In English, the verb "do" can be a transitive verb whose object stands for an action. So, we English speakers can "do a somersault," "do a back flip," and "do a cartwheel." The productivity of this construction is limited, but it does suggest a question: are there any languages in which a small and (synchronically) fixed inventory of verbs is used with other parts of speech (for example, nouns that stand for actions or states) to create phrasal equivalents to the many verbs that are found in most languages?

  • 2
    This isn't exactly what you're asking for, but Basque has two types of verbs, those which have both finite and non-finite forms (of which there are usually only 10 or so) and those which only have finite forms, which need to be inflected periphrastically using verbs of the previous category as auxiliaries. I 'm pretty sure that both are still considered verbs, though. Jul 6, 2012 at 19:43
  • 1
    @thylacine222: did you flip "finite" and "non-finite" in your comment? Feb 23, 2015 at 5:21

8 Answers 8


Yes: Japanese, Farsi, and Basque are well-known examples. Japanese verbs (and adjectives) are closed class, with new verbal senses almost exclusively expressed by “do verbal noun”, as in 勉強する benkyō suru (studying do) for “study”. This is conspicuous in Japanese due to the large number of Sino-Japanese words, with verbal senses expressed in this way.

Not as extreme as Jingulu, admittedly, but a bit better-known.

Note: Japanese verbs have opened slightly in recent years, where the suffix 〜る -ru is added informally as a verbifier, primarily to recent borrowed words, but this is both recent and marginal: it is not found in the Sino-Japanese lexicon. The best-known and oldest example is サボる sabo-ru “cut class, cut work”, from French sabotage, which dates from circa 1920, and this appears to be the model for the more recent ones.

If this word-formation pattern persists, this may be an interesting case of a word class opening up, or may instead re-analyzing 〜る as a separate element, similar to existing する suru “do”, which is the existing verb used in these constructions.

  • 5
    I immediately thought of Japanese as well upon seeing this question. However, I think it's valuable to note that even in this case it's a tendency, not an absolute prohibition on new verbs: here's an interesting blog post on "guguru", "saboru" and the like: guidetojapanese.org/blog/2005/03/14/… Feb 23, 2015 at 5:26
  • 3
    Yes, informally verbing with -r-u is quite productive—there are a lot of established examples and informally, including nonce coinages, there are thousands of examples. Typically the source words are clipped and any long vowels are shortened, and final ru is added if it isn't already present; these words always conjugate as consonant stem verbs, even if they fit the shape of possible vowel stem verbs.
    – user2081
    Feb 23, 2015 at 7:13
  • Thanks both, that’s a good and valid point. To underline: traditionally verbs (and even more adjectives) have been a closed class in Japanese. These recent 〜る -ru verbs are marginal. Perhaps they will persist and Japanese verbs will be open, perhaps this suffix will be re-analyzed. Updated post to reflect your comment. Feb 25, 2015 at 5:18
  • @snailplane, there are also quite a few i-verbs that came from loans in Japanese at the moment. Some sound rather humorous but others are pretty much accepted in the lexicon. It seems to be a somewhat productive process to reinterpret na-nominals as i-verbs.
    – Zorf
    Jul 7, 2020 at 16:19
  • Do you have a citation? Oct 29, 2020 at 21:43

Yes, there are at least two: Jingulu and Igbo. Quoting (Soschen, 2008):

Certain languages have a very restricted number of verbs. For example, the aboriginal language Jingulu spoken in Australia has only three verbs: do, go, and come. Igbo (Ibo), a language of approximately 18 million speakers in Nigeria, does not have verbs at all. Instead, Igbo uses clusters termed ‘inherent complement verbs’ (ICV) that have the structure –gbá plus a noun. The root gbá is the only root in Igbo “devoid of meaning”, and the most productive one (Chinedu Uchechukwu, p.c.; see also Uchechukwu 2004). Here are some examples of Igbo clusters: –gbá egwú ‘dance a dance’ — egwú ‘dance’, –gbá igwè ‘ride a bicycle’ — igwè ‘bicycle’, –gbá ákụ́/egbè ‘shoot’ — àkụ́ ‘arrow’/égbè ‘gun’, gbá ụkwụ́ ‘kick’ — ụ́kwụ ‘foot’, –gbá ọsọ ‘run a race’ — ọsọ ‘race’, –gbá motò ‘travel with a vehicle’ — motò ‘vehicle’, etc.

  • 2
    The Igbo claim is preposterous.
    – user483
    Jul 6, 2012 at 18:47
  • 5
    I mean Soschen's claim that it "does not have verbs at all."
    – user483
    Jul 6, 2012 at 19:20

With regard to this statement: "Igbo (Ibo), a language of approximately 18 million speakers in Nigeria, does not have verbs at all." There is a serious misunderstanding somewhere. Please note the following groups of verbs in the Igbo language

(1) a limited number of simple verb roots

(2) a combination of these "simple verb roots" to form "compound verbs"

(3) a combination of a simple verb root with a suffix to form a "complex verb"

(4) a combination of a simple verb root with a noun/prepositional phrase schematically summarized as [VERB+NP/PP]

It is the last group that is called the INHERENT COMPLEMENT VERB structure and which were cited by Soshen above. Many Igbo verbs are formed from these three, but the inherent complement structure is the most productive. The earlier investigators of the structure noted this fact of ONE verb root combining with MANY NP/PPs to form what in an average European language would be expressed through a single verb, for example –gbá ọsọ simply means ‘run', but also includes 'run a race’. For this particular reason, the earlier investigators then called these formations from one verb the clusters of that verb. That is how the name of, for example, -gba cluster arose referring to ALL the inherent complement verbs formed by the verb root -gba as belonging to the -gba cluster. Each verb root therefore was understood as having its own cluster of inherent complement verbs formed from it.

Chinedu Uchechukwu

Department of Linguuistics

Nnamdi Azikiwe University, Awka

Anmabra State


  • 4
    Welcome to Linguistics.SE. I can see that you are providing important information here. However, you are only commenting on one of the other answers; you are not answering the question above. That is discouraged on Stackexchange sites. If limit your answer to less than 600 characters, I can convert it to a comment on the appropriate answer. Otherwise, I'm afraid your answer will soon get flagged for deletion, for not answering the question.
    – prash
    Jun 26, 2015 at 12:03

There is a term for this, namely "light verb", which may help you to Google examples. This is actually separate from the question of verbs being a closed class. Such constructions are common in Kurdish and Indic languages (and Farsi per Nils von Barth, so see this for focus on Indic), and some Bantu languages. Korean starkly contrasts with Bantu languages in the potential for loans. Apparently Korean absolutely resists borrowing verbs, but Bantu languages so not, though both have light verb constructions. So I think the question of being "closed class" is different from having relatively open light verb constructions.


There are many languages wherein the part of speech category 'verb' (inflecting for person-number and TAM) is a closed class. These languages are common in north Australia, some having as few as five verbs and others up to 250. In New Guinea this is a common feature of languages of the Trans New Guinea language family, which typically have 60–150 verbs.

Verbs with closed verb classes make up for this lack by having large, open classes of complex predicates created by combining verbs with other parts of speech. Many north Australian languages have an open POS 'coverb' (sometimes called 'preverb'), which is a non-inflecting lexeme that combines with an inflecting verb to form a specific semantic unit. The Trans New Guinea language Kalam has an open class POS known as 'verb adjunct' which operates in a similar way to the north Australian coverbs, combining with inflecting verbs to provide specific meanings. Kalam also employs a rich system of serial verb constructions, which provide another way of representing events.

Pawley (2006) provides an overview of complex predicates in Jaminjung (north Australia) and Kalam (Trans New Guinea family) and compares them with English, finding striking similarities in regard to the frequency and meaning of particular concepts expressed by inflecting verbs.

  • To be fair though, te definition “(inflecting for person-number and TAM)” is somewhat circular, because it's not called “tense”, “aspect”, or “mood” if it not be on a verb. Quite a few languages for instance express the same semantics with an adverb that is called a mood in another language — nouns also inflect for person and number in many languages. I don't believe there is actually a rigorous unified definition of “verb”, and many languages have parts of speech about which one might debate whether they are verbs or not.
    – Zorf
    Jul 7, 2020 at 16:14
  • @Zorf I'm not saying 'inflecting for TAM' is a universal definition of verb, but rather that in the languages I'm discussing there is a closed category which behaves in that way and have the appropriate semantics and combinatorial behaviour such that they are given the label 'verb'. But it's true there is no universal, cross-language definition of 'verb' and I don't think any linguists think there could be such a definition. There is indeed continuing debate about POS categories, especially in those languages which seem to have precategorial lexemes, eg Samoan. Jul 8, 2020 at 5:08

Like what Nils has said , this too is pretty common in the korean language.. due to welll.. large amounts of sino words imported from chinese.

the korean equivalent of "to do study"

study do practice do. 공부하다. 연습하다.

Edit: i was told by curiousdanni that the above mentioned might not constitute as a close verb

  • 1
    Just because the translation of "to study" into Korean is literally "to do study" does not mean that the verb class is closed! All it means is that one particular action is expressed with the Korean equivalent of 'to do'! Please edit this to add much more evidence that it is in fact closed.
    – curiousdannii
    Mar 2, 2015 at 5:20

Turkish has lots of those kind of verbs: kılmak, etmek, eylemek, yapmak, olmak, saymak. They can be used for all kinds of nouns, but especially they are used with loan words to indicate verbal form of it. Like: farz etmek - to assume. (farz is an arabic loanword meaning assuming)

Although it should be noted that most of the loanwords have Turkic equivalent. But both loan and Turkic words are used for different situation.

  • I think I understand what you mean, but you might want to clarify your answer. Does Turkish have a fixed number of verbs whose meanings are completed by an open class of nouns that serve as the verbs' complements? May 7, 2017 at 17:17
  • I didn't understand whether you mean number of words in the whole language, or the verbs with uncleared meaning are fixed. I assumed second one. For the second one the words I listed are the only ones. They don't have clear meaning (some of them almost meaningless) without a noun.
    – kabraxis
    May 7, 2017 at 20:40

Hardly any language in modern times will have an entire verb lexicon to be a closed class (or more accurately, a finite set) semantically and pragmatically. Just like nouns, it's just not in the nature of verbs to be so.

In Yorùbá, there is a closed class of cannonical verbs only identifiable by morphology, namely a CV structure (e.g. bà, bá, ge, sè, rí, kà, gbó, etc... note that [gb] is one sound, a voiced labio-velar plosive). Accounting for the tone on V and possible C+V combinations, they total 540 (Déchaine, 2015).

These are probably the first ever predicates of basic states and events in the language from a diachronic point of view. They are highly and richly polysemous as well as combinatorial (with nouns, with other verbs in a serial verb construction, etc), which can cause a lot of headache for a foreign learner.

The issue with the Igbo language is totally wrong, as has been clarified for us by Chinedu Uchechukwu above.

  • 1
    If you were wondering about the comment about deletion, you can read the tour here or the help center here about the website. Since this is a question and answer site, non-answers are routinely deleted, not as a punishment against the answerer.
    – b a
    Jun 10, 2018 at 18:21
  • This is clearly an answer and not just a comment. Have we? No, we haven't and it's exceedingly unlikely (implied: a certain pattern of verbal phrase may count as verb; granted, this could use commitment to a clarification, but maybe it was better to leave it open, both literally and figuratively speaking).
    – vectory
    Jul 5, 2020 at 8:57

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.