Abstract
This paper describes the speech timing of Modern Standard Arabic (MSA) using rhythmic features. Two main approaches are used to compute rhythm metrics: Interval Measures and Pairwise variability indices. The first approach consists of a comparison of MSA rhythm metrics computed from the ALGerian Arabic Speech Database with those resulting from using the West Point Arabic Speech corpus. The second approach compares these results with the rhythm measurements of six Arabic dialects. Many unexpected results are observed concerning rhythm variation within MSA and the dialects according to speakers' localities.