Focus Marking in Focus Marking in Prosodic Marking of Focus Domains German German Categorial or Gradient Categorial and Categorial and gradient prosody gradient prosody (1) a. Q: Who did you call? (Baumann et al. (Baumann et al. b. A: [I called] background [MAry] F 2006) 2006) (2) a. Q: Did you call John? Results and Results and discussion Focus Marking in German b. A: No, [I called] background [MAry] F discussion Articulatory Articulatory (3) a. Q: What happened? gestures and focus gestures and focus b. A: [I called MAry] F marking marking The experiment The experiment ◮ The differences between answers in (1b), (2b), and (3b) Results and Results and are discrete: discussion discussion ◮ it is either MAry or I called Mary which is in focus and ◮ MAry is either contrasted with another specific person, Kordula De Kuthy or is singled out from a larger set. HS Neuere Arbeiten zur Fokusprojektion WS 09/10 ◮ Are these differences marked prosodically, and February 2, 2010 ◮ does the prosodic marking involve ◮ discrete means, i.e. phonological categories such as pitch accent type, or ◮ gradient means, such as duration, or F0 timing and scaling differences (which do not lead to a difference in phonological categories. 1 / 17 2 / 17 Focus Marking in Focus Marking in Prosodic marking of broad vs narrow focus in Production experiment German German Baumann et al. (2006) design a production experiment to German Categorial and Categorial and investigate whether gradient prosody gradient prosody (Baumann et al. (Baumann et al. ◮ prosodic means are used in German to differentiate 2006) 2006) ◮ F´ ery (1993) looked for categorial distinctions in the between three sizes of focus domains involving focus Results and Results and discussion discussion prosodic marking of broad versus narrow focus in projection and Articulatory Articulatory ◮ between these and narrow focus and gestures and focus gestures and focus German. marking marking ◮ between narrow focus and contrastive focus ◮ The result of an production experiment revealed that The experiment The experiment Questions: Results and Results and speakers used the same nuclear pitch accent type discussion discussion 1. Was gibt’s Neues? What’s new? (H*L) in both broad and narrow focus as in (4) and (5). 2. Was gibt’s Neues von Manuela? What about Manuela? 3. Was will Manuela? What does Manuela want? 4. Was will Manuela malen? What does Manuela want to (4) a. Q: Was ist los? paint? 5. Manuela will Gesichter malen? Manuela wants to paint faces? b. A: [ANna ist weggelaufen.] F Answers: Manuela will Blumen malen. (5) a. Q: Wer ist weggelaufen? 1. [ ] focus broad 2. [ ] focus b. A: [ANna] F [ist weggelaufen.] background [ ] focus 3. 4. [ ] focus narrow 5. Nein, [ ] focus contrastive lit.: Manuela wants flowers paint 3 / 17 4 / 17
Focus Marking in Focus Marking in Labeling of the resulting data Results and discussion German German Categorial means Categorial and Categorial and gradient prosody gradient prosody ◮ Contrary to predictions in the literature, both the size of (Baumann et al. (Baumann et al. L+ H* !H* 2006) 2006) the focus domain and type of focus affect the choice of broad focus Results and Results and accent type on the focus exponent: discussion discussion ◮ In broad focus structures a downstepped nuclear accent Articulatory Articulatory gestures and focus gestures and focus was produces in 42% of all cases. marking marking ◮ in narrower focus domains fewer downsteps occurred. The experiment The experiment ◮ in contrastively focused utterances no downstep was Results and Results and discussion produced at all. discussion L+ H* H* 100 narrow focus 80 nuclear 60 pitch accent 40 downstep type (%) 20 no downstep 0 Figure 1: Example F0 contours for broad and narrow focus 1 2 3 4 5 (answers 1 and 4, speaker CB) sentence type Figure 2: Differences in nuclear pitch accent type in relation to sentence type, all speakers (N=120) 5 / 17 6 / 17 Focus Marking in Focus Marking in Gradient means Increased duration German German Categorial and Categorial and gradient prosody gradient prosody (Baumann et al. (Baumann et al. ◮ Across all speakers, duration of the focus exponent 2006) 2006) varied consistently with the size of the focus domain. Results and Results and discussion discussion As the focus domains narrows in the examples, the use of ◮ But it did not distinguish between contrast and Articulatory Articulatory different gradient means was observed: gestures and focus gestures and focus non-contrast narrow focus. marking marking ◮ increased duration of the focus exponent The experiment The experiment 450 ◮ higher peak on the nuclear accent (marking the focus Results and Results and discussion duration of discussion 420 exponent) all speakers focus 390 exponent ◮ greater pitch excursion to the peak of the nuclear accent speaker NP (ms) 360 (female) ◮ delay in the nuclear accent peak 330 1 2 3 4 5 sentence type Figure 3: Differences in duration of focus exponent (mean values) in relation to sentence type, speaker NP 7 / 17 8 / 17
Focus Marking in Focus Marking in Higher accent peaks Greater pitch excursion German German ◮ Perceived prominence is very often not a correlate of Categorial and Categorial and gradient prosody gradient prosody pitch height , but of relative pitch excursion . (Baumann et al. (Baumann et al. ◮ The data from the production experiment show that for 2006) 2006) ◮ Two speaker show a highly significant effect of nuclear Results and Results and all speakers the nuclear rise excursion in sentence type discussion discussion accent pitch hight on sentence type. 1 (broad focus) is significantly smaller than the rise in Articulatory Articulatory gestures and focus gestures and focus sentence type 5 (contrastive focus). marking marking 300 ◮ The data for speaker CB indicate tendencies towards a The experiment The experiment nuclear 200 Results and gradual increase of the pitch excursion as the focus Results and discussion discussion accent peak domain narrows, plus a sharp increase from narrow to speaker NP 100 (Hz) (female) contrastive focus. 0 1 2 3 4 5 12 sentence type 10 pitch ex- Figure 4: Differences in pitch height of nuclear accent peak 8 cursion (mean values) in relation to sentence type, speaker NP nuclear 6 speaker CB rise (semi- 4 (male) tones) 2 0 1 2 3 4 5 sentence type Figure 5: Differences in pitch excursion of nuclear accent rise (mean values) in relation to sentence type, speaker CB 9 / 17 10 / 17 Focus Marking in Focus Marking in Late accent peaks Articulatory gestures and focus marking German German Categorial and Categorial and gradient prosody gradient prosody ◮ Another indicator of prominence consists in late accent (Baumann et al. (Baumann et al. 2006) 2006) peaks. Results and Results and discussion ◮ Two speakers showed such a trend: the smaller the discussion ◮ The study in Hermes et al. (2008) reports on a Articulatory Articulatory focus domain, the later the peak measured in ms from gestures and focus gestures and focus production experiment investigating tonal and marking marking the accented syllable onset. The experiment articulatory means of encoding different focus The experiment 300 Results and Results and structures in German. discussion discussion nuclear 250 peak ◮ The movements of the upper and lower lips during the 200 speaker NP position 150 (female) production of target words occurring in four different 100 (ms from speaker MG 50 word onset) focus conditions were examined. (female) 0 1 2 3 4 5 sentence type Figure 6: Differences in nuclear peak position (mean values) in relation to sentence type, speaker MG 11 / 17 12 / 17
Recommend
More recommend
