In the linguistics topic of language typology, Japanese is often included in lists of agglutinative (or agglutinating) languages, but when learning or reading about Japanese grammar exclusively this is rarely if ever mentioned. Other examples of agglutinative languages are Turkish, Finnish, Hungarian, and Basque.

Languages always considered agglutinative usually talk about things like lots of case inflections on nouns or lots of "slots" for various infixes and affixes in the potentially long endings of both verbs and nouns. Japanese in contrast usually talks about lots of particles and lots of verb endings only.

So is it true that Japanese is an agglutinative language and how should we regard this in relation to how we normally discuss Japanese grammar?

  • 3
    How two languages are traditionally taught may not correlate with their linguistic similarity--i.e. just because Japanese textbooks don't use the word "slots" doesn't mean that they don't exist.
    – Amanda S
    Jun 25, 2011 at 20:15
  • 1
    @Amanda: Indeed I would like to ask that question but I thought it prudent to ask this question before getting there. It does seem that Japanese textbooks take a traditional approach to Japanese grammar whereas linguistics textbooks might cover a broader range of different analyses, but those are harder to find in Wikipedia or by Googling (I've tried). Jun 25, 2011 at 20:20
  • 1
    One thing that obscures the issue is the mismatch of Western linguistic thought being imposed on Japanese linguistic phenomena. Compare "学校文法" (and its domestic theoretical basis) with, say, a Western university course in Japanese linguistics. The Western method focuses on tree diagrams extracted from a linear parsing, whereas the domestic method views the clause-end verb as the root and everything else as an equally-relevant optional addition. That interchangeability may be the equivalent of the "slots" you mention. Dec 24, 2016 at 20:03

4 Answers 4


Agglutinative languages are somewhat harder to understand than other categories. it's easy to see what the difference between synthetic fusional languages (like Latin or Russian) to isolating ones (such as Chinese or English): in isolating languages you only have words mixed with each other in various ways, but no morphology (or at least not very much of it, since no language is pure).

But why shouldn't we count agglutinative languages as isolating? If they just add suffixes (or prefixes, to be inclusive: Bantu languages such as Swahili and Zulu use prefixes more than suffixes for indicating things). You can treat every suffix as yet another word and then they're all just a few words bunched together, no different than in Chinese.

But the truth is that agglutinative particles aren't words. Unlike "pre"-positions, they can't stand alone and they can't put themselves wantonly in both sides of words ("understand" vs "stand under") or entirely removed from the word they describe ("Which city do you live in?"). In fact, agglutinative particles are so tied to the "full" words they are attached to that they really are, for all intents and purposes, affixes. In most agglutinative languages, they are indeed written as prefixes and suffixes (i.e. there's no space between them and the words they describe) but in Japanese romanization (especially when targeted toward foreigners) they are not. This is mostly due to a westernized perception of Japanese in my opinion.

For the record, the modern language most similar to Korean, which is given as an example for slot, is most probably Japanese. They both have a verb conjugation strategy which can be explained through slots (though it's arguably simpler in Japanese than in Korean), but you won't find the word "slots" used in textbooks in neither of these languages (at least not the ones I've read).

Actually, "slots" is just a more disciplined way to view what is essentially a standard order for suffixes. In Japanese, as well as in Korean, certain things have to come before others. If your verb specifies negation for instance, it should definitely come after volition (~たい), but also before the past tense marker (~た). If you fulfill every possible category you can also count slots in Japanese, but it's less useful than in Korean, since in Korean some suffixes that are not attached directly to each other still need to come together in some of the cases, so it might make talking about slots more helpful, even for students (but it isn't - most Korean textbooks I've read are quite bad at making things simple :/)

Deep inside, Korean uses verbal bases, just like Japanese, and each suffix is attached to one of these verbal bases and can be conjugated to any of them (so the next suffix could be attached to it). The same thing can be said about Japanese, where the auxiliary verb だす is attached to the I-base of verbs (e.g. kaki for kaku) and the auxiliary verb あげる can only be attached to the TE-form. They both have further bases of their own, of course. Some other suffixes (such as ~ない and ~たい) behave more like i-adjectives, but as we've seen before, i-adjectives are, in fact, verbs.

Case inflections on noun are actually quite rare. As far as I know, the only major (or perhaps we should say: commonly mentioned) agglutinative languages to ever have cases are Finnish and Estonian. Some languages, such as Turkish, Hungarian or even Japanese are claimed to have cases, since there are particles that mark things like direct object (accusative) or the instrument of a verb (instrumental). These should only be called cases, however, if they are used for noun agreement, especially with adjectives agreeing to their nouns. In Latin they do (compare De Novo Mundo about the new world to Novi Mundi of the new world), and in Finnish they also do. Japanese is obviously different, but the same can be said for most other agglutinative languages.

In fact, there is probably no single grammatical category (be it case, gender, tense, person, number, volition, negation or whatever) that is done through agglutinative suffixes in all agglutinative languages. Being an agglutinative language has nothing to do with what is expressed through agglutination, but rather with how it is expressed.

  • Hmm I have always read of Hungarian as having many cases just like Finnish. I'm also finding a number of pages counting many Japanese particles as case markers, but going into more depth unfortunately. I don't find much of what you have to say about adpositions to match the accepted views. Rather than a "standard order" of affixes I'm sure I'd only heard of "rigid order" of affixes as part of the definition of agglutination. Earlier today I read that Estonian was diverging from Finnish to the point of becoming an isolating language. I'm interested to read more tying "case" to "agreement". Jun 25, 2011 at 20:36
  • 1
    @hippietrail: It's often said about Hungarian, but Hungarian grammatical tradition was probably highly influenced by Latin. The fact is that adjectives in Hungarian don't agree by case, while Finnish adjectives do. The Hungarian adjectives may have done the same thing in the past (since both languages are part of the same Finno-Ugric family), but I don't know enough about these languages to tell.
    – Boaz Yaniv
    Jun 25, 2011 at 20:41
  • 1
    And more about case: one of the most confusing things here is that some generative linguists use "case" an entirely different (and separate) sense from "morphological case". I'm not strong on generative linguistics (frankly, it kinda bores the heck out of me :)), but from what I understand those cases really refer to the syntactical role of a noun phrase (NP in generative lingo) in the sentence. This is similar to what a morphological case usually does, but for the generative syntacticians they exist in every language - they are just marked differently.
    – Boaz Yaniv
    Jun 25, 2011 at 20:49
  • 1
    @hippietrail: I mostly judge on what I know. The only agglutinative languages I know to a satisfying degree (to make some assumptions about them) are Japanese, Korean, Swahili and Quechua (and I know Japanese far better than the other three :)). It might make a my view a bit biased, but that would be true for just about anyone. Anyway, linguists usually tend to have many different opinions on the same subject, though when you inspect things closely, you'll often find the difference is not so huge.
    – Boaz Yaniv
    Jun 25, 2011 at 20:53
  • 1
    @hippietrail: Concerning Hungarian, the use of "case" in Hungarian grammar could be challenged as Boaz said because the inflexion is not aligned on epithets, except in minor situations, but it is not completely wrong and has been internalized by speakers. Hungarian has changed a lot during the last nine centuries, notably verb grammar, and has been heavily influenced by latin, german, slavic languages and turkish. Speaking of Georgian: really another ball game altogether. The agglutinate aspects are just one part of it. Good luck to you.
    – ogerard
    Jun 25, 2011 at 22:11

According to the Wikipedia page on Inflection:

Japanese shows a high degree of overt inflection of verbs, less so of adjectives, and very little of nouns, but it is mostly strictly agglutinative and extremely regular. Some fusion of morphemes does take place (e.g. causative-passive -sare- as in ikasareru "is made to go", and non-past progressive -ter- as in tabeteru "is eating"), but this is rare. Formally, every noun phrase must be marked for case, but this is done by invariable particles (clitic postpositions). (Many grammarians consider Japanese particles to be separate words, and therefore not an inflection, while others consider agglutination a type of overt inflection, and therefore consider Japanese nouns as overtly inflected.)

So while Japanese is primarily agglutinative, it is not purely so. It is worth noting that no language is purely any one type, as languages are constantly evolving (and doing so in a cycle, e.g. of synthetic⇔analytic inflection1).

  • The points on nouns, particles, clitics, postpositions is interesting and compliments the other answer by Amanda S nicely. The part about agglutination and overt inflection is confusing though. And the part about adjectives being less overtly inflected may inspire another question. Jun 25, 2011 at 19:47
  • I wonder if "less so of adjectives" refers to i-adjectives inflecting - like verbs and na-adjectives not inflecting - like nouns? Jun 25, 2011 at 19:59
  • @hippietrail That is likely one aspect of it, yes.
    – rintaun
    Jun 25, 2011 at 20:01
  • 3
    @hippietrail: nouns in Japanese are not internally inflected, but they are agglutinated. -no, -mo, -de, -ni even the copula are all really just suffixes, and they form a single unit with the nouns they follow. The only difference between nouns and verbs is that nouns have just a single base to which all these suffixes attach, while verbs have several different bases (such as Mizenkei, Rentaikei, etc.)
    – Boaz Yaniv
    Jun 25, 2011 at 20:24
  • 1
    @Boaz: Can multiple of such suffixes be added to a single noun? Is the order rigid? What is the most that can be added? Jun 25, 2011 at 20:38

From the Examples of agglutinative languages section of the Wikipedia page on Agglutination:

Japanese is also an agglutinating language, adding information such as negation, passive voice, past tense, honorific degree and causality in the verb form. Common examples would be hatarakaseraretara (働かせられたら), which combines causative, passive or potential, and conditional conjugations to arrive at two meanings depending on context "if (subject) had been made to work..." and "if (subject) could make (object) work", and tabetakunakatta (食べたくなかった), which combines desire, negation, and past tense conjugations to mean "(subject) did not want to eat".

In short, Japanese is considered an agglutinative language because what other languages express with helping verbs and other additional words (was not red, could not be made to eat), Japanese expresses with suffixes that are added to the word root (赤くなかった, 食べさせられなかった).

Nouns are usually not counted as agglutinative because they are conjugated with the copula and not with a suffix, but if one considers particles to be noun suffixes and not separate words, one could make the case that nouns are also agglutinative, especially since some particles can be agglutinated (私にも, 僕だけには).

  • Hmm it's not totally convincing since it doesn't break the verb endings into pieces which fit into slots and uses the word "combine" which seems more apt for describing the endings in inflectional languages. Your own comment only contrasts Japanese with isolating/analytic languages and not with inflected languages which is the usual contrast to agglutinating languages. Also no mention is made of whether or not nouns are also agglutinative... but I guess it's a good start until we get an answer from a linguist. Jun 25, 2011 at 19:21
  • This answer doesn't explain why the suffixes shouldn't be considered independent words, in which case Japanese could be called analytic! Jul 13, 2011 at 22:32
  • The reason is that nothing can come between them: たべさせられ早くなかった does not make sense. Also, say A says もう食べたかった? and B says いいえ、たかくなかった, it doesn't make sense, proving たい cannot be a separate word.
    – ithisa
    Aug 29, 2013 at 15:46

In some agglutinative languages, such as Hungarian, there are not just suffixes, but pre-fixes and post-fixes. All may be used in the same word expressed. The spoken language is not according to rule, but the intent what a person want to express. Some cases an entire sentence can be expressed by a single word. Yes, some words can be indeed very long, or a lot can be expressed by a short single word, that are formed by agglutination. That is why it's hard to learn it. Unless picked up at a very early age. The expressed or spoken language formed by the mind of the speaker, not by following any set rules..

  • What do you mean by "not just suffixes ... and post-fixes"? Your description doesn't seem to be particularly about agglutinative languages, either... Dec 25, 2016 at 7:26
  • Georgian also has circumfixes. I believe there are languages which also have infixes and even interfixes. Dec 26, 2016 at 5:23

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .