Hadrami Arabic
Native toYemen
Speakers5.1 million (2020)[1]
Arabic alphabet
Language codes
ISO 639-3ayh
Distribution of Hadhrami Arabic according to Ethnologue
This article contains IPA phonetic symbols. Without proper rendering support, you may see question marks, boxes, or other symbols instead of Unicode characters. For an introductory guide on IPA symbols, see Help:IPA.

Hadhrami Arabic, or Ḥaḍrami Arabic (ḤA), is a variety of Arabic spoken by the Hadhrami people (Ḥaḍārem) living in the region of Hadhramaut in southeastern Yemen, with a small number of speakers found in Kenya.[1]


The dialect in many towns and villages in the Wādī (valley) and the coastal region is characterised by its ج //-yodization, changing the Classical Arabic reflex // to the approximant ي [j]. That resembles some Eastern Arabian and Gulf dialects, including the dialects of Basra in Iraq, Kuwait, Qatar, Bahrain other Arab Emirates. In educated speech, ج is realised as a voiced palatal plosive [ɟ] or affricate [] in some lexical items which are marked [+ religious] or [+ educated] (see ق /q/ below).

The ق /q/ reflex is pronounced as a voiced velar [ɡ] in all lexical items throughout the dialect. In some other Arabic dialects, /q/ is realised as a voiceless uvular plosive [q] in certain marked lexemes [+ religious], [+ educational]: /qurʔaːn/ “Qur’an”. With the spread of literacy and contact with speakers of other Arabic dialects, future sociolinguistic research may reveal whether using the uvular /q/ in certain lexemes and retaining the velar /ɡ/ for others will occur.


Labial Interdental Dental/Alveolar Palatal Velar Uvular Pharyngeal Glottal
plain emph. plain emph.
Nasal m n
Stop voiceless t k q ʔ
voiced b d ɟ ~ d͡ʒ ɡ
Fricative voiceless f θ s ʃ χ ħ h
voiced ð ðˤ z ʁ ʕ
Trill r
Approximant l ɫ j w


Front Back
Close i u
Open a

In non-emphatic environments, /aː/ is realised as an open front (slightly raised) unrounded [æ]. Thus, /θaːniː/ "second," which is normally realised with an [ɑː]-like quality in the Gulf dialects, is realised with an [æː].

Phoneme Allophone Notes
/i/ [ɪ] in shortened, non-emphatic environments
[ɨ] in emphatic or emphatic-like environments
[e̝] within the positions of pharyngeal fricatives
/a/ [æ] in non-emphatic environments
[ʌ] in emphatic-like environments
[ɑ] within the positions of emphatic consonants
/u/ [ə] in shortened, non-emphatic environments
[ʉ] within the positions of labial or high articulated consonants
[o] within the positions of uvular or pharyngeal consonants
/iː/ [iː] elsewhere in non-emphatic environments
[iːᵊ] diphthongization occurs when in emphatic environments
/eː/ [ɛ̝ː] elsewhere in non-emphatic environments
[ɛː], [ɛːᵊ] within the positions of emphatic environments
/aː/ [æː] elsewhere in non-emphatic environments
[ɑː] within the positions of emphatic environments
/oː/ [oː] elsewhere in non-emphatic environments
[ɔːᵊ] within the positions of emphatic environments
/uː/ [uː] elsewhere in non-emphatic environments
[uːʷ] within the positions of emphatic environments
Phoneme Allophone
/aj/ [æ̆ɪ]
/aw/ [ăʊ]
/uj/ [ɵ̆ɪ]
/uːj/ [uːɪ]
/eːw/ [eːʊ]

Distinctions ث, ت /t/, /θ/ and ذ, د /d/, /ð/ are made in Wādī, but ض // and ظ /ðˤ/ are both pronounced ظ [ðˤ]. The Coast merges all the pairs into the stops د, ت and ض ([t], [d] and []), respectively.

The dialect is characterised by not allowing final consonant clusters to occur in final position. Thus, Classical Arabic /bint/ "girl" is realised as /binit/. In initial positions, there is a difference between the Wādī and the coastal varieties. The coast has initial clusters in /bɣaː/ "he wants," /bsˤal/ "onions" and /briːd/ "mail (n.)," but Wādī realises the second and third words as /basˤal/ and /bariːd/, respectively.


When the first person singular comes as an independent subject pronoun, it is marked for gender: /anaː/ for masculine and /aniː/ for feminine. As an object pronoun, it comes as a bound morpheme: /-naː/ for masculine and /-niː/ for feminine. The first person subject plural is naḥnā.

The first person direct object plural is /naħnaː/ rather than the /-naː/ of many dialects. Thus, the cognate of the Classical Arabic /dˤarabanaː/ "he hit us" is /ðˤarab naħnaː/.

Stem VI, tC1āC2aC3, can be umlauted to tC1ēC2aC3, thus changing the pattern vowel ā to ē. That leads to a semantic change, as in /tʃaːradaw/ "they ran away suddenly" and /tʃeːradaw/ "they shirk, try to escape."

Intensive and frequentative verbs are common in the dialect. Thus /kasar/ "to break" is intensified to /kawsar/, as in /koːsar fi l-lʕib/ "he played rough." It can be metathesized to become frequentative, as in /kaswar min iðˤ-ðˤaħkaːt/ "he made a series (lit. breaks) of giggles or laughs."


The syntax has many similarities to other Peninsular Arabic dialects. However, the dialect contains a number of unique particles used for co-ordination, negation, and other sentence types. Examples in coordination include /kann, laːkan/ "but, nevertheless, though," /maː/ (Classical Arabic /ammaː/) "as for…," and /walla/ "or."

Like many other dialects, apophonic or ablaut passive (as in /kutib/ "it was written") is not very common, and is mainly confined to clichés and proverbs from other dialects, including Classical Arabic.

The particle /qad/ developed semantically in the dialect to /kuð/ or /ɡuð/ "yet, already, almost, nearly" and /ɡad/ or /ɡid/ "maybe, perhaps."


There are a few lexical items that are shared with Modern South Arabian languages, which perhaps distinguish this dialect from other neighbouring Peninsular dialects. The effect of Hadhrami emigration to Southeast Asia (see Arab Indonesians and Arab Singaporeans), the Indian subcontinent and East Africa is clear in the vocabulary especially in certain registers like types of food and dress: /sˤaːruːn/ "sarong." Many loanwords are listed in al-Saqqaf (2006).[3]

See also


  1. ^ a b Hadrami Arabic at Ethnologue (25th ed., 2022) Closed access icon
  2. ^ Al-Saqqaf, Abdullah Hassan Shaikh (1999). A descriptive linguistic study of the spoken Arabic of Wādī Ḥaḍramawt, Yemen. University of Exeter.
  3. ^ Al-Saqqaf, Abdullah Hassan (15 January 2006). "The Linguistics of Loanwords in Hadrami Arabic". International Journal of Bilingual Education and Bilingualism. 9 (1): 75–93. doi:10.1080/13670050608668631. S2CID 145299220.