Note: this research is published in Anthropological Lingusitics in 2003, Volume 45, No.4, p 413-425. This journal is published by Indiana University –Bloomington-USA, and it is international, indexed and refereed journal. Thus, any citation from this research has to indicate the author. Copyright reserved to Anthropological Linguistics.
The Use of the [q] Variant in the Arabic Dialect of Tirat Haifa
MAHMOUD EL SALMAN
Al Balqa' Applied University, Jordan
Abstract. This article is a sociolinguistic study of Palestinians now living in Irbid, Jordan, who originally migrated from the village of Tirat Haifa, Palestine. A key feature of the Arabic dialect spoken by members of this group is the [q] variant of the (Q) variable. In this study, variation in the use of [q] in the speech of informants from this group is examined to identify possible correlations with selected extralinguistic factors.
1. Introduction. This article examines the use of the [q] variant of the (Q)
variable in the Arabic dialect of Palestinians living in Irbid, Jordan, who
migrated from Tirat Haifa, Palestine. The dialect of this group is one of very few
Arabic dialects in the area whose phonemic and phonetic inventory still has the
variant [q] of the (Q) variable—a phenomenon of central debate in Arabic socio-
In the standard configuration, well known in the Levant, Iraq, and Egypt, the variant [q] is profiled for higher register, formality, education, Arabic nationhood, whereas other variants ([k], [?], [g], etc.) are ethnically specific, or associated with local prestige. What happens, however, when the local ethnic group itself has [q] as their native variant? Indeed, it is this question that I try to answer in this article—a question that rarely arises, since few groups in these areas have such a variable.
Section 2 below gives background information about the village of al-Tira (Tirat Haifa) and its people. Section 3 discusses methodology used in the study. Section 4 discusses the general significance of the (Q) variable. Section 5 treats the use of the (Q) variable by people from al-Tira before they were forced to leave their village in 1948, and section 6 their use of the (Q) variable, after they settled in Irbid, Jordan. Section 7 is a conclusion.
2. Background: al-Tira and its people. The village of al-Tira, also referred
to as Tirat Haifa (i.e., "the Tira of Haifa") to distinguish it from similar place
names, is located on the northwestern slopes of Mount Carmel, Palestine,
overlooking the coastal plain. At its widest point, it is three kilometers across. It
is bounded by Mount Carmel to the north, by the Mediterranean Sea to the west,
and by two neighboring villages, Ayn Hawd to the south and Osfiyah to the east.
Two valleys converge at the plains to the southwest of al-Tira (El Salman
Before 1948, al-Tira was linked to Haifa by a spur to a coastal highway northwest of the village. Al-Tira was one of the most important villages in the
Haifa district; it was the district's most populous village and the second largest in the area after Ijzim (Khalidi et al. 1992:195). The Crusaders called it St. Yohan de Tire. In 1596, al-Tira was a village in the Nahiya of Shafa (Liwa' of Lajjun), with a population of 286 (Khalidi et al. 1992:196).
Tirawis, or people from al-Tira, began migrating to Jordan in large numbers in 1948 as a result of the Arab-Israeli war when the village was occupied by the Israelis. The vast majority of these migrants settled in the city of Irbid in northern Jordan.
3. Methodology. This study is empirical in its methodology in that it is fully dependent on natural speech. All of the interviews were conducted in Irbid, Jordan, in a face-to-face setting to obtain data with regard to linguistic varia¬tion. The study included a total of forty-eight informants, comprising Jordanians originally from al-Tira in Palestine, who migrated to the city of Irbid as a result of the Arab-Israeli war in 1948, and their descendants. The size of the sample should be acceptable, since it is comparable to the number of informants in many similar studies. (For example, Trudgill  used sixty informants in a study of a city with a population of 118,610, and Daher's study , conducted in Damascus, used forty-six.) Labov also notes that "the structure of social and stylistic variation of language can be studied through samples considerably smaller than those required for the study of other forms of social behavior" (1966:638). My own large social network enabled me to draw my sample from different kinds of people and from both sexes. The fact that I myself belong to this group helped me to avoid the inherent difficulty of a male-to-female interview; hence, I was not obliged to engage a female to help me in carrying out the interviews. This is a very problematic issue in the Arab world, which has led some sociolinguists to exclude females from their study, as did Al-Jehani (1985) in his study of Makkan Arabic (Saudi Arabia), because of the cultural norms restricting women's contact with men. Other researchers have engaged a female to help in conducting their interviews with females, as did Al-Khatib (1988) in his study in Irbid, Jordan, and Daher (1998) in his study in Damascus. In my study, I was not obliged to exclude females or engage a female to help in conducting interviews on my behalf.
This study was conducted in light of the Labovian paradigm. The interviews were carried out in shops, offices, houses, and workshops. This was determined by the possible availability of the informants. The social network framework was followed and the informants were approached in the capacity of a friend of a friend or in some cases a friend of a friend of a friend (Milroy and Milroy 1978). I also tape-recorded group conversations among Tirawis, when this was possible, to use as a "a supplementary check" on the face-to-face interviews (see Labov
The questions asked in the interviews were designed to provide a challenge for the interviewees' memories in that they depended solely on the personal
knowledge of the Tirawis, and dealt with what they still remembered about the village or remembered hearing from their parents. This aspect of the research design, intended to ensure that they remembered more about the village than anyone else, reduced the attention paid to the linguistic terms used to describe the reminiscences. This, I believe, succeeded in helping me to deal with the "observer's paradox," by reducing the attention paid to the language (see Labov 2001:36). Thus, I can claim that at least in most parts of the interviews I suc¬ceeded in determining how Tirawis speak when they are not being systema¬tically observed (see Labov 1972a:l8l). Nevertheless, I also discovered that the "observer's paradox" appeared to be difficult to solve at least at the beginning of the interviews. Indeed, the interviewees' attempts to make their speech conform to Standard Arabic are clearly seen at the beginning of many of the interviews.
I also asked additional questions after the main interview. These questions were designed to measure informants' perception of different variants in the speech of Tirawis.
I divided the sample into three kinds of groups, according to education, sex, and age. In order to gain a statistically accurate view of the use of the variables, a univariate analysis is employed. For purposes of the univariate analysis, each group is presented as a number. On the parameter of level of education, the un¬educated group is presented as 1 and the educated group as 2. On the parameter of sex, the male group is 1 and female group is 2. On the parameter of age, the young age group is 1, the middle age group is 2, and the old group is 3.
4. The variable (Q). Abdel-Jawad states that "because of the social and geo¬graphical importance of this variable [(Q)] as a carrier of local or regional loyalties, it has often been used by dialectologists as the main criterion for esta¬blishing the dialect boundaries or isoglosses in the Arabic dialects" (1981:159). The variable (Q) has four major variants: the voiceless uvular stop [q], the voiced velar stop [g], the voiceless velar stop [k], and the glottal stop [?]. For the pur¬pose of this study, the following three variants will be considered: the voiceless uvular stop [q], which is in general considered the Standard Arabic variant, while in this study it is also part of the native Tirawi dialect; the voiced velar stop [g], which is the local variant of Irbid; and the glottal stop , which is an important urban variant and has also begun to be adopted by some Tirawis. The voiceless velar stop [k] will not be considered, as it is neither part of the Tirawi dialect nor part of the speech of Horanis, the local people of Irbid (Al Khatib 1988:89). Nor is the [k] variant used by any of the informants. It is typically found in the rural dialects of central Palestine (Al Khatib 1988:82; Amara, Spolsky, and Tushyeh 1999:67; Al-Zu'bi 2001:93).
The voiceless uvular stop [q] is the key feature of the Tirawi dialect. Since the Tirawi variant corresponds to the Standard Arabic variant, education will not be a factor in the use of [q], as this is the local variant for all Tirawis—they use [q] both for colloquial words with the (Q) variable and for words adopted from Standard Arabic. This contrasts with speakers of dialects that have a variant other than [q] for the variable (Q). These speakers use the non-[q] vari¬ant in pure colloquial words, but might use [q] in words adopted from Standard Arabic. Most Tirawis who still preserve the local [q] variant are illiterate. Unlike most of the studies conducted in the Arab world, in our study the use of [q] is considered part of the colloquial dialect used by rural Tirawis and cannot indi¬cate level of education. The Statistical Program for Social Sciences (SPSS) used for data analysis shows that education is not significant in the use of the [q] variant among Tirawis (the threshold for significance is p < 0.05, while p < 0.504 for education in the use of the [q] variant). Previous studies considered the vari¬ant [q] in its capacity as the Standard Arabic variant. Furthermore, it is con¬sidered an indicator of the level of education in that it is typically only attainable through education (see Abdel-Jawad 1981; Al Khatib 1988; Kanakri 1988; Al-Wer 1991, 1999, 2000; Walters 1991; Daher 1998; Amara, Spolsky, and Tushyeh 1999). The [q] used in my study identifies a speaker as a Tirawi and is not correlated with the level of education. As noted above and shown here in figure 1, most of the [q]-preservers are illiterate or have the minimum level of educa¬tion. (Here, the uneducated group is 1.00 and the educated group is 2.00.)
This, however, does not rule out the possibility that educated Tirawis, espe¬cially those who have abandoned the [q] variant, might use [q] as a result of an attempt to standardize their speech in some domains. For example, as the Standard Arabic variant, its use among most educated people might be pre¬ferable. The use of [q] may also be an attempt to make their speech conform to the original dialect of their village. When used by the educated to show a high level of education, the [q] variant can be related to certain words classified as Standard Arabic (see Abdel-Jawad 1981:208; Al Khatib 1988:99). This type of [q] is used at a relatively low rate in speech and rarely throughout the interview, as it is not the native variant of these speakers. Words, mentioned in other studies (e.g., Al Khatib  and Al-Wer ), that are used in the speech of my informants with the [q] variant and are classified as Standard Arabic include qaraar 'decision', qaanuun 'law', and faqat 'only'. Educated Tirawis use the [q] variant with words classified as Standard Arabic or with words classified as purely colloquial, just as uneducated Tirawis do. Examples of words classified as purely colloquial include izzalaqah (the name of a very well-known place in the village—repeated five times in the speech of the informants with the [q] variant regardless of the level of education of the speaker); quf (the name of a hill); qa9quur, alqit9ah, and qdeesih (names of agricultural places); and qaraaniif'(a place name). Other examples include certain words taken from the environment of al-Tira, such as bulluq 'a kind of plough (used by Tirawis before 1948)', qirtyaan 'a kind of a stone (found in al-Tira)', qunza9ah 'the lid of the bread oven', miqlaa9 'a stick (used to move the bread in the oven)', miqHaar 'a stick (used to prepare the bread in the oven)', and qumbaaz 'a kind of loose-fitting garment (formerly worn by men in al-Tira, especially on important occasions)'. The use of some of these words related to the indigenous local life of the Tirawis, and, in particular, the use of such place names in al-Tira by some Tirawis who appear to have abandoned their native variant [q] might be the result of their having been borrowed in full with their Tirawi pronunciation. Abdel-Jawad also notes that "speakers usually perceive . . . colloquial items in their local pro¬nunciation because they feel that these words exist in their local variety in this phonetic shape" (1981:218). In other words, when such words are used with the [q] variant by Tirawis who appear to have generally abandoned their native dialect, these Tirawis are probably using this variant in order to demonstrate their membership of the village community, rather than attempting to apply the [q] standardization rule.
Thus, while Tirawis who have abandoned the [q] variant might use this variant as an attempt to standardize their level of speech, what is peculiar to this group is that their use of [q] might also be an attempt to identify themselves with their village, especially when this pronunciation is found in words known to be related to their village and its environment, such as those mentioned above. This is a feature of the speech of both educated and uneducated Tirawis.
5. The (Q) variable in al-Tira before 1948. The people of al-Tira had major external and economic relations with the nearest city, Haifa. Most linguistic anecdotes of these people were centered around the voiceless uvular stop [q], as the key feature of the al-Tira dialect, and the glottal stop [?], as the key feature of the Haifa dialect. The fact that the [q] variant was considered the variant of the rural people of al-Tira, while the [?] variant was considered the variant of the urban people of Haifa, meant that the [q] variant was perceived as the unrefined, coarse, or "tough" variant. From this "toughness," the people of al-
Tira were stereotypically viewed as stubborn. The [?] variant, by contrast, was perceived as the soft variant (Al-Wer 1991; Daher 1998). Because of this, the people of al-Tira regarded the use of the [?] variant as unsuitable for "real" men and any attempt by any person from al-Tira to use the [?] variant would have been perceived as an attempt to dissociate themselves from the rural group. This, I believe, was the main reason for women's general reluctance to use the [?] variant in al-Tira before 1948, even though it reflects softness and urbaniza¬tion, features regarded as suiting the nature of women in some parts of the Arab world (Al-Wer 1999:41; Amara, Spolsky, and Tushyeh 1999:70).
The social meanings of these linguistic features came to be manipulated by both the people of al-Tira and those of Haifa to describe one another. The people of Haifa used to employ the [q] variant of the al-Tira dialect humorously to de¬scribe the stubbornness and toughness of the Tirawis. For instance, if a man from Haifa wanted to know if somebody was from al-Tira, he might humorously ask him inta min illi qawwasu ilbaHar? 'Are you one of those who shot the sea?' And, if the answer was "yes," it was not uncommon for this question to be fol¬lowed by a second typical question: qarqa9 babuur ittiirih willa ba9duh? 'Has the grinding machine of al-Tira started making a noise, or not yet?' Notice that the key feature [q] of this dialect is used in these phrases to convey this humor¬ous and friendly social meaning. The word qarqa9 'made a grinding noise' has the [q] variant twice, confirming the speaker's intention to signal a correlation between his identity as a Tirawi and his use of [q].
My point in this discussion is to show that the key feature [q] became indissolubly linked with the Tirawis and has come to be identified with them whether they use it or not.
Since the large population center nearest to al-Tira is Haifa and the people of Haifa were specifically known to use the urban variant [?], Tirawis subcon-siously relate the social connotations of this variant as the variant of urbaniza¬tion and softness to Haifa. This impression was confirmed by discussions I had with my informants after the informal interviews. It is also something I have heard many times from other Tirawis over the years. The three cities of Haifa, Yafa, and Aka (Acre) played a major role in spreading the urban variant [?] to Jordan and other places, in that most of the people of these three cities emigrated to Jordan in 1948 before the connotations of the [?] variant were widely known. In addition, the people of Jordan established good relations with these urban centers even before 1948. Haifa was a major port and had a large oil refinery. These two factors made Haifa commercially prosperous and, thus, a suitable place for those seeking work. Moreover, since Haifa was an important urban center, its dialect must have been known to most of the people of the area, including the people of Jordan, before 1948.
Because social values associated with particular groups in society tend to become attached to the linguistic forms used by the members of those groups and values attached to language usage tend to vary with social background
(Wolfram 1997:120; Blom and Gumperz 1972:421), the urban dialect of the cities of Haifa, Aka, and Yafa was perceived positively by the people of Jordan and enhanced the association between urbanization, softness, and the  variant.
6. The (Q) variable used by Tirawis in Irbid: findings and discus¬sion. In this section, I discuss the use of the (Q) variable by the Tirawis who have recently come to live in Irbid. Table 1 shows the distribution of informants by sex and age. In terms of the realization of (Q), the members of the study are divided into [q]-preservers, [g]-adopters, and [?]-adopters.
Table 1. Distribution of Informants by Sex and Age
AGE MALE FEMALE TOTAL
YOUNG 8 8 16
MIDDLE 8 8 16
OLD 8 8 16
TOTAL 24 24 48
6.1. [q]-preservers. Recent immigrants to Irbid who retain frequent use of the [q] variant can be called [q]-preservers. These people are not affected by the powerful local variant [g] or by the newly spreading urban variant [?].
Table 2 shows that the [q] variant is used in 100 percent of occurrences of the (Q) variable among old males and in 98 percent of occurrences among old females. In the remaining 2 percent of cases, old females use the [?] variant of the (Q) variable. Thus, innovation in this area can be seen to have begun among old females, even though it is only marginally attested in this case. SPSS analysis also shows that age is very significant in the use of the [q] variant (the threshold for significance is p < 0.05, while p < 0.00 for age with regard to the use of the [q]).
Table 2. The Distribution of the Variable (Q) by Age and Sex
AGE MALE FEMALE
[q]% [g)% [?]% N [q]% [g]% [?]% N
YOUNG 6 94 0 497 3 2 95 410
MIDDLE 10 90 0 500 6 0 94 400
OLD 100 0 0 590 98 0 2 514
Figure 2 also displays this pattern for the use of [qj. It shows that there is a general increase in the use of [q] from young to old speakers and that the peak is located in the old age group. (Here, the young age group is 1.00, the middle age group is 2.00, and the old age group is 3.00.)
Linguistically, both females and males of the old age group behaved similarly in that they categorically preserved the [q] variant. SPSS analysis shows that sex, by contrast, has a very low significance in the use of the [q] variant (the thresh¬old for significance is p < 0.05, while p < 0.278 for sex in the use of the [q] variant). The crucial factor, therefore, is age. It is a very well-known fact that the old are subject to much more pressure from their society to preserve their native dialect than the young (Al-Wer 1991:91; Walters 1991:201). Eckert also reports that "community studies of variation frequently show that increasing age correlates with increasing conservatism in speech" (1997:157). In the case of women, we can add the factor of contact. Generally, in Arab societies, women have less contact with the outside world than men and the social contact between them is greater than that existing between men. Bakir notes that Arab women, "being restricted to domestic chores and each other's company, have a narrower social network and do not have the opportunity or need to use the standard forms" (Bakir 1986; cited by Mazraani 1997:3).
Table 2 shows that there is a drastic decrease in the use of the native [q] variant among the members of the middle age group. The rate is 10 percent among males and 6 percent among females. The drastic decrease is due to the fact that most of the people in the middle age group were born in Jordan (those
who are less than fifty-three years old). In addition, a small number in this age group were born before the 1948 war, but at that time they were less than four years old. In other words, they were too young at that time to have mastered their native dialect and for this dialect to have settled. Thus, they do not differ linguistically from other people in their age group who were born in Jordan after the war. In contrast, all the members of the old age group were born in the original village where the native dialect was dominant. As a result, the likeli¬hood of using the original vernacular forms should be higher among the old age
group. In addition, the concern with prestige should to be greater among mem¬bers of the old age group than it is among members of the middle age group, since there is greater pressure on the old.
The young do not preserve the native [q] variant, as is shown by its use in 6 percent of occurrences among young males and 3 percent among young females. The young, in general, try to follow the patterns that they believe have the social meanings they seek. The fact that there is less social pressure on the young than on the old in Jordanian societies (Al-Wer 1991) also helps the young to lead in sound change, since such innovative behavior patterns could not pass without much criticism by the community.
6.2. [g]-adopters. The [g] variant is used in 94 percent of the total number of occurrences of the variable (Q) in the speech of young males. It is the most sali¬ent feature that carries a social meaning of locality and serves as a symbol of Jordanian identity (Al-Wer 1991:75).
The [q] variant is always perceived as the Standard Arabic variant by non-Tirawis. Many studies in the Arab world (e.g., Abdel Jawad 1981:209; Daher 1998:194; Amara, Spolsky, and Tushyeh 1999:69; Walters 1991:211) demon¬strate the existence of a direct link between level of education and the use of [q]. Among non-Tirawis, [q] is always used by those who wish to present themselves as educated. As a result, its use by Tirawis might not be understood in the manner intended and might lead them to be ridiculed, especially if it is used in a domain that is not suitable for the Standard Arabic [q], such as the marketplace or one's local neighborhood. The [q] of the Tirawi dialect thus would be a marked variant in most contexts. In addition, in Irbid the Tirawi [q] is readily identified with the code of al-Tira, which is stigmatized in the Irbid social context as it is the code of immigrant rural people. Thus, the use of [q] by Tirawis in Irbid does not identify them with the prestige of Standard Arabic, as the [q] is here typi¬cally associated with Tirawis. As a result, the young avoid it. The [q] variant is also avoided on the grounds that it is the variant of rural groups in other studies in the area (Daher 1998). According to Daher, in Damascus "[q] is associated with a rural background, lacking in what is perceived to be the superior status of city life" (1998:198).
This stigma might also occur as a result of the use of [q] by Tirawis in words that do not favor the [q] variant (see Abdel-Jawad I98I; Al Khatib 1988), since these words are classified as neither Standard nor nearly Standard Arabic, but rather as purely colloquial. This makes the use of [q] in such words seem odd. According to Abdel-Jawad, "the lexical status of the word in which the variable occurs is very significant to the application of the (Q) standardization rule with standard words favoring it most, while colloquial words heavily disfavor the rule" (1981:209). Some sort of stigma would result if, in speaking to their local peers, young Tirawis showed the same readiness as older Tirawis to use the [q] variant with colloquial items, such as qabaday 'courageous' (basically a Turkish
loanword borrowed from a language other than Standard Arabic [Al Khatib 1988:101]), or an item like qahwaji 'coffee maker', which is classified in by Al Khatib, as a colloquial item "where parts of the form are borrowed, but where the meaning is native" (1988:101).
Young males typically used the local variant [g]. As the most salient feature with a social meaning of locality, it is not unexpected that Tirawis should adopt it, since, as Milroy and Milroy claim, males in general seem "to favor more localized variants, which carry some kind of identity-based social meaning in the local community" (1997:55). For these Tirawis to live a socially normal life, they have to integrate with the local groups, including in their linguistic behavior. Consequently, the powerful local variant [g] is adopted and the nonlocal variant [q] is abandoned. According to Myers-Scotton, "a major motivation for using one variety rather than another as a medium of an interaction is the extent to which this choice minimizes costs and maximizes rewards for the speaker" (1995:100). The cost and reward principle goes beyond assessing things materially to assessing them psychologically.
6.3. [?]-adopters. Young Tirawi females used the [?] variant in 95 percent of the total number of occurrences of the (Q) variable in their speech. It is also used in 94 percent of occurrences for the middle age group. SPSS analysis shows that sex has a very powerful significance in the use of the [?] variant (the threshold for significance is p < 0.05, while p < 0.00 for sex in the use of the [?] variant). It also shows that the interaction between sex and age is significant in the use of [?] (p < 0.029 with regard to the interaction between sex and age). Figure 3 shows that females (2.00) use the [?] variant more than males (1.00).
It is normal for females to adopt a prestigious urban variant, as they tend to use the code identified as the code of prestige. Given that the [g] and [?] variants are widely known to indicate toughness and softness, respectively, it is not unexpected for the [g] variant to be adopted by males and the [?] variant to be adopted by females. Because of this, Tirawi females do not adopt the [g] variant as their male counterparts do.
7. Conclusion. Given the above findings, I can conclude that the key feature [q] of the dialect of al-Tira is preserved only by the old. The younger generations have started to adopt new variants as an alternative to the native variant. Men tend to adopt the local variant [g], while women tend to adopt the nonlocal urban variant [?]. This choice depends to a great extent on what the society expects from the two sexes. The masculine stereotype demands toughness and strength. Thus, men choose the variant that embodies this toughness—i.e., the local vari¬ant [g], which was once the symbol of rurality and the harshness of Bedouin life.
Women are expected socially to reflect softness and urbanization. Thus, they have adopted the urban variant [?], which best reflects these features. In addi¬tion, it should be clarified that, although women from al-Tira only adopted the urban variant [?] after they migrated to Jordan, they comprehended its social meaning even when they lived in al-Tira itself, as it is a feature of the dialect of the major urban center to which their village was attached and with which they had the strongest contacts. In other words, the people of al-Tira perceived the urban dialect as a dialect of prestige because it was related first to Haifa.
Following their migration to Jordan, the new status of Tirawis as Tirat Haifans created a new psychological framework through which the urban dialect became an acceptable model. Hence, Haifa should be considered to have a signi¬ficant role in the spreading of an urbanized dialect in the region—a role that stands apart from any other city.
Thus, the native variant [q] of al-Tira has been abandoned by both sexes in favor of the two variants [g] and [?]. More generally, when the [q] variant is used by educated people whose native dialect has [q], as in the case of the Tirawis, its use may be an attempt to standardize their speech to conform with Standard Arabic, in line with the behavior of most educated people. However, it may also be an attempt to make their speech conform to the original dialect of their village, as spoken by uneducated people, to demonstrate their membership within the specific community that has the [q] variant.
Transcription. The most noteworthy symbols used in transcribing Arabic words given in this article are ? glottal stop, q voiceless uvular stop, g voiced velar stop, T emphatic voiceless alveolar (or dentialveolar) stop, D emphatic voiced alveolar (or dentialveolar) stop, tt voiceless interdental fricative, G voiced alveolar (or palatoalveolar) affricate, H voiceless pharyngeal fricative, and 9 voiced pharyngeal fricative.
1981 Lexical and Phonological Variation in Spoken Arabic in Amman. Ph.D.diss., University of Pennsylvania.
1985 Sociostylistic Stratification of Arabic in Makkah. Ph.D. diss., University of Michigan.
Al Khatib, Mahmoud
1988 Sociolinguistic Change in an Expanding Urban Context: A Case Study ofIrbid City, Jordan. Ph.D. diss., University of Durham.
1991 Phonological Variation in the Speech of Women from Three Urban Areas inJordan. Ph.D. diss., University of Essex.
1999 Why Do Different Variables Behave Differently? Data from Arabic. In Language and Society in the Middle East and North Africa: Studies in Variation and Identity, Yasir Suleiman, 38-57. Surrey: Curzon Press.
2000 Jordanian and Palestinian Dialects in Contact: Vowel Raising in Amman. Essex Research Report in Linguistics 18:26-49.
2001 Al Taghayyir al-Taarikhi lil-'aswaaT fi l-'arabiyya w-al-lughaat as-
saamiyya (diraasa muqaaranah). Ph.D. diss.', University of Jordan.
Amara, Muhammad, Bernard Spolsky, and Hana Tushyeh
1999 Sociolinguistics of Socio-Political Patterns in Bethlehem: PreliminaryStudies. In Language and Society in the Middle East and North Africa, Studies in Variation and Identity, edited by Yasir Suleiman, 58-80. Surrey: Curzon Press.
1986 Sex Differences in the Approximation to Standard Arabic: A Case Study.Anthropological Linguistics 28(l):3-9.
Blom, Jan-Petter, and John J. Gumperz
1972 Social Meaning in Linguistic Structures: Code-switching in Norway. InDirections in Sociolinguistics, edited by John J. Gumperz and Dell H. Hymes, 407-34- New York: Oxford Press.
1998 Gender in Linguistic Variation: The Variable (Q) in Damascus Arabic. InPerspectives on Arabic Linguistics XI, edited by Elabbas Benmamoun, Mushira Eid, and Niloofar Haeri, 183-205. Amsterdam: John Benjamins.
1997 Age as a Sociolinguistic Variable. In The Handbook of Sociolinguistics,edited by Florian Coulmas, 151-68. Oxford: Blackwell.
El Salman, Mahmoud
1991 Tirat Haifa between 1900-1948. Irbid: Qudsiyyih Press.
1988 Style and Style Shifting in the Speech of Educated Speakers of Jordanian Arabic. Ph.D. diss., University of Wisconsin-Madison.
Khalidi, Walid, ed.
1992 All That Remains: The Palestinian Villages Occupied and Depopulated by Israel in 1948. Washington, D.C.: Institute of Palestinian Studies.
1966 The Social Stratification of English in New York City. Washington, D.C.:Center for Applied Linguistics.
1972a The Study of Language in Its Social Context. In Sociolinguistics: Selected Reading, edited by J. B. Pride and Janet Holmes, 180-201. Middlesex: Penguin.
1972b Sociolinguistic Patterns. Philadelphia: University of Pennsylvania Press.
2001 Principle of Linguistic Change. Vol. 2: Social Factors. Oxford: Blackwell.
1997 Aspects of Language Variation in Arabic Political Speech-Making. Richmond, Surrey: T. J. Press International.
Milroy, James, and Leslie Milroy
1978 Belfast: Change and Variation in an Urban Vernacular. In Sociolinguistic Patterns in British English, edited by Peter Trudgill, 19-36. London: Arnold. 1997 Varieties and Variation. In The Handbook of Sociolinguistics, edited by Florian Coulmas, 47-64- Oxford: Blackwell.
1995 Social Motivations for Codeswitching: Evidence from Africa. Oxford: Clarendon Press.
1974 The Social Differentiation of English in Norwich. Cambridge: Cambridge University Press.
1997 Dialect in Society. In The Handbook of Sociolinguistics, edited by Florian Coulmas, 107-26. Oxford: Blackwell. Walters, Keith
1991 Women, Men, and Linguistic Variation in the Arab World. In Perspectives on Arabic Linguistics III, edited by Bernard Comrie and Mushira Eid, 199-229. Amsterdam: John Benjamins.