Say ship. The sound just leaks out, a long hiss between the front of your tongue and the roof of your mouth, with nothing in front of it. Now say chip. Before that same hiss, your tongue jumps up and taps the roof of your mouth, sealing the air for a split second, then lets go. A tiny stop, then the hiss. That little tap at the front is the entire difference between the two words. If chip and ship come out of your mouth as the same word, your tongue is either skipping the tap or adding one where it doesn’t belong, and you are in plenty of company: this is one of the consonant pairs that a great many first languages fold into a single sound.
This particular merger sorts speakers by which half they’re missing. French has the hiss but not the tap, so chip arrives as ship and chair as share. Standard Spanish has the opposite inventory, the tap but no bare hiss, so the error runs the other way and ship comes out as chip, she as che. Greek tends to have neither and reaches for an s, turning ship toward sip. Whichever way it goes, English leans on the contrast constantly, and a tap that goes missing or shows up uninvited is enough to swap one everyday word for another.
CH and SH are not one sound said two ways. /ʃ/, the SH in ship, is a single long friction: the front of your tongue hovers near the roof of your mouth, air hisses through the narrow gap, and you can hold it for as long as your breath lasts. /tʃ/, the CH in chip, opens with a tiny /t/: the tongue touches the roof, stops the air completely, then releases into that same SH hiss. A stop and a friction fused into one quick sound you cannot stretch. Most first languages that blur the pair are missing one half, either the stop at the front of /tʃ/ or the bare hiss of /ʃ/, so the fix is to build the half you lack, not to hunt for a setting between them.
Two different sounds
The two spellings look like cousins on the page, but in the mouth the sounds are built differently, and that build is the whole story.
The SH sound is a fricative. The blade of your tongue rises toward the ridge behind your top teeth without touching it, your lips push forward into a small rounded shape, and you force air through the gap until it hisses. Nothing ever closes, so the air keeps moving and the sound keeps going. Take a breath and run shhhh until you run out, the way a librarian shushes a room, and it holds steady the whole time. That is the first tell: a clean /ʃ/ is a sound you can lean on and stretch.
The CH sound is an affricate, which is a fancy word for a stop and a fricative welded together. Your tongue first presses flat against that same ridge and seals the air off completely, almost the way it does for a /t/, though a touch further back. Pressure builds for an instant. Then, instead of releasing into open air the way a plain T does, the tongue pulls back just enough to let the trapped air burst out as a short /ʃ/. Stop, then hiss, so close together they read as one event. You cannot hold a /tʃ/ the way you can hold a /ʃ/; try to stretch it and it either freezes on the T or slides off into a bare shhhh. It happens once and it is over.
Both sounds are voiceless: your vocal cords stay quiet for each, so you cannot tell them apart by listening for which one is heavier. What separates them is that opening tap. /tʃ/ begins with the air stopped; /ʃ/ begins with the air already flowing. That single difference, a closed start or an open one, runs through pair after pair of English words: chair and share, watch and wash, catch and cash, cheap and sheep.
| /tʃ/ — tap, then hiss (CH) | /ʃ/ — hiss only (SH) |
|---|---|
| chip | ship |
| chair | share |
| watch | wash |
| catch | cash |
| cheap | sheep |
| choose | shoes |
| chew | shoe |
| cheer | sheer |
| much | mush |
| witch | wish |
Read down both columns. If the two sides land as the same word, that is the merger this article takes apart, and the repair is a single moving part you can feel: the tongue either touches the roof of the mouth before the hiss, or it doesn’t.
Why “a softer CH” is the wrong frame
If your first language runs these together, the two feel like one sound with a dial on it, and learners reach for “a softer CH” or “a sharper SH,” as if the target sat somewhere on a slider between them. There is no slider. A sound that starts with the air sealed off and a sound that starts with the air already hissing are not two settings of one thing; you cannot dial your way from one to the other. They are a closed door and an open one.
The question is not how hard or soft you say it. It is whether your tongue stops the air before the hiss, or lets the hiss run from the start. A tap, or no tap.
The hold test settles it faster than any amount of listening. Try to stretch the sound. If you can hold it, a steady shhhh that runs as long as your breath, you are on /ʃ/. If it refuses to stretch, if it fires once and quits, you are on /tʃ/, because the stop at the front will not let the sound continue. You do not need a partner or a recording for this. You can feel the difference in your own mouth, between a sound that flows and a sound that pops.
It helps to see where this pair sits in a larger family. English keeps four sounds in this corner of the mouth, two voiceless and two voiced. /tʃ/ and /ʃ/ are the voiceless pair this article is about. Switch your voice on and you get their voiced twins, /dʒ/ and /ʒ/. Many languages that struggle with the voiceless pair struggle the same way with the voiced one, which is handy, because the fix you build for chip and ship carries straight over to the voiced pair. More on that once you can make the first two cleanly.
How to make each sound
Most readers already own one of these two and are building the other from scratch. Which one you have depends on your first language, so the order below covers both.
Start with the /ʃ/, because it is the simpler of the two and the CH is built on top of it. Put the blade of your tongue up near the ridge behind your top teeth, close but not touching, leaving a thin slot. Round your lips a little and push them slightly forward. Now blow voiceless air through the slot until it hisses, and hold it: shhhh. If your hiss sounds thin or whistly, like an s, your tongue is too far forward; slide it back a hair and round your lips until the sound goes from a sharp sss to a fuller, darker shhh. The lip rounding is part of the sound, not decoration, so keep the lips pushed out the whole time.
Now build the /tʃ/ on top of that hiss. Set your mouth as if for a /t/, tongue pressed flat to the ridge but a touch further back, air sealed off behind it. Hold that seal for a beat. Then release the tongue not into open air but straight into the SH position you just practiced, so the trapped air escapes as a quick hiss. The result is t + sh fused: tch. A reliable way to find it is to say the two sounds back to back and speed them up. Say white sheep, then run the words together faster and faster until the t and the sh collide into a single tʃ: whi-cheep. That collision is the affricate. The tap is doing the work; the hiss is just where it lands.
Which half you have to build depends on the slip your language hands you:
If your chip keeps coming out as ship, you are skipping the stop. The hiss is fine; what is missing is the tongue touching the roof of the mouth before it. Exaggerate the tap at first. Press the tongue hard to the ridge, hold it a touch too long, and pop it into the hiss, so the word starts almost like t-ship. Once the stop is reliable you can make it lighter and quicker until it stops sounding forced.
If your ship keeps coming out as chip, you are adding a stop that should not be there, which is the standard Spanish-speaker pattern. Your tongue is reaching up to touch the ridge before the hiss out of habit. Keep it down. Start the word with the air already flowing through the slot, a hiss in motion before the rest of the word arrives, and never let the tongue make contact. Stretch the SH on purpose at first, shhhip, to prove to your mouth that nothing has to close.
The most common mistake in both directions is treating the stop as optional decoration. It is the sound. A /tʃ/ with no real stop is just a /ʃ/, and a /ʃ/ with a stop snuck in front is just a /tʃ/. Decide on purpose, each time, whether the tap is there.
The voiced cousins: /dʒ/ and /ʒ/
Once the voiceless pair is solid, two more sounds come almost for free. Keep everything the same and switch your voice on. A voiced /tʃ/ is /dʒ/, the J in job: choke becomes joke when the vocal cords kick in, same tap, same release, now buzzing. A voiced /ʃ/ is /ʒ/, the soft sound in the middle of measure and pleasure, the same held hiss with the voice switched on. English almost never starts a word with /ʒ/, so you mostly meet it in the middle of a word, in vision, casual, television. A few French loans carry it at the edges instead, final in garage or even word-initial in genre. If you can already feel the stop-versus-no-stop difference in the voiceless pair, you have already done the hard part for the voiced one.
Which spelling is which
The letters mostly behave. The pair ch usually spells /tʃ/: chip, chair, much, teacher, lunch. The pair sh almost always spells /ʃ/: ship, share, wash, fish, fresh. The cluster tch is just another way to write /tʃ/, the T already half-drawn into the spelling: watch, catch, match, kitchen. So far the page tells the truth.
Then English borrows words from other languages and keeps their spellings, and the ch stops meaning what you expect. In a batch of words taken from French, ch is pronounced /ʃ/, the bare hiss, with no tap at all: machine, chef, brochure, Michigan, and the city of Chicago. In another batch taken from Greek, the same ch is pronounced like a hard /k/: school, stomach, character, chemistry. There is no way to predict which is which from the letters; the word’s history decides, and you learn them one at a time.
The hiss also hides under letters that are not sh at all. A whole family of common endings turn a t, c, or s into the hiss before a weak vowel, with two endings that land on the tap instead:
| Spelled with | Pronounced | Examples |
|---|---|---|
| -tion, -tial, -tient | /ʃ/ | nation, station, patient, partial |
| -cial, -cean, -cious | /ʃ/ | special, ocean, delicious |
| -ssion, -ssure, -nsion | /ʃ/ | mission, pressure, tension |
| -ture | /tʃ/ | nature, picture, future |
| -stion | /tʃ/ | question, suggestion |
The last two rows are the quiet ones. The ending -ture is not a hiss but a full tap-then-hiss /tʃ/, so nature is nay-cher and picture is pik-cher, the same CH that opens chip, hiding inside a T. The cluster -stion does the same after an s, so question comes out kwes-chun. Once you notice this tap, you hear it everywhere in careful American speech: culture, future, century, actually.
Two everyday words bend the rule in their own direction. sugar and sure open with a su that says /ʃ/ even though the vowel right after it is stressed, not weak. They are a tiny set to memorize, not a pattern: almost every other su word keeps a plain s, as in super and suit.
Train the ear before the mouth
You cannot reliably produce a contrast you cannot hear. Plenty of learners can place a clean /tʃ/ and a clean /ʃ/ when they slow down and think about each one, then lose the difference the instant a sentence moves at speed, because the ear never learned to flag which one just went by. Perception comes first, and for this pair it comes quickly, because the two carry a clear signature: a /ʃ/ is a hiss that keeps going, and a /tʃ/ is a hiss with a little click of a stop punched onto the front.
The drill is minimal pairs fed to you out of order. Have a partner or a text-to-speech voice say one word from a pair at random: chair or share, watch or wash. You only sort, naming which one you heard, no speaking yet. Our chip vs ship comparison page puts the two side by side with audio if you want a ready-made voice to sort against. When you can label fifteen in a row without stalling, your ear has built the category and your mouth has a target. A quieter version needs no partner. Take a minute of American speech with a transcript, an interview or a clip from a show, and mark every ch and sh word in it. Replay each one and answer one question: tap, or no tap? You are teaching your ear to stop filing the two under a single label, which is the step that makes the mouth-work hold.
One more ear-trap is worth naming. If your first language leans on s, you may be hearing and saying s where English wants /ʃ/, so ship drifts toward sip and she toward see. The cure is the lips. A real /ʃ/ rounds and pushes them forward; an s leaves them flat. Watch your mouth in a mirror on the word ship: if your lips stay slack, you are on an s, and pushing them out is half the repair.
Practice phrases
Read these out loud, twice each. Most lines make your mouth switch between /tʃ/ and /ʃ/, which is harder, and more useful, than drilling either sound alone. Slow down until each word lands on the start you meant, tap or no tap, then bring the pace back up. Two lines break the pattern on purpose: the seventh is all /ʃ/, every ch a bare hiss, and the last is all /tʃ/, every ch a tap. Take both of those slowly.
- If the mug has a chip, don't ship it. If the mug has a chip, don't ship it.
- Watch her wash the dishes. Watch her wash the dishes.
- Catch the cash before it blows off the table. Catch the cash before it blows off the table.
- Share the chair by the window. Share the chair by the window.
- Choose the shoes that match. Choose the shoes that match.
- The cheap wool came from one sheep. The cheap wool came from one sheep.
- The chef fixed a machine in Chicago. The chef fixed a machine in Chicago.
- Are you sure such sugar is cheap? Are you sure such sugar is cheap?
- Each picture tells a fresh story. Each picture tells a fresh story.
- Which chair did the teacher choose? Which chair did the teacher choose?
If a line trips you at speed, that is the point of stacking both sounds into one breath. The tongue has to touch the roof of the mouth for the CH and stay clear of it for the SH, several times a sentence, and getting that timing automatic is the real drill. Watch the words chef and machine in line seven: both are spelled with ch and both are pure SH, no tap, a useful reminder that the spelling is not always in charge.
How different first languages handle this
Where you start depends on which half your first language gave you. Some hand you the hiss, some hand you the tap, a few give you neither and one gives you both. None of this is a deficiency; it is just the shape of the gap you are closing.
| Your L1 | Has both already? | Default slip | What to focus on |
|---|---|---|---|
| French | ~ Hiss only | chip → ship | You own a clean /ʃ/ already. Build the stop at the front of /tʃ/: touch the ridge, seal the air, then release into your hiss. |
| Portuguese | ✓ Mostly | spelling: ch read as /ʃ/ | You have /ʃ/, and Brazilian Portuguese adds a /tʃ/ before i (tia). The real trap is the spelling: Portuguese ch is /ʃ/, but English ch is usually /tʃ/, so give chip the tap. |
| Spanish (most dialects) | ~ Tap only | ship → chip | You have a firm /tʃ/ (mucho) but no bare /ʃ/. Build the hiss with no contact: air flowing, tongue down, lips pushed forward. |
| Greek | ✗ Neither | ship → sip; chip → tsip | Build the SH back from your s: same air, but slide the tongue back and round the lips. Then stack a stop on it for CH. |
| Dutch | ~ Marginal | chip softens to ship | Both sounds live mainly in loanwords. Your /ʃ/ is usually fine, but keep the sj habit from splitting it into a two-part s-y. The real gap is the affricate /tʃ/: build a firm tap and release it straight into the hiss so chip does not soften to ship. |
| Japanese | ✓ Yes, palatal | mostly fine; see → she | You have close versions in shi (し) and chi (ち). The trap runs the other way: keep plain s from turning into SH before i, so see stays see and does not drift toward she. |
| Korean | ~ Partial | show → so | Your s shifts toward SH before i but stays a plain s before other vowels, with no separate /ʃ/. Your affricates already sit close to /tʃ/; build a full /ʃ/ before every vowel, lips rounded. |
| Mandarin Chinese | ✓ Yes | mostly fine; watch retroflex coloring | You have rich versions in both sh/ch (retroflex) and x/q (palatal). Aim for the middle, blade up near the ridge, without curling the tongue back. |
| Hindi, Urdu | ✓ Yes | mostly fine | Both /tʃ/ (च) and /ʃ/ (श) live in your language. Mostly a mapping job, plus the English spelling traps. |
| German | ✓ Yes | mostly fine | Your sch is a native /ʃ/, and deutsch and Quatsch give you /tʃ/. Spend your time on the spelling, where English buries /ʃ/ in -tion and French ch. |
FAQ
Almost always because your first language has only one of the two sounds and uses it for both English words. If you speak French or European Portuguese, you have the SH hiss but no native CH, so chip comes out as ship. If you speak Spanish, you have the CH tap but no bare SH, so ship comes out as chip. The CH in chip is /tʃ/, a tiny /t/ stop released into a hiss, while the SH in ship is /ʃ/, the hiss alone. The fix is to build whichever half you are missing: the stop at the front, or the bare friction with no stop.
/ʃ/ (SH) is a fricative: the front of your tongue hovers near the roof of your mouth, air hisses through the gap, and the sound holds for as long as your breath lasts. /tʃ/ (CH) is an affricate: your tongue first touches the roof and stops the air completely, like a /t/, then releases into that same SH hiss, so it is one quick stop-plus-hiss you cannot stretch. Both are voiceless, so the difference is not how hard or soft the sound is. It is whether the air is stopped before the hiss (CH) or flowing from the start (SH).
Your tongue is touching the roof of your mouth before the hiss when it should stay clear of it, which adds a CH stop to a word that has none. This is the usual Spanish-speaker pattern, since Spanish has /tʃ/ but not a bare /ʃ/. Start the word with the air already hissing through a narrow slot, lips rounded and pushed forward, and do not let the tongue make contact. Stretching the SH on purpose at first, shhhip, proves to your mouth that nothing has to close.
Because English borrowed those words from French and kept the French spelling, where ch stands for /ʃ/, the plain hiss with no tap. The same thing happens in chef, brochure, Michigan, and champagne. A different set of words came from Greek and pronounce ch like a hard /k/ instead, as in school, stomach, and chemistry. The spelling cannot tell you which is which; the word’s origin decides, so these are learned one at a time.
Almost always, yes. In endings like -tion, -tial, and -tient, the t is pronounced /ʃ/, so nation is nay-shun, station is stay-shun, and patient is pay-shunt. The related endings -cial, -cean, and -cious do the same thing with a c: special, ocean, delicious. Two endings break toward the CH sound /tʃ/ instead: -stion after an s, so question is kwes-chun, and -ture, so nature is nay-cher and picture is pik-cher.
Yes. /dʒ/, the J in job, is a voiced /tʃ/: the same tap-then-hiss with your vocal cords switched on, which is why choke and joke differ only by voice. /ʒ/, the sound in the middle of measure and pleasure, is a voiced /ʃ/: the same held hiss with voice added. So the stop-versus-no-stop difference you learn for chip and ship carries straight over to the voiced pair, and the work mostly transfers.
The CH/SH merger is one of the more common consonant confusions in English, and one of the more satisfying to clear, because the whole difference is a single moving part you can feel: the tongue either taps the roof of your mouth before the hiss, or it doesn’t. Spend a few days only hearing the contrast, then a week running your mouth between the two through the lines above. Your tongue only has to learn that the hiss comes two ways, with a tap in front and without, and to decide, each time, which one you mean.
