A double negative occurs when two forms of negation are used in the same sentence. Multiple negation is the more general term referring to the occurrence of more than one negative in a clause.
In most logics and some languages, double negatives cancel one another and produce an affirmative sense; in other languages, doubled negatives intensify the negation. Languages where multiple negatives intensify each other are said to have negative concord. Portuguese, French, Persian, Russian, Ukrainian, and Spanish are examples of negative-concord languages, while Latin and German do not have negative concord. Standard English lacks negative concord, but it was normal in Old English and Middle English, and some modern dialects do have it (e.g. African American Vernacular English and Cockney), although its usage in English is often stigmatized.
Languages without negative concord typically have negative polarity items that are used in place of additional negatives when another negating word already occurs. Examples are "ever", "anything" and "anyone" in the sentence "I haven't ever owed anything to anyone" (cf. "I haven't never owed nothing to no one" in negative-concord dialects of English, and "Nunca devi nada a ninguém" in Portuguese, lit. "Never have I owed nothing to no one"). Note that negative polarity can be triggered not only by direct negatives such as "not" or "never", but by words such as "doubt" or "hardly" ("I doubt he has ever owed anything to anyone" or "He has hardly ever owed anything to anyone").
In standard written English, when two negatives are used in one sentence, the negatives are understood to cancel one another and produce a weakened affirmative. However, in many dialects, the second negative is employed as an intensifier and should be understood as strengthening the negation rather than removing it.
Two negatives resolving to a positive
In Standard English, two negatives are understood to resolve to a positive. This rule was observed as early as 1762, when Bishop Robert Lowth wrote A Short Introduction to English Grammar with Critical Notes. For instance, "I do not disagree" could mean "I certainly agree". Further statements may be necessary to resolve which particular meaning was intended.
Because of this ambiguity, double negatives are frequently employed when making back-handed compliments. The phrase "Mr. Jones was not incompetent." will seldom mean "Mr. Jones was very competent" since the speaker would have found a more flattering way to say so. Instead, some kind of problem is implied, though Mr. Jones possesses basic competence at his tasks.
Two or more negatives resolving to a negative
Although they are uncommon in written English, double negatives are employed as a normal part of the grammar of Southern American English, African American Vernacular English, and most British regional dialects, particularly the East London and East Anglian dialects. Dialects which use double negatives do so consistently and follow a different set of descriptive linguistic rules.
Because of their non-standard nature, such double negatives are often employed in literature and the performing art as part of characterization, particularly to establish a speaker's lower-class or uneducated status. In the film Mary Poppins, the chimney sweep Bert employs a double negative when he says, "If you don't want to go nowhere..." Another is used by the bandits in the "Stinking Badges" scene of John Huston's The Treasure of the Sierra Madre: "Badges? We ain't got no badges. We don't need no badges!" More recently,[when?] the British television show EastEnders has received some publicity over the Estuary accent of character Dot Branning, who speaks with double and triple negatives ("I ain't never heard of no license.").. In the Harry Enfield sketch "Mr Cholmondley-Warner's Guide to the Working-Class", a stereotypical Cockney employs a septuple-negative: "Inside toilet? I ain't never not heard of one of them nor I ain't nor nothing."
In music, double negatives can be employed to similar effect (as in Pink Floyd's "Another Brick in the Wall", in which schoolchildren chant "We don't need no education / We don't need no thought control") or used to establish a frank and informal tone (as in The Rolling Stones' "(I Can't Get No) Satisfaction.").
Historically, Chaucer made extensive use of double, triple, and even quadruple negatives in his Canterbury Tales. About the Friar, he writes "Ther nas no man no wher so vertuous" ("There never was no man nowhere so virtuous"). About the Knight, "He nevere yet no vileynye ne sayde / In all his lyf unto no maner wight" ("He never yet no vileness didn't say / In all his life to no manner of man").
Following the battle of Marston Moor, Oliver Cromwell quoted his nephew's dying words in a letter to the boy's father Valentine Walton: "A little after, he said one thing lay upon his spirit. I asked him what it was. He told me it was that God had not suffered him to be no more the executioner of His enemies." Although this particular letter has often been reprinted, it is frequently changed to read "not ... any" instead.
Double negation is uncommon in other West Germanic languages. A notable exception is Afrikaans, where it is mandatory. (For example, "He cannot speak Afrikaans" becomes Hy kan nie Afrikaans praat nie, "He cannot Afrikaans speak not.") Dialectal Dutch, French and San have been suggested as possible origins for this trait. Its proper use follows a set of fairly complex rules as in these examples provided by Bruce Donaldson:
- Ek het nie geweet dat hy sou kom nie. ("I did not know that he would be coming.")
- Ek het geweet dat hy nie sou kom nie. ("I knew that he would not be coming.")
- Hy sal nie kom nie, want hy is siek. ("He will not be coming because he is sick.")
- Dis nie so moeilik om Afrikaans te leer nie. ("It is not so difficult to learn Afrikaans.")
Another point of view is that this construction is not really an example of a "double negative" but simply a grammatical template for negation. The second nie cannot be understood as a noun or adverb (as can, e.g., pas in French), and cannot be substituted by any part of speech other than itself with the sentence remaining grammatical. It is a grammatical particle with no independent meaning that happens to be spelled and pronounced the same as the embedded nie, meaning "not", through historical accident.
The second nie is used if and only if the sentence or phrase doesn't already end with nie or another negating adverb.
- Ek sien jou nie. ("I don't see you")
- Ek sien jou nooit. ("I never see you")
Afrikaans shares with English the property that two negatives make a positive. For example
- Ek stem nie met jou saam nie. ("I don't agree with you." )
- Ek stem nie nié met jou saam nie. ("I don't not agree with you," i.e., I agree with you.)
While double negation is still found in the Low Franconian dialects of west Flanders (e.g., Ik ne willen da nie doen, "I do not want to do that") and in some villages in the central Netherlands such as Garderen, it takes a different form than that found in Afrikaans. In Belgian Dutch dialects, however, there are still some widely used expressions like nooit niet ("never not") for "never".
Similar to some dialectal English, Bavarian employs both single and double negation, with the latter denoting special emphasis. For example, compare the Bavarian Des hob i no nia ned g'hört ("This have I yet never not heard") with the standard German "Das habe ich noch nie gehört".
In Romance languages, negation is generally expressed by placing a negative adverb (non in Latin and Italian, no in Spanish and Catalan, não in Portuguese, ne in French, nu in Romanian) before the verb, but more negative adverbs or pronouns may appear elsewhere to indicate what kind of negation is being made.
In French, a second negative particle pas is normally employed in simple negation. Standard Catalan also uses the word "pas," but only to express emphasis or reversal of one's expectations. In Latin, passus was the word for "step", so originally, French Je ne marche pas and Catalan No camino pas meant "I will not walk a single step." In French, this initially emphatic usage spread so thoroughly that in colloquial speech, ne is often left out with pas serving as the sole negating element, such as "Je sais pas" or "sais pas" meaning "I don't know." A similar practice occurs in Northern Catalan, which omits no, and Occitan, which uses non only as a short answer to questions. In Venetian, the double negation no ... mìa can likewise lose the first particle and rely only on the second: magno mìa ("I eat not") and vegno mìa ("I know not"). The French language actually has in total five negative words and expressions. Ne pas literally means not and is used in sentences such as "Je ne sais pas" (grammatically correct usage of "I don't know"), ne jamais is the French negative word for "never" or "not ever" as in "On ne sait jamais" ("we never know"), ne rien coming from the French word "rien" meaning nothing either means no or nothing as in "Il n'y a rien," ("There is nothing" or "There are none"), "Rien n'est arrivé" ("nothing happened"), or "Non, Je ne regrette rien" ("No, I regret nothing"). Ne rien comes from the French word "rien" literally meaning "none" or "nothing." Ne plus in a neagative context means "no longer" or "no more" such as "Il n'y a plus de baguettes" ("There aren't any more baguettes"). Ne plus comes from the French word plus meaning "more," and finally ne personne means "no one" or "nobody" as in "Personne ne sait" ("Nobody knows").
In Italian, a second negative particle usually turns the phrase into a positive one, but with a different meaning. For instance Voglio mangiare ("I want to eat") and Non voglio non mangiare ("I don't want not to eat") mean "I want to eat", but the second one means more precisely "I'd prefer to eat".
Colloquial Brazilian Portuguese and Romanian often employ doubled negative correlatives. Portuguese Não vi nada, não ("I did not see nothing, no"), and Romanian Nu văd nimic ("I do not see nothing") are used to express "No, I did not see anything".
Other Romance languages employ double negatives less regularly. In Asturian, an extra negative particle is used with negative adverbs: Yo nunca nun lu viera ("I had not never seen him") means "I have never seen him" and A mi tampoco nun me presta ("I neither do not like it") means "I do not like it either". Standard Catalan also used to possess a tendency to double no with other negatives, so Jo tampoc no l'he vista ("I neither have not seen her") meant "I have not seen her either". That practice is dying out.
In spoken Welsh, the word ddim (not) often occurs with a prefixed or mutated verb form that is negative in meaning: Dydy hi ddim yma (word-for-word, "Not-is she not here") expresses "She is not here" and Chaiff Aled ddim mynd (word-for-word, "Not-will-get Aled not go") expresses "Aled is not allowed to go".
Negative correlatives can also occur with already negative verb forms. In literary Welsh, the mutated verb form is caused by an initial negative particle, ni or nid. The particle is usually omitted in speech but the mutation remains: [Ni] wyddai neb (word-for-word, "[Not] not-knew nobody") means "Nobody knew" and [Ni] chaiff Aled fawr o bres (word-for-word, "[Not] not-will-get Aled lots of money") means "Aled will not get much money". This is not usually regarded as three negative markers, however, because the negative mutation is really just an effect of the initial particle on the following word.
Doubled negatives are perfectly correct in Ancient Greek. With few exceptions, a simple negative (οὐ or μή) following another negative (e.g., οὐδείς, no one) results in an affirmation: οὐδείς οὐκ ἔπασχε τι ("No one was not suffering") means more simply "Everyone was suffering". Meanwhile, a compound negative following a negative strengthens the negation: μὴ θορυβήσῃ μηδείς ("Do not permit no one to raise an uproar") means "Let not a single one among them raise an uproar".
These constructions apply only when the negatives all refer to the same word or expression. Otherwise, the negatives simply work independently of one another: οὐ διὰ τὸ μὴ ἀκοντίζειν οὐκ ἔβαλον αὐτόν means "It was not on account of their not throwing that they did not hit him", and one should not blame them for not trying.
Modern Greek prefers double negation (Κανείς δεν μίλησε, "No one did not talk") to single (here, Ουδείς μίλησε, "None talked").
In Slavic languages other than Slavonic, multiple negatives are grammatically correct ways to express negation, and a single negative is often incorrect. In complex sentences, every part that could be grammatically negated should be negative. For example, in Serbian, Niko nikada nigde ništa nije uradio ("Nobody never did not do nothing nowhere") means "Nobody has ever done anything, anywhere", and Nisam tamo nikad išla ("Never I did not go there") means "I have never been there".
A single negation, while syntactically correct, may result in a very unusual meaning or make no sense at all. Saying "I saw nobody" in Polish (Widziałem nikogo) instead of the more usual "I did not see nobody" (Nikogo nie widziałem) might mean "I saw an instance of nobody" or "I saw Mr. Nobody" but it would not have its plain English meaning. Likewise, in Slovenian, saying "I do not know anyone" (Ne poznam kogarkoli) in place of "I do not know no one" (Ne poznam nikogar) has the connotation "I do not know just anyone": I know someone important or special.
As with most synthetic satem languages double negative is mandatory in Latvian and Lithuanian. Furthermore, all verbs and indefinite pronouns in a given statement must be negated, so it could be said that multiple negative is mandatory in Latvian.
For instance, a statement "I have not ever owed anything to anyone" would be rendered as es nekad nevienam neko neesmu bijis parādā. The only alternative would be using a negating subordinate clause and subjunctive in the main clause, which could be approximated in English as "there has not ever been an instance that I would have owed anything to anyone" (nav bijis tā, ka es kādreiz būtu kādam bijis kaut ko parādā), where negative pronouns (nekad, neviens, nekas) are replaced by indefinite pronouns (kādreiz, kāds, kaut kas) more in line with the English "ever, any" indefinite pronoun structures.
Double or multiple negatives are grammatically required in Hungarian with negative pronouns: Nincs semmim ("I do not have nothing") means "I do not have anything". Negative pronouns are constructed by means of adding the prefixes se-, sem-, and sen- to interrogative pronouns.
Double negation is required also in Finnish, which uses the auxiliary verb ei to express negation. Negative pronouns are constructed by adding one of the suffixes -an, -än, -kaan, or -kään to interrogative pronouns: Kukaan ei soittanut minulle ("No one did not call me") means "No one called me".
Japanese employs litotes to phrase ideas in a more indirect and polite manner. Thus, one can indicate necessity by emphasizing that not doing something would not be proper. For instance, しなければならない (shinakereba naranai, "must") literally means "not doing [it] would not be proper". しなければいけません (shinakereba ikemasen, also "must") similarly means "not doing [it] cannot go forward".
Of course, indirectness can also be employed to put an edge on one's rudeness as well. "He has studied Japanese, so he should be able to write kanji" can be phrased 彼は日本語を勉強したから漢字で書けないわけがありません (kare wa nihongo o benkyō shita kara kanji de kakenai wake ga arimasen), there is a rather harsher idea: "As he has studied Japanese, the reasoning that he cannot write Kanji does not exist".
Mandarin Chinese also employs litotes in a like manner. One common construction is 不得不 (Pinyin: bùdébù, "cannot not"), which is used to express (or feign) a necessity more regretful and polite than that expressed by 必须 (bìxū). Compared with "我必须走" (Wǒ bìxū zǒu, "I need to go"), "我不得不走" (Wǒ bùdébù zǒu) tries to emphasize that the situation is out of the speaker's hands and that the speaker has no choice in the matter: "Unfortunately, I got to go". Similarly, "没有人不知道" (Méiyǒu rén bù zhīdào, "There is not a person who does not know") is a more emphatic way to express "Everyone knows".
Double negatives nearly always resolve to a positive meaning even in colloquial speech and, triple negatives can as well. Saying "我不相信没人不来" (Wǒ bù xiāngxìn méi rén bù lái, "I do not believe no one will not come") means "I think everyone will come". However, triple and multiple negatives are considered obscure and are typically avoided.
History of languages
Many languages, including all living Germanic languages, French, Welsh and some Berber and Arabic dialects, have gone through a process known as Jespersen's cycle, where an original negative particle is replaced by another, passing through an intermediate stage employing two particles (e.g. Old French jeo ne dis → Modern Standard French je ne dis pas → Modern Colloquial French je dis pas "I don't say").
In many cases, the original sense of the new negative particle is not negative per se (thus in French pas "step", originally "not a step" = "not a bit"), but in Germanic languages, such as English and German the intermediate stage was a case of double negation, as the current negatives not and nicht in these languages originally meant "nothing": e.g. Old English ic ne seah "I didn't see" >> Middle English I ne saugh nawiht, lit. "I didn't see nothing" >> Early Modern English I saw not.
A similar development to a circumfix from double negation can be seen in non-Indo-European languages, too: for example, in Maltese, kiel "he ate" is negated as ma kielx "he did not eat", where the verb is preceded by a negative particle ma- "not" and followed by the particle -x, which was originally a shortened form of xejn "nothing" - thus, "he didn't eat nothing".
- Fromkin, Victoria; Rodman, Robert & Hyams, Nina (2002). An Introduction to Language, Seventh Edition. Heinle. p. 15. ISBN 0-15-508481-X.
- Merriam-Webster's Collegiate Dictionary, 11th Edition. Op. cit. Merriam-Webster Online. Accessed 29 Sept. 2010.
- The American Heritage Dictionary of the English Language, 4th ed. Houghton Mifflin Co, 2009. Accessed 29 Sept. 2010.
- "double negative". Memidex/WordNet Dictionary. Retrieved 2012-06-11.
- Fraser, Antonia, Cromwell Lord Protector, at 129, Primus, New York, NY 1973 ISBN 0-917657-90-X.
- John Forster, STATESMEN OF THE COMMONWEALTH OF ENGLAND WITH A TREATISE ON THE POPULAR PROGRESS IN ENGLISH HISTORY, pp. 139–140.
- A Grammar of Afrikaans, Bruce C. Donaldson, Walter de Gruyter, 1993, page 404
- Borsley, Robert; Tallerman, M & Willis, D (2007). "7. Syntax and mutation". The Syntax of Welsh. Cambridge University Press. ISBN 978-0-521-83630-2.
- Kastovsky, Dieter. 1991. Historical English syntax. P. 452
- Van Gelderen, Elly. 2006. A history of the English language. P. 130
- Grazio Falzon. Basic Maltese Grammar