Voiced vs Voiceless Consonants

A consonant is voiced if your vocal cords vibrate while you make it, and voiceless if they stay still, and you can feel the difference in a second with a hand on your throat. That one idea explains a large part of the consonant system, because most English consonants come in pairs that are identical except for this single switch. Type sip and zip into the pronunciation Studio, play them back to back, and you are hearing voicing at work.

Get this rule and you get a lot for free: eight consonant pairs, the spelling of plurals, and the sound of every past-tense verb, all from one small distinction.

The one test that tells you everything

Rest two fingers on the front of your throat and make a long “sss”. Nothing happens under your fingers. Now, without moving your mouth, switch to “zzz”. Your throat springs to life. That buzz is your vocal cords vibrating, and its presence or absence is the whole of voicing.

Try it with other sounds. Hold “fff” then “vvv”, or “sh” then the sound in the middle of measure. In each case the mouth barely moves and the only thing that changes is whether the throat is switched on.

If your throat is too quiet to feel, cover your ears instead and make the sound. A voiced consonant produces a low hum inside your head; a voiceless one stays silent. Either test works. This on-or-off vibration is what phonetics calls voicing, and it is one of the main features used to describe every consonant.

The eight consonant pairs

Most English consonants travel in pairs. The two sounds in a pair are made in the same place, in the same way, and differ only in voicing. Learning them side by side is much easier than treating each as a separate sound.

You can even feel the logic of the list. /p/ and /b/ are both made by closing the lips. /t/ and /d/ are both made with the tongue tip behind the top teeth. /k/ and /ɡ/ are both made at the back of the mouth. In each pair, your mouth sets up in one position and the only decision left is whether to switch the voice on. That is why these eight pairs cover so much ground: you are really learning eight mouth positions, each with a voice-off and a voice-on version.

The consonants with no partner

Not every consonant has a twin. Seven English consonants are always voiced and have no voiceless version: /m/ (me), /n/ (no), /ŋ/ (the “ng” in sing /sɪŋ/), /l/ (lie), /r/ (row), /w/ (we), and /j/ (the “y” in yes /jes/).

One goes the other way. /h/, as in hat /hæt/, is voiceless and has no voiced partner in English. These loners are worth knowing so you are not left hunting for a pair that does not exist.

Why voicing actually matters

This is the part that pays off every day, so it is worth slowing down for. Two of the most common word endings in English change their sound based on voicing, and almost no one is taught why.

Take the plural and verb ending -s. It is not always said the same way. After a voiceless sound it comes out as /s/, and after a voiced sound it comes out as /z/. Your mouth does this automatically in your first language; in English it follows the same logic.

The past-tense ending -ed splits the same way. After a voiceless sound it is /t/, and after a voiced sound it is /d/, even though the spelling never changes.

Say walked out loud. There is no “ed” in it at all; the ending is a clean /t/ because /k/ is voiceless. Now say played, and the ending is a /d/ because the vowel before it is voiced. One rule explains the sound of thousands of verbs.

There is a third case worth a line: when the base already ends in a /t/ or /d/, the ending becomes a full extra syllable, /ɪd/, as in wanted /ˈwɒntɪd/. That is the only time you actually hear an “ed”.

A quick word on aspiration

One more detail rounds out the picture. At the start of a stressed syllable, the voiceless sounds /p/, /t/, and /k/ come with a little puff of air. In narrow transcription this is written with a small raised mark, [pʰ tʰ kʰ].

Hold your hand in front of your mouth and say pin, then spin. You feel a puff on pin and almost nothing on spin. It never changes the meaning of an English word, so dictionaries leave it out, but it is why an unaspirated English p can sound a bit like a b to some ears.

Voicing quietly changes the vowel before it

Here is a subtle effect that native speakers use without knowing it. A vowel is noticeably longer before a voiced consonant than before a voiceless one. The consonants are the pair; the giveaway is often the length of the vowel in front of them.

Say bad /bæd/ and bat /bæt/ one after the other. The vowel in bad is held a fraction longer, because the /d/ at the end is voiced. The same pattern shows up in seed /siːd/ against seat /siːt/, and bag /bæɡ/ against back /bæk/.

This matters for being understood. If a listener cannot quite hear whether your final consonant is voiced, the length of the vowel before it often settles the question for them. Stretching the vowel a touch before a voiced ending can make the word land more clearly than hammering the consonant itself.

Final devoicing, a common and normal transfer

Several major languages, including German, Russian, Dutch, and Polish, do not allow a voiced consonant at the end of a word. In those languages, a final voiced sound is automatically said as its voiceless partner. Speakers of these languages often carry the habit into English, so dog comes out closer to dock and is closer to iss.

This is a predictable, normal transfer, not a flaw. The fix is simply to keep the voice running through the end of the word: let the throat stay switched on all the way to the final /ɡ/, /z/, or /d/, rather than shutting it off early. Pairs like bag and back, or buzz and bus /bʌs/, are useful to practise for exactly this.

The same letter, both sounds

Voicing also explains why a single letter can stand for either sound depending on the word. The letter s is voiceless /s/ in bus /bʌs/ and dose /doʊs/, but voiced /z/ in busy /ˈbɪzi/ and rose /roʊz/. The spelling gives you an s every time; only the sound, and the IPA, tell you which one to say.

This is the same lesson as the endings, seen from another angle. English spelling was never built to track voicing, so it reuses one letter for a voiced and a voiceless sound and leaves you to work out which. A transcription settles it in a single symbol, which is a good reason to check the IPA when a word is new to you.

Practice: voiced or voiceless?

Say each consonant sound with a hand on your throat and decide. Cover the answers first.

1. /v/
2. /k/
3. /ʃ/
4. /dʒ/
5. /n/

One rule, many sounds

Everything here comes back to a single switch: throat on for voiced, throat off for voiceless. From that one distinction you get eight consonant pairs, the correct sound for every plural, and the hidden /t/ or /d/ at the end of past-tense verbs.

The clearest single example of the switch is the pair of TH sounds, so if you want to feel voicing in isolation, read the guide to /θ/ and /ð/. Keep the Studio handy to play any pair on demand, and if the symbols are still settling, how to read IPA covers them from the start.