Abstract
Conference Title: 2014 International Conference on Computer, Communications, and Control Technology (I4CT) Conference Start Date: 2014, Sept. 2 Conference End Date: 2014, Sept. 4 Conference Location: Langkawi, Malaysia Arabic is one of the most widely spoken languages in the world, but little attention has been paid to detect pronunciation errors of non-Arabs from different nationalities. In this paper, the speech of four nationalities of Asian non-Arabs speaking different mother languages is analyzed to identify their pronunciation errors while learning Arabic as a foreign language. Two human experts have evaluated the speech of all speakers for mispronunciation, and evaluation show that the nature of errors is almost the same for all nationalities under investigation with the fact that pharyngeal, alveodental, and interdental sounds are difficult to pronounce by learners of Arabic. Some of the errors are due to sounds that are not present in mother languages or the learner is unable to pronounce the phoneme correctly due to similar place of production or/and manner of articulation. An interesting observation is that pronunciation errors of some learners are due to switching pairs of phonemes.