Originally published in 2016; rebuilt in 2026. This page began as a short Primary English pronunciation lesson about the, three and tree. Its original RFE remains useful: help learners hear and produce sounds that can become confused in English. The replacement keeps that job, removes obsolete tuition/contact material and unrelated travel imagery, and reframes pronunciation around intelligibility, listening discrimination and self-correction rather than accent erasure.
Quick Read
the usually begins with the voiced dental fricative /ð/; three begins with the voiceless dental fricative /θ/; tree begins with /t/. These sounds differ in tongue position, airflow and voicing. Learners improve fastest when they first learn to hear the contrast, then produce it slowly, then use it inside phrases and ordinary speech.
One-sentence answer: do not train the word alone—train the sound contrast from ear → mouth position → word → phrase → conversation.
The RFE of this page
This URL owns one narrow English-learning job: help learners distinguish and produce /ð/, /θ/ and /t/ clearly enough that listeners can reconstruct the intended word.
It is not a general oral-examination page and it is not a current Punggol tuition page.
Start with the three target words
| Word | Typical pronunciation | Opening sound |
|---|---|---|
| the | /ðə/ or /ðiː/ depending context and accent | /ð/ — voiced dental fricative |
| three | /θriː/ | /θ/ — voiceless dental fricative |
| tree | /triː/ | /t/ — voiceless alveolar stop |
The important difference is not the spelling. It is the physical action used to create the opening sound.
/θ/: the opening sound in “three”
For /θ/, the tongue tip or front edge comes lightly between, or very close to, the upper and lower front teeth. Air continues to flow through the narrow gap.
- tongue: near or lightly between the teeth;
- air: continuous;
- voice box: normally not vibrating;
- sound type: fricative.
Try holding the sound: thhhhh. If the sound can continue while air flows, the learner is producing a fricative rather than a stop.
Then add the rest of the word slowly:
/θ/ → /θr/ → /θriː/ → three.
/ð/: the opening sound in “the”
/ð/ uses a similar tongue position to /θ/, but the vocal folds vibrate.
- tongue: near or lightly between the teeth;
- air: continuous;
- voice box: vibrating;
- sound type: voiced fricative.
Place two fingers lightly on the throat and compare:
- /θ/ — little or no voicing vibration;
- /ð/ — clear vibration.
This voiced/unvoiced contrast is useful because the mouth position is similar while one important variable changes.
/t/: the opening sound in “tree”
/t/ is made differently. The tongue contacts the ridge just behind the upper front teeth, temporarily blocks the airflow, then releases it.
- tongue: behind the upper teeth at the alveolar ridge;
- air: briefly stopped, then released;
- voice box: normally not vibrating;
- sound type: stop or plosive.
This means three and tree differ not only in tongue location but also in what happens to the airflow.
A useful physical contrast
| Sound | Tongue | Airflow | Voicing |
|---|---|---|---|
| /θ/ | At/near teeth | Continuous | No |
| /ð/ | At/near teeth | Continuous | Yes |
| /t/ | Behind upper teeth | Stopped then released | No |
This table gives the learner three controllable variables: place, airflow and voice.
Why learners may substitute one sound for another
Speech sounds are learned through long exposure to the sound system of the languages and varieties around us. If a learner’s familiar language system does not use a particular contrast in the same way, the ear may initially map the unfamiliar sound onto a familiar category.
This is not evidence of low ability. It is a speech-learning problem: the learner must build a new auditory and motor distinction.
For some English speakers in Singapore and elsewhere, /θ/ or /ð/ may be realised differently in everyday speech. Accent variation is normal. The educational goal here is not to erase a local identity. It is to give learners control: they should be able to hear the standard contrast and produce a clear version when intelligibility, formal speaking or assessment requires it.
Hear before correcting the mouth
If the learner cannot reliably hear the difference between two sounds, repeated mouth instructions may produce only temporary correction.
Start with listening discrimination.
- three / tree;
- thin / tin;
- thank / tank;
- thought / taught;
- then / den;
- they / day;
- those / dose.
Some pairs also differ in other phonetic details depending on accent, so use them as listening practice rather than assuming every pair differs in only one perfect laboratory variable.
The minimal-pair learning loop
- Teacher says one of two words.
- Learner identifies which word was heard.
- Learner repeats the word.
- Teacher checks the opening sound only.
- Learner uses the word in a short phrase.
- The words are later mixed into ordinary sentences.
Do not stay at the isolated-word stage for too long. The learner needs the sound to survive normal speaking speed.
A mirror exercise for /θ/ and /ð/
A mirror can provide useful visual feedback.
- Relax the jaw.
- Bring the tongue tip lightly toward or between the front teeth.
- Blow a continuous stream of air.
- For /θ/, keep the voice off.
- For /ð/, add voicing while keeping a similar tongue position.
- Move into the following vowel without pulling the tongue back too early.
The tongue should not need exaggerated force. The goal is controlled contact and airflow, not muscular strain.
“The” has more than one common vowel form
The definite article the is commonly pronounced with a weak vowel, approximately /ðə/, before many consonant sounds. A form approximately /ðiː/ is commonly heard before vowel sounds or when the word is strongly emphasised.
Examples:
- the book;
- the tree;
- the answer;
- the person I meant.
Actual fluent speech varies across accents and contexts. The important lesson for a Primary learner is first to stabilise the initial /ð/ and then notice how function words become weaker or stronger in connected speech.
From sound to phrase
Once the learner can say the words independently, move into short phrase chains:
- three trees;
- the three trees;
- the three tall trees;
- I can see the three trees;
- The three trees are near the road.
The original 2016 page used “The Three Trees at Waterway Point are Free!” as a playful practice line. It still works as an articulation challenge because it alternates /ð/, /θ/, /t/ and /f/.
Do not correct every sound in the same sentence
Too much simultaneous correction overloads the learner and can make speaking feel unsafe.
If the target today is /θ/ versus /t/, let minor unrelated pronunciation differences pass unless they block meaning.
A useful correction sequence is:
hear → attempt → identify one difference → adjust → retry → use in context.
Intelligibility before imitation
English has many legitimate accents. A Singaporean learner does not need to imitate a British or American identity to speak good English.
The practical questions are:
- Can the listener identify the intended word?
- Can the learner distinguish important sound contrasts?
- Can the learner adjust pronunciation when misunderstanding occurs?
- Can formal speaking remain clear and controlled?
Accent is part of identity. Intelligibility is part of communication. The two do not need to be enemies.
Connected speech changes sounds
Students sometimes pronounce a target word perfectly in isolation but lose the sound inside a sentence. That is normal: connected speech adds speed, rhythm, stress and neighbouring sounds.
Practise in three stages:
- careful speech: slow and accurate;
- phrase speech: natural grouping;
- conversation: spontaneous use while attention shifts to meaning.
The final stage matters most because communication requires the sound to remain available while the learner is thinking about ideas.
Self-recording creates a feedback loop
A phone recording can help a learner compare prediction with outcome.
- Predict which word may be difficult.
- Record a short sentence.
- Listen once for meaning.
- Listen again only for the target sound.
- Compare the mouth action with the intended /θ/, /ð/ or /t/.
- Record a second attempt.
The learner should learn to detect their own mismatch rather than depend permanently on a teacher saying “wrong”.
For Primary oral work
Clear pronunciation can support oral communication, but good oral performance is larger than individual sounds.
- understand the prompt;
- organise ideas;
- speak clearly enough to be understood;
- use appropriate pace and volume;
- listen and respond;
- repair misunderstandings when necessary.
A learner should not become so anxious about one consonant that the whole message disappears.
Parents: a five-minute practice
- Say three and tree in random order five times; child points to the word heard.
- Child produces /θ/ and /t/ slowly.
- Child says three, tree, three trees.
- Child says one natural sentence.
- Stop while the practice is still successful.
Short, repeated practice is often more useful than turning pronunciation into a long correction session.
Teachers and tutors: diagnose the actual contrast
Do not assume every learner has the same pronunciation problem.
| Observed issue | Likely teaching job |
|---|---|
| Cannot hear three/tree difference | Auditory discrimination |
| Hears it but cannot produce /θ/ | Tongue placement and airflow |
| Produces /θ/ in isolation but loses it in sentences | Connected-speech transfer |
| /θ/ and /ð/ are confused | Voicing distinction |
| Clear sound but speech remains hard to understand | Look beyond one consonant: rhythm, stress, phrasing or language formulation |
AI and pronunciation practice
Speech tools and AI can provide models and transcription, but automatic speech recognition is not a perfect pronunciation judge. It may successfully transcribe an accented word even when a human listener would need more effort—or fail on a perfectly intelligible local accent.
Use technology as one feedback source, not the final authority.
- listen to several reliable models;
- record the learner;
- compare the target contrast;
- check whether real listeners understand;
- retain the learner’s natural accent while strengthening control.
The deeper principle
Pronunciation improves when the learner can connect three forms of evidence:
what I intended to say → what my mouth actually did → what the listener actually received.
That is the useful calibration loop. Hear the contrast, produce it deliberately, transfer it into real speech and keep enough flexibility to repair the message when communication fails.
Historical note: first published 10 March 2016 as a Punggol Primary English pronunciation lesson on “The”, “Three” and “Tree”. The 2026 rebuild retires obsolete contact and promotional material, removes unrelated imagery, preserves the original pronunciation RFE and expands it into an intelligibility-first guide to /ð/, /θ/ and /t/.