The Secret Life of Suprasegmentals

The Secret Life of Suprasegmentals

First Author: JoAnne Arciuli -- University of Sydney
Abstract / Summary: 

Reading and spelling are processes that involve more than decoding and encoding of segmental information (i.e., phonemes). Suprasegmental information also plays an important role in reading and spelling. This symposium will showcase the methods currently being used to advance our understanding of the link between orthography and suprasegmental information. The presentations cover a range of topics including special populations such as children with autism spectrum disorder, specific language impairment, and dyslexia, as well as non-native speakers of English. The four presentations result from research conducted in different labs across the USA, Australia, and Taiwan and our Discussant is from the UK. Presenters are a mix of early career and senior researchers. There is emphasis on new theoretical contributions as well as new data in the presentations include in this symposium – something which will be emphasised by our discussant (Professor Clare Wood, Coventry University, UK).

Symposium Papers: 

Prosody and Nonword Repetition in Children with Language and Reading Disorders

First Author/Chair:Alison Hendricks -- University of South Carolina
Additional authors/chairs: 
Dan Fogerty; Suzanne Adlof; Lesly Wade-Woolley

Purpose: Children with Specific language impairment (SLI) struggle with non-word repetition tasks compared to typically developing peers. It is not well known how the comorbidity of SLI and dyslexia affects performance on non-word repetition tasks. Performance on non-word repetition tasks is typically measured as correct reproduction of phonemes, either with binary accuracy scores or percent phonemes correct. Less is known about the prosodic performance on non-word repetition tasks.

Method: Eighty second grade students (Mean age = 8;0) were randomly selected from a larger, population-based study on word learning. Participants fell into one of four sub-groups: SLI-only (SLI), Dyslexia (DYS), SLI and dyslexia (SLI-DYS), and typically developing children (TD). All participants completed the first 25 items of the non-word repetition sub-test of the Comprehensive Test of Phonological Processing- second edition.

Results: Analysis of the number of items correct reveals that the SLI-DYS children perform significantly lower than any of the other sub-groups. However the SLI and DYS groups do not differ from each other or from the TD children. In order to examine prosodic differences, acoustic analyses of accurate items were conducted. All files were run through a noise reduction algorithm using Audacity prior to analysis. Comparisons were made between children’s production and the CD master file that the children heard. Preliminary analyses include differences in duration, phonation time, and amplitude peaks.

Conclusion: Analysis is ongoing; we anticipate that children with both SLI and dyslexia will reproduce prosodic contours that are different on these parameters than children with single or no deficits.

The relationship between prosodic awareness and spelling accuracy in children with Autism Spectrum Disorder

First Author/Chair:Benjamin Bailey -- Centre for Disability Research and Policy, The University of Sydney
Additional authors/chairs: 
Joanne Arciuli

Purpose – Children with Autism Spectrum Disorder (ASD) often present with difficulties in prosodic awareness and literacy. However, the relationship between these abilities is poorly understood. Nash and Arciuli (2014) identified links between prosodic awareness and reading accuracy in early readers with ASD. The current study sought to extend on these findings by investigating the relationship between prosodic awareness and spelling accuracy in children with ASD.

Method –Twenty-five children with ASD (2 females) aged 5-12 years were given a standardised test of word spelling accuracy (WRAT-4; Wilkinson & Robertson, 2006) and two measures of prosodic awareness: the Mispronunciation task (Holliman, Wood, & Sheehy, 2010), a measure of lexical stress awareness; and the Compound Noun task (Whalley & Hansen, 2006), a measure of metrical stress awareness. Participants were also given standardised tests of vocabulary (PPVT-4; Dunn & Dunn, 2007) and phonological awareness (CTOPP-2; Wagner, Torgesen, Rashotte, & Pearson, 2013).

Results – A multiple regression analysis showed that age, vocabulary, phonological awareness, and lexical stress awareness were significant predictors of word spelling accuracy. Collectively, these factors accounted for 78% of variance in word spelling ability.

Conclusion – These results show that prosody – specifically, awareness of lexical stress – contributes to word spelling accuracy for children with ASD. More broadly, the findings suggest similarities in the spelling profiles of children with ASD and neurotypical children. Pedagogical implications will be discussed.

Amplitude envelope onset, native prosodic and phonological awareness, and nonnative word learning.

First Author/Chair:Wei-Lun Chung -- Department of Applied Chinese Language and Culture, National Taiwan Normal University
Additional authors/chairs: 
Linda Jarmulowicz; Gavin M. Bidelman

The study explored the contributions of prosodic awareness (speech and non-speech stimuli) to English word learning over time.

Sixty-one Mandarin-speaking children were given following tasks: (1)rise time discrimination (rate of intensity change at tone onset), (2)Mandarin tone perception, (3)Mandarin phonological awareness (syllable onset), and (4)English word learning, measured by the number of trials needed to recall and produce three spoken words (i.e., termite, crocus, and macaw) associated with pictures, at Time 1 and Time 2 (one week later). Children’s parents filled out a questionnaire regarding their years of English learning.

After controlling for years of English experience, the ability to identify phoneme sequences in the first language (Mandarin PA) was important to English word learning, regardless of time tested. At Time 1, when English word learning was most novel, the ability to hear and differentiate syllable-like beats in non-linguistic stimuli (rise time discrimination) was a predictor. By Time 2, a different prosodic measure, Mandarin tone perception, emerged as important for English word learning.

For Mandarin children, phonological awareness is consistently important for English word learning. Additionally, the ability to distinguish syllable boundaries is important for initial word learning in English (in which syllables carry relative stress information); however, upon reintroduction of the same words one week later, an association with the ability to perceive L1 prosodic patterns emerges. Overall, results suggest that to learn L2 English words, Mandarin-speaking children rely on phonological awareness, but recruit different types of prosodic awareness depending on word familiarity.

Does value-driven attentional capture explain stress effects in reading?

First Author/Chair:Lindsay N. Harris -- Northern Illinois University
Additional authors/chairs: 
Charles A. Perfetti

Purpose: This talk is primarily theoretical in nature. No new analyses have been conducted, but we have developed a novel interpretation of published findings. Data collected in two previous studies suggest that lexical stress information activated during silent reading influences a reader’s perception of the string itself. However, the mechanism by which phonological information might influence a reader’s perception of visual information is unclear. The purpose of this talk is to introduce one possible mechanism of this influence.

Methods: In Experiment 1, participants (N=145) completed a spelling decision task in which 40 items were systematically misspelled in stressed and unstressed syllables. In Experiment 2, the same items appeared in a narrative passage and a second set of participants (N=209) were instructed to proofread for spelling errors.

Results: Repeated-measures ANOVAs were used to examine stress effects. In both experiments, items were significantly more likely to be identified as misspelled when misspellings occurred in stressed than in unstressed syllables.

Conclusion: We propose value-driven attentional capture (Anderson, Laurent, & Yantis, 2011) as the mechanism of the stress effects observed by ourselves and others. Because lexical stress is a high-value feature of speech during spoken language perception, and because what is valuable in spoken language tends to be valuable in written language, we argue that stress fits Anderson et al.’s definition of a “nonsalient, task-irrelevant stimul[us] previously associated with reward” that captures attention during visual activity—in this case, reading.


First Author/Chair:Discussant Clare Wood