The Phonetics and Phonology of Ḥarsūsi: An Instrumental Phonetics Study

Ḥarsūsi is one of six Modern South Arabian Languages (henceforth MSAL), spoken in Jiddat Al-Harasis /d͡ʒiddɐt el ħɐrɐːsiːs/ in Central Oman. The MSAL branch of languages came to the attention of European scholars around 1830s, with Ḥarsūsi first mentioned in scholarly works by Thomas Bertram in 1929. Other work was conducted on Ḥarsūsi by Thomas Bertram in the 1930s and by Thomas Johnstone in the 1970s. To date, however, scholarly works on Ḥarsūsi remain scarce. Ḥarsūsi is labelled as a definitely endangered language by UNESCO (Moseley, 2010).
This thesis is the first analysis of the Ḥarsūsi sound system based on first-hand field data since Johnstone’s work conducted over 43 years ago, and constitutes the first step towards a full linguistic documentation of Ḥarsūsi.
Qualitative and quantitative methods are applied to the investigation of the phonetics and phonology of Ḥarsūsi. A database of audio, video and written text materials of transcribed and analysed word lists, narratives and conversations was produced for linguistic analysis. Acoustic analyses were done using the linguistic analysis programme PRAAT (Boersma & Weenink, 2020). Linear-mixed model tests were run using the statistical programme R to provide statistically founded results. Chapter One: Introduction introduces the Ḥarsūsi language, then presents the aims and objectives of the study, research questions and the structure of the thesis. Chapter Two: Literature Review presents the literature review in terms of linguistic diversity in Oman, the MSAL, Ḥarsūsi, and emphatics in Semitic. Chapter Three: Methodology describes the methodology in terms of fieldwork, consultants, data collection, transcription, and acoustic and statistical analysis.
Chapter Four: Phonetics and Phonology of Ḥarsūsi presents an overview of the Ḥarsūsi sound system. It considers the consonant phonemes and their allophones, provides an overview of the vowel phonemes and their allophones, and lays particular focus on the class of consonants known in the literature on Semitic languages as ‘emphatics’. The chapter demonstrates that the emphatics pattern phonologically with the voiced consonants as opposed to the voiceless consonants.
Chapter Five: The Phonetic Realisation of Ḥarsūsi Stops investigates the Ḥarsūsi stops by examining temporal and non-temporal acoustic parameters in an attempt to establish a) how the emphatic stops are realised phonetically, b) how the emphatic stops differ from their non-emphatic counterparts, and c) the extent to which the phonological patterning of the emphatic and voiced stops is reflected in their acoustic phonetics. The temporal parameters examined are Voice Onset Time (henceforth VOT) and oral closure duration, while the non-temporal parameter is the presence or absence of glottal closures exhibited by each segment across the data.
Chapter Six: The Phonetic Realisation of Ḥarsūsi Fricatives investigates the fricatives in Ḥarsūsi. As for the stops, temporal and non-temporal acoustic parameters are examined in an attempt to establish a) how the emphatic fricatives are realised phonetically, b) how the emphatic fricatives differ from their plain counterparts, and c) the extent to which the phonological patterning of the emphatic and voiced fricatives is reflected in the acoustic phonetics. In terms of temporal parameters, frication duration and overall segment duration are measured to establish how the emphatic fricatives differ from their plain counterparts. In terms of non-temporal parameters, it looks at the number of pre- and post-frication silent lags, frication intensity and Centre of Gravity of the spectrum (henceforth CoG) to establish whether Ḥarsūsi emphatic fricatives are realised typically as ejectives or as backed consonants.
Drawing evidence from the acoustic characteristics of the emphatics and their plain counterparts and from their phonological patterning, the thesis presents overall results that support the laryngeal categorisation of Ḥarsūsi consonants as one based on airflow (breath), as argued by Heselwood (2020) and Watson & Heselwood (2016) for the sister languages of Ḥarsūsi, Mehri and Śḥerɛ̄t, rather than the traditional categorisation based on voicing. The emphatics and canonically voiced consonants are ‘unbreathed’, while the consonants traditionally termed ‘voiceless’ are ‘breathed’.