The discussion on stress in Arabic in the literature has primarily focused on the phonological constraints/rules behind the grammar of stress in Arabic. A rather limited number of studies delve into the topic of stress acquisition. The objective of this study is to investigate into the acquisition of stress and how it is biased by frequency. Following the usage-based approach to language acquisition, it is predicted that the more frequent a stress pattern is, the better the performance of the learners.

To examine this prediction, I conducted an experiment consisting of three tasks: stress identification, lexical decision, and stress production. The stimuli implemented in the experiment are real and nonsense words that have contrastive frequency of stress pattern. By exposing these stimuli to the learners of Arabic in the experiment, a biased performance preferring frequent stress pattern is predicted.

The results confirm this prediction, as the participants react to words with frequent pattern with significantly faster reaction time and accuracy in the production task. Additionally, it is found that the frequency of the stress pattern seems to interfere with word recognition as nonsense words with frequent stress pattern tend to be perceived as real words by the participants. These findings strongly support the role of frequency in language learning and processing.