In several Indo-European languages the verb that denotes possession (to have) is also used to construct verb tenses.

Some examples:

I have seen ... I have a dog. (English)

Am văzut ... Am un câine. (Romanian, am = I have. Note: this past tense in Romanian doesn't correspond directly to present perfect in English, but it's still an example of a tense formed with the equivalent to "to have".)

Έχω δει ... Έχω ένα σκύλο. (Greek, έχω = I have)

(And many others, such as Italian, French, Norwegian, etc. All sentences above have the same meaning, except for the noted difference in the precise tense.)

Since all of these languages are related, it is not that surprising that in all of them the verb to have shares these functions. I used to think that this is just an accidental idiosyncrasy of Indo-European languages, and there's no fundamental reason why it should be so.

But it turns out that Chinese (an unrelated language) is also like this:

没(有)看到 ... 我 一只狗。 (有 yǒu = have, 没有 méi yǒu = haven't)

Wǒ méi (yǒu) kàn dào... Wǒ yǒu yī zhi gǒu. (Meaning: I haven't seen ... I have a dog.)

Hungarian (my mother tongue) doesn't have anything like this (in fact it doesn't even have a verb for possession), so this double function of have is surprising and counterintuitive to me.

Question: Is there any fundamental reason why several unrelated languages would all use to have in these two seemingly separate and unrelated roles?

  • Please help with tagging this correctly.
    – Szabolcs
    Commented May 15, 2014 at 17:42
  • 1
    @hippietrail, as you might know, in (spoken) German and French, the have+past-participle form is losing (has lost?) its aspectual meaning and is being used as a pure past tense.
    – dainichi
    Commented May 19, 2014 at 5:08
  • 1
    @hippietrail, yay I found an isogloss map. Green line in map in section 3. wals.info/chapter/68
    – dainichi
    Commented May 19, 2014 at 7:59
  • 2
    The use of 有 as an auxiliary verb has a very marginal existence in Mandarin. 没(有) is used as the negative imperfective, specifically negating the completion of an event (as summarised by Li & Thompson). Hence it is incompatible with the perfective marker 了, which refers to bounded events. Constructions with 有 in the positive, which are more characteristic of southern Mandarin, are generally more common as responses to assertions with 没有, much as the English "do" auxiliary is.
    – Michaelyus
    Commented May 21, 2014 at 10:02
  • 1
    @dainichi: Ah yes that's exactly what I had felt but not known for sure about French and German! Commented May 21, 2014 at 13:57

4 Answers 4


The World Atlas of Language Structures (WALS) provides information that is relevant to this question. Feature 68A (The Perfect) is divided into perfect from possessive (i.e. from have or the like), from finish or already, as well as a category for other perfects and one for languages that don't have a perfect.

The perfect from 'have' is relatively rare with only seven related and exclusively European languages. It developed from structures like I have the work done, which were later reanalysed as having the meaning of the perfect. Of course not all languages are part of WALS, but Mandarin is, and it is categorised as not having a perfect. I'm not saying that your example from Mandarin Chinese is wrong. It's possible - though perhaps a bit unlikely - that WALS is wrong, or that your example is not a perfect after all.

The perfect from 'finish' or 'already' is in fact more frequent, with 21 languages, and distributed over different language families.

  • 3
    The WALS categorisation seems arbitrarily restrictive. It identifies perfects from possessives in Icelandic and Swedish, but not Norwegian, Danish and Faroese, all of which form perfects exactly the same way; and in German, but not Dutch, which again forms perfects the same way. It’s also completely missing Irish (and I believe Scottish) Gaelic, which can use the same construction for perfects as for possessives (tá sé agam = I have it; tá sé ráite agam = I have said it). Arguably Japanese too (私は [N]が あります = I have [noun]; 私は [Vた こと]が あります = I have [verb]ed). Commented Mar 4, 2023 at 1:02
  • 2
    The Japanese and Chinese forms are not specifically perfect, which is perhaps why they don’t appear in WALS here, but they are past and experientive in nature and play a similar role to traditional Western European perfects. Commented Mar 4, 2023 at 1:04

Robert's answer leaves us with a puzzle. Since this construction of a perfect with "have" is so rare, it would be a very strange coincidence that it is present in French, German, English, Italian, etc. just by chance. The obvious solution would be to assume that this construction in those language appeared only once, in a language parent of these modern language, that is in Indo-European or at least some dialect of Indo-European of which those languages come form. However, an instant of reflexion shows that this solution does not work. In Latin, common ancestor of French and Italian, there was no composed perfect, in particular, no perfect with "have".

To solve this puzzle, the great French linguist A. Meillet has proposed that this construction of a perfect with "have" amongst the above said-language was developed only once, probably in germanic languages, and then transmitted not genetically, but laterally, that is by borrowing, into French, Italian, etc. This may seem surprising, as borrowing usually concerns words (mostly technical words), not whole grammatical structures. However, Meillet argues, this happened in a special situation, where an unusually large part of the population in western Europe was bilingual, speaking both a germanic language and a latin language: between the fifth and ninth or tenth century AD, after the invasion of the Roman empire by Germanic tribes. We can see everyday that bilingual people tends to borrow some grammatical or syntactic construction from one of their language to another (my daughter does that all the time, between English and French). It is thus plausible that if sufficiently many people in a country are bilingual, some grammatical construction (the perfect with "have") can be borrowed to one language to the other.

  • Good to see you on linguistics.SE, Joël.
    – Olivier
    Commented May 21, 2014 at 19:42
  • Nice to see you there as well, Olivier
    – Joël
    Commented May 27, 2014 at 4:07

These analytical tense forms (perfect tenses) you refer to descend from free word combinations, only later, due to their frequent usage, they got gramaticalized and turned into the verb forms. Even now in English some traces of the expressions of have + Past Participle which are not a Perfect tense still can be found.

So, "I have written the letter." comes from the older "I have the letter written." (predicates are in bold type). This older form was something like today we can say "I have my car repaired." which is not the same now as "I have repaired my car." Generally speaking, the idea of Perfect is "as for now, the thing is done, I possess (hold in my hands, i.e. have) the thing made, the job finished."

Also note that some languages, like French and German, form the Perfect tenses from some verbs using 'to be', not 'to have', and all the Slavic languages originally had Perfect only with 'to be', never with 'to have'.

  • 1
    This is interesting, but it doesn't do anything for answering the question. Szabolcs is asking why these two seemingly-unrelated meanings are expressed with the same word in so many languages.
    – Joe
    Commented May 15, 2014 at 21:46
  • @joe - the "so many languages" mantioned are actually just 4. As for Chinese, I don't speak it, the other 3 are related, Romanian and Greek even belong to the same Sprachbund, and I'm sure they had the same evolution of the free phrases into the tense forms. Generally speaking, the idea of Perfect is "as for now, the thing is done, I possess (hold in my hands, i.e. have) the thing made, the job finished." Added that to my answer.
    – Yellow Sky
    Commented May 15, 2014 at 21:57
  • @YellowSky Romanian and Greek use "to have" in to form different tenses. Romanian uses it like Italian and Greek uses it like English. I don't believe the Balkan Sprachbund had an converging effect on this particular aspect of these languages.
    – Szabolcs
    Commented May 15, 2014 at 22:16
  • @Szabolcs - Romanian has the Perfect the same way as English, "have + Past Participle". Am I right? If no, where am I wrong about it?
    – Yellow Sky
    Commented May 15, 2014 at 22:39
  • @YellowSky The difference is in the meaning, not the form. The tense constructed that way in Romanian is used in different situation than present perfect in English. It does not focus on the outcome of the action (see my comment on the main question).
    – Szabolcs
    Commented May 16, 2014 at 17:11

I'm not a linguist, but I am a Chinese native speaker.

The "have-perfect tense" in European languages is due to the European sprachbund, as described in other answers here and in Is it unusual that English uses possessive for past tense? As mentioned by others, this feature of using the possesive verb to describe the perfect tense is rare and doesn't seem to exist outside of European languages.

As for Chinese:

The use of 有(to have) as a sort of auxilary verb is not a really standard grammatical pattern and the resemblance to European languages is a coincidence.

I think sentences like your example 我没有看到 (lit. "I no-have seen" = I haven't seen) developed as answers to questions like 你有没有看到? (lit. "you have-not-have seen?" = have you seen?), which come from the use of the expression 有没有 (lit. "have-not-have" = "do [you] have?") as a generic question word.

In other words, it seems this grammatical construct emerged in Chinese because the expression "do you have?" evolved into a generic question word that could be attached to verbs as well as nouns.

  • Thanks for the comments! So are you saying that 有没有 has no implication of some sort of "past tense", and it's merely a way to ask a question? Can you give an example of a question with 有没有 which clearly does not refer to anything in the past, and which also does not refer to possession?
    – Szabolcs
    Commented Nov 19, 2020 at 9:23
  • Well, 有没有 is not a truly general question word. It can be used with nouns and verbs, but not adjectives. With verbs, it generally refers to past/perfect but can sometimes refer to the present. For example 你有没有觉得奇怪 ?(lit. "you have? think weird?" = "do you think something is weird?")
    – Aqualone
    Commented Nov 19, 2020 at 9:32
  • 1
    Although, the more I think of it, 有没有 used in a question with a verb always carries a hint of a past tense nuance. In my example above there is a vague underlying meaning of "do you feel something is weird and did you notice something weird as the situation was developing". Again, I'm not a linguist, but my guess is this developed because Chinese doesn't have other ways of marking past/perfect tense so 有没有 was appropriated in this way (and ultimately there might be some deeper psychological/philosophical connection with the European usage: to have done = to be in possession of the action)
    – Aqualone
    Commented Nov 19, 2020 at 9:41

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.