The Consonants (and the r-sound) | Japanese Grammar

Most Japanese consonants have a close English cousin: か is your k, ま is your m, な is your n. If you simply used your English versions, you would be understood almost everywhere. But four spots on the map are genuinely different — the r-row, the f in ふ, and the palatal sounds in し, ち, つ — and these are exactly where English speakers give themselves away. This page focuses on those, because fixing them is where the payoff is.

The inventory at a glance

Row	Kana	Sound	Note
k	かきくけこ	[k]	like English k
g	がぎぐげご	[g]	like English g (often softens to a nasal [ŋ] mid-word)
s	さしすせそ	[s], し = [ɕ]	し is a soft palatal "sh"
z	ざじずぜぞ	[z], じ = [dʑ]/[ʑ]	じ is a palatal "j"
t	たちつてと	[t], ち = [tɕ], つ = [ts]	ち = "ch", つ = "ts"
d	だぢづでど	[d]	like English d
n	なにぬねの	[n]	like English n
h	はひふへほ	[h], ひ = [ç], ふ = [ɸ]	ふ is a bilabial f
b	ばびぶべぼ	[b]	like English b
p	ぱぴぷぺぽ	[p]	like English p
m	まみむめも	[m]	like English m
y	やゆよ	[j]	like English y
r	らりるれろ	[ɾ]	a single tap — see below
w	わ	[w]	like English w

Consonants come with a vowel attached

Before the individual sounds, one structural rule that shapes everything: a Japanese consonant is almost always followed by a vowel. There are no clusters like English spl-, -nks, or str-. A word is a chain of open beats — か-さ (kasa), て-が-み (tegami) — each ending in a vowel.

There are exactly two exceptions, and both get their own beat:

ん — the moraic nasal, which can end a syllable, as in 日本（にほん）. See The Moraic ん.
っ — the small tsu, which doubles the following consonant, as in 学校（がっこう）. See Geminate Consonants.

Every other consonant drags a vowel along behind it.

The r-row: one quick tap

This is the sound that worries English speakers most, and paradoxically it is easy — once you stop trying so hard. ら, り, る, れ, ろ are not the English r (the bunched, tongue-back sound in red) and not the English l (a held sound with the tongue pressed to the ridge). They are a single, light tap tip of the tongue flicks the ridge behind your top teeth once, quickly, and lets go.

You already make this sound. It is the middle consonant in the American English words butter, water, ladder, and kitty — that fast little "d/t" flick. It is also the single r of Spanish pero. Steal that flick and you have the Japanese r.

さくら

sa ku ra

cherry blossom — the ら is a quick tap, like the tt in 'butter'.

これはいくらですか。

kore wa ikura desu ka

How much is this? — kore and ikura: light taps, never a heavy English r.

車で行きます。

kuruma de ikimasu

I'll go by car. — 車（くるま）: the る is a single flick.

Here is the liberating part. Because Japanese has no l/r contrast, there is no wrong choice to make. A Japanese listener hears your tap as somewhere between r, l, and d, and files it as the one Japanese phoneme either way. You are not being asked to produce a precise "r" or a precise "l" — you are being asked to stop producing either one and just tap. The only real mistake is over-articulating: bunching your tongue for a full English r, or planting it for a long English l.

💡

Everyone worries about telling り (ri) from a would-be "li". Relax: there is no "li" in Japanese, so there is nothing to distinguish it from. One light tap covers the entire r-row. Stop reaching for the perfect r — a quick, lazy flick is exactly right.

ふ: an f made with the lips, not the teeth

English f is made by putting your top teeth on your bottom lip and blowing — [f]. Japanese ふ is made with both lips, teeth nowhere near: you bring your lips close together and blow gently through the gap, as if quietly blowing out a candle or cooling a spoonful of soup. Linguists write it [ɸ], a voiceless bilabial fricative.

It is softer and airier than English f, almost halfway to an h. ふ is the only kana in the は-row that behaves this way; は, へ, ほ are a plain [h].

富士山に登りたいです。

Fujisan ni noboritai desu

I want to climb Mt. Fuji. — 富士山（ふじさん）: the ふ is a soft, lip-blown sound.

お風呂に入ります。

ofuro ni hairimasu

I'm going to take a bath. — お風呂（おふろ）: gentle bilabial f, no teeth.

冬は寒いです。

fuyu wa samui desu

Winter is cold. — 冬（ふゆ）: blow the ふ through your lips.

The palatal set: し, ち, つ, じ

Look at the s-row and t-row again: they are sa-shi-su-se-so and ta-chi-tsu-te-to, not the tidy "si" and "ti/tu" a beginner expects. This is not a spelling quirk — it reflects a real sound shift. Before certain vowels, these consonants palatalize (the tongue rides up toward the hard palate). The odd romanizations shi, chi, tsu are honest records of what actually comes out of a Japanese mouth.

し [ɕ] — a soft "sh", but more forward and less rounded than English sh in ship. Lips stay relaxed, tongue near the hard palate.
ち [tɕ] — a "ch", palatal like し, as in 口（くち）"mouth".
じ [dʑ]/[ʑ] — a "j", the voiced partner of ち/し, as in 時間（じかん）"time".
つ [ts] — the tricky one. This is the "ts" of English cats or pizza, but at the start of a beat, which English never does. English speakers reflexively drop the t ("soo") or insert a vowel. Say cats, isolate the final "ts", then glue a vowel onto it: ts + u.

お寿司を食べに行きましょう。

osushi o tabe ni ikimashō

Let's go eat sushi. — 寿司（すし）: forward, unrounded 'sh' in し.

月がきれいですね。

tsuki ga kirei desu ne

The moon is beautiful, isn't it? — 月（つき）'moon': lead with a clean 'ts'.

靴を脱いでください。

kutsu o nuide kudasai

Please take off your shoes. — 靴（くつ）'shoes': don't drop the t — ku-tsu.

お茶はいかがですか。

ocha wa ikaga desu ka

Would you like some tea? (polite) — お茶（おちゃ）: palatal 'ch' in ち/ちゃ.

💡

These palatal sounds are why you never see "si", "ti", or "tu" in Japanese: the sounds simply shift when they meet those vowels. So the moment you learn the kana grid, the "irregular" romaji shi, chi, tsu stop being exceptions and become the predictable, regular result of palatalization.

Common mistakes

りんご

❌ RING-go (heavy English r)

Wrong: bunching the tongue into a full English 'r' for り.

りんご

✅ ringo (single light tap)

Right: り is one quick flick, like the tt in 'kitty'.

富士山

❌ FOO-jee-san (teeth-on-lip f)

Wrong: making ふ with the top teeth on the lower lip, like English f.

富士山

✅ Fujisan (lips only, gentle blow)

Right: bilabial [ɸ] — blow through both lips, no teeth.

靴

❌ koo-soo (t dropped from つ)

Wrong: reducing つ to a plain 'su' because English lacks initial 'ts'.

靴

✅ kutsu (crisp 'ts' onset)

Right: keep the t — 'ku-tsu', the ts as in 'cats'.

しお

❌ SHOW (rounded, back English sh)

Wrong: an English 'sh' with rounded lips and a back tongue.

しお

✅ shio (forward, unrounded [ɕ])

Right: 塩（しお）'salt' — a soft, forward 'sh', lips relaxed, then a clean o.

れい

❌ lay (held English l)

Wrong: planting the tongue for a long English 'l'.

れい

✅ rei (quick tap, then long e)

Right: 礼（れい）'a bow / thanks' — one flick, don't hold it as an 'l'.

Key takeaways

Most consonants are close to English — you could get by with your native versions.
Every consonant carries a vowel behind it; only ん and っ can stand alone, each as its own beat.
The r-row is one light tap [ɾ], like the tt in butter. There is no l/r contrast to get right — just stop over-articulating.
ふ is a bilabial f [ɸ], blown through the lips with no teeth.
し, ち, つ, じ are palatal; the odd spellings shi/chi/tsu/ji record real sound shifts, and initial つ ("ts") is the one to drill.

Now practice Japanese

Reading grammar gets you part of the way. The exercises are where it sticks — free, no signup needed.

Start learning Japanese→