Phonetic Properties of Nuclear Prominence in Japanese
Yoshihisa Kitagawa, Shinichiro Ishihara & Shigeto Kawahara Indiana University / Lund University / Keio University
([email protected] / [email protected] / [email protected])
This research was partially supported by the National Science Foundation under Grant No. 0650415.
Graphics provided by Sekitei
1. Introduction — Nuclear Prominence:
Research Setting: Investigation of prosody-information synchronization Goal: Quantitatively confirm the prosody of nuclear prominence in Japanese
Independent of phonetic effects of focus Nuclear Prominence:
Eng: Phonologically derived/constrained default accent in a sentence (Chomsky
& Halle 1968, Lieberman 1975, Cinque 1993, Truckenbrodt 1995, et al.) Jpn: Only sporadically and informally observed (Ishihara 2001, Sato 2012) (1) a. Taroo-ga (kyoo) hon-o katta
Taro-NOM today book-ACC bought b. hon-o Taroo-ga thon-o katta
book-ACC Taro-NOM bought
c. hon-o Taroo-ga kyoo thon-o katta book-ACC Taro-NOM today bought
'Taro bought a book (today).' Ishihara (2001:145)
2. Production Experiment:
(2) a. Participants: 12 Tokyo Japanese speakers (10M/2F; Age18~20; paid) b. Procedure:
(i) Experimental sentences presented on a computer screen in a pseudo-randomized order.
(ii) After silently reading and understanding (i), participants recited
them with a natural and smooth intonation reflecting the interpretation.
c. Stimuli:
(i) Sentences with 2 independent variables:
3 distinct word order — (3a-c)
• 3 distinct sentence types — (4a-c)
(ii) 3 distinct lexical sets and 6 recordings (out of 7):
• 3 × 3 × 3 × 6 = 162 tokens
Factor 1: Word order
(3) a. Noun-wa Noun-de Noun-ni Noun-o Verb
-TOP -at -DAT -ACC
b. Noun-wa Noun-de Noun-o Noun-ni Verb -TOP -at -ACC -DAT
c. Noun-wa Noun-ni Noun-o Noun-de Verb -TOP -DAT -ACC -at
Factor 2: Sentence types (with distinct information structures) (4a) All-new Declarative without any context:
Nori'yuki-wa Yona'guni-de Nao'miti-ni awa'mori-o azu'keta. Noriyuki-TOP Yonaguni-at Naomichi-DAT rice.brandy-ACC entrusted 'Naoyuki entrusted awamori (rice brandy) to Naomichi at Yanaguni.'
All nominals lexically accented on the second of 5 moras—LH'LLL (4b) Verum Focus on Verb:
(4a)? + soretomo azuke'-nakat-ta? or entrust-not-PAST
'Did Naoyuki entrust awamori to Naomichi at Yanaguni or did he not?' (4c) Constituent Focus (answer to a wh-question):
Q: pro1 pro2 pro3 na'ni-o azu'keta? 'What did he1 entrust what-ACC entrusted to him3 there2?'
A: Nori'yuki-wa Yona'guni-de Nao'miti-ni aWA'mori-o azu'keta. Noriyuki-TOP Yonaguni-at Naomichi-DAT rice.brandy-ACC entrusted 'Naoyuki entrusted awamori (rice brandy) to Naomichi at Yanaguni.'
3. Results:
Figure 1: the size of the initial rises (i.e., the "LH" part of the LHLLL contour)
Zooming in!
Two Observations:
Preverbal elements always show higher initial rise than
the preceding elements
However, there is a difference in magnitude:
Constituent Focus > All-new Declarative > Verum Focus
The patterns are consistent across the conditions.
Final phrase = Acc Obj Final Phrase = Dat Obj Final Phrase = Adjunct
We assessed whether rises are larger in phrase 4 than in phrase 3,
by building linear mixed models (Random factors: speaker, item, repetition).
Linear mixed model with PHRASE (the difference btw phrase 3 & phrase 4), SENTENCE TYPE, WORD ORDER, their interactions as the fixed factors.
• PHRASE only has a marginal effect (t=1.77, p=0.08)
• This was because the interaction between SENTENCE TYPE and PHRASE was highly significant (t=9.01, p<.001)
The prominence of phrase 4 depends on the sentence types to a large degree.
Linear mixed models for each sentence type, with PHRASE and WORD ORDER and their interaction as fixed factors (i.e., testing the effect of word order on the prominence)
All-new Declarative sentence:
• Rises were larger in phrase 4 than phrase 3 (t=5.83, p<.001).
• This effect did not interact with WORD ORDER (adjunct-final vs. dative-final: t=.65, n.s.; adjunct-final vs. accusative-final: t=.95, n.s.).
Regardless of word order, the preverbal prominence is observed.
Verum-focus sentences:
• Rises were larger in phrase 4 than phrase 3 (t=6.9, p<.001).
• The effect was consistent across the three constructions except:
The prominence of phrase 4 was significantly different between adjunct-final sentences and dative-final sentences (t=-2.47, p<.05).
No significant difference between adjunct-final sentences and accusative-final sentences (t=-1.60, n.s.).
Constituent focus sentences:
• Rises were larger in phrase 4 than phrase 3 (t=18.06, p<.001)
• The interaction terms were not significant (adjunct-final vs. dative-final: t=0.04, n.s.; adjunct-final vs. accusative-final: t=-1.13, n.s.).
Regardless of word order, the preverbal prominence is observed.
Linear mixed models for each word order, with PHRASE, SENTENCE TYPE, and their interaction as the fixed factors (i.e., testing the effect of sentence type on the prominence)
• The interaction terms were significant for the model comparing all- new declarative and constituent focus (t=14.49, p<.001) and the one comparing verum focus and constituent focus (t=16.09, p<.001).
• The interaction was not significant in the model comparing all-new declarative and verum focus (t=-1.85, n.s.).
Prominence of phrase 4 is more pronounced in the constituent focus sentences than the other two sentences.
The results are summarized below:
(5) a. Phrase 4 showed a larger initial rise than phrase 3 in all of the stimulus sentences.
The phonetic prominence of the preverbal element in Japanese holds across all three sentence types, even along with verum focus.
b. All types of phrases (de-adjunct, ni-dative, o-accusative) received phonetic prominence in phrase 4.
The prominence of phrase 4 does not interact with syntactic category or grammatical function of the preverbal elements.
c. The size of phonetic prominence in phrase 4 is larger when it received constituent focus, as in (4c), than the other two conditions.
Constituent focus has an additional boosting effect that is
independent of the prominence assigned to the preverbal element.
4. Nuclear Prominence and Information Packaging:
(6) Nuclear prominence in Japanese:
a. All-new declarative sentence ((4a))
Nuclear Prominence without alternatives involved
b. Verum focus sentence ((4b))
Nuclear Prominence in addition to the focus realized on the verb:
Magnitude: [ Pre-V in All-new declarative > Pre-V in Verum focus ]
c. Constituent focus sentence ((4cA)) vs. all-new declarative sentence ((4a))
Magnitude: [ Focus prominence > Nuclear prominence ]
(7) Conclusion:
Nuclear Prominence is independent of Focus Prominence and the two must be distinguished.
(4a) All-new Declarative without any context:
Nori'yuki-wa Yona'guni-de Nao'miti-ni awa'mori-o azu'keta. Noriyuki-TOP Yonaguni-at Naomichi-DAT rice.brandy-ACC entrusted 'Naoyuki entrusted awamori (rice brandy) to Naomichi at Yanaguni.'
(4b) Verum Focus on Verb:
(4a)? + soretomo azuke'-nakat-ta? or entrust-not-PAST
'Did Naoyuki entrust awamori to Naomichi at Yanaguni or did he not?'
(4c) Constituent Focus (answer to a wh-question): Q: pro1 pro2 pro3 na'ni-o azu'keta? what-ACC entrusted 'What did he1 entrust to him3 there2?'
A: Nori'yuki-wa Yona'guni-de Nao'miti-ni aWA'mori-o azu'keta. Noriyuki-TOP Yonaguni-at Naomichi-DAT rice.brandy-ACC entrusted 'Naoyuki entrusted awamori (rice brandy) to Naomichi at Yanaguni.'
(8) Contrast with the previous claims:
a. Jackendoff (1972), Truckenbrodt (1995):
Nuclear prominence must be contained in a focused constituent.
But nuclear prominence not contained in the verum focus realized on the verb ((4b))
b. Krifka (2008: 251): (cf. Klein–von Stutterheim 1987; van Kuppevelt 1994; Roberts 1996; Büring 2003)
The most important or new part of the utterance as presentational/ informational focus ⇒ Induces alternatives as an answer to an implicit wh-question suggested by the context.
Nuclear prominence (4a) and Focus prominence (4cA) would be indistinguishable.
⇒ Nuclear prominence in (4a) ≠ Answer to the wh-question:
"What did Noriyuyki entrust to Naomichi at Yonaguni?" (= (4cQ))
(9) Informally perceived interpretive effect of Nuclear Prominence: Indicates an item carrying somewhat greater informational salience/ weight than other (preceding) elements in a sentence.
Potentially affects the content of common ground (CG) of the discourse (Stalnaker 1974, Karttunen 1974) without inducing alternatives.
(10) Nuclear Prominence in Japanese:
a. Induces synchronized prominence in prosody and interpretation. b. Distinct from focus prominence both prosodically and interpretively.
(11) Further research agenda:
(i) Need to verify the interpretive effects of Nuclear Prominence. (ii) Need to find out the exact nature of this information packaging. cf. information focus (É. Kiss 1998)
presentational focus (Selkirk 2002)
non-focal but discourse new element (Selkirk 2008)
< Finis >
Thank you very much!
5. Appendix
(12) Nori'yuki-wa Yona'guni-de Nao'miti-ni awa'mori-o azu'keta. Noriyuki-TOP Yonaguni-at Naomichi-DAT rice.brandy-ACC entrusted 'Naoyuki entrusted awamori (rice brandy) to Naomichi at Yanaguni.' (紀之
ゆき
与那国で 直 泡盛
あわも
預
あず
｡)
(13) Nao'yuki-wa Mori'oka-de umaiti-ni ane'ue-o mane'ita. Naoyuki-TOP Morioka-at horse.festival-DAT sister-ACC invited 'Naoyuki invited his sister to the horse festival in Morioka.'
(直之
ゆき
盛岡
も
で 馬市
う い
姉上
あ うえ
招い ｡)
(14) Nao'yuki-wa Ao'yama-de Ane'tati-ni eri'maki-o era'nda Naoyuki-TOP Aoyama-at sisters-DAT scarf-ACC chose 'Naoyuki chose scarfs for his sisters at Aoyama.
(直之
ゆき
青山
あ や
で 姉
あ
襟巻
え き
選
え
｡)
6. References:
Boersma, Paul, and David Weenink (2015) Praat: doing phonetics by computers [Computer program]. URL http://www.praat.org.
Büring, Daniel 2003. On D-trees, beans, and B-accents. In: Linguistics and Philosophy 26 : 511–45.
Chomsky, Noam and Morris Halle (1968) The Sound Pattern of English, Harper
& Row, New York.
Cinque, Guelielmo (1993) "A Null Theory of Phrase and Compound Stress," Linguistic Inquiry, 24, 239–298.
É. Kiss, Katalin (1998) "Identification focus and information focus." Language 74:245–341.
Ishihara, Shinichiro (2001) "Stress, focus, and scrambling in Japanese," in Elena Guerzoni and Ora Matushansky eds., MIT Working Papers in Linguistics 39, MITWPL, Cambridge, MA, 151–185.
Jackendoff, Ray (1972) Semantic Interpretation in Generative Grammar, The MIT Press, Cambridge, MA.
Klein, Wolfgang and Christiane von Stutterheim (1987) "Quaestio und
referentielle Bewegung in Erz.hlungen." In: Linguistische Berichte 109: 163– 83.
Karttunen, Lauri (1974) "Presuppositions and linguistic context, "Theoretical Linguistics 1, 181–194.
Krifka, Manfred (2008) "Basic Notions of Information Structure," Acta Linguistica Hungarica 55, 243–276.
Liberman, Mark (1975) The intonational system of English. Ph.D. dissertation, MIT.
Poser, Williams J. (1984) The Phonetics and Phonology of Tone and Intonation in Japanese, Ph. D. dissertation, MIT.
Roberts, Craige (1996) "Information structure in discourse: Towards an
integrated formal theory of pragmatics." In: Jae Hak Yoon – Andreas Kathol (eds): OSU Working Papers in Linguistics 49: Papers in Semantics, 91–136. The Ohio State University, Columbus.
Sato, Yosuke. 2012. "Phonological interpretation by phase: Sentential stress, domain encapsulation, and edge sensitivity," in Ángel J. Gallego ed., Phases: Developing the framework, Mouton de Gruyter, Berlin, 283–307.
Selkirk, Elisabeth (2002) "Contrastive FOCUS vs. presentational focus: Prosodic evidence from right node raising in English." In B. Bel and I. Marlien eds. Speech Prosody 2002, 643–646.
Selkirk, Elisabeth (2008) "Contrastive focus, givenness and the unmarked status of "discourse-new." Acta Linguistica Hungarica 55, 331–346.
Stalnaker, Robert (1974) "Pragmatic presuppositions." In M. Munitz and P. Under eds. Semantics and philosophy, 197–213. New York University Press, New York.
Truckenbrodt, Hubert (1995) Phonological phrases: Their relation to syntax, focus, and prominence. Ph.D. dissertation, MIT.
van Kuppevelt, Jan (1994) "Topic and comment." In: Ronald E. Asher (ed.): The encyclopedia of language and linguistics, 4629–33. Pergamon Press, Oxford.
Table of Contents
1. Introduction — Nuclear Prominence: ... 1
2. Production Experiment: ... 2
3. Results: ... 5
4. Nuclear Prominence and Information Packaging: ... 13
5. Appendix ... 18
6. References: ... 19
