1. Defining the values

This map shows the nature of morphemes signalling clausal negation in declarative sentences. By clausal negation is meant the simple negation of an entire clause; (1b) is the negative clause corresponding to the affirmative clause in (1a) 

(1) 

 

Not considered here are noun phrase negation (No students were present), negative pronouns (Nobody came), or negative adverbs (She never eats pizza) (though see Map 115 on negative pronouns). The map is based on the expression of negation in declarative sentences; thus, where a language employs a distinct construction in imperative sentences, that construction is ignored here (see Map 71).

All of the ways of indicating negation involve negative morphemes. This contrasts, for example, with a variety of ways of signalling a polar question (Map 116) that do not involve interrogative morphemes, such as changes in word order, intonation, and the complete absence of any signal that a sentence is a question. There are no known instances of languages in which negation is realized by a change in word order or by intonation, and all languages have negative morphemes.

The first type shown on the map consists of languages that express negation by means of a negative affix attached to the verb, as in (2) from Kolyma Yukaghir (Siberia, Russia):

(2) Kolyma Yukaghir (Maslova 2003a: 492) 

 

The next three types all involve languages which express negation by separate words. The first type of negative word is a negative particle, like the English word not or the word pày in (3) from Musgu (Chadic, Afro-Asiatic; Cameroon).

(3) Musgu (Meyer-Bahlburg 1972: 186) 

 

The second type of negative word is a word that inflects as a verb, and which can be considered a type of auxiliary verb, since it normally must accompany another verb. In Finnish, for example, the negative word is an auxiliary verb, and inflects for the person and number of the subject, as in (4), while what is semantically the main verb occurs in a nonfinite participle form.

(4) Finnish (Sulkala and Karjalainen 1992: 115) 

 

In Grebo (Kru, Niger-Congo; Liberia), the negative word's status as a verb is reflected by the fact that it takes the tense inflection that occurs on the verb in the corresponding affirmative sentence. For example, in the affirmative sentence in (5a), the verb du¹ ‘pound’ occurs with a past-before-yesterday tense suffix -da². In the corresponding negative sentence in (5b), this tense suffix occurs on the negative word yi²¹ instead.

(5) Grebo (Innes 1966: 55, 106)

 

For many languages, it is easy to decide whether a negative word is a particle or a verb: if the negative word does not occur with inflectional morphology associated with verbs in the language, it is treated here as a particle, while if it does occur with inflectional morphology associated with verbs in the language, then it is treated here as a negative auxiliary verb. In some languages, however, it is more difficult to decide whether the negative word should be considered a verb or not, if verbs occur with little or no inflectional morphology or if the inflectional morphology that does occur on verbs may not be semantically appropriate for a negative word even if that negative word is a verb. Such languages are treated here as instances of the fourth type shown on the map, as languages with negative words, but where it is unclear if the negative is a verb or a particle. For example, in Maori (Polynesian; New Zealand), the negative is uninflected, but so is the verb, as illustrated in (6).

(6) Maori (Bauer 1993: 140) 

 

Because of the absence of inflectional morphology on verbs in Maori, it is difficult to decide on the basis of superficial evidence whether the negative word should be considered a verb or not. In some languages, there may be good syntactic reasons for classifying a negative word as a verb. For example, in Boumaa Fijian (Oceanic), the negative word has the syntactic properties of a main verb, and the accompanying predication is placed in a complement clause that functions as the subject of the negative verb, as in (7). 

(7) Boumaa Fijian (Dixon 1988: 40) 

 

However, in many languages with little or no inflectional morphology on verbs, it is not obvious whether or not the negative word should be considered a verb even by syntactic criteria; morever, the criteria assumed for this map are purely morphological. Thus even if there seems to be good syntactic (but not morphological) reason to say that the negative word in a language is a verb, it is classified as a language with a negative word where it is unclear if the negative word is a verb or particle. 

The fifth type consists of languages which have more than one negative construction, one in which the negative is a separate word, and one in which it is an affix. For example, Rama (Chibchan; Nicaragua) has two negative constructions, one with a preverbal negative particle, as in (8a), the other with a negative suffix on the verb, as in (8b). 

(8) Rama (Grinevald n.d.: 183, 185) 

 

Languages like Rama are coded as languages with variation between a negative word and a negative affix.

The last type shown on the map is languages with double negation, where the negative construction involves two simultaneous morphemes, one preceding the verb or verb stem, the other following. In Standard French, for example, the negative construction involves both a preverbal negative particle ne and a postverbal negative particle pas, as in (9).

(9) French 

 

In some languages with double negation, the negative construction involves a prefix and a suffix on the verb, as in (10) from Izi (Igboid, Niger-Congo; Nigeria). 

(10) Izi (Meier et al. 1975: 218) 

 

In still other languages, one of the negative morphemes is an affix on the verb while the other is a separate word, as in (11), from Ma (Adamawa-Ubangi, Niger-Congo; Democratic Republic of Congo). 

(11) Ma (Tucker and Bryan 1966: 130) 

 

No attempt is made on the map to distinguish the different types of negative morphemes that occur in languages with double negation, i.e. whether the negative morphemes involved are negative affixes or negative words or a combination of these. 

Some languages employ double negation only under certain circumstances. For example, Mupun (Chadic; Nigeria) employs an obligatory clause-final negative word as well as an optional clause-initial negative word, as illustrated in (12), where the parentheses around the initial word indicate its optionality. 

(12) Mupun (Frajzyngier 1993: 353) 

Languages like Mupun with optional double negation are not shown on the map as involving double negation but rather according to the properties of the obligatory negative mopheme. Similarly, French is coded on the map as employing a negative particle rather than as having double negation, since the preverbal particle is optional in colloquial French. 

In some languages, the morphemes that code negation also code other grammatical features of the clause as well. For instance, the example in (13) from Rumu (Turama-Kikorian, Trans-New Guinea; Papua New Guinea) illustrates a verbal suffix that codes both negation and future tense. 

(13) Rumu (Petterson 1999: 24) 

 

Negative particles also occasionally vary for tense/aspect, as in the examples in (14) from Puluwat (Oceanic; Micronesia); contrast the affirmative perfective marker ya in (14a) with the negative perfective marker há in (14b).

(14) Puluwat (Elbert 1974: 85) 

 

In Hausa (Chadic), the negative particles vary not only for tense and aspect but also for pronominal features of the subject as well. For example, the first word ban in the main clause in (15) codes negation, completive aspect, and first singular subject, while the word bài in the relative clause codes negation, completive aspect and third singular masculine subject.

(15) Hausa (Newman 2000: 359)

 

One might think that negative words like those in Puluwat and Hausa are reminiscent of negative auxiliary verbs, as illustrated in (4) and (5) above. However in (4) and (5), the inflections on the negative word are inflections which otherwise occur on verbs in the language, so that they provide a basis for saying that the negative word is a verb. In (14) and (15), in contrast, verbs do not in general inflect for tense/aspect or for subject; hence the negative words are best viewed as a set of nonverbal words that vary for more than one grammatical feature of the clause, one of which is negation. 

There are many other dimensions to the typology of negation than those covered by this map. See Maps 113 and 114 for additional aspects of the typology of negation.

2. Geographical distribution

The geographical patterns are somewhat less striking here than on some other maps. Both negative particles and negative affixes are widely distributed throughout the world. Languages with negative auxiliaries are less common, but show a striking frequency across northern Eurasia, stretching from Finland to western Siberia. Languages with other negative words are common in Southeast Asia, but this largely reflects the isolating nature of languages in this area, making it difficult to determine whether negative words are verbs or not. Double negation is also widespread, except for Europe and much of Asia, but seems particularly frequent in New Guinea and parts of Africa.