LINGUIST List 8.289

Thu Feb 27 1997

Sum: Negation/Affirmation

Editor for this issue: T. Daniel Seely <>


  1. Matti Miestamo, Negation/Affirmation

Message 1: Negation/Affirmation

Date: Tue, 25 Feb 1997 13:42:01 +0200
From: Matti Miestamo <>
Subject: Negation/Affirmation


Last month I had a query on the list concerning the relationship between
negation and affirmation. My original posting was as follows:

I am doing my MA thesis on the relationship between negation and
affirmation. Especially I am interested in the way the marking of
tense/aspect/mood or other categories is affected by negation. I would
be grateful, if you could point out to me languages where for instance
a tense/aspect/mood-distinction made in the positive is neutralised in
the negative (or vice versa!) or any languages that are interesting as
far as the relationship between negation and affirmation is
concerned. I am already aware of a number of languages that could be
relevant for my study, but any information is welcome. Also, I would
appreciate information on books and articles relevant for my study. I
will post a summary.

I thank the following persons for their responses:
Marc Picard, John E. Koontz, David Houghton, Yasuhiko Kato, Pierre Larrivee,
Tej Bhatia, Elena Koutsomitopoulou, Miao-Ling Hsieh, Larry Horn,
Claus-Dieter Pusch, Kean Kaufmann, Marina Yaguello, Ann Wehmeyer, Oliver
Baumann, Carsten Peust, Antoine Lonnet, Isao Honda, Knut Olawsky and Nancy

If you have responded, but your name is not on the list, please re-send me
your mail. We had some e-mail problems at the time my query appeared on


Isao Honda:

Your topic seems to be similar to the one that I dealt with in
my unpublished Ph. D dissertation "Negation: a cross-linguistic
study". In this dissertation, I examined the structures of
negative sentences in languages around the world (I checked
more than 900 languages' reference grammars) , and investigated
in what ways they differ from those of the corresponding
affirmative sentences. The primary purpose of this study is to
study what kinds of linguistic patterns exist in languages of the
world, and to find out which patterns are cross-linguistically
common and which are not. The outline of the dissertation is as

Ch. 1.Introduction
Ch. 2. Three types of negative sentences
In this Chapter, three types of negative sentences was
discussed. This typology is primarily based on Dahl (1979),
which uses the notion of 'finite element' as criterial. The
three types are: Type I) a negative sentence employs the
same finite element as the one in the corresponding
affirmative sentence; Type II) in negative sentences an
auxiliary verb is introduced as a finite element (e.g., English,
Korean, Ainu, Ewondo, Achumawi, Siroi, +Cariban, Awarak,
Ika, +Trans-New Guinean, Mparntwe Arrernte, +Tanoan,
+Numic, Apalachee, Chukchi, Nanay, Japanese, Bobangi,
+Rashad, Nama: + sign indicates language group name, but not
language name. It should be noted that by listing these
language groups I do not indicate that the pattern mentioned
is found in all of the languages belonging to the language
group), Type III) a negative morpheme itself is a finite
element (e.g., +Tungus, +Uralic, Gilyak, +Nilo-Saharan,
+Kru, +Mba, +Kordofanian, +Yuman, +Pomo, +Austronesian,
Ch. 3. Changes in the form of the verb
In this Chapter, I examined various phenomena where the
form of a verb is changed when the sentence is negated.
Those phenomena are classified into the following four
types according to the nature of the verb form used in
negative sentences; 1), transitive verbs appearing in
negative sentences are affixed with one of the person
markers which in affirmative sentences are used with
intransitive verbs (e.g., Yukaghir, +Muskogean), 2) lexical
verbs appearing in negative sentences are inflected like
nouns (e.g., Hixkaryana, Thompson), 3) lexical verbs
appearing in negative sentences are marked with a kind
of nominalizer (e.g., +Salishan, +Australian, +Trans-New
Guinean, Korean, +Uto-Aztecan), 4) lexical verbs appearing
in negative sentences are in a form identical or similar to
the form used in passive sentences (e.g., Luwo, Palauan,
Ch. 4. Tense and aspect in negative sentences
Ch. 5. Participants of events described in negative sentences
In these 2 Chapters, I discussed various phenomena where
certain other categories are modified when sentences are
negated. Chapter 4 dealt with languages in which tense and
aspect categories are modified when sentences are negated
(e.g., +Bantu, +Adamawa-Ubangian, Geleba, Tzotzil, Russian,
Copala Trique, +Maipuran Arawakan), and Chapter 5 dealt
with languages where person/number category or, more
broadly, the way in which participants of the described
event are expressed, is modified when sentences are negated
(e.g., +Cushitic, +Slavic, Acoma).
Ch. 6. Negative sentences with a marker which is also used in
sentences expressing irrealis events
In this Chapter, I examined languages in which negative
sentences differ from the corresponding affirmative
sentences in containing a marker that is also used in
question sentences, sentences expressing hypothetical
events or events in future, etc. (e.g., +Salishan, Tlingit,
Blackfoot, Cherokee, Diegueno, +Mayan, Bukiyip, Alamblak,
Ch. 7. Conclusion

This dissertation is available through UMI dissertation
service. UMI number is 9617870. This includes a reference
which lists a number of papers dealing with "negation" (but
does not include ones in formal theory), and Appendix, which
list the form of the negative morpheme in more than 800
languages. I hope that you will find something interesting in it.

Finally, let me mention briefly my opinion about tense/
aspect "neutralization". In my view, it is not appropriate to use
the term "neutralization", because this term (and notion) is
based on the following assumptions
1.The proposition expressed in a negative sentence is
identical to the one expressed in the "corresponding"
affirmative sentence.
2."Affirmative" and "negative" form a pair (in other words,
there is a universal category "polarity", which consists of
"affirmative" and "negative".
3.Domains of tense/aspect are uniform (e.g., "past" expressed
in affirmative is identical to the one expressed in negative.

I also object to the view that the occurrence of neutralization in tense
/aspect in negative sentences is due to the fact that
negative is a "marked" category (unless it is further explained
what is meant by "marked" and "unmarked").


Marc Picard:

Here's something you might be interested in. French has two ways of
indicating future tense, i.e. a simple and a periphrastic construction
very much equivalent to English WILL and GOING TO, e.g.:

(1) Je partirai 'I will leave'
(2) Je vais partir 'I'm going to leave'

The negative equivalents are:

(3) Je (ne) partirai pas
(4) Je (ne) vais pas partir

 In theory, and perhaps historically, all four forms could be used
interchangeably as it were. However, in Canadian French, it's now
pretty much obligatory to say:

(5) Je vais partir
(6) Je partirai pas

In other words, in practical terms, one can almost star:

(7) *Je partirai
(8) *Je vais pas partir

 Only in certain types of constructions would these latter types of
structures occur. For example, in the affirmative, you could have a
sort of imperative like:

(9) - Il va partir si tu t'excuses pas 'he's going to leave if you
 don't apologize'
(10) - Il partira! 'let him leave'

while in the negative, you'll hear things like:

(11) - Tu vas pas partir tout de suite 'don't tell me you're
 going to leave right now'

where you have a type of rhetorical question (I'm just a lowly
phonologist so I don't really know the technical term for this).
 Anyway, you can get most of the gory details on this in:

Emirkanian, Louisette, and David Sankoff (1985) "Le futur simple et le
MONTREAL, Monique Lemieux and Henrietta J. Cedergren, eds. Quebec:
Office de la Langue Francaise, pp. 189-201.


John E. Koontz:

I seem to recollect that negation is a mood or interacts with mood marking
in Eskimo languages. It has been a while since I looked at this, so I can't
guarantee anything interesting.

However, perhaps this is relevant. In Omaha-Ponca and other Dhegiha Siouan
languages, negation interacts with the morphology of person marking.
Specifically, negation is basically marked by an enclitic (a)z^i. The a
vowel is a component that inserted, or, of the final vowel of the verb is e,
it replaces the e. This alternation of e and a in final position in stems
is called ablaut in SIouan studies.

The forms of the negative vary with person, under the following paradigm:

 positive negative
 1s 0 =m=az^i
 2s 0 =(a)z^i
 3s obv 0 =(a)z^i
 3s prox =(b)i =b=az^i

 12 =(b)i =b=az^i
 2p =(b)i =b=az^i
 3p =(b)i =b=az^i

The =(b)(i) enclitic here is the plural (or augment) marker, which marks the
presence of additional persons over the basic pronominal reference, and also
of the third person proximate. The normal personal inflection is prefixal,
e.g., for regular agent (transitive subject and active intransitive subject):

 1 a
 2 dha
 3 0
 12 aN

The source of the m first person marker is an irregular active auxiliary maN
'I am thus; I use', stem aN. First and second forms of this stem regularly
follow two other adverbial enclitics, to wit, =xti 'very, real' and =(s^)naN
'to the exclusion of other activies, only, habitually'. For example:

 1 =xti=maN
 2 =xti=z^aN
 3 =xti
 12 =xti

In the negative the auxiliary precedes, and only the first person is found.

The same auxiliary aN also seems to be a fixed part of the enclitic =s^te
'soever', in the variant =s^teaN. I don't know what conditions =s^te vs.

The negative enclitic =(a)z^i seems to be comparable to the =s^(i)
adversative (=contrastive) enclitic of Dakotan Siouan: miN=s^ (me, not
someone else). The Dakotan negative is =s^niN, uninflected, possibly from
*=s^(i)=niN, where *niN looks like the real negative, based on comparison
with other Siouan languages.


David Houghton:

Welsh lacks words equivalent to the English
'yes' and 'no'. The following illustrates the range of possible
affirmative and negative answers.

I. If the yes/no question has the canonical VSO structure and uses one of
a set of auxiliary verbs one responds with a responsive construction
consisting of the verb, -- I should have said this concerns the present
tense -- preceded by the responsive negative particle 'na(c)', if
appropriate, and, occassionally, to indicate emphasis, followed by the
appropriate pronoun. Thus,

Elli di weld y gath?
Can you see the cat?

Gallaf/Na allaf

Dych chi wedi gweld y gath?
Have you seen the cat?

Ydw/Nac ydw


If the sentence uses an inflected verb outside of this class, one can use
a form of 'gwneud' "do/make".

Weli di'r gath?

Gwnaf/Na wnaf

If the question preposes some constituent, whatever the form of the verb,
one responds with the particles 'ie' or 'nace'.

Y gath welsoch chi?
Did you see the CAT?


If the verb in the verb in the question is past tense (and the verb is
first), one responds with the particles 'do' or 'naddo'.

Welsoch chi'r gath?
Did you see the cat?


That's a partial exposition of affirmative and negative responses.
The best reference grammar by far is Gramadeg y Gymraeg by Peter Wynn
Thomas, but I'm not sure whether that has been published in English
translation. For literary Welsh, you could consult A Welsh Grammar, by
Peter J. Williams. For colloquial Welsh, there's Modern Welsh, by Gareth


Pierre Larrivee recommends the following book, which "provides a good and
extensive crosslinguistic study of elements whose form is modified under
negative contexts."

Robert Forest. 1993. Negation: essai de syntaxe et de typologie
linguistique. Paris: Klincksieck. 158 p.


Tej Bhatia:

Negation triggers a number of deletion rules in South Asian languages. The
result of deletion such as Aux deletion is neutralization in Negitivized
structures. Another interesting aspect is the deletion in Serial verbs in
these languages leading to neutralization of simple and serial verbs in
negatived structures. These phenomena are discussed in detail in my book

Bhatia, Tej K. 1995. Negation in South Asian Languages. New Delhi/Patiala:
Indian Institute of Language Studies.


Elena Koutsomitopoulou:

Modern Greek:
-[theleis na ertheis sto cinema?] = would you like coming to the cinema?
-[Oxi, den thelo] = no, i wouldn't

In MG you will never find something like the English 'never found' (a
negative word with an affirmative verb); once a negative word is in a
sentence you must also have the negative 'den' together with the verb
('den tha to vris pouthena' = *you will not find it nowhere* = you will
never find it).

('den' is the negative word like the 'do not' of English)


According to Miao-Ling Hsieh Chinese is worth considering.


Larry Horn:

I've written a bit on these matters, especially in the first part of
Chapter 7 of my book A Natural History of Negation (U. of Chicago Press, 1989),
but that wasn't really my focus. (Chapter 3, on the psychology of negation,
and Chapter 1, on the historical background, also deal extensively with the
asymmetry between affirmative and negative statements, thoughts, information,


Claus Dieter Pusch:

I would like to point your attention
to Gascon, a language forming part of the Occitan diasystem and spoken in
Southwestern France. This language has a curious system of affirmative
markers called "enunciatives" (a subject which I am actually working on
for my PhD dissertation). They normally precede all inflected verbal
forms in affirmative mood, but disappear in negative sentecences (except
for some local varieties which tolerate enunciatives and negators). Thus
the Gascon of "Pierre speaks Gascon well" (= has a good knowledge of...)
would be:
[1] Peire que parla plan lo gascon. but:
[2] Peire non parla (pas) lo gascon.
which would be the negative form. (Note that the enunciative "que" in the
affirmative sentence is not to be interpreted as a relative pronoun!)
The status of enunciatives and especially the question of whether they
should be considered part of Gascon verbal morphology was (and still is)
object of controversy, but at least one should consider it as an element
contributing to finiteness in Gascon. Thus it should be subsumed with TMA
In negative form Italian imperatives loose all marks of finiteness and
become infinitives.


Kean Kaufmann:

I don't have data at hand to give examples, but you might want to look at
Tamil. Negation is normally expressed by a negative suffix on the
infinitive form; this can refer to past, present or future. To express
explicit negation of past or present, you use a nominalized form of the
tensed verb plus the negative suffix. In the future tense, though, things
get really interesting: there's a separate suffix for negative neuter
future, as opposed to positive neuter future. It's worth noting that the
neuter future is often used for generic statements, cf. English: "A cow
will (always) eat grass," "A lawyer will (always) lie."

In my own M.A. thesis, I argue that the Infinitive is actually an
aspectual suffix -- marking irrealis/imperfective/generic events -- and
that all negative and future forms can be analyzed as containing the
Infinitive. But this is a controversial claim.

However you analyze it, the data are fascinating.


Marina Yaguello:

Wolof (spoken in Senegal and the Gambia) is such a language. You can find
the information in Leopold Diouf and Marina Yaguello ,Damay jang wolof ,
Paris: Karthala 1991 or in a doctoral dissertation by Stephane Robert: Approche
enonciative du systeme verbal wolof , Paris:CNRS


Ann Wehmeyer:

Korean seems to exhibit some interesting effects in the interaction of
negation with the system of tense/aspect/mood. There is a recent
dissertation on negation in Korean which you might find useful:

Kim, Jinkyoung. 1996. Negation in Korean: A Functional and Discourse
Approach. University of Florida PhD dissertation.


Oliver Baumann:

The first thing I remembered was Lyon's (1968) French example
"Je crois qu'il vient" vs. "Je ne crois pas qu'il vienne", where
the negation triggers subjunctive in the embedded sentence
[Lyons, John (1968): Introduction to Theoretial Linguistics.
pp 312ff]. However I'm sure you are aware of this, are'nt you?
The same I'm sure you are aware of all the phenomena in Finnish,
according to this topic. I mean, first, the nominative/partitive
alternation in imperatives, in existential sentences and in the
unpersonal passive (all triggered by mood resp. default case
assigning mechanism, not by negation), s. among others Milsark,
G. (1985): Case Theory and the Grammar of Finnish. Proceedings
of 15th NELS. However, what you probably don't know is that there
is an analo mechanism in Breton (s. Anderson, S.R. (1982): Where's
Morphology? LI 13, 571-612). In Basque the negation triggers
absolutive/partitive alternation, for example:
(i) ez du gizonak ikusi ikaslea
 NEG has man-ERG seen student-ABS
 "the man didn't see the/a student"
(ii) ez du gizonak ikusi ikaslerik
 NEG has man-ERG seen student-PART
 "the man didn't see any students/student at all"
(s. here Levin, B. (1983): Unaccusative Verbs in Basque. Proceedings
of 13th NELS. pp. 132ff; or if you are interrested in what triggers
negation syntacticaly (Auxiliar shift by negation): Laka, I. (1990):
Negation in Syntax).
That's the same in Russian where you can mark definiteness by
case alternation that is if you choose accusative as direct object
case in a sentence with sentential negation, then the object is
interpreted definite. Alternatively the genitive denotes indefinite-
ness. Remember that there's no determiner (article) in Russian. In
Polish you don't have the possibility to change. Always genitive occurs.
If you need bibliographic references or examples with respect to Russian,
Polish, Czech etc., tell me.
Interrestingly enough, Basque doesn't need the negation
to trigger this alternation. It use also mood instead as licenser.
So you can have partitive in the environments "question" and "desire":
(i) ez du etxerik erosi
 NEG has house-PART bought
 "she bought no house/didn't buy any house"
(ii) *etxerik erosi du (out!)
 house-PART bought has
 "she bought a house"
(iii) etxerik erosi du?
 "Did she buy a house?"
(iv) etxerik erosiko balu
 "Would she have bought a house ..."
(s. Laka (1990, 37ff). And so forth.)


Carsten Peust:

You might have a look into "Late Egyptian" (this is the language of
around 1000 BC in Egypt). There is only one present tense in
affirmative sentences, but two ("neg. present" and "neg. habitual")
in negative sentences:

tw=i Sm "I am going, I go"
bn tw=i Sm "I am not going"
bn j.jr-i Sm "I dont (usually) go".

In the later periods, Egyptian develops also a positive habitualis,
so in Coptic (Egyptian of the Christian period) we have:

ti bok "I am going"
Saf bok "I use to go"
nti bok an "I am not going"
mei bok "I dont usually go"

The most important reference grammar of Late Egyptian is from
J. Cerny and S. Groll. You might also have a look into P. Frandsen
"Late Egyptian Verbal System".


Antoine Lonnet:

I quote you :

> Especially I am interested in the way the marking of
> tense/aspect/mood or other categories is affected by negation.

It is remarkable that Arabic (classical Arabic) shows the reverse :

 negation is affected by tense/aspect/mood.


Knut J. Olawsky:

Dagbani is a Gur language of Northern Ghana and has a particle /ku/ which
expresses negation + future.
1) o chaN Tamali. (He went to Tamale)
 3sg go (place name)

2) o bi chaN Tamali. (He did not go to T.)
/bi/ is the negative marker for non-future

3) o ni chaN Tamali. (He will go to T.)
/ni/ is the future marker

4) o ku chaN Tamali. (He will not go to T.)


Nancy Torres:

Pour parler de la negation il faudrait toucher les modes et les temps
(subjonctif et futur) les malentendus, des periphrases qui remplissent le
sens negatif, il faut toucher aussi la logique et la philosophie.
Je vous donne quelque bibliographie pour vous aider:
Culioli Antoine, Pour une linguistique de l'enonciation, Ophrys, Paris 1990
(Il y a une grosse partie sur la
J.M.Zemb, Les occurrences phematiques, rhematiques et thematiques des
archilexemes modaux
Aristote, Organon VI, Les refutations sophistiques.
Aristote, De la interpretation. Traite des categories
Ducrot, Oswald, Dire et ne pas dire, Colection Savoir, Hermann, 1980
Bourdieu, P., Ce que parler veut dire, Paris, Fayard, 1982
J.M.Adam.Borel,Calame, Le discours anthropologique, Klincksieck, Paris 1990
Molho, Maurice, De la interpretation en espagnol, Melanges Marcel Bataillon,
Revue L'information Grammaticale No.55.1992,Octobre,Paris
Robert Martin, Langage et croyance. Pierre Mardaga editeur, Bruxelles, 1987,


Yasuhiko Kato:

On your query on Linguists, the following might be
Hasegawa, N. 1991. Affirmative Polarity items and
 Negation in Japanese. Interdisciplinary Approaches
 to Language. Kluwer, pp.271-285.
 [H argues that API sensitivity is neutralized in
 conditionals and questions, which also induces an
 interesting ambiguity. A critique of H's analysis
 is found in my paper below.]
Kato, Y. 1994. Negative Polarity and Movement. MIT
 Working papers in Linguistics 24.

- Matti Miestamo
 University of Turku
Mail to author|Respond to list|Read more issues|LINGUIST home page|Top of issue