The prosody of Focus and Givenness in Hindi and Indian English

This study investigates the production and perception of prosodic cues to realize narrow and corrective focus in Urdu/Hindi. We recorded SOV sentences with the target constituents at the preverbal position. Our results show that correctively focused nouns have longer syllable duration, wider F0 range, early alignment of F0 peaks, variant production of downtrend, and less steep post-focal compression as compared to narrowly focused nouns. We further set up a perception experiment to investigate if the difference in syllable duration of narrowly and correctively focused constituents is perceptible to Urdu/Hindi speakers. We manipulated syllable duration of the target constituents and presented them in contexts via a webbased interface. Twenty-nine respondents rated the naturalness of manipulated sentences in the given contexts. The analysis of respondents' ratings indicated that while they accepted both long and short durations in narrow focus, they rated long duration significantly better in corrective focus. Our results lend support to earlier claims about the prosodic cues of corrective focus in Urdu/Hindi [1, 2] and bring new evidence regarding the perceptual relevance of duration to cue corrective focus.

Section: F0 (Contour and Alignment)supporting

confidence: 88%

“…Thus, correctively focused constituents were the longest, whereas selective and new information focus had the same relative constituent duration. While these results offer an interesting interplay of prosodic cues, a recent research by [11] has further added to the mass of contradictory findings.…”

Section: Focus and Prosody In Urdu/hindimentioning

confidence: 96%

Production and Perception of Prosodic Cues in Narrow & Corrective Focus in Urdu/Hindi

Jabeen¹,

Braun²

2018

“…A speaker with a rhythmic cadence was found to rely entirely on duration variation suggesting that specific prosody deficits may be identifiable through specific signal characteristics. Given that phrase boundary cues have been found to replace pitch accent cues in focused words in Indian English dialects [31], the acoustic correlates of phrase boundaries and the interaction of phrasing and prominence are also topics for future work. Finally, it is the prediction of goodness of reading rendition that is important.…”

Section: Discussionmentioning

confidence: 99%

“…Compensating for this ambiguity improved the average inter-rater agreement from kappa of 0.36 to a substantially better 0.64. This could also possibly indicate the frequent use of "prosodic words" by the speakers where a function word is prosodically attached to an adjacent content word [31].…”

Section: Dataset and Annotationmentioning

confidence: 99%

Detection of Prominent Words in Oral Reading by Children

Sabu¹,

Rao²

2018

The evaluation of oral reading skills is considered an important component of language education in school. Compared with word decoding skill, prosodic fluency typically takes children much longer to achieve. Prosodic fluency, however, is linked to comprehension making its evaluation very useful in an automatic reading assessment system. We consider the detection of prominent words in recordings of oral reading by children, who display good word recognition but varying degrees of prosodic fluency. The manual annotation of prominent words proves to be relatively challenging, likely due to inconsistencies by our speakers with respect to top-down lexical cues. Acousticprosodic features drawn from prominence detection research on adult speech are tested on the annotated data using a random forest classifier. Normalized maximum syllable duration and F0 (fundamental frequency) derived features turn out to be important predictors of word prominence with their relative importances being highly speaker dependent.

“…[11] refrained from assigning a tonal category and referred to the rise produced by Hindi English speakers as an LH contour. [26] proposed the same analysis for Hindi and IndE spoken by L1 Hindi speakers, and treated both the L and the H target as phrase tones that demarcate the edges of a minor prosodic unit. It is important to note that despite the presence of a "repetitive rising contour" in South Asian languages, tonal alignment of the rise is said to be a major source of cross-linguistic variation due to prominence patterns, segmental and lexical features across the languages spoken in the subcontinent [2].…”

Section: Introductionmentioning

confidence: 99%

Pitch accent types and tonal alignment of the accentual rise in Indian English(es)

Maxwell¹,

Payne²

2018

The paper presents an analysis of pitch accent inventory and tonal alignment of the accentual rise in the speech of university educated Indian English speakers from four L1 backgrounds (Hindi, Bengali, Tamil and Telugu). The results reveal that all speakers produced a high and a rising pitch accent, but with differences in the distribution of H* vs. L*+H among speakers. In addition, L1 speakers of Hindi and Bengali used L*+H in both nuclear and prenuclear contexts, contrary to the speech of L1 Telugu and Tamil speakers, who frequently produced the H* accent on nuclear accented words. An examination of peak alignment provides further evidence that the accentual rise corresponds to a bitonal pitch accent; the alignment of the H tone however shows a large degree of intra-and inter-speaker variation, and earlier timed peaks relative to the accented syllable offset for L1 Tamil speakers.