Czech ToBI - WordPress.com
1
3 Creating a Czech ToBI Czech ToBI Andrea Pešková References Beňuš, Š. & U. Reichel & K. Mády. 2014. Modeling Accentual Phrase Intonation in Slovak and Hungarian. In: Veselovská, L.; Janebová, M. (Hrsg.): Complex Visibles Out There. Olomouc modern language series, Bd. 4. Olomouc, Czech Republic: Palacký University. S. 677-689. Daneš, F. 1957. Intonace a věta ve spisovné češtině. Praha: Nakladatelství Československé akademie věd. Duběda, T. 2011. Towards an inventory of pitch accent for read Czech. Slovo a slovesnost 72, 3–11. Gussenhoven, C. 2004. The Phonology of Tone and Intonation. Cambridge: CUP. Hayes, B. & A. Lahiri. 1991. Bengali Intonational Phonology. Natural Language and Linguistic Theory 9: 47–96. Pierrehumbert, J. B. 1980. The Phonology and Phonetics of English Intonation. Cambridge, MA: MIT Press. Mathesius, V. 1937. Mluvní takt a některé problémy příbuzné. Naše řeč 33(4), 193-199. Ondráčková, J. 1954. O mluvním rytmu v češtině. Slovo a slovesnost 15(4), 145-157. Petřík, S. 1935. K intonaci věty. Intonace pozdravů a jiných výrazů zdvořilostních. Naše řeč 33(3-4), 62-68. Prieto P., et al. 2015. Intonational phonology of Catalan and its dialectal varieties. In: Intonational Variation in Romance (eds. S. Frota & P. Prieto), 9–62. Sengar, A. & R. Mannell. 2012. A Preliminary Study of Hindi Intonation. ASSTA. 149–152. *Note: This is a second version of the poster, which I did after the PaPE conference. I would like to thank to the audience for critical questions, comments and suggestions. 1 Previous research on Czech intonation • Descriptive works: - Petřík 1935; Mathesius 1937; Ondráčková 1954; Daneš 1957; etc. • Duběda 2014: Czech intonational grammar (AM-Model; see Pierrehumbert 1980; Gussenhoven 2004): 2 Some (basic) prosodic properties • Stress rule: fixed word-initial stress • Acoustic parameters for stress: - *intensity, *F0-rise on a stressed σ, *V length or *V quantity Ø What is the most important cue for recognition of stress? > AP ? • Isolated p-words: v Falling pattern: H* L • P-words in a context: v prenuclear (non-final) non-focal element: L* Ha v non-final focal element: L+H* L- v Observation: the alignment changes according to the length of the (p-)word (see Section 3) • Nuclear stress rule: right-most position (default rheme-position) > no (prosodic) prominence. • Tonal compression: Acknowledgments Trudel Meisenburg Tomáš Duběda Wendy-Elvira García Paolo Roseano Radek Skarnitzl Rafèu Sichel-Bazin Maria del Mar Vanrell all participants Phonetics and Phonology in Europe 2017 Inventory of Czech nuclear configurations, and their use in sentence types 1 L* / L*+H (L)H% Neutral, echo, confirmatory, imperative, and polite yes-no question. 2 L*+H / L+H* LH% Counter-expectative yes-no question. 3 H* / L* L% Broad focus statement. Exhort / request (neutral). Order / imperative. Exclamative wh-question. Exhortative wh-question. 4 L+H* / L*+H L% Narrow focus statement (Fc = object). 5 L*+H L% Emphatic exclamative. Obviousness statement. Order / imperative. Hesitation statement. Exhort / request (neutral). 6 L* / H* L% Neutral and imperative wh-question. 7 L*+H / L+H* !H% Neutral and insistent vocative. 8 L*+H / H* !Hː%/Lː% Exhort / insistent request (familiar). Pitch accents (nuclear): L*+H L+H* L* H* Boundary tones (BI 4): (L)H% L% !H% !Hː% j a n j a n a m i l e n a mahulena literatura spolupratsovɲ itse 4 Remaining issues • Determination of the prosodic hierarchy: IP > ip > AP > ω ? • Closer defining of AP = possible to have more than one prosodic word ? • Further strategies (e.g. particles) which impact on pragmatic meaning ? • Modeling and explaining the observed variation in the data. SVO: F0 of subjects with different number of σ: black line (broad focus), red line (narrow focus): SVO: F0 of objects with different number of σ: black line (broad focus), red line (narrow focus): L+H* LH% L+H* LH% <had> (snake) <Jana> (Jane) h a t j a n a 0 500 100 200 300 400 Time (s) 0 1.236 L+H* LH% *σ *σ σ v i.e. no truncation in a context where two or more tones are associated with the same segmental element (e.g. echo- surprise y/n question) → • Methodology - Data: Discourse completion task, supplementary reading materials (more than 300 sentences) - 10 speakers: South Moravian dialect - Acoustic analysis: Praat (Boersma & Weenink 2016) ˈma li ni H* L% l o m j a n a m i l a n a literaturu mahulenu spolupratsovɲ itsi PROPOSAL: Czech is an accentual (phrase) language (cf. French, Hindi, Bengalí, Slovak, Hungarian / see e.g. Jun & Fletcher 2014; Beñuš et al. 2014 ) its prosody is left-headed AP = rhythmic group from *σ to *σ/BI3/BI4; one pitch accent Hindi type? or Bengali type? Default tonal pattern of the AP: (optional initial high BT) L* Ha (Hayes & Lahiri 1991) (Sengar & Mannell 2012) L* H ϕ L* H ϕ (H*/L*) (L ϕ ) | | | [ [word_1] AP [word_2] AP [word_N] AP ] BT% L*+H L*+H (H*/L*) | | | [ [word_1] AP [word_2] AP [word_N] AP ] BT% 50 400 100 200 300 Time (s) 0 1.232 50 400 100 200 300 Time (s) 0 1.242 ‘Barbara loves raspberries’ [ F SVO] [ F S]VO Barbara miluje maliny L*+H L+H* v Preference for syntactic marking of focus, but also prosodic marking of elements in situ is possible (e.g. SVO) → L- a ‘a’ = AP final BT 4 7 2 1 5 3 6 8
Transcript of Czech ToBI - WordPress.com
3 Creating a Czech ToBI
Czech ToBIAndrea Pešková
ReferencesBeňuš, Š. & U. Reichel & K. Mády. 2014. Modeling Accentual Phrase Intonation in Slovak and Hungarian. In: Veselovská, L.; Janebová, M. (Hrsg.): Complex Visibles Out There. Olomouc modernlanguage series, Bd. 4. Olomouc, Czech Republic: Palacký University. S. 677-689. Daneš, F. 1957. Intonace a věta ve spisovné češtině. Praha: Nakladatelství Československé akademie věd. Duběda, T. 2011. Towards an inventory of pitch accent for read Czech. Slovo a slovesnost 72, 3–11. Gussenhoven, C. 2004. The Phonology of Tone and Intonation. Cambridge: CUP. Hayes, B. & A. Lahiri. 1991. Bengali Intonational Phonology. Natural Language and Linguistic Theory 9: 47–96. Pierrehumbert, J. B. 1980. The Phonology and Phonetics of English Intonation. Cambridge, MA: MIT Press.Mathesius, V. 1937. Mluvní takt a některé problémy příbuzné. Naše řeč 33(4), 193-199. Ondráčková, J. 1954. O mluvním rytmu v češtině. Slovo a slovesnost 15(4), 145-157. Petřík, S. 1935. K intonacivěty. Intonace pozdravů a jiných výrazů zdvořilostních. Naše řeč 33(3-4), 62-68. Prieto P., et al. 2015. Intonational phonology of Catalan and its dialectal varieties. In: Intonational Variation in Romance (eds. S. Frota & P. Prieto), 9–62. Sengar, A. & R. Mannell. 2012. A Preliminary Study of Hindi Intonation. ASSTA. 149–152. *Note: This is a second version of the poster, which I did after the PaPE conference. I would like to thank to the audience for critical questions, comments and suggestions.
1 Previous research on Czech intonation• Descriptive works:- Petřík 1935; Mathesius 1937; Ondráčková 1954; Daneš 1957; etc.
• Duběda 2014: Czech intonational grammar (AM-Model; see Pierrehumbert 1980; Gussenhoven 2004):
2 Some (basic) prosodic properties• Stress rule: fixed word-initial stress• Acoustic parameters for stress:
- *intensity, *F0-rise on a stressed σ, *V length or *V quantity Ø What is the most important cue for recognition of stress? > AP ?
• Isolated p-words:v Falling pattern: H* L
• P-words in a context: v prenuclear (non-final) non-focal element: L* Hav non-final focal element: L+H* L-
v Observation: the alignment changes according to the length of the (p-)word (see Section 3)
• Nuclear stress rule: right-most position (default rheme-position) > no (prosodic) prominence.
• Tonal compression:
AcknowledgmentsTrudel Meisenburg
Tomáš DubědaWendy-Elvira García
Paolo RoseanoRadek Skarnitzl
Rafèu Sichel-BazinMaria del Mar Vanrell
all participants
Phonetics and Phonology in Europe 2017
Inventory of Czech nuclear configurations, and their use in sentence types
1 L* / L*+H (L)H% Neutral, echo, confirmatory, imperative, and polite yes-noquestion.
2 L*+H / L+H* LH% Counter-expectative yes-no question.
3 H* / L* L% Broad focus statement. Exhort / request (neutral). Order /imperative. Exclamative wh-question. Exhortative wh-question.
4 L+H* / L*+H L% Narrow focus statement (Fc = object).
5 L*+H L% Emphatic exclamative. Obviousness statement. Order / imperative.Hesitation statement. Exhort / request (neutral).
6 L* / H* L% Neutral and imperative wh-question.
7 L*+H / L+H* !H% Neutral and insistent vocative.
8 L*+H / H* !Hː% / Lː% Exhort / insistent request (familiar).
Pitch accents (nuclear):
L*+H L+H* L* H*
Boundary tones (BI 4):
(L)H% L% !H% !Hː%
j a n j a n a m i l e n a
m a h u l e n a l i t e r a t u r a spo lupra tsovɲ i tse
elektromagnetismus
j a n j a n a m i l e n a
m a h u l e n a l i t e r a t u r a spo lupra tsovɲ i tse
elektromagnetismus
j a n j a n a m i l e n a
m a h u l e n a l i t e r a t u r a spo lupra tsovɲ i tse
elektromagnetismus
j a n j a n a m i l e n a
m a h u l e n a l i t e r a t u r a spo lupra tsovɲ i tse
elektromagnetismus
j a n j a n a m i l e n a
m a h u l e n a l i t e r a t u r a spo lupra tsovɲ i tse
elektromagnetismus
j a n j a n a m i l e n a
m a h u l e n a l i t e r a t u r a spo lupra tsovɲ i tse
elektromagnetismus
4 Remaining issues• Determination of the prosodic hierarchy: IP > ip > AP > ω ?• Closer defining of AP = possible to have more than one prosodic word ?• Further strategies (e.g. particles) which impact on pragmatic meaning ?• Modeling and explaining the observed variation in the data.
SVO: F0 of subjects with different number of σ: black line (broad focus), red line (narrow focus):
SVO: F0 of objects with different number of σ: black line (broad focus), red line (narrow focus):
L+H* LH% L+H* LH%<had> (snake) <Jana> (Jane)
h a t j a n a
0
500
100200300400
Pitch
(Hz)
Time (s)0 1.236
L+H* LH%
*σ *σ σ
v i.e. no truncation in a context where two or more tones are associated with the same segmental element (e.g. echo-surprise y/n question)
→
• Methodology- Data: Discourse completion task, supplementary reading materials
(more than 300 sentences)- 10 speakers: South Moravian dialect- Acoustic analysis: Praat (Boersma & Weenink 2016)
ˈma li ni(’raspberries’)
70
300
100150200250
Pitch
(Hz)
Time (s)0 0.9865
H* L%
Time (s)0 0.2609Pit
ch (s
emito
nes re 1
00 H
z)
-12
30
Time (s)0 0.2318Pit
ch (s
emito
nes re 1
00 H
z)
-12
30
Time (s)0 0.2597Pit
ch (s
emito
nes re 1
00 H
z)
-12
30
Time (s)0 0.2435Pit
ch (s
emito
nes re 1
00 H
z)
-12
30
Time (s)0 0.4401Pit
ch (s
emito
nes re 1
00 H
z)
-12
30
Time (s)0 0.4398Pit
ch (s
emito
nes re 1
00 H
z)
-12
30
Time (s)0 0.419Pit
ch (s
emito
nes re 1
00 H
z)
-12
30
Time (s)0 0.5246Pit
ch (s
emito
nes re 1
00 H
z)
-12
30
Time (s)0 0.648Pit
ch (s
emito
nes re 1
00 H
z)
-12
30
Time (s)0 0.6455Pit
ch (s
emito
nes re 1
00 H
z)
-12
30
Time (s)0 1.066Pit
ch (s
emito
nes re 1
00 H
z)
-12
30
Time (s)0 1.117Pit
ch (s
emito
nes re 1
00 H
z)
-12
30
Time (s)0 1.204Pit
ch (s
emito
nes re 1
00 H
z)
-12
30
Time (s)0 1.258Pit
ch (s
emito
nes re 1
00 H
z)
-12
30
l o m j a n a m i l a n a
m a h u l e n u l i t e r a t u r u spo lup ra tsovɲ i ts i
elektromagnetismus
Time (s)0 0.2609Pit
ch (s
emito
nes re 1
00 H
z)
-12
30
Time (s)0 0.2318Pit
ch (s
emito
nes re 1
00 H
z)
-12
30
Time (s)0 0.2597Pit
ch (s
emito
nes re 1
00 H
z)
-12
30
Time (s)0 0.2435Pit
ch (s
emito
nes re 1
00 H
z)
-12
30
Time (s)0 0.4401Pit
ch (s
emito
nes re 1
00 H
z)
-12
30
Time (s)0 0.4398Pit
ch (s
emito
nes re 1
00 H
z)
-12
30
Time (s)0 0.419Pit
ch (s
emito
nes re 1
00 H
z)
-12
30
Time (s)0 0.5246Pit
ch (s
emito
nes re 1
00 H
z)
-12
30
Time (s)0 0.648Pit
ch (s
emito
nes re 1
00 H
z)
-12
30
Time (s)0 0.6455Pit
ch (s
emito
nes re 1
00 H
z)
-12
30
Time (s)0 1.066Pit
ch (s
emito
nes re 1
00 H
z)
-12
30
Time (s)0 1.117Pit
ch (s
emito
nes re 1
00 H
z)
-12
30
Time (s)0 1.204Pit
ch (s
emito
nes re 1
00 H
z)
-12
30
Time (s)0 1.258Pit
ch (s
emito
nes re 1
00 H
z)
-12
30
l o m j a n a m i l a n a
m a h u l e n u l i t e r a t u r u spo lup ra tsovɲ i ts i
elektromagnetismus
Time (s)0 0.2609Pit
ch (s
emito
nes re 1
00 H
z)
-12
30
Time (s)0 0.2318Pit
ch (s
emito
nes re 1
00 H
z)
-12
30
Time (s)0 0.2597Pit
ch (s
emito
nes re 1
00 H
z)
-12
30
Time (s)0 0.2435Pit
ch (s
emito
nes re 1
00 H
z)
-12
30
Time (s)0 0.4401Pit
ch (s
emito
nes re 1
00 H
z)
-12
30
Time (s)0 0.4398Pit
ch (s
emito
nes re 1
00 H
z)
-12
30
Time (s)0 0.419Pit
ch (s
emito
nes re 1
00 H
z)
-12
30
Time (s)0 0.5246Pit
ch (s
emito
nes re 1
00 H
z)
-12
30
Time (s)0 0.648Pit
ch (s
emito
nes re 1
00 H
z)
-12
30
Time (s)0 0.6455Pit
ch (s
emito
nes re 1
00 H
z)
-12
30
Time (s)0 1.066Pit
ch (s
emito
nes re 1
00 H
z)-12
30
Time (s)0 1.117Pit
ch (s
emito
nes re 1
00 H
z)-12
30
Time (s)0 1.204Pit
ch (s
emito
nes re 1
00 H
z)
-12
30
Time (s)0 1.258Pit
ch (s
emito
nes re 1
00 H
z)
-12
30
l o m j a n a m i l a n a
m a h u l e n u l i t e r a t u r u spo lup ra tsovɲ i ts i
elektromagnetismus
Time (s)0 0.2609Pit
ch (s
emito
nes re 1
00 H
z)
-12
30
Time (s)0 0.2318Pit
ch (s
emito
nes re 1
00 H
z)
-12
30
Time (s)0 0.2597Pit
ch (s
emito
nes re 1
00 H
z)
-12
30
Time (s)0 0.2435Pit
ch (s
emito
nes re 1
00 H
z)
-12
30
Time (s)0 0.4401Pit
ch (s
emito
nes re 1
00 H
z)
-12
30
Time (s)0 0.4398Pit
ch (s
emito
nes re 1
00 H
z)
-12
30
Time (s)0 0.419Pit
ch (s
emito
nes re 1
00 H
z)
-12
30
Time (s)0 0.5246Pit
ch (s
emito
nes re 1
00 H
z)
-12
30
Time (s)0 0.648Pit
ch (s
emito
nes re 1
00 H
z)
-12
30
Time (s)0 0.6455Pit
ch (s
emito
nes re 1
00 H
z)
-12
30
Time (s)0 1.066Pit
ch (s
emito
nes re 1
00 H
z)
-12
30
Time (s)0 1.117Pit
ch (s
emito
nes re 1
00 H
z)
-12
30
Time (s)0 1.204Pit
ch (s
emito
nes re 1
00 H
z)
-12
30
Time (s)0 1.258Pit
ch (s
emito
nes re 1
00 H
z)
-12
30
l o m j a n a m i l a n a
m a h u l e n u l i t e r a t u r u spo lup ra tsovɲ i ts i
elektromagnetismus
Time (s)0 0.2609Pit
ch (s
emito
nes re 1
00 H
z)
-12
30
Time (s)0 0.2318Pit
ch (s
emito
nes re 1
00 H
z)
-12
30
Time (s)0 0.2597Pit
ch (s
emito
nes re 1
00 H
z)
-12
30
Time (s)0 0.2435Pit
ch (s
emito
nes re 1
00 H
z)
-12
30
Time (s)0 0.4401Pit
ch (s
emito
nes re 1
00 H
z)
-12
30
Time (s)0 0.4398Pit
ch (s
emito
nes re 1
00 H
z)
-12
30
Time (s)0 0.419Pit
ch (s
emito
nes re 1
00 H
z)
-12
30
Time (s)0 0.5246Pit
ch (s
emito
nes re 1
00 H
z)
-12
30
Time (s)0 0.648Pitch
(sem
itone
s re 1
00 H
z)
-12
30
Time (s)0 0.6455Pitch
(sem
itone
s re 1
00 H
z)
-12
30
Time (s)0 1.066Pit
ch (s
emito
nes re 1
00 H
z)
-12
30
Time (s)0 1.117Pit
ch (s
emito
nes re 1
00 H
z)
-12
30
Time (s)0 1.204Pit
ch (s
emito
nes re 1
00 H
z)
-12
30
Time (s)0 1.258Pit
ch (s
emito
nes re 1
00 H
z)
-12
30
l o m j a n a m i l a n a
m a h u l e n u l i t e r a t u r u spo lup ra tsovɲ i ts i
elektromagnetismus
Time (s)0 0.2609Pit
ch (s
emito
nes re 1
00 H
z)
-12
30
Time (s)0 0.2318Pit
ch (s
emito
nes re 1
00 H
z)
-12
30
Time (s)0 0.2597Pit
ch (s
emito
nes re 1
00 H
z)
-12
30
Time (s)0 0.2435Pit
ch (s
emito
nes re 1
00 H
z)
-12
30
Time (s)0 0.4401Pit
ch (s
emito
nes re 1
00 H
z)
-12
30
Time (s)0 0.4398Pit
ch (s
emito
nes re 1
00 H
z)
-12
30
Time (s)0 0.419Pit
ch (s
emito
nes re 1
00 H
z)
-12
30
Time (s)0 0.5246Pit
ch (s
emito
nes re 1
00 H
z)
-12
30
Time (s)0 0.648Pit
ch (s
emito
nes re 1
00 H
z)
-12
30
Time (s)0 0.6455Pit
ch (s
emito
nes re 1
00 H
z)
-12
30
Time (s)0 1.066Pit
ch (s
emito
nes re 1
00 H
z)
-12
30
Time (s)0 1.117Pit
ch (s
emito
nes re 1
00 H
z)
-12
30
Time (s)0 1.204Pit
ch (s
emito
nes re 1
00 H
z)
-12
30
Time (s)0 1.258Pit
ch (s
emito
nes re 1
00 H
z)
-12
30
l o m j a n a m i l a n a
m a h u l e n u l i t e r a t u r u spo lup ra tsovɲ i ts i
elektromagnetismus
PROPOSAL: Czech is an accentual (phrase) language(cf. French, Hindi, Bengalí, Slovak, Hungarian / see e.g. Jun & Fletcher 2014; Beñuš et al. 2014 )
its prosody is left-headedAP = rhythmic group from *σ to *σ/BI3/BI4; one pitch accent
Hindi type?
or
Bengali type?
Default tonal pattern of the AP: (optional initial high BT) L* Ha
(Hayes & Lahiri 1991)
(Sengar & Mannell 2012)
L* Hϕ L* Hϕ (H*/L*) (Lϕ)| | |
[ [word_1] AP [word_2] AP [word_N] AP ]BT%
L*+H L*+H (H*/L*)| | |
[ [word_1] AP [word_2] AP [word_N] AP ]BT%
Barbara miluje maliny
Barbara loves raspberries.
50
400
100
200
300
Pitch
(Hz)
Time (s)0 1.232
Barbara miluje maliny
Barbara loves raspberries.
50
400
100
200
300
Pitch
(Hz)
Time (s)0 1.242
‘Barbara loves raspberries’
[FSVO][FS]VO
Barbara miluje maliny
L*+HL+H*
v Preference for syntactic marking of focus, but also prosodic marking of elements in situ is possible (e.g. SVO) →
L-a
‘a’ = AP final BT
4
72
1
5
3
6
8
