(Fwd) RE: intonational marking of negation

[B. Woll]

I'm forwarding to the list this really interesting information which
followed a query from Ber Remijsen about intonational marking of
negation:

------- Forwarded message follows -------
From:                   "Remijsen, A.C.L."
<A.C.L.Remijsen at let.leidenuniv.nl>
To:                     "'B. Woll'" <b.woll at city.ac.uk>
Subject:                RE: intonational marking of negation
Date sent:              Thu, 22 Aug 2002 10:08:52 +0200

Hi!

Thanks for your reaction. Here is a complete summary of the
responses.

Cheers,
Bert

First of all, thanks to all of you who replied to my query! There were 15
replies, which ranged from claiming that no language has exclusively
prosodic marking of negation, to one from somebody
who expressed his surprise about my query, and said that he had
always assumed that all languages can express negation
prosodically. This range of variation assured me that I had not been
wasting your time asking you about an issue with a
self-evident answer.

A number of different phenomena are discussed in the replies:
(i) the prosodic (non-manual) marking of negation in sign languages;
(ii) tonal marking of negation in African tone languages;
(iii) intonational marking of negation in English (in combination with
segmental morpheme) and in Malay
(iv) prosodic marking of negation by means of vowel lengthening.

I will briefly summarize the reactions in the following paragraph, and
below that can find a lengthy summary of the replies.

So - are there any languages that mark negation prosodically without
additional segmental encoding?
(i) While most of the reactions discussing sign languages say that there
always is a manual marker involved, Bencie Woll argues that the
prosodic (non-manual) marker may occur on its own. So if
this claim can be substantiated with evidence, the question above can
be answered affiramtively.
(ii) There is no doubt about it that negation is marked prosodically, and
uniquely so, in a number of African tone languages. Of course, this
marking of negation is of a different nature from the examples I cited
from Papiamentu in Kuot. In the latter the prosodic component of the
marking of negation is intonational - the pitch contour that encodes the
negative is realized on whatever constituent is located in a certain
position in the phrase. In the African tone languages listed below, the
marking of negation is morphological - it is a tonal morpheme -
sometimes one of many - that is associated with the verb. The
encoding of the negative in the Creole language Ndjuka (see reply by
Huttar) may also fit in this category, and that fact may be of interest to
people who look for evidence for an Atlantic Sprachbund of Caribbean
creoles and West African languages.

(iii) No conclusive evidence was reported that negation can be marked
solely through prosody in intonational (i.e., non-lexical tone) languages.
Whereas the encoding of negation by means of intonation may
frequently or even regularly accompany the segmental marking in
English, it does not occur without the latter. If the regularity is as strong
as some say (see replies below), then one of the two markers is
functionally redundant, and would, logically, be prone to be lost. But
this does not happen, and that is surprising. Of course it may be that
we don't have the data of the intonational system that does have
negation marked solely through intonation. Another intonational
language for which negative marking through prosody was reported is
Malay (various dialects - information from David Gil and myself).
However, here the prosodic negative is limited to one or two verbs. At
least in Ambonese Malay, this contour may encode a paralinguistic
emotional meaning rather than negation as a concept of propositional
semantics. In any case, the lack of productivity of this contour is
suspicious.
(iv) A revelation for me were the reports of durational marking of
negation. While it is not reported what form this marking takes in Nanai
(reply from Roland Pfau), in the Austronesian languages cited by John
Lynch it does appear to be a matter of morphological marking, similar
to the tonal morphemes of African tone languages.

(i) Laurence Horn claims that no language exists in which negation is
marked exclusively by means of prosody, i.e., in the absence of a
segmental (morphological or syntactic) encoding. He makes this claim
in his book 'Natural History of Negation' (U. of Chicago Press, 1989;
reissue, CSLI, 2001), pp. 472-3. He notes that the claimed evidence for
marking of negation in American Sign Language involves both a
segmental and a suprasegmental dimension (research by Bridget
Copley). I.e., the marking of negation in ASL would not take place
solely through negation, but rather be similar to the phenomena I
mentioned of Kuot and Papiamentu.

Ton van der Wouden refers to another publication in which the claim is
made that such languages do not exist: John Payne: Negation, in
Timothy Shopen (ed.)  Language Typology and Syntactic Description,
Cambridge University Press 1985, 197-242. He also discusses the
counter-evidence from a sign language, in this case the Sign Language
of the Netherlands (Nederlandse Gebarentaal). He notes that the
negative morpheme is obligatorily accompanied by a specific facial
expression. The data and analysis come from the PhD diss of Jane
Coerts (Nonmanual grammatical markers. An analysis of interrogatives,
negations and topicalisations in Sign Language of the Netherlands,
Amsterdam 1992). He says that Jane Coerts argues that the facial
encoding of negation marks the scope of the negation.

Bencie Woll and Roland Pfau give more information on negation in
sign languages. Pfau: "In signed languages, prosody is (often) marked
by a non-manual modification, that is, by facial expressions and/or
head movements. In DGS it is possible to negate a sentence by adding
a side-to-side headshake which is realized simultaneously with the verb
sign."
Bencie Woll claims that negation can be expressed exclusively non-
manually or prosodically in sign languages: "As far as I know, all sign
languages use intonational marking of negation (i.e. marked by non-
manual features such as head movement, eyebrow position, etc.).
These may co-occur with segmental negation markers (e.g. NOTHING,
NOT, NEVER, NOT-YET) or occur on their own. There is a substantial
literature in the sign linguistics field on this topic.

(ii)
Bruce Connel reports that: "marking of negation by means of tonal
alternation occurs in Mambila, a Bantoid language spoken in Cameroon
and Nigeria.  In Ba-Mambila, which has four lexical tones (1 - 4 = High
to Low, verbs with inherent tone 2 or 2-1 take tone 1 in the negative,
while verbs with inherent tone 3, 3-1, or 3-4, all take tone 4 in the
negative. It seems to me it is not uncommon in Benue-Congo (and
perhaps also elsewhere in Niger-Congo) for marking of negation to be
done tonally, either accompanied by segmental marking or not, though
I don't have other exx readily to hand."

Elke Hentschel: "one of the languages you're looking for is Kele (Bantu,
Central Africa).  There is an example in Carrington, John F. (1949):
Talking Drums of Africa. Oxford: Alden Press, on p. 19: ayeke (high-
low-low): 'let him come' ayeke (high-high-low): 'don't let him come'
and an anecdote of someone unintentionally praying 'may Thy
Kingdom not come, may Thy will not be done' etc."

Roland Pfau also presents evidence of prosodic marking of negation in
African tone languages: he writes that in a survey of the literature, in
which he looked for evidence of prosodic marking of negation, he "only
came across three spoken languages in which negation is expressed by
a prosodic modification alone: Ga, spoken in Ghana (but only in the
past tense); Ogbru, spoken in the Sothern Ivory Coast (only with
monosyllabic verbs due to a tonal constraint); and Nanai (see below).
Similarly, Michael Becker writes that "in many Bantu languages,
negated verbs have various effects on prosodic phrasing."

George Huttar: "In Ndyuka, a creole of Suriname, positive vs. negative
is indicated by high vs. low tone respectively, with no segmental
difference--but only for the copula _na_:Mi na gaanman.  'I am chief.'
(low tone on _na_) Mi na gaanman.  'I am not chief.' (high tone on
_na_)See George Huttar and Mary Huttar, 1994, _Ndyuka_ (London:
Routledge), p. 134."

Janet Bing: "In Liberian Krahn negation is marked with a high tone only.
There are examples sentences (39)-(42) in the following article. 1993
Janet Bing and John Duitsman.  Postpositions and the Valency Marker
in Krahn: Monosemy Versus Polysemy. Journal of West African
Languages. 23 (1) pp. 91-104." I do not know if this a tone language -
given it's geographical position, it may well be, and that would make it a
case of morphological tone.

(iii)
Both Remy Viredaz and Michael Becker argue that the expression of
negation has a prosodic component in English: "Even in English
negation is sometimes expressed by intonation. The mini-sentences
Pronoun + Auxiliary Verb (positive or negative) like "I do", "I don't"
seem to have high tone on the positive verb and low or low-falling tone
on the negated verb. The fact may pass unnoticed in general, so I do
not know if it's general. With "I can" and "I can't" in American
pronunciation, the final plosive and the shortening are certainly
not always very distinct so you will almost have an minimal pair.
I remember that during a stay in Britain I once answered "I can" with
unvoluntary low tone on "can", and I was answered back: "You can, or
you can't?!" with the respective high and low tones on the verbs (and
that British dark long a, though the person may have thought that I
used the American æ in can't, too). That tends to mean that tone, in
such a case, is even more important than the phonemic distinction.
....
The one-word sentences Yes and No probably have those high and low
tones, too, but I have not had the opportunity to check this."
The issue of the expression of negation in English intonation was also
raised by Michael Bernstein. He notes that this is exactly what happens
in sarcasm.

David Gil: "In Jakarta Indonesian, (almost) exclusively with the verb
_tahu_"know", a specific intonation contour can be used as the sole
marker of negation, in place of the usual overt segmental negative
marker.  The intonation contour involves low tone and lengthening of
the penultimate syllable followed by a short high-tone final syllable.  I
once heard a speaker using the same construction with another verb,
_mau_ "want", to mean "not want"."

The same phenomenon takes place in other Malay dialects, among
others Ambonese Malay. Here the tone pattern on /tau/ 'know'  is high
level. Although it is risky to make such a claim as a non-native speaker,
I get the impression that there is a paralinguistic load involved, namely
the expression of impatience or the unwillingness to communicate on a
topic.

(iv)
Very interesting/surprising was the reply from John Lynch. Whereas all
the evidence of the prosodic expression of negation in spoken
language involves pitch/tone/f0 as the prosodic encoding of negation,
he tentatively expresses the opinion "that the Hula and Aroma
languages (Austronesian languages spoken just to the east of Port
Moresby in PNG) mark negation by vowel length. Subject-TAM is
marked by prefixes (partly portmanteau) to the verb; thus in Aroma:
  pana-thao  'I will go'  pono-thao 'you will go' pene-thao 'he will go' The
negative forms of these are paana-thao, poono-thao and peene-thao."
Roland Pfau mentions another language in which negation is expressed
by vowel lengthening: Nanai, spoken in Siberia.

Malcah Yaegar-Dror replied that there is an entire bibliography on the
prosodic marking of negation. WHile not listing any studies, he says
that he is willing to provide references. One place where they may be
found is the the "Journal of Pragmatics 34(10/11), and many of the
references cited in the introduction and particularly in the articles by
Yaeger-Dror & Kaufmann.




------- End of forwarded message -------
--
Professor Bencie Woll
Chair of Sign Language & Deaf Studies, Department of Language and
Communication Science
City University, Northampton Square, London EC1V 0HB
+44 (0)20 7040 8354 (tel.) +44 (0)20 7040 8314 (minicom) +44 (0)20
7040 8577 (fax)
b.woll at city.ac.uk   www.city.ac.uk/lcsDr. Bencie Woll
b.woll at city.ac.uk
Chair of Sign Language and Deaf Studies
Language and Communication Science
City University, Northampton Square
London EC1V 0HB
Tel: +44 (0)20 7040 8354 (voice) 8314 (textphone)
Fax: +44 (0)20 7040 8577