High Variability Phonetic Training: Arabic and Persian Speakers Learning Finnish Vowels /y/ and /ø/
Savo, Satu (2017-01-02)
High Variability Phonetic Training: Arabic and Persian Speakers Learning Finnish Vowels /y/ and /ø/
Savo, Satu
(02.01.2017)
Tätä artikkelia/julkaisua ei ole tallennettu UTUPubiin. Julkaisun tiedoissa voi kuitenkin olla linkki toisaalle tallennettuun artikkeliin / julkaisuun.
Kuvaus
Siirretty Doriasta
Tiivistelmä
This study examines second language acquisition and more precisely the production of difficult Finnish vowels /y/ and /ø/ by Arabic and Persian speaking asylum seekers. Thirteen Arabic speakers and five Persian speakers were recruited in reception centers in Turku area: the participants had stayed in Finland for less than six months and had not studied Finnish for more than three months. A listen-and-repeat production training protocol was used in order to study whether the participants’ production changes with the help of High Variability Phonetic Training. The target vowels were selected because they are not familiar to Arabic or Persian speakers.
The testing stimuli were recorded by two females and two males. Three out of ten recorded pseudowords were chosen as stimuli for each target vowel (/ty:ti/ and /tø:ti/). The selection was done based on the proportions of vowel duration and word duration – it must be ensured that the vowel duration is not too short or too long compared to the word duration. The pretest and posttest measurement were identical and included 24 repetitions each (12 x /ty:ti/, 12 x /tø:ti/). The training sequence was a loop consisting of four test sequences, and the experiment included two training blocks.
The results showed that, as hypothesized, training improved the production of both vowels in both subject groups. The Persian speakers improved their F₂ values of /y/ more than the Arabic speakers did, and both groups are considered to learn /y/ as a similar sound (Speech Learning Model), or possibly as a new sound. According to the classification of Perceptual Assimilation Model, it would either assimilate into a native sound within the lines of category goodness difference or it would not assimilate at all. For the Persian speakers/ø/ is either a similar or a new sound (SLM) – although the native vowel /e/ is very close to the target vowel, the combination of gestures (rounded lips, tongue in the front) is new to them. For Arabic speakers /ø/ is clearly a new, non-assimilable sound because so many participants tried to pronounce the target vowel via /o/, which does not exist in their native vowel system. HVPT appears to enhance the production of difficult nonnative vowels, which should be taken into consideration in S2 teaching.
The testing stimuli were recorded by two females and two males. Three out of ten recorded pseudowords were chosen as stimuli for each target vowel (/ty:ti/ and /tø:ti/). The selection was done based on the proportions of vowel duration and word duration – it must be ensured that the vowel duration is not too short or too long compared to the word duration. The pretest and posttest measurement were identical and included 24 repetitions each (12 x /ty:ti/, 12 x /tø:ti/). The training sequence was a loop consisting of four test sequences, and the experiment included two training blocks.
The results showed that, as hypothesized, training improved the production of both vowels in both subject groups. The Persian speakers improved their F₂ values of /y/ more than the Arabic speakers did, and both groups are considered to learn /y/ as a similar sound (Speech Learning Model), or possibly as a new sound. According to the classification of Perceptual Assimilation Model, it would either assimilate into a native sound within the lines of category goodness difference or it would not assimilate at all. For the Persian speakers/ø/ is either a similar or a new sound (SLM) – although the native vowel /e/ is very close to the target vowel, the combination of gestures (rounded lips, tongue in the front) is new to them. For Arabic speakers /ø/ is clearly a new, non-assimilable sound because so many participants tried to pronounce the target vowel via /o/, which does not exist in their native vowel system. HVPT appears to enhance the production of difficult nonnative vowels, which should be taken into consideration in S2 teaching.