Assuming that languages do not create complexities in vain, the existence in Mandarin of two different propositional negation devices - via “bù”, an adverb, and “méi” or “méiyou” (verbs) - seems to point to relevant differences between the negated 'predications' and possibly between the ontological properties of the ‘states of affairs/events’ that such predications describe. However, what such difference(s) may be is as yet unclear to me.

The grammars of Mandarin I have seen all roughly say that the time interval at which the state of affairs/event is described as not holding/having occurred is relevant, and in particular that ‘méi’ is used to negate sentences describing events/states of affairs that were not the case in the past, whereas “bù” is used to negate sentences describing events/states of affairs that are not the case at present or will not be the case at some future time.

However, I see no valid ontological difference between non-events/non-states of affairs in that respect, nor any obviously relevant grammatical difference between the predications that must be negated with "bù" and "méi", respectively. [One of the complications is that there happens to be a unique Chinese stative verb "you3" sometimes meaning 'have', 'possess', 'exist [for somebody]', but also equivalent to 'yes', 'is the case', 'is true' in other cases, that must nowadays be negated with "méi(you)", not "bù", in spite of the fact that it obviously refers to a currently holding state of affairs. Another is that the adverb "hái", meaning 'yet' as in E. "not yet", can co-occur with (precede) both "bù" and "méi(you)", although the sentence in which it appears cannot possibly refer to a non-fact of a past world].

My question, then, is this: are there any other languages in which a comparable distinction exists between the way sentences must be negated depending on whether they describe non-events of the past or the non-past, respectively?

2 Answers 2


Making no assumptions about Mandarin semantics, it is common in Bantu languages to have numerous tense inflections such as crastinal future, immediate future, immediate past, hesternal past, remote past and so on. These distinctions are encoded by selecting difference prefixes and suffixes on the verb. Usually, for any tense there are distinct affirmative vs. negative forms (negation involving some number of tense-determined special negative prefixes and often some special prefix forms regarding the marking of remote vs. immediate past). So in Bantu, negation is a package of morphemes (thought there are some languages which have a distinct and general negation word applicable to all tenses), and thus "I didn't see(remote) the child, although I did see(hesternal) the child" is not a contradiction.

In Finnic and Saami (and I think also Mari) there are similar distinctions, in that there is a negative verb which bears all of the inflection, and the form of the negative verb is tense-determined. Hausa has something like that, where the negative particle (partially) indicate tense-aspect and subject person. I suspect that such distinctions are very common in languages where negation is part of inflectional morphology.

  • Very interesting, thank you! As I unfortunately know nothing about any of those languages, I hope you will excuse me for further asking this: Is there any evidence as to whether the 'selection' preferences observed between Bantu tense-aspect suffixes and Bantu polarity prefixes correspond to (partly) parallel ontological distinctions between immediately/remote/etc. past/future/etc. states of affairs as conceptualized by Bantu speakers, or are they, on the contrary, merely grammatical properties comparable in status to, say, Latin declensions or grammatical gender?
    – user6814
    Commented Mar 12, 2015 at 20:38
  • I'm not sure I understand the question. Here are some examples from Kerewe: a-ka-téék-a 3s-remote-cook-infl.1 is "he cooked (remote past)", and to deny that you would say ti-y-áá-téék-ile neg-3s-rem-cook-infl.2 "he didn't cook (remote past)". I would say that it doesn't reflect different conceptualizations of events or negation depending on time reference, but rather tense-aspect-mood-polarity are a big inflectional jumble, so you have to select the correct morphemes according to tense, mood, polarity and so on.
    – user6726
    Commented Mar 12, 2015 at 21:06

Arabic (classical) has a lot of different ways of expressing negatives. laysa refers (usually) to the present time, lam to past time, lan to future time, lā and mā can refer to any time depending on the tense of the following verb. laysa is a conjugated verb, the others are immutable particles.

  • Thank you, too! May I further ask you the same question I asked user6726? Are there any hints that the different negation strategies of classical Arabic are motivated by different ways to conceptualize non-holding states of affairs depending on whether they are non-facts of past or present 'worlds' or prospective non-facts of future possible worlds?
    – user6814
    Commented Mar 12, 2015 at 20:46

Your Answer

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