Sonority and Cluster Reduction in Typical and Atypical Phonological Development in Farsi
Forugh Shooshtaryzadeh [+]
Imam Khomeini International University
The Sonority Sequencing Principle (SSP) (Clement, 1990; Whitney, 1865) proposes that phonemes with high sonority values are placed towards the centre of the syllable, while phonemes with low sonority values are located at the syllable margins. In accordance with SSP, the Sonority Hypothesis (SH) (Ohala, 1999) hypothesized that children reduce word-initial consonant clusters in a way that creates a maximal rise in sonority, and word-final consonant clusters are lessened to produce a minimal fall in sonority. The sonority values of special sounds and the related approaches were employed for the analysis of the articulation of consonant clusters by children with typical speech development and the children with speech impairment (e.g., Chin, 1996; Gierut, 1998; Ohala, 1999; Pater and Barlow, 2002; Vasanta, 2006; Gerlach, 2010). The majority of these studies, which are generally on the acquisition of onset clusters, assume SSP as the main motivating factor for cluster reduction because, in their studies, onset clusters are reduced to the less sonorous segment as predicted by SSP. Keeping in mind the results from previous studies, the present study examines and compares the Sonority patterns in the production of some typically developing (TD) children and the children with functional (non-organic) phonological disorder (PD) acquiring Farsi as their first language. It aims to investigate if SSP and SH applies to cluster reduction and a few other processes happened in typical and atypical children acquiring Farsi. Data for the present study were collected from five typically developing children and five phonologically disordered children. Data collection was carried out through a test called Naming-Picture Task and a 15-30 minutes free recording. There are just samples of word-medial (abutting) and word-final (coda) clusters in the data, as there is no word-initial cluster in Farsi. The collected data was considered and the results were analysed in Optimality Theory (OT). The results of this study led us to new insights on the role of sonority and the related hypothesis in cluster reduction. Unlike the previous studies, the production of coda clusters in present study says the opposite to SSP and SH. Because, in spite of the prediction of SSP and SH, the coda clusters produced by both PD and TD children acquiring Farsi do not keep the most sonorous segment. Conversely, they act as onset clusters in the aforesaid studies and reduce to the less sonorous segment in the clusters that, in general, are articulatory less marked segments too. However, though in this study no significant traces of the effects of sonority hierarchy have been observed in cluster reduction, the processes such as substitution, metathesis and voicing/devoicing modifications indicate some effects of sonority scale on these processes. The present study concludes that, both, in PD and TD children there are other factors motivating cluster reduction in coda clusters and probably in onset clusters, such as Articulatory Ease and Continuity Preservation, than sonority hierarchy. However, SSP shows some role in minimal cluster repairs.