Phonetic transfer in the production of Chinese and Arabic learners of Persian: A case study on Persian dorsal stops

This research aims at investigating the pronunciation of Persian dorsal stops in the the production of Chinese and Arabic learners of Persian. Dorsal stops are different in Persain from both Chinese and Arabic. Persian dorsal stops are pronounced as either palatal or velar depending on the front/backness of the following vowel, while dorsal stops in Chinese and Arabic are always produced as velar irrespective of the place of articulation of the following vowel. Givn this phonological difference between the Persian and the Chinese and Arabic sound systems, we asked to what extent the acoustic phonetic characteristics of Persian dorsal stops in the the production of Chinese and Arabic learners of Persian agree with those of native Persian speakers. A courpus was designed to test the research hypothesis. Three sets of spekers, namely native Persian spekers, Chinese learners of Persian and Arabic learners of Persian participated in the experiment. Data analysis included spectral examination at stop release as well as regression locus equation computed from messurement of F2 at vowel onset and the steady state of the vowel. Results suggested that while Persain speakers’ productions of dorsal stops fall within two different phonetic categories, namely palatal and velar, depending on the following vowel; Chinese and Arabic learners of Persian realize Persian dorsal stops as only velar irrespective of the vowel context. The findings are discussed with respect to the theory of Speech Learning Model (SLM), and suggestions are made as how the research findings may be applied in teaching Persian as a second language.
Extended Abstract:
The interference of native phonetics and phonology on the acquisition of non-native vowels and consonants has been studied extensively, and results typically suggest that discrimination of non-native sounds can be predicted from the perceptual relatedness of non-native categories to native categories. According to Speech Learning Model (SLM), proposed by Flege (1995), L1 and L2 sounds exist in a common phonological space, and thus influence each other. As argued by Flege et al. (2003), the interaction between the two systems involves two mechanisms, namely, category assimilation and category dissimilation. An L2 sound assimilates to an L1 sound when it is perceived as an instance of the L1 sound, despite audible differences between the two sounds. However, category dissimilation happens when learners can auditorily differentiate an L2 sound from the closest L1 sound and from the neighboring L2 sound. Flege makes the assumption that under such a condition a new phonetic category for an L2 sound can be established.
This research aims at investigating the pronunciation of Persian dorsal stops in the production of Chinese and Arabic learners of Persian. Dorsal stops are different in Persian from both Chinese and Arabic. Persian dorsal stops are pronounced as either palatal or velar depending on whether the following vowel is front or back, while dorsal stops in Chinese and Arabic are always produced as velar, irrespective of the place of articulation of the following vowel. Given this phonological difference between Persian on the one hand and Chinese and Arabic on the other hand, we asked to what extent the acoustic phonetic characteristics of Persian dorsal stops in the production of Chinese and Arabic learners of Persian agree with those of native Persian speakers.
A corpus was designed to test the research hypothesis. Three sets of speakers, namely native Persian speakers, Chinese learners of Persian and Arabic learners of Persian participated in the experiment. The participants were recorded individually in a quiet room using a digital audio recorder, Sound Blaster X- Fi 5.1, and a directional condenser microphone. The stimulus tokens were digitized at 22.05 kHz and low-pass filtered at 4.8 kHz. The output amplitude levels for each individual speaker were normalized to the maximum amplitude range.
All acoustic measurements were made using Praat acoustic software (Boersma and Weenink, 2010). Data analysis included spectral examination at stop release as well as regression locus equation computed from measurement of F2 at vowel onset and the steady state of the vowel.
Results suggested that patterns of spectral energy distribution in dorsal stops (the release section) are significantly different in the production of native Persian speakers from both Chinese and Arabic speakers, when the stop consonants are produced before front vowels, but they are rather similar when the target stops are produced before back vowels. More specifically, in back vocalic contexts, all speakers, irrespective of their mother tongue (differences in L1) pronounced the dorsal stops in such manner that the distribution of energy was rather compact, with energy being distributed in a polar fashion in three different spectral regions, namely, low formant frequencies, mid formant frequencies, and high formant frequencies. However, in the front vowel context, the native and non-native speakers of Persian differed as to whether the spectral shape of the target stops were diffuse or compact in their pronunciations. While, Chinese and Arabic speakers uttered the dorsal stops in the front vocalic context in almost the same fashion as the back context, Persian speakers pronounced the consonants such that there were less poles in the spectrum, and thus the spectral shape of the target stops were more diffuse than compact. Furthermore, the results indicated that the locus equation slopes serve as phonetic descriptors of stop place in CV utterances. Using the locus equation measurement sites for F2 onsets, dorsal stop consonants in the production of Chinese and Arabic speakers had lower locus equation slopes relative to native Persian speakers. When locus equations were derived using F2 onsets for dorsal stops that were measured closer to the stop release burst, the slopes were radically different between the native and non-native speakers of Persian.        
These results can be interpreted to suggest that while Persian speakers’ productions of dorsal stops fall within two different phonetic categories, namely palatal and velar, depending on the following vowel; Chinese and Arabic learners of Persian realize Persian dorsal stops as only velar irrespective of the vowel context. The findings are discussed with respect to the theory of Speech Learning Model (SLM), and suggestions are made as how the research findings may be applied in teaching Persian as a second language.