Page 431 - Encyclopedia of Philosophy of Language

Guarani (Tupi family)
Juan nd-o'u-i pira(vs.Ivanestrybu)
J. NEG-eats-NEG fish
(lOi)
existential verb, can occur for the negative copula 'there is not, there does not exist,' as in Turkish yok-
Notice that: (a) In English there exists a special auxiliary verb to form negative sentences, the sem- antically main verb remaining uninflected; (b) French and Guarani have the so-called 'discontinuous NEG' around the verb (:ne...pas).—French shows also a morphosyntactic alternation between de and du depending on whether the sentence is negative or posi- tive; (c) in German the NEG may be expressed by an adjectival inflected form; (d) in Finnish there is an inflected NEG verb whereas the lexical verb remains uninflected; (e) Turkish agglutinates a NEG mark -ma- j-me- to the verbal accented stem, before other stem determinations (e.g., -me+yor>-miyory, (f) the Japanese NEG is a verbal negative form -nai suffixed to the indefinite verbal base; (g) and Welsh has a particle (nid) introducing NEGsentences parallel to the particles for declaratives (y) and interrogatives (a, as in: a mae John yn bwyta pysgod?) so that NEG is a modality of the sentence. Among the eight languages under scrutiny Finnish is from the typological view- point the nearest to English.
The phenomena here underlined by no means exhaust the phenomenology of NEGsentences. Kwaa (Niger-Congo family) expresses NEG via a modi- fication of the word order. Dravidian languages have the possibility of a negative conjugation for every active, passive, 'neutral,' or even causative verb (and this may remind students of the Welsh examples). In Tamil the NEG mark is -a-; but this -a- agglutinating with the desinential vowel may disappear. The nega- tive verb is therefore characterized by the absence of a specific mark (one of the not very common examples of unmarked forms for semantically and functionally marked meanings): kan-p-Sn 'I shall see'; kan-t-Sn 'I saw'; but simply kan-sn 'I do not / 1 shall not / 1 did not see.' However NEGforms usually contain one (or more) mark(s) more in comparison with the cor- responding positive sentence, for instance (lOc) and (lOd). (In fact, from a philosophical point of view, it has been said by Bergson that, whereas the positive sentence has to be considered as a judgment, the nega- tive is more complex, being a judgment on a judgment (see example (1) above: [[it is not true that]...]).)
Basically the main strategies for sentence negation are (a) morphological: a NEG mark e.g., (lOg), (lOh) which can be integrated in the verb, e.g., (10e), (lOf); (b) morphosyntactic: NEG is expressed by a particular verbal cluster with a negative auxiliary expressing tense, aspect, mood, and person, e.g., (10a), (lOd); however, the verb categories may be distributed between the negative auxiliary and the main verb, as is the case for some Uralic languages. Purely syntactic strategies like specific word order rules as in Kwaa are rather exceptional; (c) lexical: as in (lOc). Special lexemes, different from the corresponding affirmative
versuswzr, e.g.:
Osten Dahl published in 1979 a large study on the expression of NEGin around 240 languages, which yielded the following results: NEG is most frequently expressed by either bound morphemes as part of the predicate (45 percent) or by separate particles (44.9 percent). NEG auxiliaries make out only 16 percent of Dahl's sample, while the use of dummy auxiliaries as in English is quite rare.
Tu Ekmek bread
yok
vs. Ekmek var (11) bread there is
There's bread'
there is not There's nobread'
The expression of NEG(according to Dahl 1979)
1. morphologicallyaspartofthepredi- % cate 45.0 2. morphologically in an auxiliary verb 16.7
3. by a separate negative particle
a. inpreverbalposition 12.5"
b. in preauxiliary position 20.8
c. before verbal group 2.1
d. in postverbal position 1.2
e. in postauxiliary position 3.7
= 44.9%
4. by
a. in sentence-initial position 0.4
b. in sentence-final position 4.2^
a separate negative particle
(N.B. The total runs above 100% because a number of languages have more than one type of NEG)
4. Diachronic Evolution of NEGMarkers
In several linguistic traditions a trend can be observed which moves toward the use of an invariant NEGpar- ticle (preferably in preverbal position: see below). This is actually the most natural NEG strategy and comes nearest to the transparency ('diagrammatic') principle of optimization, i.e., 'one meaning: one form.' Thus Estonian does not inflect the NEG verb ei (3so) which in this way approaches the uninflected particle con- struction type, drifting away from the auxiliary con- struction of the Finno-Ugric languages, that is, shifting from a morphological to a syntactic strategy (see John ei sod' kala 'J. does not eat fish' but also with a subject in plural Johnja Mary ei kohtunud koolis 'J. and M. did not meet at school'; also the extension of ain't in Black English as generalized NEG form).
A very interesting case is represented by the French- based Creoles:
Louisiana
mo kup pa
and mo pa kup
(12a)
(12b) my watch NEG PROGR work: 'my watch isn't working.'
mecut NEG
Mauritius
meNEGcut:'Idon'tcut.'
mo mdte pa pe
travaj
Negation
409