¹ Ingham, B., ‘Urban and rural Arabic in Khūzistān’, BSOAS, XXXVI, 3, 1973, 533–53.CrossRefGoogle Scholar

² Wirth (map 11, opp. p. 22) gives the areas of southern Iraq with a settled population before 1800 as the banks of the Shaṭṭ al-'Arab and Euphrates as far as Nāṣiriyya and also the immediate vicinity of ‘Amāra and between Ḥai and Rifā'i on the Gharrāf. Wirth, E., Agrargeographie des Irak, Hamburg, 1962.Google Scholar

³ The population of the newer towns such as Ābādān, Ahwāz, ‘Amāra, and Nāṣiriyya have a large foreign element, in Iran from the Persian-speaking areas and in Iraq from the older towns. They are therefore to some extent outside this system of classification. However, the Arabic of the majority of townspeople is largely the same as that of the ḥa ar.

⁴ See below, p. 78, n. 66.

⁵ Lorimer, J. G., Gazetteer of the Persian Gulf, Oman and central Arabia, Calcutta, 1908–1915, 11, A, p. 961.Google Scholar

⁶ An important difference between the social structure of these two areas is that, along the river banks, the basic unit of social organization is the /hōz/ or area between two irrigation channels /nahar/. Inhabitants of one /hōz/ may include members of different tribes all of whom co-operate in matters relating to irrigation. Away from the rivers, however, the population is divided into large territorially organized tribes and clans, who in some areas retain a somewhat warlike tradition. Although the former groups may band together into a confederation, as in the case of the Muḥaisin league, they are not regarded by the 'arab as constituting a tribe as such.

⁷ The original significance of the term 'arab is, of course, ‘nomad’, see Khaldūm, Ibn, Muqaddima, ed. Wāfī, A. A.. Cairo, 1957–1958, 11, 409–13Google Scholar, who uses it also to cover non-Arabic-speaking nomads such as the Turks. However, as Haim Blanc points out, in Iraq at least for Baghdādī, it is used also in reference to settled rural tribesmen (Communal dialects in Baghdad, Cambridge, Mass., 1964, p. 201, n. 175Google Scholar). A distinction between Iran and Iraq is that, at least along the Euphrates, 'arab contrasts with badu which signifies camel-owning nomads whose yearly winter migration takes them further into the Syrian desert. In Iran no badu as such exist and the terms 'arab and badu are, at least by townsmen, used synonymously. Of other writers who have mentioned this classification, von Oppenheim, M., Die Beduinen, 111, Wiesbaden, 1952, 478Google Scholar, and Drower, S. in Field, H. (ed.), Anthropology of Iraq, 1, 2, Chicago, 1949, 252Google Scholar regard the main contrast as that between nomad and sedentary. Oppenheim gives ‘Schāwije ('Arab)’ versus ‘Filḥ’. Drower contrasts ‘shāwiya’ ‘shepherd pastoralist’ with ‘hadhr’ ‘settled cultivator’. Only Dickson, H. R. P., The Arab of the desert, London, 1949, 550Google Scholar, allows that the term in opposition to ‘hadhr’ ‘cultivator’, in his case ‘Shawiyah’, may not be entirely nomadic. He describes the latter as ‘half settled cultivators and half nomadic owners of flocks’ who ‘spend the late winter, spring and early summer in the desert with their flocks leaving some of their number in the reed villages by the river to till the fields’.

⁸ As mentioned by Salira, the term Mi'dān is used by those outside the area for all groups who live in reed homes on and around the lower Tigris and Euphrates. However, within the area, the basic distinction is whether or not buffalo are kept. I follow his usage of the term ‘marsh dwellers’ to include both Mi'dān and non-Mi'dān of the area. Salim, S. M., Marsh dwellers of the Euphrates delta, London, 1962, 9.Google Scholar

⁹ The main sources consulted were Ibrāhīm al-Sāmarrā'i, al-Tauzī' al-lughawī fī 'l-'Irāq, Baghdād, 1963, particularly pp. 235–42Google Scholar, and Lorimer, D. L. R., A notebook on the Arabic of Ahimz (unpublished MS).Google ScholarEdzard, D. O., ‘Zum Vokabular der Ma'dān Araber in südlichen Iraq’, in Wiessner, G. (ed.), Festschrift für Wilhelm Eilers, Wiesbaden, 1967, 305–17Google Scholar, treats the vocabulary of the Mi'dān. However, he is primarily interested in lexis and uses a generalized notation system which does not bring out the phonological features of the dialect. Much of the material he gives is general to the rest of the area considered here. The material given by Mcissner and Weissbaeh originating from the areas of Hilla and Musayyab was also consulted for comparative purposes. Meissner, Bruno, ‘Neuarabische Geschichten aus dem Iraq’, Beiträge zur Assyriologie und Semitischen Sprachwissenschaft, v, 3, 1903, i–lviii, 1–148Google Scholar; Weissbach, F. H., Beiträge zur Kunde des Irak-Arabischen, Leipzig, 1930.Google Scholar For a reconsideration of the material of Meissner and Weissbach see also Denz, A., ‘Die Verbalsyntax des neuarabischen Dialektes von Kwayriš (Irak)’, Abhandlungen für die Künde des Morgenlandes, XL, 1, 1971.Google Scholar

¹⁰ The transcription system used follows that of the previous article, except that here only broad transcription is used. This corresponds to the system used for Arabic generally, with the exception that non-final short close to half-close vowels are all assigned to a unit /I/. The pronunciation is ‘u’-like in the environment of labials and labio-dentals, more retracted in the environment of the pharyngealized consonante, and otherwise ‘I’- or ‘ə’-like depending on whether the consonant involves spread or neutral lip position. For further details, see ‘Urban and rural Arabic in Khüzistän’, 535–7.Google Scholar A dot is used to mark pharyngealized consonants, superscript for descenders otherwise subscript, i.e. /ḅ ṃ ẉ ṭ ṣ ḷ ṛ ḳ ġ ẏ ħ ḥ/. /ħ/isa voiceless pharyngeal fricative, // a voiced pharyngeal continuant, /?/ a glottal closure, /j/ a voiced palato-alveolar affricate, /č/ its voiceless counterpart, /y/ a palatal glide, /x/ and /y/ are velar fricatives, // and /ẏ/ uvular. Otherwise all consonant symbols have approximately their IPA value. With the exception of /I/ which is mentioned above, the vowel symbols have approximately their value in standard Baghdādī colloquial.

¹¹ This contrasts somewhat with the view implied in the previous article in which a number of lexical contrasts were given correlating with the urban/rural division (pp. 537–8). In the light of the additional material from Iraq, I would now wish to regard these contrasts as of regional significance dividing regional variety 1 from regional variety 2 (see below, pp. 76–9). Within Iran, however, they do also correspond to the ḥa ar /'arab division since the majority of the ḥa ar speak variety 1, while the majority of the 'arab show certain characteristics of variety 2.

¹² For the general characteristics of this group and the classification of the dialects of Mesopotamia into ‘gelet’ and ‘qeltu’ dialects, see Blanc, , Communal dialects in Baghdad, 7.Google Scholar

¹³ For details of this historical process, see Johnstone, T. M., ‘The affrication of “käf” and “qāf” in the Arabic dialects of the Arabian peninsula’, JSS, VIII, 2, 1963, 210–26Google Scholar, also Blanc, , Communal dialects, 25–30.Google Scholar

¹⁴ This feature is also mentioned by Sāmarrā'ī as characteristic of the region of ’Amāra, , al-Tauzī' al-lughawī, 236.Google Scholar

¹⁵ Johnstone has treated the geographical distribution of /y/ as a reflex of jīm for the dialects of the Arabian peninsula in ‘The sound change j > y in the Arabic dialects of peninsular Arabia’, BSOAS, XXVIII, 2, 1965, 234–7+y+in+the+Arabic+dialects+of+peninsular+Arabia’,+BSOAS,+XXVIII,+2,+1965,+234–7>Google Scholar, also in Eastern Arabian dialect studies, London, 1967, 9–11.Google Scholar The evidence of informants in Iraq suggested that the northernmost limit of /y/ on the Tigris was somewhere between Kūt and Baghdād, while on the Euphrates it was somewhere between Batha and Samāwa. I myself heard /y/ from rural speakers of the region of Küt.

¹⁶ The active participles of the verbs ‘to come’ and ‘to go’, i.e. /ya:y/ and /ya:d/, are used in all ‘gelet’ dialects with the meaning ‘this way’ and ‘that way’. In the ‘Amāra region, however, they have lost their directional significance and mean purely ‘here’ and ‘there’.

¹⁷ The form /ke:f/ occurs in the Shaṭṭ al-'Arab area with the meaning ‘state, condition’ in such sentences as /?ana 'ke:fi 'mu: -ze:n/ ‘My condition is not good, I do not feel well’. In the ‘Amāra area, however, /či:f/ is an interrogative particle occurring in such sentences as /wilak 'či:f ma: -riħit/ ‘How is it that you did not go?’, /'či:f haliħ'ča:ya/ ‘What is this story?’.

¹⁸ This particle occurs in such forms as /'li:ya:d/ ‘in that direction’, /'li:wara/ ‘backwards’. The forms with /li:-/ are also characteristic of the dialect of Baghdād, as also is the case with /hi:č/ and /či:ma/.

¹⁹ In the Shaṭṭ al-'Arab area of Iraq the non-diminutive form /'a:na/ is used.

²⁰ In this item, as with /š'ṭi:t/ below, the final apical was not pharyngealized as is the case regularly with reflexes of Cl. /ṭ ṣ/ following /i:/, e.g. /mi'ri:ð/ ‘ill’, /ri'xi:s/ ‘cheap’. Compare Cl. /mari:, raxi:ṣ/.

²¹ Layard, , ‘Description of the province of Khuzistan’, JRGS, XVI, 1, 1846, 5, gives this as Zehiriyyah.Google Scholar

²² Compare Middle Persian Meshān, Classical Arabic Maisān, the name of the whole of lower Mesopotamia and parts of Khūzistān up till the late Middle Ages, see ‘Maisān’, Encyclopedia of Islam, 111, 146–7Google Scholar, also Hansman, J., ‘Charax and the Karkheh’, Iranica Antiqua, VII, 1967, 27.Google Scholar A number of related place-names are also found, e.g. Naisān, Shatt Naisān, Hansman, 35; the writer also visited a settlement north-east of Chibāyish called Nīshān. The form /?i:ša:n/, the name for the elevations on which many Mi'dān settlements are made, may also be traceable to Meshān /Maisān.

²³ The form /'?intam/ occurs in the ‘Amāra area and in the speech of the Kawāwila. See also Sāmarrā'ī, , al- Tauzī' al-lughawī, 237.Google Scholar

²⁴ The different syllabic structures of these two imperatives are also characteristic of these two varieties. For further examples, see ‘Urban and rural Arabic in Khūlzistān’, 542.Google Scholar

²⁵ The Ahwāz informant also gave the forms /ḅi'ḳa:n/ and /bi'ča:n/.

²⁶ This feature is also mentioned by Sāmarrā'i for ‘Amāra, al-Tauzī' al-lughawī, 242.Google Scholar

²⁷ The form /mu:š/ as a negative particle occurs in a much wider area extending northwards even to Dīwāniyya on the Euphrates. See Sāmarrā'ī, , 240.Google Scholar

²⁸ There is no comparable phrase in the dialect of the Shaṭṭ al-'Arab. The construction is reminiscent of the Persian /nīst o nābūd šod/ of the same meaning.

²⁹ The forms /?i'hiwwa, ?i'hiyya, ?i'hiṃṃa, ?i'himna/were also reported for the speech of the Banī Läm, a nomadic tribe of the area north and east of ‘Amāra. Similar forms are also given by Johnstone for Baḥrainī, EADS, 104.Google Scholar On the basis of this and other similarities, see ‘Urban and rural Arabic’, p. 544, n. 20Google Scholar; it seems likely that some Shī'ī speech of Baḥrain has its origins in southern Iraq.

³⁰ Also forms in /-a/ involving the feminine suffix /-a, -at/, the /t/ of which is absent in pause, as in the last three examples.

³¹ Forms of this type are also given by Weissbach, ‘a áḏit ’, p.7; ‘šāfít ’, p. 23; ‘tĕlággit ’, p. 8.

³² In the Shaṭṭ al-'Arab area the lexeme /aṃṃ/ occurs for ‘to hide’.

³³ Compare Classical /qismah/. In these dialects it seems possible to assign the initial consonantal articulation which may be either plosive or fricative to /y/ equating it with the /y/ in /'yisal/ ‘he washed’, compare Classical /yasala/. See also ‘Urban and rural Arabic in Khuzistan’, 537.Google Scholar

³⁴ Although Baghdādī has a deictic particle of similar function /hiy'ya:t-/, there do not seem to be parallel elements in the rest of the area considered, see Clarity, B. E., Stowasser, K., and Wolfe, R. G. (ed.), A dictionary of Iraqi Arabic, Washington, D.C., Georgetown University Press, 1964, 179.Google Scholar

³⁵ ibid., 16, the form given is /bsabab/.

³⁶ ibid., 198.

³⁷ cf Persian /hargiz/ ‘never’.

³⁸ The Eastern Arabian dialects are spoken along the west coast of the Persian Gulf. See Johnstone, EADS.Google Scholar Of particular relevance to this relationship are parts 1, 11, and 111.

³⁹ Longrigg, S. H., Four centuries of modern Iraq, Oxford, 1925, 1, 39Google Scholar, gives Zubair as a staging-post for caravans from the sixteenth century.

⁴⁰ Cases (i), (ii), (iii), (iv), and (v) all result from historical sound changes which are general to the whole area, but which have applied differently for different dialects and thus produced sets of contrasting forms; see also Johnstone, , EADS, 6–9Google Scholar, for cases (i), (ii), and (iii), also the appendix to that work and ‘Aspects of syllabication in the spoken Arabic of ‘Anaiza’, BSOAS, XXX, 1, 1967, 1–16, for case (v).Google Scholar

⁴¹ Johnstone mentions a similar fluctuation as regards forms of the type treated under case (iii) below for the Eastern Arabian dialects; EADS, 43.Google Scholar

⁴² See Labov, W., ‘Hypercorrection by the lower middle class as a factor in linguistic change’, in Bright, W. (ed.), Sociolinguistics, The Hague, 1966, 84–113, particularly 84–5.Google Scholar

⁴³ Only these two classes have the possibility of syllabic variation of this type. Of the others the ‘hollow’ and ‘doubled’ involve only two consonant articulations, while the ‘initial weak’ has only /?/, /w/, and /y/ as initial radical.

⁴⁴ Although no examples of this type of form with initial radical /y/ or /h/ were recorded, it would seem likely that they occur since they are common in nomadic speech of the Syrian desert and also occur in Weissbach's material. Compare the following examples: ‘ĭth ḏib’ p. 208; ‘ihán lan’, p. 207; ‘ěṙazûn’, p. 25; ‘īṙálub’, p. 62.Google Scholar See also ‘Urban and rural Arabic’, 543–4.Google Scholar

⁴⁵ The term ‘unaugmented’ is here used to signify that none of the internal morphological additions to the root characteristic of certain other types of form occur. Where these are present, this typo of contrast does not occur. See also ‘Urban and rural Arabic’, 540–1.Google Scholar

⁴⁶ The occurrence of /i/ or /a/ depends on the nature of the consonantal environment as specified for case (v) below, p. 75. For further details, see Johnstone, EADS, 257–9.

⁴⁷ This is given by Lorimer as ‘ightufān’, the name of a nomadic section of the ‘Anāfija, a tribe of northern Khūzistān, Gazetteer, 11, A, p. 73.Google Scholar I presume that his /t/ should be /ṭ/ following Classical Ghaṭafān. He also gives Aghzawi as a branch of the Banī Mālik, Gazetteer, 11, B, p. 1156.Google Scholar However, this form is suspect since Oppenheim gives it as Ighzēwī, Die Beduinen, TV, 1, Wiesbaden, 1967, 46.Google Scholar

⁴⁸ This form occurs as the name of a lake to the north-east of Chibāyish, /birkat byada:d/.

⁴⁹ In dialects of the ‘sedentary’ type recorded, the item /taħat/ did not occur. In these varieties the forms /jiwwa/ or /hadir/ occurred for the meaning ‘below’, the second of which is mentioned by Johnstone for the Eastern Arabian dialects, EADS, 68.Google Scholar

⁵⁰ This form is given by Lorimer as ‘raháma’ a division of the Banī Lām living as nomads to the north-east of ‘Amāra. Gazetteer. 11, B, p. 1081.Google Scholar

⁵¹ The /'caab/ ‘ankle bones (of a sheep)’ are used in a local game of that name, see Westphal-Hellbusch, S. and Westphal, H., Die Ma'dān, Kultur und Geschichte der Marschenbewohner, Berlin, 1962, 298–301.Google Scholar

⁵² Forms without the initial /y/ are also heard frequently, i.e. /(y)ti'ba:ha, (y)ta'la:bas/, although these are homophonous with the perfective.

⁵³ In ‘sedentary’ varieties /y'wa:iḅ/ or /y'di:r 'ba:la/.

⁵⁴ The verb /yida/ which in ‘sedentary’ speech means ‘to become’ does not occur in the imperative in these varieties. The normal imperative would be /ṣi:r/ from /ṣa:ṛ/ ‘to believe’.

⁵⁵ This phrase is a formula addressed by passers-by to a group sitting eating, with the implied subject /aḷḷa:h/ ‘God’. The customary reply is /'mmḥiṃ/ ‘(May you be) one of them!’ Similar forms of greeting (where the addressee is indicated by the 3rd person) are not uncommon in the speech area. Also recorded were /?aḷ'ḷa: ysa:'idḥiṃ/ ‘May God assist them!’, and /'ħayyḥiṃ/ ‘Give them life!’. The first of these is common Iraqi usage. The second is from a tape of Bedouin speech from Kuwait.

⁵⁶ Blanc, , Communal dialects, 168–70.Google Scholar For the arrival of the main Arab tribes in Kliūzistān. see Kasravī, A., Tārīkh-i pānṣadsāla-i Khūzistān, Tehran, 1934, 142Google Scholar; Lorimer, , Gazetteer, 111, p. 1627Google Scholar; Oppenheim, , Die Beduinen, IV, 1, pp. 10, 11, 15.Google Scholar

⁵⁷ In particular /?imiš/ ‘go, let us go!’ used by many speakers whose speech is primarily of the ‘sedentary’ type.

⁵⁸ The exact localities for which material was collected were Ḥauz ‘Abd al-'Azīz al-Rāshid, 13 km. south of Fāu, Fāu town, Sība, both sides of the Shaṭṭ at Abū Khaṣib, ‘Ashār, Tannūma, and Bū Wārda on Ābādān island.

⁵⁹ Niebuhr, C., Beschreibung von Arabien, Kopenhagen, 1772, 11, ch. v, p. 320Google Scholar; De Bode, G. A., Travels in Luristan and Arabistan, London, 1845, 11, ch. XX, pp. 111–12Google Scholar; Lorimer, , Gazetteer, 11, A, p. 962.Google Scholar For a more recent account of the history of the Ka'b see Perry, J. R., ‘The Banū Ka'b: an ambitious brigand state in Khūzistān’, in [Aubin, J. (ed )], Le monde iranien et l'Islam, 1, Genève, Paris, 1971, 131–52.Google Scholar

⁶⁰ Wilson, A. T., Précis of the relations of the British government with the tribes and shaikhs of Arabistan, Bushire, 1911, 67.Google Scholar

⁶¹ Lorimer, , Gazetteer, 11, A, p. 121Google Scholar, gives the Ka'b as completely sedentary. Layard, , ‘Descrip tion of the province of Khuzistan’, JUGS, XVI, 1, 1846, 1–105Google Scholar, states that the majority of the Ka'b are ‘Deh Nishíns or settlers in villages’. The accounts of Stocqueler and Niebuhr also sug gest a mainly sedentary population, see particularly Stocqueler, J. H., Fifteen months' pilgrimage through untrodden tracts of Khuzistan and Persia, London, 1832, 11, ch. viii, p. 386Google Scholar, and Niebuhr, , 11, ch. iii, p. 168.Google Scholar

⁶² The two groups of Kawāwila were settled at Chawādir in the desert south-west of Baṣra and just outside Ahwāz. Both these groups had been nomadic tent dwellers up till recently, but were associated respectively with the localities of Ma'mal Waraq between Baṣra and Gurna and Mullā ānī above Ahwāz on the Kārūn. Their speech was identical, except that the Baṣra informant showed hesitation between /y/ and /Ʒ/ in many instances.

⁶³ Dickson, , The Arab of the desert, 517–19Google Scholar; R. Montagne, ‘Contes poétiques bédouins (récuillis chez les Šammar de ezīré), BEO, v, 1935, 72.Google Scholar

⁶⁴ Montagne, , ‘Contes’, 72–3Google Scholar; Dickson, , The Arab of the desert, 512–23.Google Scholar

⁶⁵ Gazetteer, 11, A, p. 73.Google Scholar

⁶⁶ For Khūzistān in particular the evidence of earlier writers testifies to the nomadic conditions of the area. Lorimer gives the main pastoral areas as Ahwāz and Ḥuwayza. Tribes with a majority of nomadic sections, he gives as the Bāwiya and ‘Anāfija (Ahwāz), the Al Kathīr (Dizfūl and Shūshtar), and the Banī Tamīn and Banī Ṭuruf (Ḥuwayza), Gazetteer, 11, A, pp. 119–24.Google Scholar

⁶⁷ No phonological or morphological features of purely regional relevance were noted for this region. This may be partly due to the comparatively small amount of data obtained. A significant lexical feature was the occurrence of the form /čiθi:r/ for the meaning ‘many’. In the rest of the area reflexes of /wa:jid/ are used, /hwa:ya/ in the ‘Amāra area and in Baghdād /wa:yid/ along the Shaṭṭ al-'Arab and in Khüzistān.

⁶⁸ Examples of all except type (vi) of the ‘nomadic’ forms can be found in their material, e.g. from Weissbach, ‘it ádim’ ‘she serves’, p. 23; ‘ĭṭlá ,’ ‘they went out’, p. 5; ‘l-igháṷ ’ ‘coffee’, p. 1; ‘ite ârak’ ‘they fight’, p. 13Google Scholar; ‘inbadar’ ‘he went down’, p. 3.Google Scholar

⁶⁹ Dickson, , The Arab of the desert, ch. xlv, particularly p. 552–3.Google Scholaral-Ṭāhir, A. J., al-'Ashāyir al-'Irāqiyya. Beirut, 1972, particularly pp. 1–5.Google Scholar

⁷⁰ My information from the speech of the shepherd tribes comes from informants from the Budūr and Banī Ḥachīm. For the settled cultivators my informants were from the awālim and one unidentified rural informant from the Nasiriyya area.

⁷¹ This dialect division into a Euphrates variety versus one characteristic of the area of ‘Amāra and parts of Khuzistan corresponds to Salim's view of the main patterns of cultural interaction in the area, Marsh dwellers, 8.Google Scholar Salim also states (p. 9) that ‘Marsh dwellers cannot be distinguished on a linguistic basis because there are only small local changes in dialect’. Since his use of the term marsh dwellers covers all cultivators along and between the lower Euphrates and Tigris, this statement is contrary to my findings. However, it may be that he is referring to the relatively slight variation observed in comparison to the two main dialect groups of Mesopotamia, the ‘gelet’ and ‘qeltu’ dialects. Blanc, , Communal dialects, 7.Google Scholar

⁷² The relationship of the speech of the Bedouin to that of the sedentary population of the north-east of the Arabian peninsula is mentioned both by Blanc, , Communal dialects, 68Google Scholar, and by Johnstone, , EADS, 1, 2.Google Scholar Johnstone includes both sedentary and nomadic dialects under his groups (b) Shammarī and (c) ‘Anazī of the north Arabian group. The main features which distinguish the speech of the Bedouin from those of the shepherd tribes are the same as those given below as differentiating the Eastern Arabian dialects from those of Mesopotamia.

⁷³ EADS, particularly 70–92.Google Scholar

⁷⁴ The absence of the ‘nomadic’ form for case (i) agrees with the dialect of Kuwait, EADS, 70.Google Scholar

⁷⁶ Communal dialects, 55.Google Scholar

⁷⁶ [Holmes, D. and Sam'ān, S.], A handbook of Kuwaiti Arabic, London, c. 1951, 204.Google Scholar

⁷⁷ EADS, 72.Google Scholar

⁷⁸ ibid., 30.

⁷⁹ ibid., 78.

⁸⁰ ibid., 13.

⁸¹ ibid., 169.

⁸² ibid., 68.

⁸³ Johnstone, ibid., 171 notes this form for the Trucial Coast.

⁸⁴ Contrast /yahaw, yahay/, the forms common to the rest of the area, ‘Urban and rural Arabic’, 547.Google Scholar

⁸⁵ Landberg, C.Glossaire daṯînois, 1, Leiden, 1920, 311.Google Scholar

⁸⁶ Compare Kuwaiti /ko:s/ ‘north wind’, EADS, 125.Google Scholar

⁸⁷ A phonological feature shared exclusively by the Fān variety and the dialect of Kuwait, which to some extent supports the supposition of the maritime connexion in view of the lexemes involved, is the presence of nominals in final /-o:/ or /-o:h/. Johnstone mentions this feature, EADS, 84Google Scholar, and gives two examples, /ḥalwayoo/ ‘variety of fish’ and /čaftoo[h]/ ‘keelson (naut)’, also /yāmlo/ ‘rail at the poop’, Johnstone, and Muir, , ‘Some nautical terms in the Kuwaiti dialect of Arabic’, BSOAS, XXVII, 2, 1964, 322.Google Scholar My material from Fāu includes /ṭaṭi'o:h/ ‘variety of fish’ and /ṣille:'bo:/ ‘juice of the chōlān plant’, also the possibility of forming familiar forms of names by adding a suffix /-o:/ e.g. /i:'so:/ ‘Isa, /xali:'lo:/ Khalīl, /nay'mo:/ Najm, /ywa:'do:/ Jawād.

⁸⁸ Lorimer, , Gazetteer, 11, B, p. 1951.Google Scholar

⁸⁹ Johnstone mentions a number of features which point to the influence of southern Iraqi dialects on that of Kuwait, EADS, 70, 71.Google Scholar

⁹⁰ Also the town of Zubair, with the exception of case (i). See above, p. 80.

⁹¹ See above under regional variety 2, p. 77–9.