Geen diatitel

1
(No Transcript)
2
Questions Conference - University of
Liverpool10-12 November 1999Why Question
Intonations Rise and Why they Sometimes Fall

• Carlos Gussenhoven
• University of Nijmegen

3
"Be aware that intonation is important when
differentiating between questions and statements"
Draft C-level guidelines Modern Languages
Curriculum of the Scottish Consultative Council
on the Curriculum, October 1999
4
OnbeantwoordSommige hondenKunnen heel
goed vragend blaffenMaar hoe je ook
luistertWat ze eigenlijk vragen Daar kom je
niet achterRudy Kousbroek NRC Handelsblad, 2
October 1999Kinderpagina
5
UnansweredSome dogsAre good at asking
questioninglyBut however keenly you listenWhat
theyre askingYou cannot make out
6
Ladd 1981

• Universalist Hypothesis biologically determined
  pattern of rising questions and falling
  statements
• Language-specific Hypothesis grammatically
  determined, arbitrary relationship between form
  and function

7
View defended here

• Universal meaning is omnipresent
• There is a statistical bias towards natural
  phonological interpretations
• There are degrees of naturalness in
  phonological interpretations
• Unnatural representations may arise in many
  ways through language change

8
Why do questions rise and statements fall?

• Liebermans theory of physical tension
• Ohalas ethological theory

9

• Lieberman 1967. Questions are marked vis-à-vis
  Statements, and therefore require marked pitch.
  Unmarked pitch is governed by subglottal
  pressure, marked pitch by laryngeal adjustments.
• Ohala 1983,1984, 1994. Questions are one of many
  phenomena illustrating the ethologically
  developed connection between powerlessness and
  the signalling of small size.

10
Dimorphism in humans

• Male larynx almost twice the size of female
  larynx in front-back dimension
• Male vocal tract 17 cm, female vocal tract 14.5 cm

11
Threads in Ohalas theory

• Affective use of F0
• Tone in sound symbolism
• Nonhuman vocalisations
• Non-vocal correlates
• Facial expressions
• Consonants and vowels in sound symbolism

12
Affective use of F0

• Global F0 raising ? less potent, more nervous,
  less confident, etc.
• Mean F0 Japanese women
  highest Dutch women Dutch
  men Japanese men lowest

13
The ideal man (van Bezooijen 1995)
14
The ideal woman
15
Tone in sound symbolism

• Ewe k??????k?????? small
  gba?gba?gba? large (Ohala 1983)
• High-toned diminutives in Lahu, Yi, Wu, Yue, etc.
  (Matisoff 1994, Lapolla 1994)

16
VERB STEM EXTENTIVE DIMINUTIVE be this
only this ?? big tShi h? tShi
h???/h??j ma? many tShi ma tShi
ma??/ma?j ji? long tShi Si tShi
S???/S??j
Matisoff 1994
17
F0 in nonhuman vocalisation

• F0 inversely related to body size
• Same creature - high F0 ? whine, simper,
  yelp - low F0 ? growl

18
Non-vocal correlates

• Nonvocal features suggesting large size -
  erected feathers - permanent features, like
  humps, mane, peripheral facial hair

19
Facial expressions
O-face aggression Smile submission Pouting the
lips enlarges the vocal tract, creating lower
formant frequencies, suggesting a larger organism
20
Vowels in sound symbolismhigh vowels small
sizeopen vowels large size(cf. Hinton,
Nichols Ohala 1994Sound symbolism, CUP)
21
Ohalas Frequency Code
Low and falling pitch signal protectiveness,
assertiveness and confidence high and rising
pitch signal vulnerability, submissiveness and
uncertainty
22
Phonological interpretations of natural
statements
Falling and low pitch
H Li Passim
H !H Li Passim
23
So now for rising and high pitch...
24
Phonological interpretations of natural
questions
HL Hi English, Dutch, German nonfinal
H Hi (Am) English, Dutch, German
LH Hi (Br) English, Dutch, German
L Hi Dutch, German
25
Unnatural phonological interpretations of
question intonation
LH Li Southern Italian
L HiLi Greek, Hungarian,
Romanian Bengali Limburgian Dutch
26
Explanations for unnatural questions
By-effects of natural behaviour
Ohalas Frequency code
27
One type of phonetic interrogative
intonation Raise final pitch peak
Declarative
Interrogative
28
Hz
Raised and delayed
time
29
Liberman Pierrehumbert 1984
30
The aLARM went off
A This was an everyday occurrence B This
was an unusual experience
Ladd Morton 1997
31
Late peak
Early peak
Ladd Morton 1997
32
Neapolitan Italian (DImperio 1999)
Mamma ballava da Lalla HL HL Li Mamma
ballava da Lalla ? HL LH Li Mum used to
dance at Lallas
33
With in broad focus
Mamma ballava da Lalla HL !HL Li Mamma
ballava da Lalla ? HL LH Li Mum used to
dance at Lallas
34
Unnatural phonological interpretations of
statement intonation
L Hi Belfast English LH
Li Amsterdam Dutch H Hi Chickasaw
35
So why do statements rise?
36
Explanations for unnatural contours
By-effects of natural behaviour
Explicit pronunciations
Ohalas Frequency code
37
Amsterdam Dutch
Het Openbaar Vervoer is in staking gegaan
Li HL HL Li Het
Openbaar Vervoer is in staking gegaan
Li HL LH Li
The public transport system has gone on strike
38
Full course
Truncation
End of utterance
39
Belfast English
Who do you think I met in the market today?
Mr Leaner! Mr Leaner?
Courtesy Esther Grabe
40
?
!
41
Explanations for unnatural contours
By-effects of natural behaviour
Incidental circumstances
Explicit pronunciations
Ohalas Frequency code
42
Now for a reconstructed case of a specific change
in Greek splitting a general HL into
HL vs L HiLi
43
Greek declarative H Li Broad focus
accent all lexical words Deaccenting for
-focus Greek interrogative L
HiLi Broad focus accent V (SVO) Accent O if
narrow -focus
44
A?fto to ?ifazma za?roni Li H H H
Li This fabric creases
mi?lai i ma?rina Li H Li speaking is
Marina Marina is on the PHONE
Courtesy Ineke Mennen
45
a???????? le?mo?a Li L Hi Li bought-you
lemons Did you buy lemons
? psaxnis ?ia ve?lo - na Li H L
Hi Li Are you looking for a NEEDLE
46
Simplistic proto-Greek
Declarative
!
Interrogative
!
Li HL HL Li
speaking is Marina
47
Dutch learners of Greek use higher F0 peaks in
Greek questions than do native speakers, despite
the fact that Dutch seems dull compared to
Greek (Mennen 1999) Hypothesis Dutch learners
interpret boundary peak as HL and raise it to
signal interrogative (Greeks dont do this, of
course)
48
!
Li HL HL Li
Suppose their neighbours stressed the verb in
questions...
Li L Hi Li
Now redo the phonetics...
!
milai i marina?
49
An alternative analysis as HL HiLi is unlikely,
because tones will pile up at the end if accent
is final
HL HiLi
HLHiLi
50
An even more specific case of a development
leading to falling questions and rising
statements (in addition to rising questions and
falling statements)
Roermond Dutch
51
Polytonic area
52
Earliest representation of Accent 2 (1) a. dax
b. da?x H
day days Brief account of
its origin (2) SG PL Middle High German
dax da?? OSL dax da???
Apocope dax da?x Fake AL ?da?x
da?x Interpret duration as H dax
da?x H
53
Current representation of Accent 2 (1) a.
da?x b. da?x H
day days and
many other minimal pairs...
54
The contrast between Acc 1 and Acc 2 occurs on
stressed syllables containing two sonorant
moras VV a?, ??, ... V Son ??, ??,
...
55
Maastricht bEin2 bEin1 leg legs
bQ??x2 - bQ??x1 mountain mountains
pe??t2 pe??t1 horse horses
stEin2 stEin1 stone stones
Tongeren b?i?f2 b?i?f1 letter
letters ki??2 ki?l1 shirt
shirts kna?jn2 kna?jn1 rabbit
rabbits nO?t2 nO?t(?)1 nut nuts
56
Simplistic Middle High German
a.
b. Ich seh nen MANN Sieste
nen MANN?
H Li L Hi I see
a man Do you see a man?
Enter lexical tone...
man1 basket man2 man
57
Declarative Acc1 and Acc 2
a. b. Ich
seh nen ??? Ich seh nen ???

H Li HH Li I see a
basket I see a man
58
Maastricht
Iech höb e bal H Li I have a
party Iech höb ne bal H H Li I
have a ball
59
Interrogative Acc 1 and Acc 2
a. b. Sieste nen ???
Sieste nen ???
L Hi LH Hi Do you
see a basket? Do you see a man?
60
Reproduce the firmness of Acc 1 in Interrogative
a. b. Sieste nen ???
Sieste nen ???
L Hi Li LH Hi Li Do you
see a basket? Do you see a man? (Truncati
on)
61
Problems in nonfinal position...
a. b. Wenn da eine ???? ist
Wenn da eine???? ist \
\ L
HiLi LH HiLi When there is a
pigeon When there is a baptism
62
Solve by assimilation of LH to LL
a. b. Wenn da eine ???? ist
Wenn da eine???? ist \
\ L
HiLi LL HiLi When there is a
pigeon When there is a baptism
63
Roermond
Zitte dien BEIN aan dien veut? Are your legs
attached to your feet? Zit diene BEIN aan
diene voot? Is your leg attached to your foot?
64
New representation of final interrogative Acc 2
Sieste nen mAn
\ LL HiLi H-assimilation H ? L
/ (L ? )?
65
Now for rising statements...
66
Three phonetic implementations of HHLi
a. b. c. mAn mAn
mAn
HH Li HH Li HH
Li man
67
Get HLH by placing lexical tone outside Ti
Declarative Interrogative
a. b. m A n m A n

HLiH LLi H
68
Final interrogative

• Old situation L H HiLi
• New order L HiLi H
• Assimilate L LiLi H
• Identity L Li H

And so were back to natural intonation for
this one...
69
Ik gaef Toon eine knien \ Li
HH HLiH I give Tony a rabbit
Ister get gebeurd met diene bein?
\ Li LLiH Has something
happened to your leg?
70
So a rising declarative and a rising
interrogative for Acc 2, but a falling
declarative and a falling interrogative for Acc1
71
Ik gaef Toon twee knien Li
HH HLi I give Tony two rabbits
Ister get gebeurd met dien bein?
\ Li L HiLi Has something
happened to your legs?
72
Prediction of omnipresence
73
Conclusions

• The Frequency Code is reponsible for the bias
  towards high/rising pitch in Question
  intonations.
• The Frequency Code indirectly explains why peaks
  in Questions are late

74

• Explicitness is indirectly responsible for late
  peaks in significant statements
• Late peaks may develop into rises through
  truncation
• Specific circumstances may cause unnatural
  patterns to develop...
• but the Frequency Code will always be always with
  us.

75
The End
76
(No Transcript)