Phonological Transfer in Chinese Persian Language Learners: Vocalic Space in Focus

The so called “foreign accent” is one manifestation of phonological transfer. The current study aimed to study: first, “if a part of Chinese PFL learners’ foreign accent is caused by errors in pronouncing Persian vowels”, and second “if Chinese PFL learners transfer the acoustic features of the vowels in their mother tongue to their Persian speech”. Through a process of convenience sampling, 42 Persian and Mandarin Chinese native speakers were included in the present study. F1 and F2 of The Persian vowels, the Mandarin vowels, and the Persian vowels produced by the native Mandarin PFL speakers were extracted and compared. Data analysis showed that, firstly, F2 of the vowel /e/ in the Persian speech of the native Mandarin speakers is meaningfully more fronted than it is in Persian. Hence, this error in vowel pronunciation might be a cause of the apparent foreign accent in the Persian speech of native Mandarin PFL speakers. Secondly, a comparison of Persian and Mandarin Chinese vocalic spaces showed that in the same position where the Persian /e/ is located, Mandarin Chinese has four similar front vowels (1 phoneme and 3 allophones); it seems as if the Chinese PFL learners exceedingly front the Persian vowel /e/ to contrast it with these similar vowels in their mother tongue. These findings corroborate the “modified transfer hypothesis” since the source of error in this particular case is not direct phonological transfer but the mother tongue seems to indirectly affect the pronunciation of the foreign language. 
Extended Abstract: 
Statement of the Problem: The so called “foreign accent” is one manifestation of phonological transfer. The problems investigated in the present study are: first, “if a part of Chinese Persian as a Foreign Language (PFL) learners’ foreign accent is caused by errors in pronouncing Persian vowels”, and second “if Chinese PFL learners transfer the acoustic features of the vowels in their mother tongue to their Persian speech”. 
Theoretical Framework: Knight(2013, p. 16) classifies the status of the speech sounds of a foreign language compared with the corresponding speech sounds in the mother tongue in the three classes of “new” (as completely different), “identical” (as completely the same), and “similar” (as similar sound with slight acoustic differences). The identical speech sounds in the foreign language are mastered with ease by being put in the same class as the phonemes in the mother tongue. The pronunciation of the new speech sounds improve as new phoneme categories are made in mind of the second language (L2) learners through sufficient practice (Fromkin, Rodman, & Hyams, 2014, p. 432). Among these three classes, the similar speech sounds are the most difficult to master nativelikely by the L2 learners and the slight acoustic differences they have might be taken as a major source of foreign accent in the target language (Best & Strange, 1992, p. 105). 
Methodology: Through a process of convenience sampling, 42 Persian and Mandarin Chinese native speakers (11 female and 13 male native Persian speakers, and 8 female and 10 male native Mandarin Chinese PFL learners) were included in the present study. F1 and F2 of The Persian vowels, the Mandarin vowels, and the Persian vowels produced by the native Mandarin PFL speakers were extracted in a laboratory situation and the data was statistically compared. For calculating the mean F1 and F2, all the 24 native Persian speakers and 18 native Mandarin Chinese PFL learners were included; in the second phase 8 native Persian speakers and 2 native Mandarin Chinese PFL learners were randomly left out to conduct the statistical analyses (independent samples t-test for normally distributed data and Mann–Whitney for abnormally distributed data) on quantitatively similar data samples from both genders. 
Results and Discussion: in accord with what Knight(2013, p. 16) claims about the so called “identical” speech sound, among the 5 articulatory similar vowels in Persian and Mandarin Chinese, 4 of them turned to be acoustically “identical” without causing any meaningful pronunciation errors. Only the F2 of the Persian vowel /e/ was observed to be meaningfully different from the corresponding vowel in Mandarin Chinese. However, in contrast with the findings of Best & Strange (1992, p.105), the native Mandarin Chinese PFL learners did not put this “similar” speech sound in a corresponding category of a similar vowel from their mother tongue; instead of transferring the acoustic features of a similar vowel in their mother tongue to the pronunciation of the Persian /e/, the data revealed that the Mandarin Chinese PFL learners treated the Persian /e/ as a “new” speech sound and pronounce it differently from all the similar corresponding vowels in their mother tongue. The vowel /e/ in the Persian speech of the Mandarin Chinese PFL learners, however, was also different from the native Persian /e/; a pronunciation error which might serve as a source of their foreign Persian accent. 
A comparison of the vocalic spaces of Persian and Mandarin Chinese shows that in the area where the Persian /e/ is expected to be located, Mandarin Chinese already has 4 closely similar front vowels (phonemes /e/, /y/, and /i/ and the allophone [E]). It seems as if the Mandarin Chinese PFL learners recognize that the Persian /e/ is fronter than the three /e/-like vowels (namely /e/, /y/, and [E]) and lower than the similar vowel /i/ in their mother tongue. However, it seems that in order to both compensate for this difference in frontness and make the new vowel also different from the corresponding similar vowels in their mother tongue, they exaggerate in the frontness of their Persian /e/ and that could be the reason behind this pronunciation error causing their foreign Persian accent. 
Conclusions: the findings of the current study show that the error in pronouncing the Persian /e/ in the speech of Mandarin Chinese PFL learners cannot be taken as a manifestation of direct negative language transfer; this is in contrast with the suggestions of the strong version of language transfer in general and Chen & Wang (2011), Wang & Heuven (2006), and Knight (2013) in particular. The Mandarin Chinese PFL learners have not activated the Persian /e/ as a part of the so called innate linguistic competence, but in their Persian speech, they pronounce an /e/ which is different from both native Persian and Mandarin Chinese corresponding vowels; then, these findings do not also confirm the weak version of language transfer. However, the results of the present study seem to corroborate the modified version of the language transfer, since the observed error in the pronunciation of /e/ in the Persian speech of these PFL learners could be taken as a kind of indirect negative interference of the similar linguistic patterns of the mother tongue in L2.