Why a Rising Tone is Falling in Mandarin Sentences

1
Why a Rising Tone is Falling in Mandarin Sentences
Word Accents and Tones in Sentence Perspective A
symposium in conjunction with the 60th birthday
of Professor Gösta Bruce

• Chilin Shih
• University of Illinois at Urbana-Champaign

January 10, 2007
Lund, Sweden
2
Generated by WordsEye from text description.
Under development at SemanticLight, Inc.
3
Outline

• What we know
• Chinese is a lexical tone language.
• Surprise!
• Tones in sentences may deviate considerably from
  their lexical specifications.
• Research question
• Explain the difference between lexical tones and
  the observed sentence production.
• Implication
• A simulation model linking phonology to
  phonetics.

4
Chinese Lexical Tones
Tone shapes differentiate lexical meaning.
Ma1 mother Ma2 hemp Ma3 horse Ma4 to scold
5
Chinese Sentences
Ma1-ma0 ma4 ma3. Mother scolds the horse.
Ma3 ma4 ma1-ma0. The horse scolds mother.
6
Chinese Intonation Types (Data from JiahongYuan)
Statement

• Li3bai4wu3 Luo2yan4 yao4 mai3 lu4.
• On Friday Luoyan wants to buy a deer.

Question
7
Classification of Tone Shapes
Tone 1 High level
Tone 2 Rising
Tone 3 Low falling
Tone 4 High falling
8
Cause of Tonal Distortion

• Ease of articulatory effort
• Balancing articulatory effort and communication
  need

9
Physiological constraints
Communication errors

• When you say what you think
• you are saying

• When you are not saying want you
• think you are saying

10
Ease of Articulatory EffortI
11
Ease of Articulatory EffortII
12
Ease of Articulatory EffortIII
13
Production of Rising and Falling Tones
14
Severe Tonal DistortionI
15
People Talk Nearly As Fast As Possible
16
Severe Tonal DistortionII
17
Local distortion is predictable from global
optimization
18
A Racing Game
19
Adjusting the Best Path
20
Best Path in Tonal Production
0.5
1.0
1.0
1.0
0.0
21
Stem-ML

• The prosodic modeling is based on Stem-ML
• (Soft Template Mark-up Language).

• Stem-ML consists of a set of mathematically
• defined tags with value attributes.

For example Tone prosodic strength

• Allowing user-defined accent shapes, phrase
• curves, and other speaker specific parameters.

Kochanski and Shih (2003), Prosody modeling with
soft templates, Speech Communication V. 39. Shih
(in preparation), Prosody Learning and
Generation, Springer.
22
Basic Assumptions

• Pre-planning.
• Balance articulatory effort and communication
  needs (Lindblom, Ohala).
• A dynamical model for the muscles that control f0
  (Hill).

23
We further propose

• Speaker shifts weights dynamically
• as they speak.

• This is the prosodic strength,
• which reflects the articulatory effort.

24
Linking Phonology and Phonetics

• A model is a sequence of templates (i.e. points
  representing tone/accent shapes). The templates
  encodes phonological information.
• For tone languages, there is one template per
  tone. Templates are stretched to fit duration.
• Each template has a strength. The strength value
  determines phonetic variation.

25
Representation

• Surface F0 contours are coded as a set of
• Template strength

T11.0 T3 0.3 T4 1.2 T5 0.8 T21.0 T1 0.5

• Generation Template strength ? F0

• Learning Template, F0 ? Template strength

26
Modeling Math (Credit to Greg Kochanski)
Effort
is the muscle tension (frequency) at time t.
Each target encodes some linguistic information,
ri is the error of the ith target, and si is its
importance.
Error
y is the ith pitch target and a bar denotes an
average over a target.
27
Representing F0 As Tone Strength
28
Simulation of Tonal ProductionI
29
Simulation of Tonal ProductionII
30
Model Fits to Mandarin Chinese
0.61 free parameters per syllable, 13 Hz RMS
error.
31
Works for English
The highest f0 is on a weak, unaccented word.
would
I
like
Uhm
A flight to Seattle
from Albuquerque
32
Muscle Dynamics
Interpolation
33
Discourse Functions

• Topic initialization
• Discourse structure
• Phrasing
• Emphasis
• New vs. old information
• Other communicative means

34
How Do They Fit Together?
35
Conclusion

• Speech is a communication system. Speakers
  balance articulatory effort and communication
  needs.
• We need a representation that encodes
• Accent template
• Articulatory effort
• Emotional State
• We present a computational simulation model that
  generate surface phonetic variations from this
  representation.