Turkish pronunciation skills of Arabic speakers in early and late language learning processes

In this study, it is aimed to identify the pronunciation errors made by basic level Arabic speakers learning Turkish as a foreign language during oral reading and speaking activities and to examine these errors on the axis of mother tongue influence, gender and age variables. The participants of the research consist of 91 students who learned Turkish in basic level courses at Azez and Afrin Yunus Emre Turkish Cultural Centers, selected by homogeneous sampling method. Since the study aimed to determine the effect of age and gender variables, as well as native language, on pronunciation skills, 58 adults (32 males-26 females) and 33 children (18 males-15 females) participated in the study. In order to detect pronunciation errors in segmental-suprasegmental units, the reading texts were also pronounced by native speakers (6 adults-6 children). The way for foreign language learners to communicate with target language speakers is through the ability to speak clearly, fluently and accurately. For this reason, in the Common European Framework of Reference for Languages (CEFR), the achievements for gaining pronunciation skills are presented under the title of qualitative features of spoken language and phonological control scale. Accordingly, the learner can only pronounce the sounds of the target language clearly at the C2 level. However, achievements that indicate that individuals can use the segmental and suprasegmental features of the target language from the basic level, even in a limited area, are also included. The way to understand whether the pronunciation gains in question have been achieved by the learner or not; is possible with valid and reliable research. For this reason, 12 reading texts and 4 independent speaking questions selected from the basic level Turkish teaching sets were presented to the learners, and the learners' oral reading and speaking performances were recorded. Praat (Boersma and Weenink, 2024) audio analysis program was used to analyze the data. The data obtained from the voice recordings of the learners and native speakers were graphed with the Python programming language and analyzed and interpreted using descriptive analysis and content analysis methods, which are qualitative research designs. As a result of the research, it was determined that the learners' pronunciation errors in segmented units supported the critical period hypothesis. It was found that children were more successful than adults in pronouncing syllable units formed with vowels and consonants of different qualities, but when examined in terms of gender variable, it was found that boys in adults and girls in children had less difficulty in producing Turkish phonemes. When analyzed in terms of native language disability, it was seen that the learners had difficulty in producing syllable units formed with /ü/, then /ö/, then /ı/ and finally /o/, among the sounds that are not in the phonetic inventory of their native language. In addition, the learners made negative transfers from their native language when emphasizing Turkish words and sentences. Recommendations are presented to instructors, program designers and researchers so that learners can achieve the goal of a socially understandable accent and overcome communication barriers.