Abstract
This study investigates /d((sic))/-variations in Saudi Arabian newscasting under different in-word positions and different phonetic environments in 20 randomly selected newscasts. The study provides a thorough quantitative analysis of the frequency of occurrence of the two coexisting variants [d((sic))] and [o((sic))], and concludes that neither in-word position, nor the phonetic environment is able to predict the phonetic behavior of the sound. In a total of 734 /d((sic))/-lexical items, the sound was found to be properly realized as [d((sic))] in 519 instances (70.7%), and improperly realized as [o((sic))] in 215 tokens (29.3%). The very same percentages were found to quantify the sound variations under each linguistic condition, and the statistical analysis used reveals the irrelevance of each condition to predicting the sound variations. Also, the sound proves to be unpredictable in content and stressed words.