Phonology Accent of the Japanese verbs" />

Phonology Accent of the Japanese verbs

Accent of the Japanese verbs

The accent of the verbs is very simple. Basically, there are only two patterns: accented and unaccented, and they have nothing to do with whether a verb is godan or ichidan.

Reminder about verb groups


The radical of ichidan verbs ends with a vowel and its infinitive (or dictionary form) is made up of radical+ru.

For example:
tabe (to eat)  tabe-ru tabe-nai tabe-masu

mi (to see)  mi-ru mi-nai mi-masu

The radical of godan verbs ends with a consonant and its infinitive is made up of radical+u.

For example:

hanas (to speak) hanas-u hanas-anai hanas-imasu

nobor (to climb) nobor-u nobor-anai nobor-imasu

In Japanese linguistics, it is often considered that the radical of verbs like “kau” or “omou” is actually “kaw” and “omow”, that is to say that they end with a “w”.
In regards to the number of moras

Unaccented verbs

No matter how many moras there are, the first mora is low while all the other moras are high like this: xXXX.

Accented verbs

2 moras Xx
3 moras xXx
4 moras xXXx
5 moras xXXXx

As shown above, the common point of accented verbs is that the accent always comes on the penultimate mora, i.e. the second mora starting from the end.

To sum it up、the accent of a verb has absolutely no link with its length (number of moras) nor with its conjugation (ichidan or godan), a verb can only be either unaccented or accented and even if it is accented the accent can only come on the penultimate mora!

If we look at it from a different point of view, when upon hearing a verb, if Japanese native speakers think “Ah, that’s a verb!”, it is not just because it ends with a “u” but also because it follows one of the two accent patterns above.

For example,

it is because verbs like “utau” or “nemuru” become respectively

uTAU and neMURU

while verbs like “hoeru” and “hikkaku” become respectively

hoEru and hiKKAku,

that they are identified as verbs.

On the other hand, words like “saikokineshisu” (psychokinesis) and “sutemitakkuru” (Double-Edge) may end with a “u” like verbs, however since they are pronounced as

SAIKOKIneshisu and suTEMITAkkuru

with the accent on the preantepenultimate mora (i.e. the fourth mora starting from the end) they are not identified as verbs.

Hypothetically speaking, if they were pronounced as


they would feel kind of verb-like to native speakers.

(But even if they are pronounced unaccented as SAIKOKINASHISU and suTEMITAKKURU, they don’t feel like verbs. There are also words like “chôhatsu” (provocation) and “meisô” (めいそう, meditation) that end with “u” and what’s more they are already unaccented but I can only see them as nouns. I don’t know why. Actually, if someone knows please tell me why‼)

However, there are some counter-examples such as “kaeru” (to come back) that gives KAeru, but it is because of the diphthong “ae” which is a special mora.
Special moras are made up of the second vowel in a diphthong, the first part of a geminate consonant (written with a small っ, also called “sokuon”), the second part of a long vowel (often written ー) and the nasal mora (written ん and also called “hatsuon”). These special moras cannot be accented and so, should the accent fall on them, it gets advanced to the preceding mora.
On the other hand, the ichidan verb “kaeru” (to change) is unaccented and so there is no change because of the diphthong: kaERU.



Even if a verb can be used on its own, it can also be followed by an auxiliary or an adverb. In this section we will talk about how the accent changes in these cases.

Basically, with a form like verb+suffix, the accent is decided by the suffix.

The accent pattern of every suffix has to be learnt separately but accented and unaccented verbs both follow the accentuation rule of the suffixes.

In this section we will also consider flexion to be a suffix just like auxiliaries and adverbs.

Let’s have a look at the following examples.

Unaccented verbs

korogaru (koROGARU) 〈unaccented〉

No matter the length of an unaccented verb, they will behave like the above example.
Let’s have a look at other suffixes.

With the auxiliary “reru/rareru”, unaccented verbs remain unacccented: koROGASARERU.

So here we have a rule that states that when the auxiliary “reru/rareru”is affixed to an unaccented verb, it always results in the unaccented forms RERU and RARERU.

Other auxiliaries and adverbs all have their own accent pattern. But that specific accent pattern will always be the same no matter how long the verb is.

Accented verbs

kamitsuku (kaMITSUku) 〈accented〉

Like for unaccented verbs, accented verbs all follow the same pattern no matter their number of moras.
If we add the auxiliary “reru/rareru” to an accented verb, we get kaMITSUKAREru.

That’s because when suffixing “reru/rareru” to an accented verb it becomes REru and RAREru.

As it was said above with unaccented verbs, depending on the suffix the accent pattern changes however the same suffix will always have the same accent pattern no matter the length of the verb.

What I want you to pay extra attention to is that what decides the accent pattern here is the suffix rather than the conjugation and type of the verb: with “nai” unaccented verbs stay unaccented and accented verbs are accented on the third mora from the end; with “masu” both accented and unaccented verbs are accented on the last but one mora; with “te” unaccented verbs stay unaccented and accented verbs are accented on the third mora from the end; etc…

Introduction to the accent patterns of different suffixes

I will now introduce the different existing accent patterns for suffixes.
(The accent pattern depends of course on the suffix but this does not mean that it depends only on the suffix. The accent of the verb is not always dominated by the accent of the suffix.)

In the following section suffixes are separated in three groups. For each group I will explain both accented and unaccented patterns.

By the way, this classification is adapted from the one given by the accent dictionary of the NHK.

After comparing the NHK and the Shinmeikai accent dictionaries, I thought that the explanations given in the NHK one were easier to understand so I used this one as base for this section.

① Suffixes with a weak accent

GA, kaRA, KEredo(mo), SHI, SHIka, ↑t↓te, TO, NAra, NOda, NOde, NOni, WA, BA, daKE, hoDO, TA, DA, TE, DE, NI
※some can be suffixed to the dictionary form (shûshikei) or to other forms, but the form of the verb has no influence on the accent pattern.
・when the above suffixes are attached to a verb, they become weak.
・the accent of the verb is prioritized.
・many of them are one mora long or unaccented.

☆With unaccented verbs

With the suffixes
GA, kaRA, KEredo(mo), SHI, SHIka, ↑t↓te, TO, NAra, NOda, NOde, NOni, WA, BA
the accent comes on the last mora of the verb:

(ex) neMURUKAraneMURUkara

※But the suffixes daKE, hoDO, TA, DA, TE, DE, NI are exceptions and when they are attached to a verb, they become unaccented!
…XX+●● → …XXXX


☆With accented verbs

The above suffixes follow roughly the rule below when attached to accented verbs:



Only the accent of toKEru is used.

In short, the accent pattern of the verb is preserved.
The accent of the suffixes is so weeeeak!

When “te”, “de”, “ta” or “da” are attached to an accented ichidan verb, the position of the accent is shifted by one mora.
That’s probably because since they are directly attached to the radical, the number of mora of the verb changes.

That is to say that after deleting the “ru” suffix, a verb that used to be 3 mora long becomes 2 mora long.
So it may not be appropriate to say that the accent of the verb stays exactly the same.

★Ichidan verbs
…XXx+● → …Xxxx
toKEru + ta → TOketa

Instead of saying that “the accent of the verb remains the same”, it is better to say that “the position of the accent of the verb is preserved”.


★ Godan verbs
……XXx+● → …XXxx
iBAru + ta → iBAtta

Since the last mora changes instead of being deleted, the number of moras doesn’t change and the rule given at the start of this subsection can be applied as is!
Right now, clever readers may be thinking “If we just summarize it as the accent coming on the penultimate mora, we can explain everything so far. Happy end, yay!”. However. Then it would be difficult to explain cases where several suffixes are attached to a verb. For further details see section 5.2 (The case of ubawaretanowa)

② Suffixes with a strong accent

GUrai, KUrai, SOoda, DOkoroga, BAkari, MAde, MItaida, YOoda, YOri, RAshii, TAra, DAra, TAri, DAri, daROo, deSHOo

They are all at least two mora long and accented, the elite of the suffixes.

☆ With unaccented verbs

With the exception of daROo and deSHOo, they all follow the pattern below.

The accent of the verb is obliterated and the accent of the suffix reigns without mercy.

If the accent is on the penultimate mora (second mora starting from the end) of the suffix, then the overall accent is also on the penultimate mora.

XX+Xx…→ …XXXx

neMURU + MAde → neMURUMAde

If the accent is on the antepenultimate mora of the suffix, the overall accent will also be on the antepenultimate mora.

XX + Xxx → …XXXxx

neMURU + BAkari → neMURUBAkari

※However, for daROo and deSHOo, the pitch tends to fall twice:

XX+ xXx → …XXxXx

neMURU + daROo  neMURUdaROo

If the pitch falls twice making it sound like there are two accents, it’s because the verb and the suffix don’t get parsed as one word but as two and there is an initial pitch rise on the suffix.

☆ With accented verbs

When the above suffixes are attached to an accented verb, both accents remain.


Xx + Xx→…XxXx

toKEru + MAde → toKEruMAde

XxXxx → …XxXxx

toKEru + BAkari → toKEruBAkari

③ Special cases


I will now introduce suffixes that don’t follow any of the previous rules.

★ Types where the verb is stronger!

“senai”, “sasenai”, “seru”, “saseru”, “sooda”, “tai”, “nagara”, “reru”, “rareru”

A, Unaccented → Unaccented

neMURU+ seru → neMURASERU

B, Accented → Accented (accent on the penultimate)

toKEru + seru → toKESASEru


★ Types where neither is stronger!

Accent of verbs followed by the negation suffix “nai”


A, NAi (final form, shûshikei)

Unaccented → Unaccented


…XX+●● → …XXXX


Accented → Accented on the last mora of the verb

Xx+●● → …XXxx

toKEru +NAi → toKEnai


B, Nai+suffix

Accented verbs follow the pattern in A.

When followed by NAide, NAkatta, NAkute, NAkereba


◎ Unaccented verbs

neMURU + NAide → neMURANAide

The accent comes on the first mora of the suffix: right on the “na”!


◎ Accented verbs

Xx + Xxx → …XXxxx

toKEru + NAide → toKEnaide

The accent comes on the last mora of the verb: just before the “na”!


★ Types where the suffix is stronger!

The suffix annihilates the accent of verb and its accent is the only one to remain.

Oo, YOo, TAkatta, TAGARAnai, TAGAru, TAKUte, TAkereba, TSUtsu, NASAi, NIkui, MAshita, MASHOo, MAsu, MASEn, YASUi


☆ Unaccented verbs

XX +●● → …XXXx

neMURU + Oo → neMUROo

☆Accented verbs

Xx + ●● → …XXXx

toKEru + YOo → toKEYOo

Accent of compound verbs

The accent of compound verbs is extremely simple. It’s really stupidly simple.


The first verb is the one to seize the power!

When the first verb is unaccented → the accent goes on the penultimate mora (the pitch is high until the last mora)
When the first verb is accented → the compound is unaccented
The compound takes the opposite pattern of the first verb. What a tsundere.

First verb is unaccented

As I said earlier, the compound is accented on the penultimate mora.


…XX + …XX → …XXXx


…XX + …Xx → …XXXx

suU + TOru → suITOru

When the first verb is accented


As I said earlier, the compound is unaccented!


However, that is true mostly for elderly people. Nowadays younger generations tend to put the accent on the penultimate mora even when the first verb is accented!


Accented × Unaccented


Old pattern
Xx + …XX → …XXXX

tsuKEru + aGARU → tsuKEAGARU

New pattern
Xx + …XX → …XXXx

tsuKEru + aGARU → tsuKEAGAru

Accented × Accented


Old pattern
Xx + …Xx → …XXXX

KIru + SAku → kiRISAKU

New pattern
Xx + …Xx → …XXXx

KIru + SAku → kiRISAku

To put it simply,

younger generations pronounce all compound verbs as accented.
You just need to remember that compound verbs are unaccented only for older generations when the first verb is accented. Also, for noun+verb compounds and adjective+verb compounds, the accent is on the penultimate mora most of the time!



I’ve been thinking of writing about the accent of the verbs for a long time but I could never find the motivation to. But recently, I just happened to get asked by a reader about it so it was the occasion to address it in this blog.

Here’s the question I was asked:

In cases such as

“iiwatasaretekara”, “ubawaretanowa”, “kuzureteikimasu”

where there are several suffixes next to each other, what happens to the accent?


In the case of “iiwatasaretekara”

It is made up of iu+watasu+reru+te+kara.

The compound verb iiwatasu is made up of

unaccented iU and unaccented waTASU so it is accented: iIWATAsu.

Then following the rules given above, with the suffix reru the accent comes on the penultimate mora (in this case “re”):


Then for “te”, since iiwatasareru is accented, we follow the rule for accented verbs and put the accent on the penultimate mora of the radical of the verb (in this case “sa”):


Finally, when adding “kara” we get


with the accent at the same place as iIWATASArete.

That is because like I said earlier, with “kara” the accent of the verb is protected.


In the case of “ubawaretanowa”

uBAu (accented) + REruuBAWAREru

For accented verbs, the accent comes on the antepenultimate mora when suffixing “ta” (here it comes on the “wa”).


There was nothing in my sources about “nowa” but I think it’s okay to think of it as one of the weak accent suffixes. And so, we get the following accent:


(If it was with an unaccented verb like aBARERU for example, we would get aBARERUnowa)

※I wrote in section ① that saying that the accent comes on the penultimate mora with weak suffixes in accented cases was a bad idea because then cases where there are several suffixes would become hard to explain. This is exactly the case here. In this example, when “nowa” follows “ubawareta”, the accent comes on the antepenultimate mora of the verb radical and NOT on the penultimate. This also happens with other suffixes such as “kara”: uBAWAretakara is accented on the “wa”.

However, if instead of “ta”, it is the nonfinal tense uBAWARErunowa, the accent comes on the penultimate mora of the verb. This is why I think it is better to say that the respective accents of uBAWAreta and uBAWAREru remain.

The case of “kuzureteikimasu”

It is made up of kuSUREru (accented) + te and iKU (unaccented)+masu
And so, the accent comes on the penultimate mora for both of them:



Then all we have to do is to juxtapose them:


Since there are two accents, you can just think that the two verbs are following each other rather than being one compound.




NHK新アクセント辞典 新版