting. If the intended word is consistently understood across listeners and the target no longer creates confusion, further accent modification may have low educational value. Time may be better spent on vocabulary, grammar, oral organisation or reading.

A Fifteen-Minute Daily Pronunciation Cycle

• 3 minutes — listen: identify /θ/, /ð/ or /t/ without reading.
• 3 minutes — place: practise the sound slowly with a mirror or tactile cue.
• 3 minutes — word: produce a small contrast set.
• 3 minutes — phrase: use the target in natural chunks.
• 3 minutes — transfer: read or speak one short sentence without pausing for the sound.

Short repeated practice is usually more useful than one long weekly drilling session because motor speech control depends on frequent retrieval. The exact schedule should fit the learner; the principle is repeated contact with feedback.

A Ninety-Minute English Lesson With Pronunciation Integrated

Pronunciation should not consume every English lesson. An illustrative class can spend ten minutes on discrimination and production, then move into reading, comprehension, writing or oral communication. Target words from the reading can be recycled during the pronunciation segment. Later oral discussion provides a natural transfer check.

In a three-student group, one learner may need /θ/ placement, another may already be clear and need oral organisation, and a third may need reading fluency. Shared text allows common discussion while individual pronunciation work remains targeted. Students should not be forced into identical accent training simply because they share a class.

AI and Speech Tools Can Help, but Human Listening Still Matters

Speech-recognition and pronunciation apps can provide repeated models and visual feedback. They can also misrecognise accents or reward one narrow acoustic pattern. Use them as one source of feedback, not a final judge of good English. The most important test is whether real listeners can understand the intended words in natural speech.

A useful workflow is student attempt, human or tool feedback, focused repair, then a new sentence. If the app says a word is wrong but several competent listeners understand it easily, inspect whether the tool is overfitting to a preferred accent. Technology should widen practice, not redefine intelligibility as imitation.

Frequently Asked Questions

Must the tongue visibly stick out for /θ/ and /ð/? The tongue often approaches or lightly contacts the teeth, and a small amount may be visible. Exaggeration can help beginners feel the placement, but natural production usually becomes smaller and faster once control develops.

Is replacing /θ/ with /t/ always a serious error? It matters most when it reduces intelligibility or changes words. Accent variation exists. The educational goal is reliable communication and register-appropriate speech, not elimination of every non-prestige feature.

Should pronunciation be corrected during every conversation? No. Constant interruption can destroy fluency. Choose high-value recurring errors, practise them separately, then use conversation as a transfer check.

What is the strongest sign of progress? The learner hears the distinction, produces it in unfamiliar words and keeps enough control during spontaneous speech that listeners understand without special effort.

The Pronunciation Receipt

A learner has moved beyond a pronunciation tip when they can hear the contrast, explain the placement, produce the sound in a new word and retain it in natural speech without monitoring every mouth movement. The system has transferred from conscious instruction into usable communication.

That is the deeper lesson behind “the”, “three” and “tree”: pronunciation is not a list of corrected words. It is a listening-and-motor system that should become increasingly automatic while the speaker’s own accent identity remains intact.

Troubleshooting: Where the Pronunciation Chain Breaks

If a learner still says tree for three, identify the first failure. Can they hear the difference when another person says the words? Can they produce /θ/ in isolation? Can they keep the tongue and airflow correct before a vowel? Can they maintain the sound inside a cluster such as /θr/? Can they still do it at normal speech speed? Each stage requires a different repair.

A student who hears the contrast but cannot produce it needs motor practice. A student who produces it alone but loses it in three needs cluster practice. A student who succeeds in slow reading but not conversation needs transfer. This ladder prevents teachers from repeatedly drilling isolated /θ/ when the real problem occurs later in the chain.

Initial, Medial and Final Position

Practise the target sounds in more than one word position. Initial /θ/ appears in words such as thin and think; medial /θ/ can appear in words such as healthy; final /θ/ appears in both or path depending on accent and word choice. /ð/ appears initially in common function words such as this and those, medially in words such as mother, and finally in a smaller set of forms in some accents.

The point is not to memorise a complete phonological dictionary. It is to prevent one successful word from becoming false evidence that the sound is mastered everywhere. Position changes the neighbouring sounds and therefore the motor demand.

Sentence-Level Contrast Without Tongue Twisters

Use ordinary sentences such as “Three students thanked the teacher,” “Those trees are taller,” and “The third task is shorter.” Ask the learner to speak for meaning first, then repeat while monitoring one or two target words. Sentences should remain understandable and age-appropriate; a tongue twister that is nearly impossible even for fluent speakers may train frustration more than clear speech.

After controlled sentences, ask a genuine question containing the target word: “Which three things would you take on a trip?” The learner now has to retrieve content and pronunciation at the same time. This is a stronger transfer test than repetition.

Pronunciation and Oral Confidence Should Be Separated

A hesitant child may speak unclearly because the sound is unfamiliar, because the idea itself is hard to formulate or because the learner is afraid of being corrected publicly. Pronunciation correction should not be used as the only explanation for quiet speech. Give low-pressure attempts, private rehearsal where useful and enough response time.

The final goal is a speaker who can communicate without monitoring every sound consciously. If correction makes the learner so self-conscious that oral participation falls, reduce the target load and rebuild confidence alongside intelligibility.

A Final Transfer Test for /th/

Ask the learner to speak for thirty to sixty seconds on an ordinary topic containing several target words, such as “three things I would take on a trip” or “the best thing about Thursday lessons”. Do not interrupt. Afterwards, replay or reflect on only the target contrast. Can the learner identify which words were clear and which shifted toward /t/ or /d/?

Then repeat the same communicative job with a different topic. If the sound remains clear while the learner is thinking about ideas rather than mouth position, the pronunciation has moved closer to automatic use. That is the final destination: intelligibility that survives real speech.

Pronunciation Should End in Meaningful Speech

After the final sound drill, ask the learner to tell a short story, explain a picture or answer a genuine question without being reminded to produce /θ/ or /ð/. Listen for intelligibility, not perfection. If the sound remains clear often enough that meaning is not disrupted, the learner is ready to move attention toward broader oral communication while keeping occasional retrieval practice.

This fading step matters. Pronunciation teaching succeeds when conscious placement becomes less necessary and the student can focus again on ideas, vocabulary and the receiver.