Linguistics Of Negation Research Paper

(1) The King of France is not bald.

If one has any occasion to use sentence (1), one will use it to deny or contradict the view that the King of France is bald. One communicates this negative meaning with a linguistic sign, namely, with the word not. This word can also be called a ‘negation,’ or a ‘negative marker’ or ‘negator.’ Obviously, the meaning and the form are related: one could say that the English word not is dedicated to the expression of a negative meaning. However, the dedication is not complete. Not every use of a negative marker actually conveys a negative meaning.

(2)       Je crains que Jean ne vienne.

I fear that Jean NEG come. SUBJ.PRS.3SG

‘I fear that John is coming.’

[Abbreviations: ACC accusative; AFF affirmative; COMP completive; DISTPST distant past; FFNLV fixed form of negative lexical verb; GEN genitive; NEG negation; OBJ object; PFV perfective; PL plural; POT potential; PRS present; PRT particle; PST past; SBJ subject; SG singular; SUBJ subjunctive.]

In French, as in sentence (2), the complement of a verb of fearing may contain the negative element ne, but what the speaker fears in (2) is not that John would not be coming, but rather that he would. Conversely, not every transmission of a negative meaning needs a unique dedicated negative marker. In (3) the same negative meaning allows two strategies, and one is not more dedicated than the order.

(3)       (a) I saw nothing.

(b) I didn’t see anything.

In what follows, the focus will first be on meaning, and then on form.

1. Meaning

1.1 Scope And Focus

In Shakespeare’s Julius Caesar, Brutus has to explain to the citizens of Rome why he killed Caesar. His speech includes the following words:

(4) Not that I loved Caesar less, but that I loved Rome more.

Sentence (4) starts with the negator not, followed by that I loved Caesar less, which is a complete statement or proposition. This proposition is said to be ‘within the scope of negation.’ It is a statement that some citizens of Rome might have made in the context of Brutus’ murder of Caesar: if one murders someone, one probably does not care for the victim too much. We here hit upon a typical feature of negation: when a proposition is negated, it is usually entertained in the context or even mentioned explicitly.

Sentence (4) illustrates another typical feature: After the negative sentence, i.e., after establishing that lack of love was not that motive for the assassination, Brutus goes on to identify the real motive, the love of Rome. Without the part of the sentence that starts with but, the audience would not be satisfied. In general, people are more interested in what is the case than in what is not the case, and in the context at hand, the Romans are more interested in what motivated Brutus to kill Caesar than in what did not motivate him.

Finally, note that though not has scope over an entire statement, the negation does not really concern each constituent. The but that I loved Rome more part still involves Brutus as a subject and it still involves Brutus’ love for someone or something. What is denied is only the degree of Brutus’ love for Caesar, compared to his love for Rome. If one wants, one could call the degree of love the ‘focus’ of the negation, but note that the focus is not actually reflected in the syntax. In (4) the syntax is vague with respect to the focus of the negation. And even when we integrate the negation into the sentence, as in (5), the focus may still be vague syntactically.

(5)       I did not love Caesar less,

(a) … I loved Rome more.

(b) … Cassius did.

(c) … I loved Antony less.

Even though in (5) the negation is associated syntactically with the verb, it does not necessarily negate the verb. It remains a ‘sentence negation.’ This is different in (6), which is part of what Mark Antony has to say to the Romans:

(6) I come to bury Caesar, not to praise him.

In (6) the scope is still a whole sentence: Mark Antony denies the truth of I came to praise Caesar. But this time the focus of the negation is much clearer: it is the constituent to praise Caesar. This use of the negation can be called ‘constituent negation.’

The study of negation and its relation to the truth of the sentence is not only a concern for linguists. It has been on the agenda of logic for as long as the discipline has existed. Within logic, one can approach negation from the perspective of scope and then one is necessarily doing propositional logic, and negation will be a propositional operator. When one looks at the focus of negation, however, one can enter the internal structure of the proposition and can enrich the study of quantifiers (all, some) with the study of negative quantifiers (not all or no, not some). A term for the part of the scopal proposition that is out of the focus of negation is ‘presupposition.’ The term was introduced into twentieth-century philosophy of language by Strawson (1950). It was a much debated notion in the 1970s, and attempts were made to turn it into a pivotal notion in the study of the interface between semantics and pragmatics (e.g., Levinson 1983).

1.2 Contrariness And Contradiction

(7)       (a) Mark Antony was unhappy.

(b) Mark Antony was not happy.

In both versions of the sentence in (7) arguably we are dealing with constituent negation: in both, the speaker denies the appropriateness of the predicate ‘happy.’ However, the two versions can differ in meaning. In (7a) Mark Antony is reported as being sad. (7b) is vague: Mark Antony could indeed be unhappy too, but he could also be neither happy nor unhappy. Both ‘unhappy’ and ‘neither happy nor unhappy’ are instantiations of ‘not happy,’ and ‘not happy’ is called the ‘contradictory negation’ of ‘happy.’ On the other hand, ‘unhappy’ is the ‘contrary’ negation of ‘happy.’

Like the distinction between sentence negation and constituent negation, the one between contradiction and contrariness has existed since the beginnings of logic, most prominently since the formulation of Aristotle’s Square of Oppositions.’ A version of this square, as applied to quantifiers, is shown in Fig. 1.

Just as ‘unhappy’ is the contrary of ‘happy,’ ‘all not’ is the contrary of ‘all,’ and just as ‘not happy’ and ‘happy’ exhaust all logical possibilities and are thus each other’s contradictories, so ‘not all’ and ‘all’ are contradictories.

Notions such as contrariness and contradictoriness have been a concern to both logicians and linguists. What interests linguists is that these notions offer a framework for the study of lexicalization patterns. Just as English unhappy is a lexicalization of the contrary of happy, English all has its contrary lexicalized as no. The contradictory ‘not happy’ does not allow for a lexicalization, and neither does ‘all not.’ What these examples illustrate is that contrariness lexicalizes more easily than contradictoriness, which is not to say that the latter does not lexicalize at all (thus English non-, different from both unand in-, often produces contradictories, e.g., nonscientific). Linguists have also made it clear that contradictory notions often change their meanings into contrariness. A classic case, known since Tobler (1882), is French il ne faut pas. It used to mean ‘it is not required,’ which is the contradictory of ‘it is required,’ but it now only means ‘it is forbidden,’ which is the contrary of ‘it is required.’ The linguist who has done most for making the logical tradition relevant for the linguistics of negation, and for discussing issues of lexicalization and polysemy, is Laurence R. Horn (esp. 1989). Horn has also looked at what has come to be known as ‘negative raising’ from this perspective. What happens with negative raising is that a negation whose meaning is felt to belong to a complement clause is actually expressed with the complement-taking main verb.

(8)       (a) I don’t suppose that you believe him.

(b) I suppose that you don’t believe him.

Using a transformational perspective, the negation can be seen as being ‘raised’ from the lower clause (you belie e him) to the higher clause (I suppose). The point that is relevant here is that the two versions of the sentence can be characterized in terms of the distinction between contrariness and contradictoriness. The raised (8a) is literally a contradictory negation (either one supposes something or one does not), but it is normally equivalent to the non-raised (8b), and it then conveys a contrary meaning (instead of supposing p and supposing not p, one could also suppose neither p nor not p).

2. Form

The distinction between unhappy and not happy has a formal dimension, too. In unhappy the negation is expressed with morphology, more specifically with a prefix. In not happy the negative marker is an independent word, an adverb or particle—one can speak about ‘syntactic negation.’ Semantics and form are

 again related. With unhappy vs. not happy, affixal negation is associated with contrariness, and syntactic negation is associated with contradictoriness. The relation is not very tight, however. The syntactically negated il ne faut pas que, after all, expresses a contrary meaning in present-day French. Conversely, in nonscientific, the prefixal non-strategy results in contradictoriness.

Cross-linguistically, negating with morphology and negating with a separate syntactic element are the two basic strategies. Sound or word order differences may have a function too, but they rarely or never mark negation by themselves. Both morphological and syntactic negation exhibit subtypes. The linguist’s increased understanding of the typology of morphological and syntactic negation is related directly to the increased importance of typology in general, in the wake of Greenberg (1966). Important works in the typology of negation are Dahl (1979), Dryer (1988), Forest (1993), Honda (1996), Kahrel (1996), Miestamo (2000), and, somewhat indirectly, Haspelmath (1997).

As to syntactic negation, there are two subtypes: either the negator is an uninflected element (adverb or particle) or it is a verb (and, depending on the criteria used, an auxiliary verb). English not is an example of an uninflected element. For an example of a negative verb, one can turn to the Tungusic language, Evenki. Evenki has a negative verb, whose stem is e-, which shows agreement with the subject in person and number, and which furthermore inflects for tense, aspect, and mood. It combines with what Nedyalkov (1994) calls the ‘fixed form of the negated lexical verb’ (FFNLV).

(9)       Si min-e e-che-s doldi-ra.

You I-ACC NEG-PST-2SG hear-FFNLV

‘You did not hear me.’ (Nedyalkov, 1994 p. 16.)

Negative verb forms are arguably also the English form don’t, as in (10), or the classical Latin negative imperative construction with noli in (11).

(10)     I don’t like reggae.

(11)     Noli venire.

‘Don’t come.’

The Evenki negative verb differs from the English and Latin ones, however, in that a morphological analysis of the latter will allow one to identify a morphological negator, too. The element don’t contains the morphological (enclitic) negator n’t, and noli is the imperative of the negative nolle ‘not want,’ which is a contraction of velle and a proclitic or prefixal ne-. This does not detract from the fact that don’t and noli count as syntactic negators. To negate (12) and (13), the morphological negator cannot attach to the main verb, nor can it be replaced by a negative particle.

(12)     (a) I like reggae.

(b) *I liken’t reggae.

(c) *I like not reggae.

(d) *I not like reggae.

(13)     (a) Veni.

‘Come’

(b) *Neveni.

(c) *Non veni.

(d) *Veni non.

It has been argued that if a language has a special negative verb, frequently it will be an existential verb or a verb of wanting, and that these specialized constructions may lose their verbal properties and thus turn into particles (Payne 1985 p. 222, Croft 1991). It has also become clear (esp. through Honda 1996) that when a language employs a special verb to express negation, this special verb does not itself have to be negative. In the Carib language, Hixkaryana, negation can be expressed with an affix on the verb, but this construction further requires a positive copula ‘be.’

(14)     (a) Koso w-ono-ye

deer 1SBJ 2OBJ-eat-DISTPST

‘I ate the deer.’

(b) Koso y-ono-hra w-exe-ye.

deer 3 eat-NEG 1-be-DISTPST

‘I didn’t eat the deer.’ (Derbyshire 1979

48. 48.)

A literal paraphrase of (14b) is ‘I was not without eating the deer,’ and there is a sense in which the sentence is not really negative.

Note that each of the syntactic negators illustrated thus far precedes the main verb. There seems to be a consensus that languages generally prefer to have their syntactic negation in front of the main verb (Dryer 1988, Dahl 1979). However, a postverbal position is by no means rare, as in Dutch (15), or as when negation is sentence-final, i.e., post-everything, and hence also postverbal, e.g., in the Chadic language, Ngizim, in (16).

(15)     Ik hou niet van reggae.

I hold NEG of reggae

‘I don’t like reggae.’

(16)     Ii naa duuka bai.

I have horse NEG

‘I don’t have a horse.’ (Thompson 1998 p. 328)

Syntactic negation may also involve a part that precedes and a part that follows. This is often called ‘discontinuous’ or ‘embracing’ negation. It is illustrated with French (17), be it that the preverbal part ne is a clitic is thus intermediate between morphology and syntax.

(17)     Je ne le lui donne pas.

I NEG it him give NEG

‘I don’t give it to him.’

Discontinuous negation is particularly perspicuous and well-studied in European languages (Bernini and Ramat 1996). The locus classicus is Jespersen (1917), and what has come to be known as ‘Jespersen’s cycle’ is the view that languages may go from a preverbal system to a postverbal system via an embracing negation stage, in which the postverbal element functions originally as a reinforcement.

Whether the Negation-before-Verb preference also holds for morphological negation is not clear (Dahl 1979 p. 82 vs. Bybee 1985 p. 177). In any case, there are again three types: (a) a prefixal type, already illustrated with Latin noli, (b) an affixal type, as in English isn’t, and (c) a circumfixal type, as in Cairene Arabic.

(18)     Ma-katab-m-s.

NEG-write. PFV-3PL-NEG

‘They didn’t write.’ (Bernini and Ramat 1996

47. 47.)

A peculiar manifestation of morphological negation is the ‘zero morpheme.’ In South Dravidian languages, negation can be expressed by the mere absence of a tense suffix. In Havyaka Kannada, spoken in the coastal areas of the Indian state of Karnataka, these negative forms are frequent and have furthermore acquired a modal meaning. The Havyaka verbal paradigm makes a distinction between indicative and subjunctive, and between past and nonpast. In (16a–b) the verbal forms have a tense mood slot filled by the nonpast indicative and subjunctive forms. In (16c) this slot is unfilled, and this in itself marks the structure as negative subjunctive (van der Auwera and Bhat 1999).

19        (a) ma:sTrakko kate o:du-tt-avu

teachers story read-IND.NPST-3PL

‘The teachers are reading a story.’

(b) ma:sTrakko kate o:du-g-u

teachersI story read-SUBJ.NPST-3PL

‘The teachers may be reading a story.’

(c) ma:sTrakko kate o:d-Ø-avu

teachers story read-SUBJ.NEG-3PL

‘The teachers may not be reading a story.’

Zero negatives may be rare, but they are not exclusive to Dravidian. They have also been reported for the Tupı language Karitiana, spoken in the Amazon Basin. According to Landin (1984), affirmative sentences mark verbs with a preverbal particle or clitic expressing affirmation. In negative sentences, however, there is no such prefix or clitic.

(20)     (a) Y ta-oty-j yn

I AFF-bathe-tense I

‘I will bathe.’

(b) Y oty yn

I bath I

‘I will not bathe.’ (Landin 1984 p. 237.)

Note that the verbs in the Karitiana and Havyaka Kannada negative sentences lack tense markers. This touches upon a large area of investigation: the relation between negation and other categories. A forceful illustration of the intricacy of this relation is provided by the Mixtecan language, Copala Trique. The illustration makes the following points: (a) a language may have more than one negative marker, (b) the choice between them may be related to the choice of other categories, in this case a tense–aspect–mood (TAM) category, and (c) the compositionality of joining the negative marker and the TAM marker need not be very transparent. In (21a) the prefix gmarks completiv e TAM. The language has two negative markers, ne³ and ze⁴. The construction in (21a) combines only with ze⁴, but then the TAM changes: as (21b) shows, the completive becomes what Hollenbach (1976 p. 126) calls ‘potential’—the superscripts in the examples mark tone contours.

(21)     (a) g-ucuh³⁴ zini³ yuwe²¹ ah

COMP-lay boy palm.mat PRT

‘The boy laid the palm mat d own.’

(b) ze⁴ g-ucuh³⁴ zini³ yuwe²¹ ah

NEG COMP-lay boy palm.mat PRT

‘The boy won’t lay the palm mat down.’

To reach the meaning ‘The boy didn’t lay down the palm mat, ’ one has to start from a potential aspect.

(21)     (c) ne³ g-ucuh⁴ zini³ yuwe²¹ ah

NEG POT-lay.POT boy palm.mat PRT

‘The boy didn’t lay the palm mat down.

Negation may also relate to the shape and the meaning of referring expressions. In Russian, for instance, objects that take Accusative cases in positive environments may allow or prefer Genitive case in negative environments.

(22)     (a) On soxranil podlinnik pis’m-a

he preserved original.ACC letter-GEN

‘He preserved the original of the letter.’

(b)       On ne soxranil

he NEG preserved

podlinnik/podlinnik-a pis’ma

original.ACC/original-GEN letter-GEN

‘He didn’t preserve the original of the

letter.’ (Timberlake 1993 p. 869.)

Or consider English (23): the word something is normally replaced by (c) did not … anything or by (d) nothing, a strategy that seems typical for Europe (Kahrel 1996, Haspelmath 1997), and even by (e) didn’t … nothing, which is substandard in English but grammatical in many other languages.

(24)     (a) I heard something.

(b) ?I didn’t hear something.

(c) I didn’t hear anything.

(d) I heard nothing.

(e) I didn’t hear nothing.

This leads to a topic that has received a lot of attention from semanticists, namely, ‘positive polarity,’ the preference words and constructions may exhibit for nonnegative environments (as with some-thing in (23a)) and especially ‘negative polarity,’ the preference for negative (and quasi-negative) constructions (as with anything) (van der Wouden 1997).

Bibliography:

1. Bernini G, Ramat P 1996 Negative Sentences in the Languages of Europe. Mouton de Gruyter, Berlin
2. Bybee J L 1985 Morphology. A Study of the Relation Between Meaning and Form. Benjamins, Amsterdam
3. Croft W 1991 The evolution of negation. Journal of Linguistics. 27: 1–27
4. Dahl O 1979 Typology of sentence negation. Linguistics 17: 79–106
5. Derbyshire D 1979 Hixkaryana: Lingua Descriptive Studies. North-Holland, Amsterdam
6. Dryer M S 1988 Universals of negative position. In: Hammond M, Moravcsik E, Wirth J (eds.) Studies in Syntactic Typology. Benjamins, Amsterdam
7. Forest R 1993 Negations: Essai de syntaxe et de typologie linguistique. Klincksieck, Paris
8. Greenberg J H (ed.) 1966 Universals of Grammar. MIT Press, Cambridge, MA
9. Haspelmath M 1997 Indefinite Pronouns. Oxford University Press, Oxford, UK
10. Hollenbach B E 1976 Tense-negation interplay in Copala Trique. International Journal of American Linguistics 42: 126–32
11. Honda I 1996 Negation: A Cross-linguistic Study. PhD diss. University of New York, Buffalo, NY
12. Horn L R 1989 A Natural History of Negation. University of Chicago Press, Chicago
13. Jespersen O 1917 Negation in English and Other Languages. A. F. Host, Copenhagen, Denmark
14. Kahrel P 1996 Aspects of Negation. PhD diss. University of Amsterdam
15. Landin D J 1984 An outline of the syntactic structure of Karitiana sentences. In: Dooley R A (ed.) Estudos Sobre Linguas Tupi do Brasil. Summer Institute of Linguistics, Brasilia, Brazil
16. Levinson S C 1983 Pragmatics. Cambridge University Press, Cambridge, UK
17. Miestamo M 2000 Toward a typology of standard negation. Nordic Journal of Linguistics 23: 65–88
18. Nedyalkov I 1994 Evenki. In: Kahrel P, Van den Berg R (eds.) Typological Studies in Negation. Benjamins, Amsterdam
19. Payne J R 1985 Negation. In: Shopen T (ed.) Language Typology and Syntactic Description. Vol. 1. Clause Structure. Cambridge University Press, Cambridge, UK
20. Strawson P F 1950 On referring. Mind 69: 320–44
21. Thompson S A 1998 A discourse explanation for the cross-linguistic differences in the grammar of interrogation and negation. In: Siewierska A, Song J J (eds.) Case, Typology and Grammar. In Honor of Barry J. Blake. Benjamins, Amsterdam
22. Timberlake A 1993 Russian. In: Comrie B, Corbett G C (eds.) The Slavonic Languages. Routledge, London
23. Tobler A 1882 Vermische Beitrage zur franzosischen Grammatik. Erste Reihe. Hirzel, Leipzig, Germany
24. van der Auwera J, Bhat D N S 1999 Havyaka Kannada: Negation and modality. Indian Linguistics 60: 1–26
25. van der Wouden T 1997 Negative Contexts: Collocation, Polarity and Multiple Negation. Routledge, London