Talk:Agglutinative language

Active discussions
WikiProject Linguistics / Theoretical Linguistics  (Rated Start-class, High-importance)
This article is within the scope of WikiProject Linguistics, a collaborative effort to improve the coverage of linguistics on Wikipedia. If you would like to participate, please visit the project page, where you can join the discussion and see a list of open tasks.
Start-Class article Start  This article has been rated as Start-Class on the project's quality scale.
 High  This article has been rated as High-importance on the project's importance scale.
Taskforce icon
This article is supported by Theoretical Linguistics Task Force.

Article titleEdit

Shouldn't this be agglutinative language? --Chuck Smith

-- Probably yes. 'Agglutinative language' language gives 800 hits in Google, while 'agglutinating language' only gives 200 hits. --HJH

Len I didn't miss that; but Wikipedia is not authoritative, I took the time to read the other hits Mintguy

Take the time to minor in Linguistics and learn about the various types of polysynthetic languages, like me. The error you are seeing on the web is the near-ubiquitous one of mistaking "agglutinative" for a synonym of "(poly)synthetic". It is not. (you might find the article on fusional language, skeletal as it is, rather instructive.) --User:LenBudney

I see that German is given as an example of an Agglutinative_language and also as an example of a fusional language. Which one is it? The text implies that if it's one then it can't be the other. Or can it? --Zoicon5

I see that each type also claims Esperanto as an example. -- Zoicon5

I've moved German and Esperanto to the back, saying 'and to a lesser extent'. More seriously, I've changed 'polysynthetic' to 'synthetic'. The term 'polysynthetic' refers to languages like those in North America that incorporate both the noun and verb into a single phonological word. The term 'agglutinative' covers two things: (a) multiple morphemes in the same word, and that's in core inflection of nouns and verbs (so a few English examples like kind+ness+es don't count); and (b) each morpheme phonetically separate, as opposed to other synthetic types. In fact, the core morphology of German is fusional: -es is gen.+sg.+m/n., so it really shouldn't be in the list. Gritchka 17:43 25 Jul 2003 (UTC)~

I would like to agree with you (Gritchka) about the idea that 'polysynthetic' => noun incorporation. However, several months ago I had a conversation about this on a mailing list, and at least etymologically (going back to Sapir, I think) 'polysynthetic' simply means "extremely agglutinative" or "containing in a single word what would be complete sentences in other languages." Of course the latter is vague (you can arguably do this in a pro-drop fusional language), but the general idea is that there are a whole lot of morphemes in a single word. And this seems to be a (not necessarily the only) widespread idea of what 'polysynthetic' means. (I've actually never heard the term 'synthetic' used this way, but maybe I have just overlooked it.)
Mcswell 22:55, 22 March 2006 (UTC)


Could we have some examples of this type of formation in action? I've read the article, but I still have only a vague idea of what this is!

Here are some examples in some languages I'm familiar with:
ka:?a:?a:kó:khá:r?a (Wichita)
(ka:? - a:? - a: - ki - wakhahr - ?a) (six morphemes)
intensive - quotative - third person subject and preverb - aorist - patient is an activity - come
long ago it came to pass that... (used to start a story) (Wichita is basically agglutinative; the differences between the morphemic gloss and the surface form are governed by a few basic phonological rules.)
aqhjazbacr'aghawdætwaaylafaq'ayt'madaqh (Ubykh)
(a-qhja-z-bacr'a-gha-w-dæ-tw-aay-la-fa-q'a-yt'-ma-da-qh) (16 morphemes)
them - benefactive - me - under - ablative - you -causative - take - again - exhaustive - able - past - imperfective - negative - conditional - optative
if only you had not been able to make him take it all out from under me again for them

thefamouseccles 00:37 13 Oct 2003 (UTC)

I'd like to give an example from Turkish, a word primary-school kids like to play on:

Çekoslovakyalılaştırabildiklerimizden misiniz?

This is one word, and one sentence, actually, meaning: are you among the ones whom we were able to turn into a Czechlovakian? (politically outdated, though)

Çekoslovakyalı[Czechoslovakian] + -laştır[turn into] + abil[ability = to be able to] + -dik[past tense = were] + -ler[plural suffix for the object -> so that it becomes first person plural; "we" namely] + -imiz[this doesn't mean "we" actually, but indicates that the object performing the action is "we", first person plural] + -den[among] + -mi[question suffix] + -siniz [are you? -> makes the subject the second person plural]

The longest known word in Finnish (and possibly in the whole world) is 103 letters long:


As its creator, professor Artturi Kannisto, has died, nobody has been able to actually interpret its meaning -- or even tell how many morphemes there are in it. However, it is definitely a real word.

Um, if no one knows what it means, or even what the morphemes are, how can we be sure it's a real word? I'm not slagging, I'm seriously curious. I could just as easily say that freemjaboodinalitatiousnessism is a new long word, and as no one could say what it means, then by the logic apparent above, it would be a "real" word...? Or am I missing something in your argument?
Also, not to rain on any parades, but my (probably limited) understanding was that the longest words on record were chemical names. Due to the complicated atomic structure of some rather large molecules, it becomes possible to come up with some truly ridiculously long formal names for these compounds (such as that listed here at, sporting 1,185 letters).
In that case, the parts (individual names of the chemicals) sould be considered separate and separately they are not so long. Also in some languages like Finnish or German, the spaces from between words that describe a single object are left out (like "ferry boat deck paint" -> "ferryboatdeckpaint") but these are not considered to be long words as in the article, since there is no upper limit on how long a word you could form this way.

No, the spaces aren'T "left out". English has those spaces, German has not. In fact, the ability to form compound nouns is rsther unique to German since you can also change adjectives and Verbs to fit into the word.

Cheers, Eiríkr Útlendi 23:21, 14 October 2005 (UTC)
Hi, I am a native Finnish speaker, and to me it looks sort of intelligible, but definitely it's too complex a compound to understand its meaning. It is a synthetic word created by a professor, so it's probably grammatically correct, but quite impossible to understand completely as the structure of the word is too complex to be understood all at once. Furthermore, grammatic shift makes it increasingly difficult to even break the compound down into all of its parts cohesively. The word begins with "kumar-", the stem of "kumara" which means bent. It's used as a verb "kumartaa" which means to bow. Now, if that would be agglutinated into "kumarretuksi" that would mean "to become bowed" (meaning the object (probably a person) becomes someting that people bow in front of). The word goes on to characterizing that this bowing-of is a tendency of the object (like one's personality). The middle part of the word twists the attibute further, revolving around the theme that the tendency is characteristic of the object person and that the object person might be prone to it. This part also negates the tendency quite a few times until it becomes quite utterly unclear (to me) as to whether the object has this tendency or lacks the tendency. All in all, the middle part is really difficult to understand. The end part ("...ttelemattomammuuksissaansakaankopahan") is quite clear, however, and it describes a wondering if an accomplishment (which would be defined in the next word) of the object person has been due to a (momentary or nondeliberate) state of lacking the tendency, further noting that the speaker does not quite firmly believe that it would be so. That's what it looks and feels like to me, anyway, but please correct me if I am wrong :) Panu-Kristian Poiksalo (talk) 08:33, 26 April 2011 (UTC)

Examples that illustrate the agglutinative structureEdit

It would be very helpful to include an example from each language which shows the full structure of its agglutination. By this I mean which agglutinative morphemes can appear in which positions.

It would be nice to have one example for a verb, and one example for a noun.

Some languages are regularly stated to be agglutinative but their agglutinative properties are seldom discussed. Japanese and Korean fall into this category. Illustrative examples for them will be valuable.

Some languages can be analysed in multiple ways with one way being more accepted, or different ways accepted by different groups. For instance, Hungarian is considered to have cases by some, and considered to have nouns built from individual particles by others. Japanese verbs are usually described as conjugating but the conjugations can be broken down into functional pieces. Japanese nouns are pronounced with their following particles attached but described in terms of the particles being separate words. It would be good to show the more agglutinating analysis here, perhaps with a note. — Hippietrail 13:38, 12 Feb 2005 (UTC)

The postposed particles in Japanese are indeed part of the same phonological word, but they're not agglutinative morphemes; they're clitics. Note also that the copula is also a clitic (I mean anything like da, desu, deshou, etc.). I don't know enough Japanese to propose good examples, but I guess I can chip in with tabetakunakatta: tabe- "eat", -ta- "want", -ku adverbial operator, -na(i) negative, -katta past tense mark for adjectives = "did not want to eat". --Pablo D. Flores 11:03, 14 Feb 2005 (UTC)
Thanks for the example. I have the Finnish structure for nouns, verbs, and participles downstairs. I think I have the Basque structure too. I'll try to post them soon.
A couple of questions, am I right in thinking the difference between the enclitics and agglutinative morphemes is that the latter would only be able to join specific word classes (nouns or verbs but not both) whereas the former can join onto any kind of word whatsoever?
Also, can clitics be added in arbitrary order or does the degree of freedom vary by word class and by language? Do you know the answer at least in the case of Japanese? — Hippietrail 12:18, 14 Feb 2005 (UTC)
A clitic is more or less defined as a morpheme that behaves syntactically like a separate word, and phonologically like a bound morpheme. That is, a clitic appears in places where you would expect a separate word, but sounds as if it were an affix (which it isn't). An affix is applied to specific word classes, as you say, while a clitic may be attached anywhere (well not anywhere, but the distribution rules are not as rigid). For example, the articles in Romance languages are proclitic: Spanish un beso "a kiss" is phonologically one word. But you can add an adjective in between: un apasionado beso "a passionate kiss". The article just cliticizes on the first word it encounters.
As for the rules: I'm sure no language allows clitics complete freedom of order (it's the same with words). In Japanese the clitics are the postpositions, the copula and some other particles, and they all come after the rest. They're not distributed arbitrarily. Suppose this (clumsy) example:
Kore wa anata no desu ka?
Phonologically those are two words: kore-wa and anata-no-desu-ka. Neither can be ordered differently: the genitive always follows the noun/pronoun, the copula or the verb comes at the end in declarative sentences, and if the sentence is an interrogative one, you add ka after the verb. --Pablo D. Flores 12:11, 15 Feb 2005 (UTC)

Number of Japanese irregular verbsEdit

At the bottom of the page, there is the statement:

"Agglutinative languages tend to have a high rate of affixes/morphemes per word, and to be very regular. For example, Japanese has only two irregular verbs, and Turkish has only one."

I believe that there are three irregular Japanese verbs:

"suru" (to do) has the stem form "shi" -- if it was regular, it would be "su" or "suri"

"kuru" (to come) has the stem form "ki" -- if it was regular, it would be "ku" or "kuri"

"iku" (to go) has the gerund form "itte" -- if it was regular, it would be "iite"

I suspect that the two noted verbs in the article are "kuru" and "suru". Is there any reason that "iku" was left out?


Some analyses of Japanese grammar don't consider "suru" to be a true verb but rather an "auxiliary". So maybe that writer was only considering "kuru" and "iku" — Hippietrail 22:59, 29 May 2004 (UTC)
Suru is an independent verb, not only an auxiliary. In the sentence "Nani o shimashou ka?" ("What shall we do?"), it is plainly not an auxiliary, as there is no other verb. It's much more likely that the verbs the author had in mind are "kuru" and "suru", because these are the only two verbs that are not classified as ichidan or godan verbs in traditional Japanese grammar. Iku is definitely irregular, though, as is "aru" (its negative is "nai", rather than the expected "aranai"), and there are other irregular verbs once you consider honorific speech (e.g., the imperative of "kudasaru" is "kudasai", the masu form of "gozaru" is "gozaimasu" and not "gozarimasu", etc.). - furrykef (Talk at me) 10:28, 4 September 2009 (UTC)

Can it really be said that Japanese has only 3 irregular verbs? That is based on the idea that the verb です - basically, "to be" - is not a verb, rather a thing that is added to end of a sentence that lacks a verb, even though it acts like a verb. An IRREGULAR verb. And there are also several other verbs formed from a noun + one of those three irregular verbs, and so they too are irregular. elvenscout742 22:55, 5 Mar 2005 (UTC)

I concur that some consideration should be given for what to do with the Japanese ending copula da / desu. However, I would not go so far as to include compound verbs formed by nouns + irregulars. On the one hand, the list gets ridiculously long pretty quickly, and on the other, they are after all compounds. Though 勉強する benkyō suru translates roughly as "study" or "to study" in English, the basic morphemes in Japanese are quite separate and distinct -- benkyō, and suru. One may as well advocate lumping together similar constructs in English as single verbs -- but then "dothelaundry" doesn't seem to work very well. :) --- Eirikr 05:42, 6 Apr 2005 (UTC)

The wiki Irregular Verbs page, which links from this page, gives the number of Japanese irregular verbs as + or - five, not the two mentioned in this article. Should a footnote--or more information--be added on either or both pages to reconcile/explain this discrepancy? For example, I found the information in this discussion to be extremely helpful. Perhaps an edited version could be included on the wiki itself, not just in the discussion? --Sabishii 06:55, 22 October 2007 (UTC)

Irregular verbs in TurkishEdit

It says there is only one. i'm a native turkish speaker but i haven't heard of it yet. could somebody tell me?

I just had a quick look with Google and found this site which lists at least 14 irregular verbs in Turkish! This seems to be restricted to the aorist. Here is another site which goes into more detail about to be, which is probably what the writer on this article had in mind. — Hippietrail 07:14, 6 Oct 2004 (UTC)

I have been to the website, but I do not see why they are irregular (the site which lists the supposedly 14 irregular verbs). I have consulted with over 5 teachers of Turkish in the United States and Europe, and they claim that Turkish does not have any irregular verbs. That is why I corrected the article, saying that Turkish has no irregular verbs.

If 'to be' isn't conjugated the same as every other verb in Turkish - in other words, if you can't take the infinitive and deduce all forms of the verb using a consistent rule - then it's irregular. By that definition it seems irregular to me, based on the link above. 01:07, 30 November 2005 (UTC)

I don't know about "to be" in Turkish, but it's quite possible that some people analyse it as a verb and others as a copula which they see as not a verb though usually with many similarities. — Hippietrail 02:08, 30 November 2005 (UTC)
If the verb in question is "one of a kind", it's a copula, and is morphosyntactically different from the others in some way, then maybe it should not be called an irregular verb. You don't call a vast plain with a single tall mountain "an irregular surface". :) --Pablo D. Flores (Talk) 22:37, 16 December 2005 (UTC)
Irregularity happens when a non-Turkic word came into usage. —Preceding unsigned comment added by (talk) 00:41, 23 March 2009 (UTC)
Turkish has native irregular verbs (and nouns for that matter): For example the aorist tense (geniş zaman) suffix -r requires an epenthetic vowel after verb stems with a final consonant. It's open (-e- or -a- depending on the vowel harmony) for monosyllabic verbs (dal-a-r) and close (-ı-,-i-,-u- or -ü-, again with vowel harmony) for polysyllabic verbs (otur-u-r). But about twenty monosyllabic verbs get close vowels (gel-i-r). It's clearly an irregularity. Also some verbs go through lenition of the last consonant, but most don't (bit-ecek, but gid-ecek, not *git-ecek). Causative form construction is also very irregular (al -> al-dır-, iç -> iç-ir, git -> götür) but it's a borderline case between inflection and derivation so I'll stop here. So I vote for changing the article as Turkish having "only a handful of irregular verbs".Cyco130 (talk) 18:03, 8 October 2013 (UTC)

Irregular verbs in NahuatlEdit

What is the reference for there being only a few irregular verbs in Nahuatl? I can't find anything on the 'Nahuatl' wiki page. Also at issue is which Nahuatl. Presumably this refers to classical Nahuatl, i.e. the form attested in various written manuscripts from the early post-conquest era. I'm working with another researcher on building a morphological transducer for several varieties of modern-day Nahuatl, and I can attest to the fact that these varieties are *full* of irregularity. (Of course some of the irregularity is predictable on a phonological basis, if you allow your phonological rules access to lexical features; but to me, needing to refer to lexical features in the phonology is synonomous with "irregular".) Mcswell 14:31, 22 March 2006 (UTC)

Comment added later: I just ran across the Wikipedia entry for 'Irregular verb', which also does not mention Nahuatl. Mcswell 20:11, 27 March 2006 (UTC)

Hungarian conjugationEdit

Some Hungarian verb endings have even 3 meanings.

-k: first person / singular / indefinite
-m: first person / singular / definite

-- 15:35, 10 May 2007 (UTC)

Hi guys, my name is plc580e. I was born in the US but hungarian was spoken in the house as my grandmother only spoke hungarian . After many years of not using the language I started to re-learn it . I visited hungary a few years ago and had no difficulty communicating with the locals but I'm sure my grammar was a riot. ( Either that or I was telling a lot of jokes I wasn't aware of). Anyhow , I am relearning the language rapidly .

   I have a gripe though.  I've been out to the net and searched "hungarian grammar",Hungarian verbs" hungarian conjugations"  and any other Ungar things I can think of.

What I found is that all the popular dictionaries and tutorials have very idiomatic translations from hun-->eng and eng--->hung. This makes it impossible to learn actual words and usages. A phrase I heard a lot when I was a kid was " adok neked nyaklevest" which I knew the literal xlation as well as the idiomatic xlation very well. All the dictionaries had such definitions as "I'll give you what for", "you'll be sorry" . How the heck is anyone one supposed to learn new works when the xlations are so misleading.

   A suggestion to all the dictionary and phrase book writers is to provide both a literal and an idiomatic xlation going in both direction.  Some of the xlations are so far afield or arcane that the xlation makes no sense what so ever in either language. To learn a language this way is pure "phrase memorization/pronunciation" with no real understanding of the words used .
   Any sources you can reccomend that are accurate and literal would be appreciated.

Regards, —Preceding unsigned comment added by (talk) 20:11, 14 September 2007 (UTC)

Fusional/inflected distinctionEdit

The article currently states, "Agglutinative is sometimes used as a synonym for synthetic, although it technically is not. When used in this way, the word embraces fusional languages and inflected languages in general." However, the first sentence of the fusional language article basically says that "fusional language" and "inflected language" are synonyms (also, Inflected language currently redirects to Fusional language).

If these terms are really synonyms, then this article shouldn't talk about them as if there's a distinction; if the terms refer to different things, then Fusional language should be fixed to reflect this. I'm not a linguist, so I'll leave the task to someone who knows. Gzabers (talk) 17:41, 27 August 2009 (UTC)

Arabic language could have agglutinative morphology.for example the single arabic word=أوفلاليستأمنوننكموهالهن awafalaliyasta'minunannakomuhalahonnaEdit

Arabic language could have agglutinative morphology.for example the single arabic word=أوفلاليستأمنوننكموهالهن awafalaliyasta'minunannakomuhalahonna meaning=

and so should you not let them insist to entrust it(feminine object) to you(plural) for them(feminine)?

Humanbyrace (talk) 11:41, 4 December 2009 (UTC)

Most languages are agglutinative. That means it's not too special to be able to construct very long words. I think there's no need to farm examples. --vuo (talk) 21:47, 4 December 2009 (UTC)

So I can add this fact to the article.

Humanbyrace (talk) 14:43, 5 December 2009 (UTC)

To clarify, no, please do not add that example. Furthermore the division between this article and the article Agglutination (check it first) is a work in progress. --vuo (talk) 22:25, 5 December 2009 (UTC)
Although this word seems to be circulating the internet as the longest word in Arabic, it's not a real Arabic word, and it makes no sense. Even "supercalifragilisticexpialidocious" is more real!! Xevorim (talk) 15:00, 22 August 2011 (UTC)
Arabic is an agglutinative language. When translating a normal sentence from Arabic to English or from Arabic to French, one doubles the number of the words. Nevertheless, since Hebrew is much like Arabic but has much western influence, one can phrase Hebrew either as an agglutinative language or not. Examples:
  • Is it as bad as it sounds? = הכצעקתה? (In Hebrew it is a one word sentence - HaKeTzeaaqata - Ha = is it, Ke = like, Tzeaaqa = scream, ta = of it).
  • I did not get them to know each other.= عرّفتهمش (In Arabic it is a one word sentence - Arraftuhumsh - Arf = A Semitic root with the meaning "knowledge", the morph doubling the r means that the verb is done by one person on other ones, the tu means "I did", the "hum" means "them", and the "sh" means "not").
* Will you remember me? = ?התזכרני (In Hebrew it is a one word sentence - HaTizkereni)
* I did not see you. = شوفتكش (In Arabic it is a one word sentence - Shuftaksh)
Arabic is an agglutinative language, and a fancy Hebrew speaker can turn Hebrew into such a language too. Please put the Arabic language on the list. Thank you. — Preceding unsigned comment added by (talk) 15:28, 3 October 2014 (UTC)
"Arabic as an agglutinative language", Arabic Natural Language Processing: Challenges and Solutions, Ali Farghaly @ Khaled Shaalan, ACM Transactions on Asian Language Information Processing (TALIP) Volume 8 Issue 4, December 2009, Article No. 14 [1] — Preceding unsigned comment added by (talk) 16:18, 3 October 2014 (UTC)
The fact that a language has some agglutinative morphology does not make it an agglutinative language. Languages on this list should be commonly described as agglutinative in reliable sources. The TALIP source is a start, but it is quite clear that it is proposing a novel way of analysing Arabic which is not commonly accepted. So no, Arabic should not be included.User:Maunus ·ʍaunus·snunɐw· 18:33, 3 October 2014 (UTC)
Arabic has much more than some agglutinative morphology. Please learn some Arabic before arguing about it. Thank you. — Preceding unsigned comment added by (talk) 19:51, 3 October 2014 (UTC)
Please provide some mainstram linguistic sources for your claim that Arabic is commonly considered an agglutinative language. Thank you very much. User:Maunus ·ʍaunus·snunɐw· 20:00, 3 October 2014 (UTC)

Persian is an agglutinative language !Edit

Persian is not an agglutinative language it is Iranic language with Indo-Earopean branch how it can be a Agglutinative language? —Preceding unsigned comment added by Snake co1 (talkcontribs) 20:40, 23 August 2010 (UTC)

Your statement is a nonsequitor. There are agglutinative langauges in many language families and it is certainly possible for a family to have both agglutinative and non-agglutinative languages in it. If there is an argument for or against Persian being agglutinative then this is the wrong argument. — Hippietrail (talk) 14:03, 3 July 2011 (UTC)
Persian is indeed agglutinative, not as much as Turkish or other "pure" examples, but there is a lot of agglutination "going on"... — Preceding unsigned comment added by (talk) 21:14, 12 March 2015 (UTC)

Agglutinative languages listEdit

"Altaic /ælˈteɪɨk/ is a proposed language family of central Eurasia."

Recently, Altaic languages was removed from the Agglutinative language article because the editor (Jijwatmaat) felt that it is "Not an actual language family".

Is it not possible to list Altaic languages with the (editorial?) NOTE from the relevant article that that "Altaic is a proposed language family of central Eurasia." ??

MaynardClark (talk) 21:05, 7 November 2014 (UTC)

List of examplesEdit

The article lists "Japonic languages" as well as "Japanese". Isn't it odd? — Preceding unsigned comment added by (talk) 01:55, 20 October 2011 (UTC)

"Korean has only ten irregular forms of conjugation." - WaitEdit

This is highly misleading, especially as an example of "agglutinative languages having regular inflection!"

Not sure why this sentence is in the introduction. Maybe some readers would imagine Korean verbs being highly regular, with only ten irregular forms you have to remember.

Any native Korean speaker can tell you that this is *definitely* not the case. What it means is that Korean has (maybe) ten different *classes* of irregular verbs: many classes have hundreds (or maybe thousands) of verbs each, and you just have to memorize whether a verb is "regular" or "ㄷ-irregular" or so on.

(It is somewhat helped by the fact that each irregular form can only appear with a particular sound, so given the dictionary form, there are at most two or three ways it can inflect, or so I think.)

And then there are hundreds of verb suffixes, which can combine in a messy way, sometimes with optional (or mandatory) contractions.

In short, Korean verbs are definitely *NOT* an example of a grammar with regular inflection. (talk) 05:42, 21 October 2016 (UTC)


Is not an agglutinative language, but f.e. Batak is. Maarten van Thiel — Preceding unsigned comment added by (talk) 18:54, 26 September 2020 (UTC)

Return to "Agglutinative language" page.