Everything the other pitch-accent pages teach — 雨 as あꜜめ, 箸 as はꜜし, the four accent shapes — is Tokyo standard (標準語). It is not the only system in Japan, and in some regions it is nearly the opposite. A learner who has grinded the Tokyo accents and then visits Osaka can have a small crisis: the melodies sound inverted. This page explains why, reassures you that the ground under your feet is solid, and shows you what the main regional systems actually do.
First, the reassurance. Tokyo accent is what NHK broadcasts, what accent dictionaries record, what teachers model, and what this guide uses. It is understood everywhere in Japan without exception. You will never be misunderstood for using it, in Osaka or anywhere else. Regional accent is something to recognize, not something you must switch to. If you have not met the accent shapes yet, read Pitch Accent Patterns before this page. Morae are marked H/L, with ꜜ at the downstep.
Kansai (Keihan) has one more dimension than Tokyo
The Kyoto–Osaka accent — 京阪式 Keihan-shiki, spoken across the Kansai region — is not just "Tokyo with different words." It is a richer system with an extra degree of freedom.
- Tokyo encodes one thing: where, if anywhere, does the pitch drop? A hard rule follows from having only that dimension — the first two morae must differ in pitch. A Tokyo word can begin L·H… (and maybe fall later) or begin H·L…, but it can never begin H·H. There is no word-initial high plateau.
- Keihan encodes two things: an initial register (式) — does the word start high (高起式) or low (低起式)? — and a downstep on top of that. Because register is independent, Kansai words freely begin H·H and hold a high plateau, something Tokyo structurally forbids.
That extra register dimension is the single most audible marker of Kansai speech to a Tokyo ear: whole phrases that "start up high and stay there," where Tokyo would have started low and climbed. It is also why the two systems can produce opposite melodies for the same word.
A word that flips: 川 (river)
The cleanest fully worked example of a Tokyo↔Kansai reversal is 川 (kawa, "river").
| Region | 川 alone | 川が (+ particle) | Pattern |
|---|---|---|---|
| Tokyo | かわ — L·H | かわꜜが — L·H·L | 尾高 odaka (drop after the word) |
| Kansai (Osaka) | かꜜわ — H·L | かꜜわが — H·L·L | 頭高 atamadaka (drop after mora 1) |
Tokyo starts 川 low and lets it fall only on the following particle; Osaka starts it high and drops immediately. The two melodies are essentially mirror images. Crucially, this is not random — it reflects the systematic register/downstep correspondences between the two dialects, which is exactly why a native speaker can hear one word and place you.
この川、昔はもっと水がきれいだったらしいよ。
kono kawa, mukashi wa motto mizu ga kirei datta rashii yo
Apparently this river used to have much cleaner water.
橋の下を大きな川が流れている。
hashi no shita o ōkina kawa ga nagarete iru
A big river flows under the bridge.
The famous 雨 / 飴 case — with an honest caveat
Learners are often told flatly that "雨 and 飴 are reversed in Kansai." The direction of the idea is right — the Tokyo pair does not map neatly onto Kansai — but the reality is messier than a clean swap, and it is worth being honest about.
In Tokyo, the contrast is sharp and stable: 雨 (rain) is あꜜめ (H·L), 飴 (candy) is あめ (L·H). In Kansai, these two do not line up the way a naive "just flip it" rule predicts, and among younger Osaka and Kyoto speakers the pair is actively merging toward a single melody. So rather than memorize a specific Kansai contour for 雨/飴, take the durable lesson: the Tokyo minimal-pair contrasts you learned are Tokyo facts. Do not assume they transfer. When in doubt about a Kansai melody, 川 above is the reliable model of the systematic reversal; 雨/飴 is the famous headline that turns out to have footnotes.
急に雨が降ってきたから、傘持ってる？
kyū ni ame ga futte kita kara, kasa motteru?
It suddenly started raining — do you have an umbrella?
のど痛いなら、この飴なめてみ。
nodo itai nara, kono ame namete mi
If your throat hurts, try one of these candies. (with a Kansai-flavored 〜てみ)
Accent as identity — the sociolinguistic payoff
Because the differences are systematic, accent works like an audible ID card. A Kansai speaker who moves to Tokyo is placed within seconds, and vice versa. Osaka and Kyoto natives are famously attached to their accent and rarely drop it, because it carries strong regional pride and warmth. This is genuinely motivating for learners: pitch is not busywork; it is the thing that makes you sound like you belong to somewhere, and native speakers read it instantly. It also means a learner who mixes Tokyo words with a stray Kansai melody sounds oddly patchwork — so pick one system and stay in it. Learn to recognize Kansai; produce Tokyo consistently. If you want to explore the vocabulary and grammar side of it, see Dialect Awareness and Dialects and the Standard Language.
Beyond Tokyo and Kansai
Japan's accent map has more than two colors:
- Accentless (無アクセント) zones. In parts of northern Kanto (Ibaraki, Tochigi), southern Tōhoku, and the Miyakonojō area of Kyushu, words have no fixed lexical pitch at all. Speakers there genuinely do not distinguish 雨 from 飴 by melody — and communication is completely unaffected, because context carries it. This is living proof that pitch, while real, is not load-bearing the way vowels are.
- Kagoshima's two-pattern system. In southern Kyushu, every word falls into just two melody classes, assigned per word, rather than Tokyo's per-mora downstep. It is a wholly different logic, not a variant of Tokyo.
- Kyushu and Tōhoku generally each have their own patterns, some closer to Tokyo, some not.
地元では、雨と飴を同じ音で言うんですよ。
jimoto de wa, ame to ame o onaji oto de iu n desu yo
Where I'm from, we say 'rain' and 'candy' with the same pitch, you know. (an accentless-region speaker)
関西の人は、標準語でもイントネーションでわかっちゃうよね。
kansai no hito wa, hyōjungo demo intonēshon de wakatchau yo ne
You can tell someone's from Kansai from their intonation even when they speak standard Japanese, can't you.
Common Mistakes
大きな川が町を流れている。
ōkina kawa ga machi o nagarete iru
❌ if said half-Tokyo, half-Osaka. Learned Tokyo かわꜜが (L·H·L)? Keep it — don't swap only 川 toward Osaka かꜜわ mid-sentence. A blended melody sounds patchwork.
雨が降っていますね。
ame ga futte imasu ne
✅ In your Tokyo baseline, 雨 stays あꜜめ (H·L). Keep it everywhere, including in Osaka — you'll be understood perfectly.
このお箸、高いね。
kono o-hashi, takai ne
❌ Assuming your Tokyo 箸/橋/端 contrasts hold nationwide. They're Tokyo facts; Kansai melodies differ. Recognize that, don't 'correct' natives.
関西弁、かっこいいから真似したいな。
kansai-ben, kakkoii kara mane shitai na
✅ Fine to admire and imitate — but learn Kansai as a whole system (accent + vocabulary), not by sprinkling one reversed word onto Tokyo speech.
Key Takeaways
- Tokyo/standard accent is universally understood; use it as your baseline without worry.
- Keihan (Kyoto–Osaka) adds a register dimension (high-start vs low-start) that Tokyo lacks, so Kansai phrases can begin on a high plateau — the giveaway sound of the region.
- 川 flips cleanly: Tokyo かわꜜが (odaka) vs Osaka かꜜわ (atamadaka). The reversals are systematic, so speakers place each other instantly.
- The famous 雨/飴 "reversal" is real in spirit but messy in fact, and even merging among younger Kansai speakers — don't over-memorize it.
- Some regions (northern Kanto, southern Tōhoku, Miyakonojō) are accentless; Kagoshima uses a two-pattern system. None of this impedes communication.
- Pick one accent system and stay consistent; recognize the others.
Now practice Japanese
Reading grammar gets you part of the way. The exercises are where it sticks — free, no signup needed.
Start learning Japanese→Related Topics
- Pitch Accent Minimal PairsN3 — Words spelled and segmented identically that Tokyo Japanese keeps apart by pitch alone — 雨 vs 飴, the 箸/橋/端 triple, 花 vs 鼻, 神 vs 髪 — and why context is not the whole story.
- The Four Pitch PatternsN3 — Standard-Tokyo words fall into four accent types — heiban, atamadaka, nakadaka, odaka — and the particle が is what tells odaka and heiban apart.