Formant frequencies are among the most commonly used and reliable parameters for forensic phonetics (FP) analysis (Lindh, 2006; Loakes, 2006), and despite an ongoing debate as to which formants are the most robust, they continue to be preferred to other parameters such as segment duration. However, most previous work on the use of formants for FP examines data from languages with a relatively crowded vowel space, such as English, which exhibits vowel closeness more than languages with a tri-vowel system, such as Modern Standard Arabic (MSA). The current experimental study investigates the carry-over of formant frequencies from two distinct Arabic dialects, Najdi Arabic (NA) and Hejazi Arabic (HA), in the production of MSA by the speakers of the two respective dialects. This provides insight into the potential for utilizing the formant frequencies and formant spacing of long vowels as parameters in Arabic FP analysis. Forty single-gender (male) participants were divided into two numerically equal groups based on their dialectal background, NA vs. HA. The participants were asked to record a list of tokens that were carefully selected and judged to occur in both HA-NA dialectal Arabic and standard Arabic, albeit with barely recognizable phonological divergences. The data were gathered in two separate sessions (dialectal production and MSA production) for each group, with a lag period of three weeks between the two sessions. The findings show that the F1 and F2 of long vowels have promising robustness in order to successfully differentiate between NA and HA speakers producing their shared standard variety (i.e., MSA). Formant spacing in particular can thus be utilized as a potential parameter for Arabic FP profiling. As anticipated, formants for all three vowels, particularly the formants of /aː/ and /uː/ in NA and the formants of /aː/, /uː/, and /iː/ in HA, appeared to be carried over from the dialects to MSA production. This indicates that, despite the limited number of participants and dialects covered in this explorative study, there seems to be potential for utilizing formant frequencies as an FP parameter for MSA speakers’ dialect profiling.
