1 / 39

When Focal Cues are Conflicting!: Focus Perception in Korean

When Focal Cues are Conflicting!: Focus Perception in Korean. Young Ah Do MIT youngah@mit.edu. Focus. Focus- ‘the information in the sentence that is assumed by the speaker not to be shared by him and the hearer’ (Jackendoff 1972) Where we can find focus? Answer to Question

Download Presentation

When Focal Cues are Conflicting!: Focus Perception in Korean

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript


  1. When Focal Cues are Conflicting!: Focus Perception in Korean Young Ah Do MIT youngah@mit.edu

  2. Focus • Focus- ‘the information in the sentence that is assumed by the speaker not to be shared by him and the hearer’ (Jackendoff 1972) • Where we can find focus? • Answer to Question • Who loves John? [Mary]F loves John. • Contrast or Correction • Sally loves John. No, [Mary]F loves John.

  3. Focus marking in Seoul Korean • Prosodic marking (Jun 2002, Jun & Lee 1998) • Mainly by post-focus dephrasing • An example • No focus: [[John-nom]AP [Mary-acc] AP [likes-decl] AP] IP • Focus on ‘John’: [John-nomMary-acc likes-decl] IP

  4. Focus marking in Seoul Korean • Structural marking? • Focus element at sentence-initial position • Can be found: • Answer to Question • Whom does Mary love? Mary loves [John]F. • Whom does Mary love? [John]F Mary loves. • Contrast or Correction • Sally loves John. No, Sally loves [Jack]F. • Sally loves John. No, [Jack]F Sally loves.

  5. Focus marking & Focus perception • Focus ‘the information in the sentence that is assumed by the speaker (Production)not to be shared by him and the hearer (Perception)’ (Jackendoff 1972) • The current study • Focus perception

  6. Key Questions Focus Perception Does the hearer notice the speaker’s assumption? What are the perceptual cues to focus? Can we predict uncertainty by relative cue weights?

  7. Overview Perception test Focus perception in focus-intended productions Perceptual cues to focus Structural cue: sentence-initial position Prosodic cue: expanded pitch range w (Structural cue) > w (Prosodic cue) Object bias

  8. PERCEPTION TEST Designing Production Frames

  9. Contrastive focus • What • Mary ate [apples]F, not bananas. • [Mary]F ate apples, not John. • Why • The 2nd phrase implies what is focused in previous phrase. • Good stimuli to check what is perceived as focus • How • Presenting listeners with spoken sentence frames • Asking them to construct contrastive focus • e.g. John ate apples (spoken frames), not _____. • “not Mary” : subject was perceived as focus • “not bananas” : object was perceived as focus

  10. Spoken sentence frames • Our assumption • Prosodic cue in Production • Structural cue in Production • 6 Types of frames varying contrast and word order SOV OSV • No contrast: Subj Obj V., Obj Subj V. • Subj contrast: Subj Obj V, not Subj., Obj Subj V, not Subj. • Obj contrast: Subj Obj V, not Obj., Obj Subj V, not Obj.

  11. Spoken sentence frames • 6 Types of frames varying contrast and word order • Contrast • Sentence-initial element • How does the speaker mark contrastive focus prosodically?

  12. Focus marking • Expanded pitch range on focused word • [Subj]expanded pitch Obj V, not Subj. • Obj [Subj] expanded pitch V, not Subj. • Word order • No big difference in general • Remark • Subj [Obj] expanded pitch V, not Objis different from [Obj] expanded pitch Subj V, not Obj • Pause in Subj # [Obj] V, not Obj

  13. Our interest: Focus perception • Cutting spoken sentence frames • Subj Obj V, not Subj. • Presentational frames • Subj Obj V (spoken), not _____ (written). • Assumption • ‘not Subj’: Subj is perceived as focus • ‘not Obj’: Obj is perceived as focus

  14. Stimuli • 7 Sentences for 6 types of frames each (42 items in total) • Recorded by a male and a female native speaker • Frames were given in random order to 24 participants

  15. PERCEPTION TEST Does the hearer notice the speaker’s intention? What are the perceptual cues to focus?

  16. Prosodic cue • Speaker • Expanded pitch range on focused word • Listener • Perceived focused word well in general • Contrastive focus realized as expected

  17. Structural cue • When is non-focus misperceived as focus? • Non-focus is at sentence-initial position • When the listeners find another conflicting cue • e.g. Frame: Subj [Obj]F V • Expanded pitch range :Obj -Prosodic Cue • Sentence-initial :Subj - Structural Cue

  18. Structural cue • Structural cue • Sentence-initial position as structural cue for focus

  19. Object bias • When is non-focus misperceived as focus? • When Subject is focused • No conflict of two cues • Frame: [Subj]F Obj V • Expanded pitch range –Subj • Sentence-initial – Subj • Listeners many times choose Object in contrastive focus phrase. • [Subj]F Obj V, not Obj

  20. Object bias • More on object bias • Neutral sentence without focus : Subj Obj V. • No prosodic cue • Sentence initial: Subj • Object bias to focus perception

  21. Summary • Assumption: the speaker marks focus by prosodic and structural cues. • The hearer perceives focus as the speaker intended. • Two crucial perceptual cues to focus • Prosodic: expanded pitch range • Structural: sentence-initial position • Object bias

  22. PERCEPTION TEST How do the cues interact? Can we predict uncertainty by relative cue weight?

  23. Weight of each cue • How important is each cue? • Logistic regression analysis • To expect object in contrastive phrase • Coefficients: (Intercept) Obj Initial Subj ex-pitch Obj ex-pitch 0.4617 1.2309 -0.7555 0.7729

  24. Weight of each cue • w( Structural cue) > w (Prosodic cue) > w (Object bias) • The listener uses each cue and bias with different weights in perception. • Question • Cue interaction • Can we predict perceptual uncertainty by relative cue weights?

  25. No conflict situation • Cues and Obj bias converge • [Obj]F Subj V, • ‘Expanded pitch range’ [prosodic] on ‘Sentence-initial [structural] ‘Object’ [bias] • Results: No uncertainty at all

  26. Conflict situation • Conflict of Cues and Obj bias • Obj [Subj]F V, • ‘Expanded pitch range’ [prosodic] on ‘Non-initial [structural] ‘Subject’ [bias] • [Subj]F Obj V, • ‘Expanded pitch range’ [prosodic] on ‘Initial [structural] ‘Subject’ [bias] • Cue conflict! – Uncertainty observed

  27. Perceptual uncertainty • Observed from: • Cue conflict cases • [Subj]F Obj V: prosodic & structure vs. Bias • Subj [Obj]F V : prosodic & bias vs. Structure • Obj [Subj]F V : structure & bias vs. prosodic

  28. Combining multiple cues • Combining multiple cues • (Prosodic + Structure) vs. (Bias) • (Prosodic + Bias) vs. (Structure) • (Structure + Bias) vs. (Prosodic) • How likely to get the Object perception • Model fitness • Null deviance – Residual deviance

  29. Predicting Uncertainty • Model fitness • Best: (Prosodic + Structure) vs. (Bias) : 86.7 • 2nd: (Structure + Bias) vs. (Prosodic) : 53.3 • Worst: (Prosodic + Bias) vs. (Structure): 34.5 • Weights of cues independently • w( Structural cue) > w (Prosodic cue) > w (Object bias) • High uncertainty when combination of less important cues are competing with strongest cue

  30. Predicting Uncertainty • Listeners • Finding cues • Cue combining & bias • Deciding one element as focus

  31. When cues are conflicting! • More on worst case • Worst: (Prosodic + Bias) vs. (Structure) • Subj [Obj]F V • Two less important cues vs. the most crucial cue • Great uncertainty

  32. When cues are conflicting! • Mix of perceiving Subj focus and Obj focus

  33. Perception of double focus? • Conceivably, both Subj and Obj are perceived as double focus. • Randomly choose one since they are allowed to make contrastive focus with only one word. • Double focus marking • Answer to Question • (Who ate What ? ) [Mary]F ate [apples] F • Contrastive Focus • [John] F ate [apples] F not Mary ate bananas.

  34. Perception of double focus? • Randomly choose one between Subj and Obj since they are allowed to make contrastive focus with only one word. • Check double focus production

  35. Double focus • Double focus • Answer to Question • (Who ate What ? ) [Mary]F ate [apples] F • Contrastive Focus • [John] F ate [apples] F not Mary ate bananas. • Double focus marking • Pause after initial Subj • More distinctive expanded pitch range on Obj

  36. Double focus • Pause as perceptual cue for double focus • Listeners consider the given frames as double focus sentences • Pause for double focus marking? • Why especially in Subj [Obj]F V, not Obj case? • Any idea?

  37. Conclusions • Focus perception • The hearer notices the speaker’s assumption • Use Production cues to Perception • Perceptual cues to focus • Structural, Prosodic cues and Object bias • Uncertainty and cue interaction • Perceptual uncertainty due to cue confliction

  38. Future Work • Structural cue in production • Scrambling due to focus marking? • Asymmetry of production & perception cues • Production : Prosody > Structure • Perception: Prosody < Structure • Double focus marking • Why does pause matter? • How does pause function as focal cue?

  39. Any Questions/Comments? • youngah@mit.edu • Many thanks to Adam Albright, Michael Kenstowicz, Donca Steriade, Edward Flemming & audience at MIT Phonology Circle

More Related