Glossary of pronunciation and language-teaching terms

One of several sounds that a language treats as the same underlying unit. English /p/ is aspirated in "pin" and unaspirated in "spin", but speakers hear both as "the same sound".

Why it matters. Learners hear through their own allophone system. An English speaker cannot hear the difference between Korean plain and aspirated stops at first, because English files both under one phoneme.

Phoneme

A puff of air released after a stop consonant, before the following vowel begins.

Why it matters. English aspirates /p t k/ at the start of stressed syllables; French, Spanish, Italian and Russian do not. Removing aspiration is the cheapest single improvement to an English speaker’s accent in those languages. Test it by holding a sheet of paper in front of your mouth.

A vowel that changes quality within a single syllable — the tongue moves from one position to another, as in English "say" /eɪ/ or "go" /oʊ/.

Why it matters. English diphthongises vowels that most other languages keep pure. English speakers import this habit everywhere, which is why Spanish *mesa* and Italian *bene* come out sounding wrong even when every consonant is right.

A consonant held for roughly twice its normal duration, functioning as a distinct sound rather than a repetition.

Why it matters. Phonemic in Italian, Finnish, Japanese, Hungarian and Arabic. Italian *nono* and *nonno* are different words; Finnish *tuli*, *tuuli* and *tulli* are three. English speakers reliably under-produce it.

Italian pronunciationFinnish pronunciation

A notation system in which each symbol represents exactly one sound, independent of any language’s spelling conventions.

Why it matters. It gives teacher and student a shared vocabulary for what went wrong. "Your /ɪ/ is too close to /i/" is actionable; "that sounded a bit off" is not.

Full IPA referenceHow to read the IPA

A unit of timing smaller than a syllable. In Japanese, a long vowel, the moraic nasal ん and the moraic obstruent っ each count as one mora.

Why it matters. Japanese rhythm is built on morae, not syllables. *Shinbun* is two syllables but four morae, and getting the timing wrong is why learners sound foreign even with correct segments.

Japanese pronunciation

The smallest unit of sound that can change meaning in a given language.

Why it matters. Phoneme inventories differ between languages, which is the root of every pronunciation difficulty. A distinction that is phonemic in one language is invisible in another.

Allophone

A system in which each word carries a fixed pitch pattern that can distinguish meaning, without the full tone system of a language like Mandarin.

Why it matters. Japanese, Swedish and Norwegian all have it, and none of them write it. Japanese *hashi* means bridge, chopsticks or edge depending on pitch alone.

The music of speech: stress, rhythm, intonation and timing, as opposed to the individual sounds.

Why it matters. Prosody matters more for intelligibility than individual segments. A learner with perfect consonants and English rhythm is harder to understand than one with the reverse.

The neutral mid-central vowel /ə/, produced with the tongue at rest.

Why it matters. The most common vowel in spoken English. Failure to reduce unstressed syllables to schwa is the main cause of a "robotic" accent in English — and over-applying it is the main cause of a foreign accent in Spanish, Italian and Finnish.

The schwa in detail

In stress-timed languages (English, German, Russian) stressed syllables recur at roughly even intervals and unstressed ones compress. In syllable-timed languages (French, Spanish, Italian) every syllable takes about the same time.

Why it matters. This is the deepest rhythmic difference between language families and the hardest habit to change. English speakers compress French syllables; French speakers give English syllables equal weight.

Pitch used to distinguish word meaning. Mandarin has four tones plus a neutral one; Thai has five; Bambara has two.

Why it matters. Tone is not intonation. English speakers overlay question intonation on top of lexical tone and destroy it — the classic beginner error in Mandarin and Thai.

Rules that change a tone depending on the tones around it.

Why it matters. In Mandarin, two third tones in a row make the first one a second tone — so *nǐ hǎo* is actually said *ní hǎo*. Learners who drill tones only in isolation never encounter this.

Mandarin pronunciation

A rule requiring the vowels within a word to share a feature — typically frontness or backness — so suffixes change form to match the root.

Why it matters. In Turkish, Finnish and Hungarian this is grammar, not decoration. Confusing Turkish ı and i does not just sound wrong, it selects the wrong suffix.

Turkish pronunciation

The syllable of a word that carries prominence, through some combination of loudness, length and pitch.

Why it matters. In English it is lexical and unpredictable, and misplacing it damages intelligibility more than most consonant errors. In Finnish, Hungarian and Czech it is always initial; in Polish, penultimate; in Turkish, usually final.

Language you can understand most but not all of — Krashen’s "i+1". The claim is that acquisition happens when you understand messages slightly above your current level.

Why it matters. It is the strongest argument for graded readers, podcasts for learners, and teachers who speak the target language at a controlled level rather than at native speed or at beginner speed.

The input problem

When an error becomes permanently embedded in a learner’s speech because it never impeded communication enough to get corrected.

Why it matters. This is the argument for correcting pronunciation early. A Spanish learner who diphthongises every vowel is still understood, so nobody corrects it, so it hardens. Ten years later it is far harder to undo than it would have been in month two.

The influence of a learner’s first language on their second — in pronunciation, this means substituting the nearest L1 sound for an unfamiliar L2 sound.

Why it matters. Almost all pronunciation errors are predictable from the learner’s L1. If you know a student speaks Japanese, you know before they open their mouth which English contrasts will be difficult.

Mixing different types of practice within a session rather than blocking them — alternating between verb forms, vocabulary and pronunciation instead of doing thirty minutes of each.

Why it matters. Blocked practice feels more productive and produces worse long-term retention. Interleaved practice feels harder and works better, which is why learners resist it.

Two words that differ by exactly one sound — "ship" and "sheep", *pero* and *perro*, *mat* and *mать*.

Why it matters. The single most efficient pronunciation drill, because it isolates one variable. If a student cannot hear the difference, production practice is premature — perception has to come first.

Minimal pairs training

Recalling information from memory rather than re-reading it. The act of retrieval is what strengthens the memory, not the exposure.

Why it matters. This is why flashcards work and re-reading a vocabulary list does not. It is also why a student who "reviews" their lesson notes by reading them has done almost nothing.

Spaced repetition explained

Speaking along with a recording in near-real time, a second or so behind, copying rhythm and intonation rather than reading a transcript.

Why it matters. It trains prosody, which conventional drills do not touch. It is also one of the few techniques that improves listening and speaking simultaneously.

The shadowing technique

Scheduling reviews at increasing intervals, timed to just before you would otherwise forget.

Why it matters. It is the difference between learning 2,000 words and retaining 2,000 words. The scheduling is the whole point — daily review of everything is both less effective and vastly more work.

Spaced repetition explainedAnki for language learning

Transliteration converts one script to another letter by letter. Transcription records how something is pronounced. Pinyin is a transliteration; IPA is a transcription.

Why it matters. Confusing the two causes real damage. Thai RTGS romanisation marks neither tone nor vowel length, so a learner relying on it is missing the two features that matter most.

The effect that assessment has on teaching and learning — what you test determines what gets studied.

Why it matters. If you never check homework, students stop doing it. If you only check vocabulary, they only study vocabulary. The thing you review at the start of the lesson is the thing that gets practised.