Talk:Clitic

Wiki Education Foundation-supported course assignment
This article is or was the subject of a Wiki Education Foundation-supported course assignment. Further details are available on the course page. Student editor(s): Grh 30.

Above undated message substituted from Template:Dashboard.wikiedu.org assignment by PrimeBOT (talk) 17:52, 16 January 2022 (UTC)

Page name
Shouldn't this entry be called "enclitic"? -- hajhouse

Well, I now see the definition given here is that of an enclitic, not clitic. I will change it to have the proper definition. -- SJK

Ambiguity
It's not clear to me what "this" refers to, at the end of the article, where it says:
 * According to some linguists, this is actually one single verb.

--24.48.223.10


 * The "this" I believe refers to the French pronouns. The author seems to be suggesting that the French pronouns are morphologically formed from a verbal root.  I don't know enough French to comment further. -- Jonadab

Greek enclitics
My first question is about the Greek enclitics. The article lists τε, δε, ουν, and γαρ, four of the major postpositives. My question then is whether the postpositives as a group are a subclass of clitics or whether these four words just happen to be both.

-- Jonadab

"'em" an enclitic?
My second question is about 'em in "Give 'em some water"; I have always taken that for the third person plural pronoun in the objective case; in this example it is functioning as the indirect object, but it is also seen as the direct object ("kill 'em"). Additionally, the singular masculine form is often similarly shortened ("Give 'im some water"), though as far as I am aware "she" and "it" are always spoken and written in their entirety. Additionally, both forms can occur disconnected from the verb by other words, especially with a preposition, as in "Give it to 'em" or "Give it to 'im". This would seem to preclude its being a clitic as defined in the article.

-- Jonadab


 * "'er" is used instead of "her" at times, for what it's worth. Kairos 08:29, 12 Apr 2004 (UTC)


 * I have noticed the following lines were deleted by 216.195.166.78 at 02:37 on August 24, 2005. Is that all right?
 * &#39;im in give&#39;im a chance
 * &#39;er in let&#39;er go
 * &#39;em in go get&#39;em
 * &#39;d in they&#39;d
 * - TAKASUGI Shinji 01:27, 24 April 2006 (UTC)

&kappa;&alpha;&iota;
One final question: if τε is from -kwe as well as -ca in Sanskrit and Latin -que, is και (the primary word for "and" in common Greek) also related, or is its similarity a coincidence? (This final question is off-topic, since kai is not a clitic, but I'm curious.) -- Jonadab


 * It's probably not related to -&tau;&epsilon;, though there don't appear to be any secure cognates from other roots. Pokorny derives it from *kai 2 (519), and relates it (uncertainly) to Old Church Slavonic ce&#780; in a ce&#780; and ce&#780; i (apparently meaning something like "also, indeed"), as well as Latin ceu "as, just as", though this is usually referred to *k'e-iwe (from the pronominal stem *k'-).  &#8212;Muke Tever 03:07, 10 Sep 2004 (UTC)

"Je ne lui le en y suis pas"?
French is the second language I learned, 32 years ago, and I speak it daily. Somebody please explain to me what this sentence is supposed to mean? D.D. 20:08 28 Jun 2003 (UTC)


 * French is my first language and I have no idea either. -- Tarquin 20:16 28 Jun 2003 (UTC)


 * I would assume that is supposed to say je ne le lui y en suis pas venu, not that that means anything either :) Adam Bishop 20:17 28 Jun 2003 (UTC)

In what does it have to be changed, then? I don't suppose we'd want something like I him him about it there am not come. in Wikipedia, even in French ;-) I have a comparable question on Talk:Polysynthetic language D.D. 20:35 28 Jun 2003 (UTC)

Genitive or possessive?
Shouldn't the genitive in "English: 's, marker of the genitive relation of a noun phrase" actually be possessive? See possessive case and genitive case. Nanobug 15:29, 21 Nov 2003 (UTC)


 * To answer your question, it's "genitive." Contrary to popular (revisionist) belief, there is no such thing as a possessive case - at least not in Germanic languages.  Possession is one function of the genitive case -- that is the proper name of the case, not "possessive."


 * In English, which began as an amalgam of "Englisc" and "Seaxonisc" (or Anglian and Saxon), and subsquently begged, borrowed and stole words from every other language under the sun (and a few under the moon, as well), intermingled differences in case endings, etc., produced an English language that uses the least common denominator of its principal contributory languages.


 * The result is that the genitive case in Modern English is currently used most commonly to indicate possession -- but that is NOT the only function of the case in Modern English.


 * The contraction -'s, by the way, is not really a marker -- it's a contraction of the ancient genitive case ending -es.


 * The full genitive case ending is still used in a small number of English words. For example:


 * The second day of the work week is Tuesday, not Tu's Day (Old English Tiwes Daeg; 'Tiw' respelled 'Tu' by the Normans after 1066, when they respelled the entire language as if English were Norman French).


 * The third day of the work week is Wednesday, not Weden's Day (Weden being the Anglo-Saxon equivalent to German 'Wotan' or Norse 'Odin' - which, by the way, is pronounced "OO-din," not "OH-din"). Wedn


 * In Modern Scots, the fourth day of the work week is Thurisday, as opposed to Thur's Day ("Thur" being the Anglian equivalent to "Thor" - which, in Old Norse and Modern Norwegian, is pronounced "Thoor," not "Thore"). Note that, while the genitive case ending is not contracted, contraction is used for the second vowel in the deity's name but without an apostrophe:  Wedenesday > Wednesday.


 * In English, the 'es' of 'Thuresday' has contract to 's', leaving us with 'Thursday' but it's still displaying the genitive (as opposed to possessive) case, albeit without using an apostrophe.


 * —The preceding unsigned comment was added by 207.200.116.133 (talk • contribs).

An example to add.
A good example of a pro clitic is the Albanian e/i/te.

Jimmy Dillies

English clitics et al
I just saw that Takasugi-san made some changes. I've partially reverted two of them since I feel they're inaccurate: one, changing "word" to "unit", and two, the claim that clitics are usually "function words". The main idea of a clitic is that it joins to something else and forms a phonological word -- "phonological unit" is an imprecise term and doesn't get to the point. Because of that, also, a clitic is not usually a function word but a function morpheme, or whatever one calls it. "clitic = function word" suggests that a clitic is a word, which is precisely what the article tries to explain a clitic really is not.

About the reordering: as it was, it seemed as if English was more important, or the paragon of clitic usage, while the other languages were placed at the end, only as "examples". This is to be avoided (see WikiProject Countering systemic bias). I've placed the whole abstract explanation first, then an examples section with references to Indo-European, then English, Romance, etc. --Pablo D. Flores 20:31, 8 Apr 2005 (UTC)


 * According to SIL Glossary of linguistic terms, clitics are words. See the Kinds section in What is a word?, and you'll find clitic.  Also, see What is a clitic? (Grammar) and you'll find "(clitics) often have grammatical rather than lexical meaning," i.e. they are often function words.  Since SIL is a very reliable linguistic source, I think we should classify clitics as words.


 * In addition, the following are wrong:
 * the clitic attaches to full words of whatever kind
 * The clitic attaches to a noun, a phrase, or even a clause in some languages. E.g. the English clitic &#39;s can attach to a noun phrase with a determiner: the girl next door's cat.


 * A clitic joined to a word forms a new phonological word
 * They are grammatically two words. They form a phonologically word-like unit. - TAKASUGI Shinji 16:43, 2005 Apr 10 (UTC)

In the subsection about word stress, the last example is clearly incorrect, suggesting the "rule" may need to be rewritten or described more precisely.


 * A stressed word cannot be changed into a clitic:
 * I don’t know who she is. (*I don't know who she’s.)
 * Have you done it? —Yes, I have. (*Yes, I’ve.)
 * He’s not a fool. —He is a fool! (*He’s a fool!)

He's a fool! is not incorrect. Cellmaker (talk) 11:48, 24 August 2012 (UTC)
 * It would be incorrect as a substitution for He is a fool, with the is stressed (as it would be in responding to the assertion that he's not a fool). That's presumably what's intended to be being said. Victor Yus (talk) 16:00, 24 August 2012 (UTC)

English genitive is a clitic
This is a reply to an unregistered user and a warning that might be useful in the future, as this tends to repeat from time to time. Please keep the warning at the top of this page.

An anonymous user inserted the following text in the intro:


 * The English genitive is sometimes regarded, illogically, as a clitic, as in the phrase "the Church of St Andrew's steeple," since the steeple belongs to the church and not to the deceased St Andrew. However, most people realize that this is an awkward construction and would phrase it as "the steeple of St Andrew's church," or "St Andrew's Church's steeple."

A clitic attaches to phrases (a phrase might consist of a single word, or more). In "the Church of St Andrew's steeple", the clitic indeed marks a possession of the church and not of St Andrew; the phrase that the possessive attaches to is "the Church of St Andrew":


 * [the Church of [St Andrew]]'s steeple

If -'s was a case suffix, as implied, then it would be awkward and wrong:


 * *the Church of [[St Andrew]'s steeple]

The problems of interpretation and ambiguity arise from the fact that English has two different genitive constructions (right-branching with of, and left-branching with -'s).

--Pablo D. Flores 22:12, 19 Jun 2005 (UTC)


 * Very true... Especially since "awkwardness" doesn't equal ungrammaticality... and there are much less awkward examples than the steeple, such as "the Queen of England's iPod broke" which is better than "the iPod of the Queen of England broke", IMO. &#8212;Muke Tever talk (la.wiktionary) 23:58, 19 Jun 2005 (UTC)

I can see that 's is a clitic because it attaches to phrases, but I'm confused. I thought a clitic was a kind of abbreviation? If it is, then what is 's short for? Sorry if i'm way off the mark!


 * Clitics are not necessarily abbreviations. The suffix -'s is not an abbreviation. In Japanese, all case markers are clitic postpositions. - TAKASUGI Shinji 23:29, 28 April 2006 (UTC)

I first noticed this peculiarity when I was a child and had the book The King of Ireland's Son on my bookshelf. I knew, of course, that it had to be [The King of Ireland]'s Son, and not The King of [Ireland's Son] (which makes no sense), but I didn't quite understand why it meant that. Pity they don't teach about phrase-level clitics in elementary schools! Angr 18:50, 6 October 2006 (UTC)

Two comments after my morphology lecture on clitics ;) Firstly in the Talk section there was some mention of clitics having to be short for something. Well, that was debunked however what was not mentioned is that there are actually 2 (or 3) types of clitics, the first type is Simple Clitics, which like 'll' and 've' are actually short for something and work like an indirect substitute. There are also Special Clitics, which either are short for something but have a different syntactic distribution eg French.
 * Jean le voit (Jean sees it)
 * Jean voit le (incorrect)
 * Jean voit le livre (Jean sees the book)

In the above example 'le' as a clitic (pronounced l') comes before the verb, but as a word (le) comes after the verb.

The other type of special clitic is those that have no comparable words. This would be genitive 's in English if 's was a clitic.

HOWEVER In his 1987 paper "Supressing the Zs" Zwicky makes the case that 's isn't a clitic but rather a phrasal affix that functions as a clitic, but which takes morphological from the phrase it attaches to. In particular when the genitive 's follows the plural "s" or present tense "s" it becomes a zero morph, whereas if it follows an "s" that is just part of the word it takes the ɪz allomorph. I'm not very good at explaining this, which is why I'm not editting the article, but as an example:
 * 'People were surprised by Katz's reactions' shows the genitive 's function much as a normal clitic, the 's is added independently. However
 * 'People were surprised by the cats' reactions' shows the genitive 's interacting with the plural form to take a zero morph.

If 's truly were a clitic both examples would be said identically (barring "the"). Clitics aren't meant to regard the morphology of the preceeding phrase. If it were purely phonological both 's would have become silent, but due to the fact that it only happens after plural and present tense, it shows interaction with the morphology. Zwicky terms 's a phrasal affix as a result (half-clitic, half-affix). Sorry if that sounded confusing.

Actually, going on with the ideas highlighted above it is altogether possible that English used to be a pure affix but over time it has become a clitic which is why it retains morphological properties. That's just personal opinion, though. Saben4 05:16, 11 October 2006 (UTC)


 * While Zwicky clearly knows his stuff, the English genitive is commonly cited as a model clitic in linguistics courses. Moreover, Zwicky's argument doesn't hold for dialects in which the 's is not dropped after a preceding /s/ sound.  Zwicky does not argue that 's is not a clitic because it changes phonological form, but because it drops (in some dialects) after /s/ or /z/, and therefore behaves as a phrase-final inflection.  In my own dialect, Katz's is pronounced /kætsəz/.  Any argument that would deny clitichood based on a phonological change is untenable, as there are languages (i.e. Irish, with its consonant mutations) where lexical items that are clearly "words" undergo phonological changes conditioned by the preceding words. Straughn 22:18, 11 October 2006 (UTC)


 * No, you misunderstand. (Don't worry, it took me a few reads to understand what was being said, too.) Zwicky's point is that "cats" and "Katz" are pronounced identically, but the genitive marker behaves differently depending which one it's attached to: it's silent in the case of "cats'", and pronounced /əz/ in the case of "Katz's". Hence, it's not due to the phonology of the word, but rather to its morphology. Ruakh 22:55, 11 October 2006 (UTC)


 * Yes and to further clarify, the classical definition of a clitic says that it is based in syntax. Being based in syntax it should not be able to interact with the preceeding morphology or even phonology. Both are below the word level whereas syntax occurs only with whole words, phonology rarely changes due to differening syntactix structure. The genitive /s/ USUALLY acts the same as a regular clitic, but it behaves in ways that clitics normally aren't able and so, Zwicky argues, it deserves it's own category (I'm not sure if there are any other examples of this). "Phrasal affix" does seem to fit the role of genitive /s/ quite well, though, as it acts as an affix to an entire phrase, attaching itself as a clitic does while being morphologically conditioned, just like an affix (I hope that was more clear this time, thanks for helping me explain, Ruakh). Saben4 05:16, 11 October 2006 (UTC)


 * So, question: Hebrew has seven one-letter prefixes, with the meanings in, the, and, like (as a preposition), to, from, and that (as a conjunction and relativizer). With the exception of the, these all tend to introduce multi-word phrases; in ancient China, for example, is in-China the-ancient. With the exception of that, they all change pronunciation depending on the first syllable of the word they introduce; and, for example, is pronounced /və/, /va/, /vi/, /vɔ/, /vɛ/, or /u/. (Well, supposedly. In practice, people no longer actually talk that way; but let's assume traditional pronunciations for the sake of argument.) Additionally, they all can engender form-changes in the noun they introduce; for example, Belgium is /'bɛlgia/, but in-Belgium is /bə-'vɛlgia/. Finally, the four prepositions merge with and, taking the consonant of the preposition and the vowel of the and. (This one affects spelling as well: the letter for and disappears.) It seems like they're not ordinary prefixes, because except for the, they attach to entire phrases or clauses; but by your description, it seems like they're not clitics, either. So would they also be phrasal affixes? Ruakh 14:58, 26 October 2006 (UTC)


 * Surely a case ending can also be a clitic? The Saxon Genitive as seen in both English and Swedish is strongly argued as a clitic, yet shows strong ties to also being a case ending. The same could be taken for many Russian case endings on masculine nouns - they add an affix without altering the stress or final sound of the original noun, namely prepositional (-ye), instrumental (-om), accusative (-oo), genitive (-a). 87.194.56.53 (talk) 21:33, 7 September 2008 (UTC)

I know things change, but back in the 1950's when I was in grade school, we were taught that the possessive of a noun that ended in an "s" or "z" was formed by simply adding an apostrophe, without the extra "s": James' book, not James's book. Wschart (talk) 01:57, 28 March 2013 (UTC)
 * Well, in the 1960's when I was in high school, on the occasion of a class trip to England, my schoolmates and I visited St James's Park (in Westminster), and for all I know it hadn't acquired that final s in the latest ten years but had always been St James's Park, not St James' Park . What I was told in school is that plurals ending in s take a plain apostrophe for the genitive (the cats', the kings', etc.) and that all other words take 's (the men's, the women's, the children's, King James's, Bob Katz's, etc.) And BTW I learnt to pronounce James's as /'ʤeimsɪz/, not anything ending in /əz/. — Tonymec (talk) 03:26, 21 April 2023 (UTC)

Mesoclitics
I don't know much about this subject, so I didn't want to add anything to the article that could be incorrect.

In Portuguese there are phenomenons known as "próclise", "ênclise" and "mesóclise". They refer to using pronouns to indicate the object of a verb, i.e., the "me" in "tell me" ("-me" in "diga-me") or "them" in "give them (something)" ("-lhes" in "dê-lhes (algo)"). The different kinds refer to the position of the pronoun in regards to the verb.

From the description in the article it seems like this is basically what it is talking about, although I was surprised it didn't mention this specific case (object pronouns).

In any case, I was thinking that, if I'm correct that "clitics" can refer to this same phenomenon, then the third case mentioned above, "mesóclise" ("mesoclitic" in English, I'd assume) should be mentioned. By the way, it means that the clitic goes in the middle (hence "meso") of the word, separated by hyphens: "dar-lhe-ia" ("(I/he) would give him (something)") is the mesoclitic form of what could otherwise be written as "daria-lhe" or "lhe daria".

So if anyone would let me know whether it's ok to include that in the article, or want to do that themselves, I'd appreciate it. --Cotoco 07:43:23, 2005-09-04 (UTC)


 * Of course, after writing this I took yet another look at the article and found reference to the object pronouns, with the example of Spanish "dámelo". I will still, though, wait a little while for input from other people before adding something about mesoclitics. I also feel that maybe more emphasis should be given to the case of object pronouns as clitics, as it's an important features of at least some languages. Well, maybe not. Maybe it would be going too deep into it for this article. --Cotoco 07:56:11, 2005-09-04 (UTC)


 * I'll take a lot at it. Probably best treated in the articles of the particular languages, though. The clitic object pronoun thing is found, I think, in all Western Romance languages. I hadn't heard the term "mesoclitic" and I'm not sure it is in wide use. In any case, what is happening in Spanish dámelo, for example, is actually a verb getting a clitic and then getting another clitic on top of it; nothing is getting in the middle of anything. Da is "give" (imperative); dame is "give me" ("me" as indirect object); dalo is "give it" ("it" as direct object); dámelo is just ''Pablo D. Flores (Talk) 15:29, 4 September 2005 (UTC)


 * Continuing this almost nine years later... :) Now mesoclitics are in the article, but for the sake of completion I wanted to add to Pablo's kind response above, for him (although he seems to rarely be on WP anymore these days) as well as any other readers: yes, in Spanish "dámelo" is not a case of mesoclisis -- and as far as I know mesoclisis is not a feature of the Spanish language -- I only mentioned that to say that I saw the article does mention pronouns as clitics. On the other hand, in mesoclisis, in Portuguese, the clitic (pronoun) goes in between the verb stem and the suffix; for example, "compraria" ("(I) would buy") is the verb without the object. "I would buy it" would could be written as "Eu o compraria" (using proclitic form), and the mesoclitic form would be "Eu comprá-lo-ia" -- note how the pronounc clitic "lo" is separating the "compra" part from the "ia" suffix (the 'r' gets dropped in this case, and the acute accent added due to the accenting rules of the language).


 * I hope this helps to clarify it.


 * Cotoco (talk) 17:13, 17 August 2014 (UTC)

Accessibility
Okay, I came to this article wanting to know what a clitic was. I left, wanting to know what a clitic was. The discussion page was more helpful than the article, but I don't have a grasp of the subject. So a clitic is a morpheme that can't stand on its own...but I still don't get what distinguishes it from an affix. NickelShoe 19:56, 30 September 2005 (UTC)

Formatting
Should the included example clitics have a dash to indicate whether they are enclitics or proclitics? For example, all the Latin clitics are enclitics, and I've typically seen them rendered -que, -ve, and -ne; however, in the examples they are simply rendered que, ve, and ne.

stress and cliticization
I've added lines explaining stress prevents cliticization in English. Could someone kindly explain why you don't say "what's it" while you can say "what's that" and "who's that girl?" Is it because it is so weak "is it" always has a stress on is? - TAKASUGI Shinji 01:14, 24 April 2006 (UTC)


 * Nice question.

You can say: "What's it doing" or "What's it doing now?" In speech, almost comes out as the single morpheme "what's-it" or "whatzit".

You can also ask regarding a book or a movie: "What's-it about?"; or an object "What's-it for?"

Note the contrast "What's it eat?" which contracts "What does it eat".

If that machine doesn't do it, what does it?
 * 1) If that machine doesn't do it, what's it? # can't say that

What's it do? What does it do? What's it doing? What is it doing?

The first utterance feels more informal to me, as if the the surprise resolution to does over is is subconsciously regarded as a little sloppy.

I think the problem is that there is no way to speak "What's it?" with complete-phrase intonation, leaving the listener to topple into the vacuum of the highly anticipated resolution word (verb or pronoun) that doesn't follow. Does this article address the interesting clitical ambiguity between does/is and dependent resolution on words to follow?

What's it amount to? What does it amount to? What's it up to? What is it up to?

The first question you might ask your insurance adjuster about the net damage. It still strikes me as hurried to the point of slurring, almost nervous in the pace of utterance, as if the answer expected won't be good. MaxEnt 03:08, 5 June 2006 (UTC)


 * There's no mention of clitics based on the verb "to do" in English. I might have added this myself, but I'm not certain it is a clitic.


 * Consider:

Wudya do that for? What did you do that for? What's it for? What is it for?


 * It could be that "what's" for "what does" is just a less obnoxious version of the same compression that turns "what did you do" into "wudya" that forms a surface masquerade as a clitic. Note that 's for does is restricted to what/how/when and can't be applied to he/she.

Rarely who:

Who's he know from that team? Who does he know from that team? Whose good plumbing should we photograph? Who is good plumbing going to satisfy?
 * 1) Who's good plumbing? #        Who does good plumbing?

The first example is the kind of sentence one might speak who is having trouble catching their breath after heavy exercise or during allergy season. The bad usage example also fights against "whose" in addition to "who is".

These 's contructs on question words seem to be governed by somewhat different rules. Are any of them proper clitics or are they all shortness of breath? MaxEnt 03:36, 5 June 2006 (UTC)

Proclitic complexities
In Cell microprocessor the first sentence reads:
 * Cell is a microprocessor architecture jointly developed by a Sony, Toshiba, and IBM alliance known as STI.

Whenever I proof that sentence I feel a mild, uncomfortable pull to change a back to an. I don't know whether that is because IBM is the third (and subconsciously dominant) serial adjective, or whether the noun target somehow expresses a slightly stronger solitary pull than the dissolute serial adjectives.

It also just occurred to me that this article does not express a common aggravation with these proclitics that manifests itself in the expressions: depending on how the author prefers to vocalize the written form. Some pedants hate the blended sequel pronunciation, but it predominates esp. in adjectival-accusative contexts such as the ever-popular your sequel statement has a syntax error. MaxEnt 02:35, 12 June 2006 (UTC)
 * an SQL server - pronounce an ess-que-ell server
 * a SQL server - prounounce a sequel server


 * If I were you, I would probably change that sentence to say something like, "Cell is a microprocessor architecture developed by an alliance known as the STI whose members are Sony, Toshiba, and IBM," if somebody hasn't already done so by now. —&thinsp;RandomDSdevel (talk) 20:59, 16 March 2014 (UTC)

Welsh
ISTR that clitics are used profusely in Welsh, but I don't know enough linguistics to add the information myself. I do know that the definite article "yr" becomes "'r" after vowels, and ISTR that verbs can be thoroughly contracted such that Yr ydych chwi yn eistedd (we are sitting) contracts to Rydych chi'n eistedd. Grutness...wha?  13:38, 21 June 2006 (UTC)

Proof
The contraction n't as in couldn't etc. has been proven to be an affix rather than a clitic (Zwicky & Pullum, 1983).
 * Mighty strong words! I would have though that such matters would always involve a fair amount of semantics. FilipeS 12:18, 2 November 2006 (UTC)
 * I find their proof very convincing. Now I have a question.  Is the phrase "didn't go" just the same as "did not go", or is it weaker like the following?
 * {| class="wikitable" border="1"

! !! &minus;do !! +do ! Affirmative ! Negative
 * I went. || I did go.
 * I didn't go. || I did not go.
 * }
 * I don't feel a big difference between the two negative sentences, but not being a native speaker, I don't trust my intuition. - TAKASUGI Shinji (talk) 17:49, 28 January 2008 (UTC)
 * Nobody has replied, but I've got the answer from Nominal TAM; didn't and did not are the same. Thank you. - TAKASUGI Shinji (talk) 17:50, 24 May 2008 (UTC)

To revisit the Zwicky-Pullum n't issue: I think it is problematic to rely solely on this article for discussion of n't, whether or not the article refers to "proof". Zwicky and Pullum write, "The negative formant n't is assumed, in most recent analyses that mention it, to be an unstressed and contracted form of the word not" (p. 502). They go on to offer an contrary analysis. Zwicky and Pullum's analysis has been cited several times since 1983 (Google Scholar suggests 457 citations), but it remains only one competing view. For instance, several works arguing or assuming a contrary view have been cited more than 100 times each. The claim on this page that n't is not an enclitic should perhaps be weakened, or at least described as somewhat controversial. Cnilep (talk) 15:06, 16 May 2010 (UTC)

New lead and reorganization
Sorry for introducing so many changes all at once. The main motivation was the previous definition of clitics as words, which I know was discussed above, but in reality people use "clitic" as a much broader descriptive term. On closer inspection, some of these clitics are in fact words, some are affixes, and some might really need an intermediate category. There are no clear (pre-theoretical) boundaries between word and clitic, and clitic and affix. The article rightly mentions some well-known proposals for imposing a clitic vs affix distinction (but it's still a controversial topic). CapnPrep 23:31, 5 November 2006 (UTC)


 * It looks good overall. Two quibbles:
 * The opening paragraph seems to suggest that clitics derive from independent words, and are destined to become morphological affixes. This isn't necessarily the case; for example, English's possessive -'s actually descends from a genitive-case affix.
 * "Orthographically, clitics can be written as separate words, or set off by punctuation (hyphen, apostrophe) from their hosts" is correct, but IMHO somewhat misleading; I think it makes it sound like those are the two possibilities, which obviously is not the case.
 * Ruakh 02:51, 6 November 2006 (UTC)


 * Thanks for your quick fixes! :-) Ruakh 13:23, 6 November 2006 (UTC)

Dutch clitics
Added ze (3rd person plural pronoun), as it's on par with je I think, although I'm not sure whether either is a true clitic since both can function as non-clitics, but they can be in the same position as ie which is clearly a clitic as it cannot stand on its own. I removed the ortography of ie as -ie including the example, as this just isn't true - common ortography does not add a hyphen, but has it stand alone ('heeft ie het gedaan?' rather than 'heeft-ie het gedaan?') JAL 85.223.48.149 20:28, 12 December 2006 (UTC)


 * Cool. Although I think this information would be more useful in Dutch grammar or Dutch declension. (It's sort of already in both, which is also a problem.) CapnPrep 21:13, 12 December 2006 (UTC)


 * I disagree. 'Je' instead of 'jij', 'we' instead of 'wij' (yes, that exists too), "'t" instead of 'het', 'ie' instead of 'hij' are simply unstressed forms of the pronouns. If those are to be considered 'clitics', then the unstressed "'m" in "I'm" or the unsteressed "'re" in "they're" ought to be considered clitics too -- and I think we have a problem here. Vidyasagara (talk) 15:01, 29 July 2013 (UTC)
 * Although I've yet to find a source that calls "we" "'t" etc. clitics (and despite my 9yo remark above, I'm now not sure "ie" is a clitic either), "m" and "re" actually are clitics, or even affixes, as this article also describes. Jalwikip (talk) 15:58, 22 January 2016 (UTC)


 * Agree with Vidyasagara. The nub of the problem seems to be this: a clitic "cannot stand on its own" but "functions syntactically like a normal word" (there are not many words that can be sentence (utterance) all by themselves except in rather contrived circumstances, but we'll let that slide for now). A de-stressed pronoun form can only occur adjacent to a word that takes the stress, and in that sense satisfy a criterion for clitic-ness (cliticality?). Syntactically, ie functions just like hij (zij, het) which everybody agrees is a pronoun and not a clitic. One could almost make the case that ie is a surface variant of hij which is only distinguished at the level of pronunciation (which indeed is what is was to begin with). But then we have to ask why we find hij and ie in alternation in written Dutch (and indeed if ie should have a place in written Dutch). The answer is that using ie for hij (or zij) conveys a degree of informality, and hence possibly of attachment or else of mockery or disrespect (depending on the speaker's relationship to the person referred to as ie). I give this longish discussion of syntactical and lexical aspects of ie to show how it is not your common garden-variety clitic, or, as Vidyasagara says, not a clitic at all. 2A01:CB0C:CD:D800:2487:4AE5:1F2C:57A2 (talk) 07:26, 6 October 2021 (UTC)

"The", "a" and "an"
"A clitic must attach to an adjacent word, known as its host." Then how can the article "the" in "the house" be a clitic? If it were, it should be "big the house" instead of "the big house", right? (like in Swedish "stora huset") Apus 11:58, 1 February 2007 (UTC)


 * The host is the first word of the noun phrase (not necessarily the noun itself). A clitic can combine syntactically with a phrase, and depend phonologically/prosodically on the leftmost (or rightmost) word of that phrase. CapnPrep 14:58, 1 February 2007 (UTC)

Say eh, what?
So what's a cletic then in the English language, from this article it seems to just be a Contraction or "a" or "the" although I'm not sure why. Is that right? I've read the article and that's all I've learnt. Could perhaps the concept be described a tad simpler before launching into linguistic lingo? Caffm8 00:25, 20 May 2007 (UTC)


 * I have to second this. The introduction is so technical as to be unintelligible to an otherwise decently educated layman. --Sylvank (talk) 06:01, 15 January 2008 (UTC)


 * In short, a clitic is a grammatically independent and phonetically dependent word. It's pronounced like a prefix or a suffix, but works at a phrase level.  The phrase my wife's brother has three phonetic groups: my, wife's, and brother, but the grammatical structure is (my wife) ' s brother.  Here,  ' s is phonetically attached to the word wife but grammatically combined with the phrase my wife. - TAKASUGI Shinji (talk) 07:27, 15 January 2008 (UTC)
 * I have added the sentence above to the article. In addition, I have changed the word many in "many clitics can be understood as elements undergoing a historical process of grammaticalization" to some.  The English  ' s is from the grammatical suffix, not from a lexical word, and most Japanese postpositions are clitics from the beginning. - TAKASUGI Shinji (talk) 14:50, 23 January 2008 (UTC)
 * Thanks, I think that's a more accessible introduction. --Sylvank (talk) 18:57, 28 January 2008 (UTC)

saxon genitive as a clitic
Is the saxon genitive involving a possession of a group (plural) of things, such as "the teachers' carpark" still called a clitic, even though it does not add a phonological unit to the word? Ninahexan (talk) 06:26, 22 June 2009 (UTC)

Clitics in Romance languages
It has been said that: According to most criteria, in fact, the pronominal clitics in most of the Romance languages have already developed into affixes. I think it's not entirely true. Joining the enclitics to the stressed word (i.e. writing, for instance, dammelo instead of da me lo in Italian) could be only an orthographic convention, probably modelled on the way the words with attached enclitics are pronounced in Latin. 89.164.229.76 (talk) 10:53, 17 September 2009 (UTC)


 * The claim in question is not based on orthographic arguments. The clitics in me lo darai also behave like affixes, even though they are written separately by convention. Anyway, the statement is sourced, so you will have to find a counter-source to challenge it. CapnPrep (talk) 19:27, 17 September 2009 (UTC)

Swedish definite articles
I've never heard of these being described as anything else than affixes before, and if no one can source this analysis I will remove the examples. --Dingbats (talk) 19:03, 9 September 2010 (UTC) ...and done. --Dingbats (talk) 18:27, 14 September 2010 (UTC)
 * That was half good and half bad done, Dingbats!


 * The origin of the Swedish determined affixes are enclitic pronons; but it is not very sensible to describe them as such in the modern language. The situation was rather different in Old Norse, and still is in modern Icelandic and Faroese. Two important differences are the following:
 * In Old Norse, and in modern Icelandic and Faroese, there was/is a pronoun, which may be employed either as an independent word, or in a suffix-like position. The independent forms are slightly longer; the enclitic forms have dropped the initial h-, and sometimes a vowel, too. However, they are declined in the same manner by gender, number, and case, whether they are independent or enclitic.
 * In Old Norse, and in modern Icelandic and Faroese, there was/is a double declination; both the noun itself and the enclitic polynomial are declined by case and number.


 * If you wish references, then look at any text-book about any of these three language forms. (As I remember it, this even is explained in the few pages about Icelandic in The World's Chief Languages by Mario Pei.)


 * The situation is radically different in modern Swedish (and modern Danish and Norwegian, too). The independent pronoun is (almost completely) lost (apart from a few frozen oldfashioned forms, like Danish Hin unge (the young one), Swedish Hin håle (the hard one, i.e., the devil), and some dialectal usage). The same goes for double case declination (again with the exception of some more or less obsolete double genitives, as in Danish livsens ondskab and Swedish mörksens makter). There still is some doubling of number declination; but the suffix plural form to some extent is adapted to the one of the noun.


 * Therefore, I think it was very appropriate to remove the Swedish example, as you did. Possibly, it could be mentioned at the end, as an example of enclitic forms developing into proper endings.
 * However, you removed the Old Norse example at the same time; and I don't agree with this.(You also didn't warn or ask about the Old Norse example; just the Swedish one.) For the reasons I've explained supra, you really can't equal the situation in Old Norse and the one in modern Swedish.
 * Thus, I'm going to restore the Old Norse paragraph, thereby partially reverting you; and add a few words about the later developments. If you disagree, you may remove it again; but I'd prefere discussing it first. Best, JoergenB (talk) 14:12, 1 July 2011 (UTC)
 * Oh, no problem, go ahead. I don't even remember making that edit, and you seem to know what you're talking about. -- Dingbats (talk) 14:30, 1 July 2011 (UTC)

Swear words as endoclitics
This may seem obtuse, but can I suggest that the allegedly rare endoclicic may actually be more common in English than we give it credit. The word F***ing (and some other swear words) can be used within words - splitting the root - in order to add extra emphasis e.g Tues-f***ing-day, which can be written without the hyphens. Admittedly this is often for comedic effect - but it is real usage (or more properly, perhaps, abusage). Irvine Welsh is a writer with a penchant for the form.

There'd again, maybe it's not worth mentioning. — Preceding unsigned comment added by Doctordjbrown (talk • contribs) 09:54, 26 August 2011 (UTC)


 * Interesting idea. I am not a linguist, but while it's clear that if this use of "fucking" is a clitic then it must be an endoclitic (making for a great pun, of course), it may actually not be a clitic at all. I think that to prove that it is an endoclitic, we would need a word such that if you split it with "fucking", the pronunciation of "fucking" changes from the normal one. But as I understand this construction, there is always a clear gap before and after "fucking". The closest thing that I found is this theoretical example: "gin-fucking-ger". A person who normally pronounces "fucking" with ng at the end might pronounce it as "gin-fuckin'-ger" instead in this particular case. Come to think of it: In the same way, we might be able prove that more standard uses of this intensifier are as a clitic. For example, a person might say "killing Germans", "singing Germans" etc., and "fucking Italians" but "fuckin' Germans" . Hans Adler 11:18, 27 January 2013 (UTC)


 * It's not a clitic, it's expletive infixation. --Florian Blaschke (talk) 02:16, 21 November 2013 (UTC)

many of these are not clitics
According to most refs, and our own def, clitics are "promiscuous" in that they are not constrained to dock to a particular part of speech. English 's is like this: it appears at the end of the noun phrase, regardless of the part of speech of the word at the end of that phrase or whether it has any semantic connection to it. Romance pronominal "clitics" are not: they only dock to verbs, and then only to the verbs they are the objects of semantically. That is, they are affixes. There's little diff between them and the pronominal affixes of polysynthetic languages. I think if we're going to define clitics according to Zwicky and other RS's, then our examples should conform to that definition, rather than being taken indiscriminately from anyone who happens to call something a "clitic". I added a recent source that notes that we cannot get a coherent understanding of clitics from looking up the word in the indexes of grammars. — kwami (talk) 22:54, 12 July 2013 (UTC)
 * I remember reading a description of clitics as affixes dependent on phrases rather than single words. That struck me as a great way to put it and a more useful starting point for a definition. I don't know the origin of this conception, though. --Florian Blaschke (talk) 02:11, 21 November 2013 (UTC)
 * The way I understand it 's is a clitic because it is essentially an affix (suffix in this case) appended to an entire phrase such as the Queen of England, which means that in practice, it is suffixed to the last word belonging to the phrase, whatever its class.
 * In contrast, the reflexive -s of Danish, like the pretty much exactly corresponding -ся/-сь of Russian (except that it is not used to form a passive), is a garden variety suffix, which is only ever appended to verb forms, and its scope is limited to that verb form it is suffixed to. --Florian Blaschke (talk) 02:36, 21 November 2013 (UTC)
 * And I agree with you, the Romance "clitic" pronouns are really affixes. However, in medieval Romance varieties such as Old Spanish, and still in literary European Portuguese, they are still more syntactically variable and can appear variously as prefixes, suffixes and even infixes, which I don't know if it is appropriate behaviour for affixes. --Florian Blaschke (talk) 02:51, 21 November 2013 (UTC)
 * Short prepositions in Russian (and other Slavic languages) are much better examples for clitics. They do not behave like prefixes syntactically (always preceding the whole phrase, even though the head noun is often not but more typically at the end), but phonologically they most certainly do (на полу /nəpɐˈlu/ "on the floor"; some propositions even consist of a single consonant only, so there's no way they could receive stress even in principle).
 * Couldn't many short, often monosyllabic, Romance prepositions also be described as (pro-)clitics? --Florian Blaschke (talk) 03:45, 21 November 2013 (UTC)
 * The Korean part implies that clitics are something different than phrasal affixes are – but what is the difference then?
 * The Gothic examples are really good, too, by the way. Also Latin -que and -ve are prototypical enclitics. English -n't, -'ve, -'re, 'em seem to be suffixes, though, as their distribution is quite limited. --Florian Blaschke (talk) 03:54, 21 November 2013 (UTC)
 * Slavic prepositions definitely are clitics because they are attached to phrases. In some languages (Russian, Czech, Serbo-Croatian) they can attract stress from certain groups of words, but still form a single prosodic unit with the host word(s). — Preceding unsigned comment added by 2A02:C7F:8E04:C600:7D0D:A52C:CEE0:E12 (talk) 15:17, 19 January 2019 (UTC)

Examples of Proclitics? Suggestion
This post is pretty much self-explanatory: can anybody provide this article with some examples of proclitics to go in the 'Examples' → 'Proclitics' section?

—&thinsp;RandomDSdevel (talk) 21:08, 16 March 2014 (UTC)

correct grammar
would be 'the king's of England horse' actually....the horse of the king-of-England, not the regent of England's horse....


 * What makes 's a clitic here is that it attaches to the phrase as a whole, while the head word remains the possessor even though it does not directly receive the 's. It is precisely because the clitic does not alter the meaning in the way you indicate that it is a clitic. If you wish to argue that it should not be clitic because you feel that forming possessives of phrases is ungainly, then I would have to agree with you on a personal level but we have to admit that the two of us are not here to proscribe usage to the community of English speakers. 2A01:CB0C:CD:D800:2487:4AE5:1F2C:57A2 (talk) 07:39, 6 October 2021 (UTC)

Fix wording of opening paragraph to clarify ambiguity
Here's how it reads now: ''In morphology and syntax, a clitic (/ˈklɪtᵻk/ from Greek κλιτικός klitikos, "inflexional") is a morpheme that has syntactic characteristics of a word, but depends phonologically on another word or phrase. In this sense, it is syntactically independent but phonologically dependent, always attached to a host.[1] The term is derived from the Greek for "leaning".[2] It is pronounced like an affix, but plays a syntactic role at the phrase level.''

The trouble is with "It" in the last sentence. Under normal rules, "It" refers to the previous sentence, so that the final sentence means: [the term] is pronounced like an affix. But the meaning really is: a clitic is pronounced like an affix. I've changed the last sentence to make that clear. Omc (talk) 00:39, 30 October 2016 (UTC)

External links modified
Hello fellow Wikipedians,

I have just modified 1 one external link on Clitic. Please take a moment to review my edit. If you have any questions, or need the bot to ignore the links, or the page altogether, please visit this simple FaQ for additional information. I made the following changes:
 * Added archive https://web.archive.org/web/20060920015915/http://www.african.gu.se:80/aa/pdfs/aa02027.pdf to http://www.african.gu.se/aa/pdfs/aa02027.pdf

When you have finished reviewing my changes, please set the checked parameter below to true or failed to let others know (documentation at ).

Cheers.— InternetArchiveBot  (Report bug) 16:37, 26 November 2016 (UTC)

External links modified
Hello fellow Wikipedians,

I have just modified 2 external links on Clitic. Please take a moment to review my edit. If you have any questions, or need the bot to ignore the links, or the page altogether, please visit this simple FaQ for additional information. I made the following changes:
 * Added tag to http://morbo.lingue.unibo.it/mmm/mm5-abstracts/kopris-davis.pdf
 * Corrected formatting/usage for http://www.african.gu.se/aa/pdfs/aa02027.pdf
 * Added archive https://web.archive.org/web/20070927013419/http://www.linguistics.uiuc.edu/jyoon/Papers/Affix-order-final-single.pdf to http://www.linguistics.uiuc.edu/jyoon/Papers/Affix-order-final-single.pdf

When you have finished reviewing my changes, you may follow the instructions on the template below to fix any issues with the URLs.

Cheers.— InternetArchiveBot  (Report bug) 15:16, 9 August 2017 (UTC)

Observations on "clitic" based on same church, divergent gospels
Herein on 30 September 2005, user NickelShoe created new Talk Page section, titled “Accessibility.” Although I don’t share the user’s seeming frustration with this article’s verbiage, I’d like to echo the apparent perplexity that the user expressed upon reading the article. Despite how I credit the article’s various contributors for their efforts, I’m not satisfied that the sum of those efforts enables lexical cogency concerning what a clitic fundamentally does and does not entail. Also, whereas I benefited from the numerous morphological, phonological, and linguistic considerations related in this article, its cobbling of too many un-sourced, paradoxical, and outmoded assertions creates an untenable notion of what a clitic is supposed to be.

Specifically, the article’s lead sentence defines “clitic” as a morpheme that “depends phonologically on another word or phrase. In this sense, it is syntactically independent but phonologically dependent—always attached to a host.” That definition is troubling in two respects: (1) it offers an idea of what a clitic does but not what it is, and (2) equivocates the definition sourced from LinguaLinks, which indeed provides that “A clitic is a morpheme that has syntactic characteristics of a word, but shows evidence of being phonologically bound to another word.” While the LingaLinks definition is too nebulous to satisfy my lexical expectations, the distinctions between it and the article’s definition are quite stark.

It might be worthwhile to say I’m invariably dissatisfied with any linguistic definition that appears to be premised from a naïve set theory rather than a linguistic paradigm. Indeed, all of the linguistic definitions at Wikipedia – despite the insights they provide –are glossed from intermixed global sources that stretch across innumerable theories relating to lexicology, morphology, phonology, semantics, etymology, etc. Thus, articles such as this one necessarily result in complexities which on one hand offer limitless bases for further consideration but on the other hand offer no dispositive notion of an article’s entry word or phrase.

For me, the relevant question here is not “What is a lexical definition of clitic?” That question can be readily answered by a dictionary, i.e., “a word that is treated in pronunciation as forming a part of a neighboring word and that is often unaccented or contracted” per Webster-Merriam). Rather, my question is, “What are users’ subjective notions about what a clitic represents?” In reply to that question, this article doesn’t disappoint. My conclusion: users here present an eclectic mix of observations that amount to a predictably informed on some levels, mistaken or misdirected on others, and inconsistent on the whole.

Having said all that, I’m nearing completion of my own linguistic paradigm that is based on axiomatic set theory. In that paradigm, my definition for clitic is “an affix entailing a short form of a word as part of a contraction.” That definition indeed contradicts many of the assertions in this article, but it is internally consistent with more than 800 other linguistic definitions within the paradigm. For anyone interested, my paradigm defines the English possessive {'s} and {s’} as affixes, not as clitics. The simple reason: etymologically, the English possessive {'s} and {s’} derive from a precursory {-es} suffix, not from any separate word or words. I hope something from this section helps at least one other person. Kent Dominic 10:07, 24 November 2019 (UTC)

"il s'est réveillé" not the best example
"Il s'est réveillé." Unless my high school French is wrong (a possibility), this sentence literally means "he woke himself." The s' functions as the reflexive pronoun (himself). It is the common way of saying he woke up in French, but it uses the reflexive and subjective pronouns both. The given translation into English ("He woke up") clouds the message because the given English translation only uses one pronoun, and it is not the one used for the clitic in French. Dcs002 (talk) 02:59, 17 December 2020 (UTC)


 * Yes, but apparently de-emphasised personal pronouns count as clitics. As for the use of reflexivity, this is a vast and complex topic dealing with agency, patiency, active/passive/ergative verb morphology and so on. We hear our friend pottering about upstairs. Aha, we say, he's up. Something or somebody induced his transition from sleeping to being awake. He woke up. He was awakened, by the alarm, some other disturbance, or simply... by himself? When the agency of the subject (I mean narrative subject, not necessarily grammatical subject) is in doubt and the true agent(s) unknown, many languages avoid the passive form and go to the reflexive. Why? Because it just seems cumbersome to have something or somebody as the (grammatical and thematic) subject of the sentence. Staying with French, examples can be multiplied at will "cela se dit ..." = that is pronounced as ... ; "se faire ..." = having ... happen to oneself. 2A01:CB0C:CD:D800:2487:4AE5:1F2C:57A2 (talk) 07:53, 6 October 2021 (UTC)

More examples
As it stands, this article reads like a conversation among academics, not an encyclopedia entry whose purpose is to enlighten the average reader. Every definition or distinction throughout this article needs at least one example if it to be understandable to those without advanced degrees in linguistics. I have a PhD in English, and I found this thicket of terminology, fine distinctions, and scholarly debate impenetrable in many places. Surely there are experts in the field with a gift for making the basics of linguistics understandable to the general public. Think Introduction to Linguistics, not graduate seminar. KC 14:30, 29 May 2021 (UTC)


 * The majority of this article already IS examples, it certainly doesn't need more of them. Some efforts could be made to make it more accessible to a broader audience, though.--Megaman en m (talk) 16:34, 29 May 2021 (UTC)

This expression in Portuguese is not as weird as the article makes it sound.
As a Brazilian Portuguese speaker, i think the whole chapter about sentences like "ele tava era gordo"  feels weird,  i disagree  with both the  meaning and explanation given.

First of all, these expressions could have their essence easily translated to English. "Ele tava era gordo", which literally translates to "he was was fat" , and means something like "Fat! That's what he was". "Ele ligou é pra Paula", which translates literally to "he called is for(pra) Paula", and means "He called Paula! that's what happened".

Putting it like this, it feels more confrontational than intense. It does feel intense, the point is strongly stated, but it has other reason to be. You may have noticed that both sentences could be interpreted as an answer to a question, and not only that, it's almost like these sentences are correcting another answer. I'm not saying another question was provided earliest at the dialog, the other answer is implicit, underlying, someone was expecting it up until the point the real answer made it obsolete .The person expecting other answer could be the person listening, someone that's not here, the person talking or everyone.

About the reason for the weird way it is arranged, and this wil be quick.

You just have to flip it: "Ele tava | era gordo" -> "era gordo +que+ ele tava" "Ele ligou | é pra ela" -> "é pra ela +que+ ele ligou"

Furthermore, you could even make sense out of the literal translation with some minor tweaks, it sounds weird only because pronouns are omitted.

(O que) ele tava era gordo -> -(What was he?) -(What) he was was fat

(Pra quem) Ele ligou foi pra Paula ->

-(Who he called?) -(Who) he called was Paula.

Hope this makes sense to someone 2804:14C:4E8:8588:80FC:DD2C:572B:D9AE (talk) 03:37, 21 May 2022 (UTC)

More examples in English
Because this is the English language Wikipedia, this article could use more examples of English language clitics before going into examples in other languages. Most readings coming to this article are not linguists and might get a bit lost in the weeds of French, Latin, Sanskrit, and other languages they do not speak or understand. Morganfitzp (talk) 13:02, 2 November 2022 (UTC)

Spanish pronoun clitics on verbs
"In Romance languages, some have treated the object personal pronoun forms as clitics, though they only attach to the verb they are the object of and so are affixed by the definition used here." This is not always true, at least in Spanish. In No puedo verla, la attaches to the verb it is the object of, but in No la puedo ver, which means the same (I can't see her), la attaches to the other verb. phma (talk) 06:57, 12 January 2023 (UTC)

italian -ale?
For example, norma is a noun, normale is an adjective but as a lemma - therefore acronim acrostic mnemonic - it means: attinente lettere elencate. It is usually defined as suffix. Sabrina.ponsi (talk) 17:05, 20 March 2023 (UTC)


 * –ale is a derivational suffix, not a clitic as I understand the word. —Tamfang (talk) 04:47, 24 September 2023 (UTC)

Endoclitic in Old Japanese?
In Japanese, the “mo” particle, expressing “even/also”, usually affects nouns, but can be attached to the connective form of a verb complemented by the auxiliary verb “suru”. It’s usually conjugated in the negative. For example:

tabe-mo shinai (eat-even doesn’t) ”doesn’t even eat”

Japanese also has compound verbs which represent a single lexical unit, such as:

toriageru (to pick up, to take up), from tori(take) + ageru(raise up) okiagaru (to get up), from oki(rise/stand up) + agaru(rise up) damarikomu (to lapse into silence), from damari(become silent) + komu(go into)

Normally, when expressing “doesn’t even” with these verbs, one would treat them like any other verb:

okiagari-mo shinai ({get up}-even doesn’t) ”doesn’t even get up”

However, in Old Japanese, instead of applying an auxliary “suru”, it could be inserted in the morpheme boundary between the two compounded verbs. This makes sense grammatically because the first verb of a compound is always in the connective form. However, in this case, the “mo” applies to the verb as a whole instead of one of its parts:

oki-mo-agarazu ({get -even- up}-not) ”doesn’t even get up”

This could arguably be classified as an endoclitic, I believe. However, it splits words at a morphological boundary rather than splitting a root. 2600:6C64:507F:AA6C:5989:A2E7:9C32:F842 (talk) 23:00, 3 April 2023 (UTC)


 * This depends on what a "particle" is in Old Japanese. If a suffix, then this is presumably a sequence of suffixes. If a clitic, then this is presumably a sequence of clitics. One of the reasons "particle" is used for joshi, I believe, is that people find it difficult to classify them grammatically. I don't think it would be warranted to give it the dubious classification of 'endoclitic' in this situation if we can't agree on what it should be classified as elsewhere. — kwami (talk) 00:26, 5 April 2023 (UTC)
 * I should clarify both “okiru” and “agaru” are verbs that conjugate. When you compound them into “okiagaru”, “okiru” is fixed in a sort of infinitive form “oki” (called renyoukei or connective/continuative form) that connects it to the verb it is compounded with, and “agaru” retains its finite verb form. The word as a whole conjugates the same as “agaru” and carries a slightly different meaning than “okiru” and “agaru” do separately.
 * In this case, neither “oki” nor “agaru” are particles, but compounded morphemes that create a word with a new meaning. “agaru” could be considered a verbal suffix, but the words it attaches to are a mostly closed set.
 * This is in contrast to “get up” in English which consists of a verb “get” and a particle “up” (arguably a predicate adjective used as the complement of get, cf “I’m up” = “I’m awake and out of bed”). So it’s not a good comparison.
 * Another example of a compound in Old Japanese that uses “agaru” is “omofiagaru” which combines the verbs “omofu” (to think) and “agaru” (to go up). The result means “to think highly of oneself; to be proud of oneself; to be self-conceited”. This meaning isn’t obvious from the component parts.
 * Particles in Japanese are similar to suffixes/clitics. I would lean towards clitic since the particle itself is not an independent word, and the resulting word after the particle attaches is not a new concept but rather the same concept marked for a particular case or semantic role in the clause. It’s very unusual for a particle to be able to split a word as “mo” does in Old Japanese, but there are other particles that do it, such as “dani” (even) (mi-dani-okuru “even see someone off”, from miokuru (miru “to see” + okuru “to guide someone somewhere; to see off”). Normally this would be “miokuri-mo su”.
 * You also see “danimo” (combination of “dani” and “mo”, more emphatic) used the same. mi-dani-mo-kaferu “to even look back”, from mikaferu (miru “to see” + kaferu “(intr) to flip over; to turn around”). Normally this would be “mikaferi-mo su”. 2600:6C64:507F:AA6C:9C99:83C2:3F12:DE84 (talk) 19:47, 6 April 2023 (UTC)
 * If we analyze the joshi as enclitics, I suppose the question is whether this is actually mo inserted into the compound verb (or rather, deverbal-verb compound) oki+agaru, or whether it is a parallel construction, deverbal oki with the NP enclitic mo followed by the verb agaru, with the same semantics as the homologous compound. That is, whether we're breaking up the compound and adding a joshi to the nominal element, oki=mo agaru. I don't know if it was true of Old Japanese, but if mo docks to the end of an NP, wouldn't that suggest the latter possibility? — kwami (talk) 00:11, 7 April 2023 (UTC)

Plea to rewrite introduction
There are a few comments below from 18 years ago to 2 years ago asking/suggesting a rewrite to the opening section. I have postgraduate education and am no stranger to technical jargon, but the opening is not easily understandable by a non linguist. I would help to have a much simpler explanation for general readers of wikipedia. Einstein said if you cant explain it to a six year old, you don't understand it yourself. I'm hoping someone who understands this subject matter can take on this challenge. — Preceding unsigned comment added by Bendav (talk • contribs) 05:06, 16 May 2024 (UTC)