Introduction

Saudi Arabia’s recent economic and infrastructural reforms have placed a great emphasis on the important role English plays in attracting foreign tourists and international investors. In the eastern part of the country, the abundant presence of oil companies and the establishment of Aramco, a large petroleum and natural gas company, have added an extra layer to the complexity of English use in the region (Elyas et al., 2021). The growth of this company has led to a substantial increase in the number of Saudis in the region who learn and use English to communicate with the large number of foreign factory workers who hail from India, the Philippines, and Pakistan. Due to foreign language policies and the role of English as a lingua franca in this region, the increased demand for English in the workplace raises questions about what roles regional varieties of English play not only in the workforce, but also in English language instruction. In particular, as more Saudis learn English, it is not clear what role Saudi English (SauE) (Al-Rawi, 2012; Mahboob & Elyas, 2014) should play in the classroom. SauE that has only recently been added to the World Englishes (WEs) canon, and there are no existing studies that explore speakers’ attitudes toward this variety, or towards other Englishes that are widely used within the region. It is important to understand how Saudi learners view these varieties in order to consider the implications for teacher education, teaching materials, and assessment.

Attitude studies toward WEs have been the front face of sociolinguistic research, and attitudes toward different varieties of English are well established in the literature. Research has been conducted on attitudes toward Englishes in inner-circle (IC) countries such as the United States, Australia, and New Zealand (Bayard et al., 2001); in outer-circle (OC) countries such as Malaysia (Ng & Diskin-Holdaway, 2021) and India (Jenkins, 2007); and in expanding-circle (EC) countries like Japan (Sasayama, 2013), and Oman (Buckingham, 2015). That said, to date, there has not been any study that investigated Saudis’ attitudes toward their own variety. Therefore, the current study aims at investigating how Saudis perceive their own variety of English, with the intention to inform English language instruction in Saudi schools. Furthermore, the study offers an innovative contribution to the study of language attitudes and WEs by using regionally-relevant Englishes in the design of the study, which is a departure from existing approaches to attitudes that compare local Englishes with IC varieties.

Language Attitude in EC Contexts

EC countries like Saudi Arabia are classified as norm dependent and English is generally treated as a foreign language in Kachru’s well-known Three Circles model (Kachru, 1985). This model has been heavily criticized by Park and Wee (2009) since it is a model that “perpetuates the very inequalities and dichotomies that it otherwise aims to combat” (p. 390), and because its concentric nature does not take into account the ever-changing social complexities surrounding English use in EC contexts like Egypt, as Park and Wee (2009) assert. Saudi Arabia is arguably no different, as English is expanding into all domains (i.e., linguistic landscapes, educational settings, public and private work settings), which lends support to Park and Wee’s critique of this model to encompass and reflect the rapid social changes associated with English use in Saudi Arabia (for other criticisms of this model refer to the work of Bruthiaux, 2003; Higgins, 2003; Mufwene, 2001). Nonetheless, I situate my examination of attitudes toward SauE within research on “expanding circle” Englishes in order to provide a rationale for the study of language attitudes in Saudi Arabia and to assert a need to regionalize language attitude studies, and thereby decenter the inner circle varieties in this line of research.

Innovating Attitude Research in WE: Toward Decentering the IC

In contexts where English is neither institutionalized nor a legacy of colonialism, language attitude research has often used Verbal Guise Tests (VGTs) to explore attitudes about local Englishes, and these are typically carried out in comparison to measuring attitudes toward inner circle varieties (e.g., McKenzie, 2008; Sasayama, 2013). There are an increasing number of studies that invite research participants to respond to regionally relevant varieties from the EC as well (e.g., Buckingham, 2015; McKenzie & Gilmore, 2015; McKenzie et al., 2016; Meer et al., 2021; Yook & Lindemann, 2013); however, these studies have demonstrated that participants favor IC varieties as models for learning. When IC varieties are included as part of the study, solidarity measures show that participants tend to view their own local Englishes more positively, while power measures reveal that IC Englishes used by educated speakers from the USA and the UK remain idealized varieties for institutional and international purposes.

A common theme among these attitude studies is that the regional variety is always compared to an IC, standard variety, which is also a dominant model in English language teaching in that region. To illustrate, Sasayama (2013) investigated the attitudes of 44 Japanese students toward Japanese English and American English using a VGT and a closed-ended attitude questionnaire. The author was interested in how Japanese students would rate Japanese English compared to American English in terms of solidarity, power, personal preference, and international acceptability. The results of the VGT demonstrated that Japanese students expressed more solidarity for their regional variety of English, whereas they attributed power traits to American English. The attitude questionnaire showed conflicting identities where some Japanese participants wanted to sound like Americans when they speak, but also did not want to be identified as Japanese when they speak English. This pattern of idealizing IC varieties was also detected in one of the Gulf countries that shares borders with Saudi Arabia. Buckingham (2015) investigated the attitudes of 349 Omani EFL students toward the accent of eight different male and female speakers, including American, British, Omani, and other speakers of Asian Englishes such as Indian, Filipino, and Bangladeshi Englishes. The results of the VGT showed that Omanis considered American and British accents models of correct pronunciation. This is not unexpected, given the prevalence of these Englishes in ELT materials and tests.

Several more recent studies include other EC varieties as well, thus providing respondents with the opportunity to consider a wider array of possibilities for acceptable Englishes. However, due to the continued hegemony of IC, ‘native’ Englishes in education and in the media, it becomes difficult to effectively measure attitudes toward local varieties in the EC if these IC varieties are always included. In other words, respondents are likely to reproduce hegemonic ideologies about which Englishes are acceptable due to the recurring language ideological messaging about what constitutes ‘good’ English (Phillipson, 1992; Tupas, 2015). Another problem is that a comparison to an IC variety such as American English or British English, including the many varieties subsumed under these broad terms such as Received Pronunciation (RP), Estuary English, as well as Mainstream US English (MUSE), is problematic in the sense that it often does not reflect the social realities of English use in contexts like Saudi Arabia. The number of American or British nationals is not comparable to Indians, Filipinos, and Pakistanis who constitute a large proportionFootnote 1 of the labor force in Saudi Arabia and other Gulf countries. Saudis generally use English to communicate with such workers in places like restaurants, factories, and shopping malls (Almegren, 2018). Therefore, rather than include an IC variety, the present study attempts to decenter IC varieties in language attitude research through its design. Participants were invited to respond to an audio stimulus featuring one of the Englishes that is part of the ecology of Saudi Arabia, Indian English (IE), alongside SauE, in an attempt to more realistically depict the use of English use in the present cultural context.

Another common theme in the majority of attitude studies is the use of indirect approaches such as a Matched Guise Test (MGT) or a VGT that uses recordings of scripted texts being read aloud by speakers from different regions. Garrett (2010) explains that language attitude research has mainly utilized three approaches to measure attitude. One is the social approach where ethnography, discourse analysis, and interview methods are used to investigate the social, political, and identarian ideologies that underly attitude (Higgins, 2003). The second approach is called the direct approach where explicit attitudes are elicited via self-reported questionnaires (Sasayama, 2013). The third approach is the indirect approach manifested in a MGTs or VGTs where a speaker or several speakers are recorded while reading a prepared text and listeners are invited to rate those speakers using a Semantic Differential Scale (SDS) for example. Understandably, the purpose of using these scripted recordings is to control for variables such as speech rate and linguistic complexity (Ball & Giles, 1982; Garrett, 2010). However, McKenzie (2008) noted that this inauthentic speech leads us to neglect several social and linguistic cues that may only unfold in spontaneous speech, such as the geographical location of the speaker or the complexity of the language being used. Similarly, in his discussion of the inherited pitfalls in the vocal presentations of language varieties, Garrett (2010) highlights the issue of “style-authenticity” in indirect measure of attitude research and states that “It may not be wise to assume that more spontaneous speech will be evaluated in the same way” (p. 59). Therefore, the present study adopted an innovative approach by using an Interactive Verbal Guise Technique (IVGT) where listeners were asked to evaluate Englishes that were used as part of a natural (spontaneous) conversation.

Saudi English (SauE)

Before turning to the study, it will be helpful to provide a description of Saudi English (SauE) based on the small body of descriptive literature on this variety. SauE is based on the linguistic features that are unique to SauE users and which are attributable to the influence of linguistic features of Standard Arabic such as phonology, syntax, morphosyntax, and Lexicon to name a few (please refer to Al-Rawi et al., 2022 for a thorough description of SauE). Al-Shurafa (2009) was among the first to analyze the syntactic features found in the speech of educated Arab speakers from different Gulf countries, including Saudi Arabia, Tunisia, and Morocco. Her analysis led her to establish what she called “Arabicised-English”. The analysis showed that Arabs tend to overuse the modifier very, the first-person pronoun, and the connector and, which both reflect some aspects of the Arabic language and culture. Interestingly, these linguistic features were also present in IE, as the author noted. Fussell (2011) focused more on what he called “Gulf English” by examining the syntactic, lexical, and phonetic features found in the English speech of Gulf citizens, including Saudi Arabia. He also noticed some linguistic features similar to those found in IE such as replacing unaspirated /d/ for word-initial /t/ and the overuse of the modal would instead of will to reference the future tense. Fussell attributed the similarities between Gulf English and IE to the linguistic influence caused by the abundant presence of Indian expatriates in Gulf countries. Al-Rawi (2012) and Mahboob and Elyas (2014) further established SauE by analyzing university and high school students’ speech as well as their English language textbooks. They noted that SauE features include deletion of the indefinite articles a,an and more use of the definite article the, more variable use of tense markers such as use of the perfect tense for the past tense, and replacing /p/ with /b/ and /v/ with /f/ due to the non-existence of /p/ and /v/ sounds in the Arabic language. More recently, Al-Rawi et al. (2022) described the morphosyntactic and lexical features of SauE that are found in the speech of educated Saudi English speakers. They found that among the most frequent occurrences, speakers use ‘me’ instead of ‘I’ in coordinated subjects and have issues with subject-verb agreement especially in the present simple tense.

Saudis’ Attitudes Toward WEs

Several studies have researched Saudis’ attitude toward different WEs, but these have excluded SauE as a relevant variety. It is important to understand perceptions of SauE as a part of establishing whether and to what degree an endonorm is developing that could serve as a basis for language learning, language teaching, and language use in professional contexts. Moreover, positive attitudes towards SauE would indicate some degree of ownership over English in the Gulf, which could in turn be beneficial for building towards forms of ELT in which English learners see themselves reflected in their teachers, textbooks, and target model speakers. This was evident in a study conducted by Buckingham (2014) in the Gulf territory among 347 Omani students where their attitudes toward the accents of five native and non-native English teachers were investigated. Not only did students favor the accents of the native English teachers, speaking the UK variety of English, but they also held positive attitudes toward Arabic accents spoken by their Omani and Syrian teachers. Students also aspired to sound like a native English speaker, but according to the author that did not intervene with how they highly rated the accents of the Arabic teachers. The author attributed these interesting findings to the possibility of Omani students showing solidarity with the accents spoken by their Arab teachers and viewing them as models of successful learning. Similar findings were found in Zoghbor (2014) where the author explored the attitudes of Emirati EFL students toward six native and nonnative English speakers. Even though students aspired to speak like the British and Canadian speakers, they have expressed positive views toward the Arab Jordanian and Indian speakers and considered their English far more comprehensible than the American speaker. They also felt it would be easier to communicate with the Arab and Indian speakers compared to the other native speakers. The author explained that positive attitudes toward the Arab speaker might be due to participants’ familiarity with a speaker from their own country and someone who would represent a good model of English use. Participants were also familiar with Indian English because they were accustomed to hearing it in local shops and various outlets in the United Arab of Emirates (UAE). It was concluded that ownership of English should influence pedagogical decisions in EFL learning and teaching in the Gulf context.

Being the most populated Gulf state and the sixth most populated country in the Middle East with a rabidly growing EFL population, Saudi Arabia has gained its own variety of English based on a small body of descriptive literature (Al-Rawi et al., 2022; Al-Rawi, 2012; Al-Shurafa, 2009; Alwazna, 2020; Bukhari, 2019; Mahboob & Elyas, 2014; Elyas et al., 2021). Attitudes toward this variety, however, remain unexplored as the existing studies on language attitudes on English in Saudi Arabia have been mostly exonormative in design. Al-Dosari (2011) investigated Saudi EFL students’ perception of two teachers speaking IE and South African English (SAE), using a VGT. The author chose these two Englishes because most foreign teachers at this context speak one of these two varieties. His findings showed that Saudis rated the SAE higher in comprehensibility, but both IE and SAE were rated equally in terms of perceived accents. The author mentioned that even though the Indian speaker was more accented than the South African speaker, participants still perceived the Indian accent positively due to their constant contact with Indians in language schools and households.

Similarly, Almegren’s (2018) study included Indian and Filipino Englishes, and compared attitudes towards these with attitudes toward IC varieties of English, such as British and Australian English. Almegren chose Indian and Filipino Englishes because these two ethnic minorities constitute a huge part of Saudi Arabia’s labor class, which reflects the importance of grounding attitude studies in more sociolinguistically authentic selections. Nonetheless, Saudis held more positive views toward the IC varieties, and interestingly, even though the study did not ask them about SauE, some of them expressed that they wanted to be taught by a Saudi teacher because they shared a similar background.

The Present Study

As seen from previous literature, most attitude studies included an IC variety alongside a regional variety which could possibly influence listeners’ attitudes to regional varieties. Also, no previous attitude study has examined Saudis’ attitudes toward their own variety. Therefore, the purpose of this study was to shed light on this gap in the language attitude research in order to examine to what degree endormative perspectives are present for the use of locally-relevant varieties of English in ELT and professional contexts in the region. As Elyas et al. (2021) noted, introducing SauE to the language curricula in public education will “lead to a more culturally, politically, and linguistically relevant model compared to the English as a native language (ENL) model, which is not practiced on a routine basis in Saudi Arabia” (p. 228). By studying responses to an IVGT involving SauE and IE, the design of the study attempts to avoid the pitfalls of including IC varieties. To assess Saudi participants’ attitudes with regard to power and solidarity, participants were invited to listen to an interaction in which SauE is used alongside a less hegemonic variety, IE. The present study thus sought answers to the following research questions:

  1. 1.

    How do Saudi university students rate a SauE speaker taking part in a conversation with an IE speaker in terms of solidarity and power?

  2. 2.

    How do Saudi university students perceive SauE in terms of preference and acceptability?

  3. 3.

    What are Saudi university students’ general attitudes toward SauE?

Method

Participants

A total number of 80 Saudis studying English at an engineering college located in the Eastern province of Saudi Arabia participated in this study. They study English in the preparatory year as a prerequisite to be admitted into a designated engineering program. Upon enrollment, students get assigned to one of four Common European Framework of Reference (CEFR) proficiency levels (A1, A2, B1, B2) based on their scores on The Cambridge English Placement Test. The participants of this study are placed in the B1 level and, therefore, have an intermediate English proficiency. Their ages range between 19–23, and most of them are recent high school graduates. All participants were males, given institutional restrictions on gender-segregated education in this specific context. Once they pass the English preparatory year, students proceed to their desired programs where they are expected to become engineers after graduation and work in oil companies or industrial factories that are ubiquitous in the Eastern side of Saudi Arabia.

Instruments

The Interactive Verbal Guise Test (IVGT)

In designing the IVGT, I searched for an existing recording featuring SauE in order to start with data that contains spontaneous speech. This kind of data was important for the task because it depicts what listeners would actually encounter in real life settings where speech is unscripted. I used an audio-only recording of an unscripted interview between a male Indian English speaker and a female Saudi English speaker. The recording was excerpted from an interview conducted by the Indian speaker, Sidhant Sibal, who is a middle-aged news reporter, working for a major Indian news channel called ‘World is One’ (WION). His speech contained features that have been documented in other studies of IE, particularly at the phonological level (Sailaja, 2012; Wiltshire, 2020) such as replacing dentals like /θ/ with dental plosives ⁄ t̪ ⁄. The Saudi speaker, Nouf Marwaai, is a middle-aged woman in her 40 s, and she is a yoga instructor. According to an article published in Arab News (2018), Nouf completed her graduate studies in India, earning a master’s degree in psychotherapy and then becoming the founder of Arab Yoga Foundation. Her English has characteristics of SauE such as unaspirated /d/ for word-initial /t/ (Fussell, 2011). The interview took place in 2018 and lasted for approximately 21 min on the topic of Yoga practices in Saudi Arabia. Since this was a prerecorded natural conversation, it was not possible to control speech-related variables. However, I tried as much as possible to ensure that (a) both speakers are nearly the same age and hold similar social status to slightly control solidarity and power judgements; (b) both speak at a similar rate (approximately 24 words per 10 s) as speech rate has an influence on listener’s judgement (Giles et al., 1990); and (c) that both speakers’ produced few fillers or pauses that could negatively influence the evaluation of listeners (Street & Hopper, 1982).

After analyzing the 21-min interview, I selected an 83 s excerpt of the interview in audio-only format as the IVGT prompt. I eliminated the visual element of the video because I wanted listeners to evaluate the Englishes spoken by the two speakers, not their visual appearances. To rate the speaker(s), VGTs are usually associated with Semantic Differential Scales (SDS), which utilize several sets of polar adjectives that denote power and solidarity traits, and which listeners can score using a Likert scale. In the present study, I adapted Sasayama’s (2013) SDS with slight changes in selecting the solidarity and power traits. In her study, Sasayama mentioned that she included traits of solidarity and power that Japanese speakers would use if they were asked to describe a person. Also, Zhang (2010) claimed that the use of traits to describe a particular construct could differ across cultures. Therefore, drawing on my insider knowledge as a Saudi, I adapted 10 different traits that I felt could be appropriate for the Saudi culture from different studies in the literature (Almegren, 2018; Buckingham, 2014; Cavallaro & Chin, 2009; Li & He, 2009; McKenzie, 2008). For solidarity I chose kind, modest, likeable, honest, interesting and for power I chose rich, confident, fluent, educated, successful. These traits were also shared with two experienced EFL teachers that work in the research site and speak Arabic as their L1 to evaluate them in terms of appropriateness and clarity to Saudi speakers. They both agreed that all traits are comprehensible and commonly used to describe a person in the Saudi culture. For each of these adjectives I provided an antonym to include at the opposite side of the SDS. Participants were asked to evaluate the speakers in the SDS using a 7-point Likert-scale where one is the least favorable rating and seven is the most favorable rating (Cavallaro & Chin, 2009; Garrett, 2010). The SDS was translated into Arabic to account for students who might find it challenging to understand some English adjectives or misinterpret the instructions. Both English and Arabic versions of the SDS are presented in “Appendix 1”, alongside the instructions.

Attitude Questionnaire

Alongside the indirect measure and to increase the reliability of my method, I included an attitude questionnaire (see “Appendix 2”) that was adapted from Sasayama (2013). This instrument is a closed-ended five-item questionnaire that taps into participants’ preference and acceptability of Saudi English. For example, item 1 (When I speak English, I want to sound like a native speaker) represents personal preference. On the other hand, item 5 (In international communication, the Saudi variety of English should be accepted as long as it is intelligible) taps into participants’ ratings of the acceptability of Saudi English. It is worth noting that Sasayama (2013) used the words “American person” when referring to native speakers. However, in devising the questionnaire for the present study, I replaced this term with a more general term “almutahadeth alasli المتحدث الاصلي”, which means the ‘native speaker’ as shown in item 1 above, to account for any IC variety. This questionnaire was scored using a five-point Likert scale with five possible options (strongly disagree, disagree, neutral, agree, strongly agree). In addition to the closed-ended questionnaire, and different from Sasayama’s design, I included six open-ended questions that were adapted from (Li & He, 2009; Ng & Diskin-Holdaway, 2021) to gain deeper insights into participants’ attitudes toward SauE. As with the SDS, both parts of the questionnaire were translated into Arabic to ensure that participants understand the items and to encourage students to be more expressive of their attitudes when writing their responses to the open-ended part of the questionnaire in Arabic. It is worth noting that throughout the translation process, I assumed that participants had not heard about the term “Saudi English” before, and therefore I provided a simplistic explanation of this term in Arabic, noting that it is the English spoken by Saudi speakers including its distinct accent and grammatical structures. The translation was confirmed by another Arabic-English bilingual language teacher to ensure its accuracy. The Arabic and English versions of the questionnaire are presented in “Appendix 2” along with the instructions.

Procedures

After obtaining informed consent, the IVGT was carried out with 80 participants during class time. The IVGT and the questionnaire were integrated into Google Forms, and the whole procedure was conducted online using Blackboard, as classes in the research site were conducted remotely due to COVID-19. The teacher in that class, who worked as a research assistant, provided the instructions to the students, first spending a few minutes familiarizing students with the study and the type of information they were asked to provide before any actual data collection began. After that, he sent the Google forms link to students’ emails and asked them to fill out the consent form and indicate their voluntary participation in the study. Once they had indicated their consent, the teacher played the audio for the first time and asked students to pay attention to the female speaker and rate her speech using the scales displayed on their screen. Once they were done rating the female speaker, the teacher played the audio for a second time and students were prompted to proceed to the next page in the Google form and rate the male speaker using another blank SDS (due to space, this data is not reported in this study). After rating these two speakers, participants were instructed to fill out the two-part questionnaire found in “Appendix 2”. I asked the teacher to play the recording himself to avoid any technical issues with embedding the audio in the Google Form and to regulate the process of listening to the audio twice before moving on to the questionnaire part. The three tasks took approximately 30 min to complete.

Results

Descriptive analysis (mean and SD), inferential analysis (a one-sample t test), and Cronbach’s alpha test of reliability for all Likert-scale instruments were calculated using Statistical Package for the Social Sciences (SPSS) version 20. A one-sample t test of means was used to compare the mean of traits on the SDS rating scale and the items of the attitude questionnaire to a pre-specified value and to test the deviation from that value. This value was determined to be 4 for the SDS ratings and 3 for the attitude questionnaire (i.e., neutral values). Meaning that these values were chosen as a test for judging the degree of response; that is if the response averages of the study sample members was less than 4 in the SDS ratings and less than 3 in the attitude questionnaire and the value of the one-sample t test was statistically significant, this indicates low ratings on the SDS and tendency to reject the meaning of the statement in the attitude questionnaire. However, if the average of ratings is above 4 and the averages of responses to the attitude questionnaire is greater than 3 and the value of the one-sample t test is statistically significant, this indicates higher ratings on SDS and tendency to accept the meaning of the statement in the attitude questionnaire. The written responses on the open-ended questions were analyzed thematically to detect saliant patterns in the data.

IVGT

The first research question aimed at exploring how Saudi students would rate a SauE speaker taking part in a conversation in terms of solidarity and power. Descriptive and inferential statistics of participants’ evaluation of the Saudi speaker in terms of solidarity and power are calculated and presented in Table 1. The solidarity and power subscales of the SDS demonstrated an acceptable Cronbach’s alpha (α = .68, .72), respectively, and the scale was considered reliable.

Table 1 Descriptive and inferential statistics for SDS responses
Full size table

Descriptive analysis has shown that the Saudi speaker was rated highly and above the midpoint of the scale for all 10 traits. Both confident and educated in the power scale constituted the highest mean scores (M = 5.60, SD = 1.42) and (M = 5.63, SD = 1.30) respectively. All other traits ranged between (M = 5.46, SD = 1.42) for honest and (M = 4.79, SD = 1.69) for modest in the solidarity subscale.

Further inferential analysis using a one-sample t test has shown that the mean differences of all traits were statistically significant at a level of significance less than (.05), except for interesting in the solidarity subscale (t(79) = 0.73, p > .05 (p = .47)). The t test values for the remaining 9 items ranged between (t(79) = 4.17, p < .05 (p = .00)) for the trait modest on the solidarity subscale and (t(79) = 11.21, p < .05 (p = .00)) for educated on the power scale.

Overall, the results of the SDS showed that Saudi participants highly attributed solidarity and power traits to the Saudi female speaker in the clip and that the mean differences for 9 out of the 10 traits were statistically significant (p < .05).

Attitude Questionnaire

The second research question was related to Saudi participants’ attitude towards SauE in terms of preference and acceptability as measured by a five-item attitude questionnaire. Participants’ responses to the five Likert-scale options assigned to each item were converted to percentages and presented in Table 2 alongside the descriptive and inferential statistics of each item. The reliability of the items on the personal preference subscale (items 1, 2) was moderate (α = .58) and acceptable (α = .76) on the international acceptability subscale (items 3, 4, 5).

Table 2 Descriptive and inferential statistics for attitude questionnaire
Full size table

Looking at (item 1) of the questionnaire, 78% of the participants wanted to sound like a native speaker when they speak English. The remainder were either uncertain (12.5%) or did not aspire to sound like a native speaker of English (4%). Conflicting responses were shown on (item 2) with the highest standard deviation among all five items (SD = 1.49), where nearly half of the participants (36%) said that they envy those who can pronounce English like a native speaker while the other half (44%) disagreed with this statement. The highest disagreement percentage (79%) is observed in (item 3) where the majority of the participants stated that they do not feel embarrassed when a Saudi speaker takes part in a fluent conversation with other speakers using SauE. A similar rate was found in (item 4) as (67.5%) of the participants also disagreed with this item stating that they do not feel embarrassed when SauE is used in international activities. The last item showed that (71%) of participants considered SauE to be an acceptable variety of English in international communication as long as it is intelligible.

Inferential analysis of the mean differences of all items were statistically significant at a level of significance less than (.05), except for item 2 (t(79) = 1.47, p > .05 (p = .16)). The t test values for the remaining 4 items ranged between (t (79) = 6.73, p < .05 (p = .00)) for item 4 and (t(79) = 9.86, p < .05 (p = .00)) for item 3.

Overall, participants mostly showed a positive attitude toward the use of SauE in terms of personal preference and international acceptability.

Open-Ended Responses

In addition to the SDS and the closed-ended part of the attitude questionnaire, I invited participants to provide responses to open-ended questions to gain deeper insights into participants’ attitudes toward SauE. This also helped me to know whether they could identify the nationality of the SauE speaker, and to broaden my understanding of what Englishes they had regular exposure to. I was also interested in how they might respond to more metalinguistic questions about Saudi English. To answer the third research question, I asked participants to provide written responses to the six open-ended questions in “Appendix 2” about the Saudi female speaker in the clip and about their perception of SauE in general.

As for the first three questions, respondents mostly mentioned Saudi Arabia, Britain, and America, among other countries as shown in percentages in Table 3. To ensure that participants were able to recognize the nationality of the female speaker, I asked them to identify the nationality of the female speaker in the clip. The data showed that the majority of participants (72%) identified the female speaker as Saudi, while 12% mentioned that she might be either American or British due to her high fluency. The other 16%, interestingly, wrote that she might be Indian, Bangladeshi, or Malaysian. By asking the second question, I was interested in knowing if students are exposed to other global Englishes in their language classrooms besides the usual American or British English, and how is being acquainted with different Englishes might influence their attitudes toward SauE. Unsurprisingly, American and British Englishes were the two dominant varieties of English (81%) usually heard in language classrooms. Nonetheless, 17% of participants reported that they also listened to SauE in their classes. The remaining 2% reported Indian, Filipino, and Egyptian Englishes as being other varieties featured in language classes. In question 3, I asked the participants about the type of English they prefer to listen to in their language classes. Almost half of the participants chose American English (52%), followed by British English (35%), while only 5% chose Saudi English. There were other interesting varieties mentioned by the participants such as Sudanese and Northern British.

Table 3 Percentage of responses to Questions 1, 2, 3
Full size table

Overall, most of the participants were able to identify the nationality of the Saudi female speaker, even though a few recognized her fluent English as being American or British. Also, it is evident that participants prefer IC varieties like American or British Englishes, but there is also acknowledgement of SauE and other Asian Englishes.

Questions four and five were related to participants’ evaluation of the Saudi female’s English and of their own accents, if they stated that they had one. For question four, 75 out of the 80 participants mentioned that her English is excellent and easy to understand. One participant wrote that her English is “excellent because she pronounces words clearly and does not stutter”. Another showed more affection and sympathy by stating that “her English is beautiful and pure and she tries her best”. The remaining five participants were neutral and rated her English as “not bad” or “too slow”.

The majority of students said that they had a Saudi accent when they speak English. Most of them perceived their accents as good and acceptable, but there were some conflicting opinions. One participant compared SauE to other Arabic accents and wrote that “I think that the Saudi accent is one of the most clear and understandable Arabic accents as it enables you to understand other speakers of similar accents and foreigners as well”. On the other hand, another participant claimed that SauE is not as intelligible as other Asian Englishes: “I think the Saudi accent is not as clear as the Indian or Pakistani accents”. Others have either wanted to learn how speak in American and British accents, or they stated that they need to work on their accents to sound more comprehensible.

Instances of ownership of SauE were noticed in participants’ responses to the sixth and last question. Higgins (2003) discussed the notion of ownership in relation to legitimacy by stating that for speakers of different Englishes, including those in the EC, “the determining factor in owning the English language is whether the speakers view the variety they use as being a legitimate variety in a social, political, and economic sense” (p. 621). Legitimacy, in my understanding of Higgins’ (2003) ideas, was expressed by one of the participants when he was asked ‘what do you think of the term Saudi English?’:

I think SauE is indispensable considering where we were born and raised, and I don’t think it’s shameful to speak English in our accent as many nonnative speakers of English has their own accents. Also as I have mentioned before, SauE is clear and easy to understand.

The participant oriented to other varieties of English as being as legitimate as the Saudi variety. He also referred to his cultural identity and that SauE should not be stigmatized because it defines who we are as Saudis. This was also evident in the responses of two other participant as one stated that having SauE is “a normal thing just like having Indian English”, and the other “It’s good to have a SauE because it says much about our identity”. Another participant legitimized the use of SauE by stating that “any country that speaks English has its own accent such as the differences between Arabic accents”. Interestingly, an incident of linguistic identity was observed in of one of the participants’ responses as he reflected on the influence of the fluent pronunciation of Arabic (Fusha, فصحى) on the way Saudis pronounce English words: “I think it’s a good thing to have SauE because the way we pronounce letters is clear due to the influence of our Arabic language which is ‘fusha’”. Five participants mentioned that SauE should not exist because English is not our mother tongue or that having the label ‘Saudi English’ is not necessary as long as we speak intelligible English. However, overall, the open-ended responses gave the participants an opportunity to express ownership and legitimacy of the existence of ‘Saudi English’, and their attitudes were predominately positive toward the term ‘Saudi English’.

Discussion

The first research question aimed at finding how Saudi EFL students would rate a Saudi female speaker taking part in an authentic conversation in terms of solidarity and power. Results from the IVGT have shown that Saudi participants rated the Saudi female speaker highly in both solidarity and power scales. These ratings were also statistically significant for all traits except interesting. These findings are in line with Sasayama’s (2013) study where she found that Japanese EFL students rated Japanese English positively on solidarity and power traits. Displaying affiliation with the regional variety was also observed with Omani students (Buckingham, 2014) and Emirati students (Zoghbor, 2014) in the Gulf region. However, the results of this study are contrary to those observed with Omani students in Buckingham (2015) where the female Omani speaker was rated the lowest among other speakers of inner and OC countries, including IE. It seems that Saudi male students, unexpectedly, are more positive toward hearing a female English voice and that gender is perhaps not a potential issue as it could with Omani students in Buckingham’s (2015) study or as was evident in Wilson and Bayard’s (1992) study of New Zealand English where female speakers were rated lower on all solidarity and power traits. Although Saudi EFL students at the specific context of this study might not interact regularly with female English teachers or even female workers outside of school, they are used to hearing female voices in their curriculum’s listening materials. Also, recent reforms in Saudi Arabia have largely targeted gender-related issues, which could have led to a change of perception to what used to be a controversial topic in the Saudi society and might have positively influenced participants’ attitude toward the female speaker as many Saudi women now occupy several job roles in the market. That said, these findings must be interpreted with caution due to the high fluency of the female speaker as was evident in participants’ comments on the open-ended survey and because there was no Saudi male comparison. Also, in Buckingham (2015), it was not only gender that might had played a factor as the female speaker had a stronger accent compared to other speakers.

In the second research question, I wanted to know participants’ attitude toward SauE by using a five-item closed-response questionnaire. Almost 78% of participants reported that they wanted to sound like a native speaker when they speak English. Similar results were found with Omani students (Buckingham, 2014, 2015), Emirati students (Zoghbor, 2014), Japanese students (Sasayama, 2013), and Chinese students (Li & He, 2009). This finding is not surprising due to the prevalence of IC varieties in EFL textbooks and the influence of western media (e.g., Hollywood movies) on youth’s perception of varieties like American English. Buckingham (2015) states that “In many EFL contexts such as the Gulf, British/US cultural production continues to dominate curricula, with, at most, token references to English language film and literature produced in African or Asian contexts” (p. 193). The aspiration of participants to speak like a native speaker, however, did not influence how highly they rated the Saudi speaker in this study or the Arab speakers in Buckingham (2014) and Zoghbor (2014). This could mean that although students perceive native varieties of English as an ideal model of pronunciation, they still view their own regional variety positively. Such finding raises doubts on the benefit of including IC varieties alongside local Englishes in attitude research, especially in regions such as the Gulf where IC varieties are less popular compared to other commonly spoken Englishes (i.e., Indian English or Filipino English). The highest disagreement percentage observed in the data (79%) was on item 3 as participants mentioned that they do not feel embarrassed when they see Saudi people speak the Saudi variety of English. This question might have provoked participants’ strong sense of national pride and affiliation with a speaker of their own, as was the case with Chinese students in Pan’s (2019) study.

Nonetheless, the findings of the self-report survey need to be interpreted with caution. A potential reason is related to questioning the effectiveness of eliciting self-report attitudes in the first place, especially as it compares to indirect measures such as a VGT in attitude research. Although it might be ‘methodologically’ appealing to compare direct with indirect measures (Meer et al., 2021; Ng & Diskin-Holdaway, 2021; Sasayama, 2013), Garrett (2010) makes the claim that people in attitude studies “operate with two value systems (or two sets of attitudes) alongside each other, while only being conscious of one of them” (p. 43). In other words, if asked explicitly, people tend provide an answer that they find appropriate or wish to be true, rather than giving an answer that represents their true feelings. Regardless of that, it seems that Saudis held a positive attitude toward SauE as long as others can perceive it as intelligible.

The third and last research question looked at Saudi students’ general attitude toward the existence of SauE and investigated how endonormative their attitudes are in this specific context. Participants were asked to provide written responses to six open-ended questions. In the first question, the majority of participants (72%) were able to identify the female speaker as Saudi. This could be due to the fact that this speaker was not strikingly regional in her speech and she did not speak with a heavy accent. One might argue that the content or the context of the interview, which was about ‘yoga’ in Saudi Arabia, might have hinted to her nationality. Although this might be true, 28% of participants still mentioned that she might be either American, British, or Indian. This indicates that the context was not entirely revealing as she never explicitly stated that she is Saudi. As for the second and third questions in the survey, most participants reported that they usually listen to and prefer American and British Englishes over any other variety. These findings coincide with most attitude studies that included an IC English variety (Almegren, 2018; Buckingham, 2014, 2015; Li & He, 2009; McKenzie, 2008; McKenzie & Gilmore, 2015; McKenzie et al., 2016; Meer et al., 2021; Yook & Lindemann, 2013, Zoghbor, 2014). The fourth question in the survey was intended to give participants the opportunity to evaluate the female speaker more openly compared to the closed-ended nature of the SDS. Nearly all participants stated that her English is excellent and considered her a good model of SauE. These findings allign with the positive attitudes that Omani students (Buckingham, 2014) and Emirati students (Zoghbor, 2014) showed toward their Arabic teachers. This could set premises to the inclusion of a comprehensible model of the regional variety, such as incorporating SauE in Saudi EFL textbooks as was advocated for in Oman (Buckingham, 2014) and UAE (Zoghbor, 2014). Elyas et al. (2021) recommended the integration of SauE in EFL education at Saudi Arabia to fulfill the needs of the Saudi speech community. The present study is the first to provide empirical evidence of Saudis’ feelings about their own variety, and it adds validity to Elyas et al.’s (2021) call for incorporating SauE into EFL curricula and other similar calls in the Gulf region (Buckingham, 2014; Zoghbor, 2014). Participants’ expressions of ownership and legitimacy of SauE in question six are also strong evidence that supports our endeavors to decenter IC varieties in our EFL textbooks and give room to a regional variety that is well-perceived by its speakers.

Conclusion

This study was the first study to investigate Saudi students’ attitude toward SauE. The IVGT was used as an innovative approach, opposed to traditional VGT methods, to account for the social and linguistic cues that may only unfold in authentic conversations, and as a means to avoid importing ideologies about IC varieties into the task. Irrespective of its popularity in measuring language attitudes, adopting a traditional VGT approach was heavily criticized in Liebscher & Dailey‐O'Cain (2009) because it separates attitudes from the language and its speakers in real-life settings. In other words, I argued that showing a Saudi speaker communicating with an Indian speaker allows for listeners to receive the data in a way that mirrors authentic interaction more and avoids an invitation to decontextualize the speakers from their interactional contexts (please refer to Liebscher & Dailey‐O'Cain, 2009 for a detailed explanation of measuring language attitudes in interaction). Following this approach, 80 Saudi male students were asked to rate a female Saudi speaker taking part in a conversation with a male Indian speaker using a SDS with ten traits of power and solidarity. The results of the SDS revealed that the participants highly rated the female speaker in both power and solidarity scales. This indirect approach was supplemented with a direct approach by asking participants five closed-ended questions about their attitudes toward SauE. Most participants considered SauE to be acceptable as long as it is understood by other speakers. The last set of data was collected via written responses to six different open-ended questions about participants’ perception of the existence of their own variety of English. The data elicited from these six questions showed various expressions of ownership and legitimacy of Saudi English.

Notwithstanding with the implications of this study, there are some limitations to be acknowledged. This is a relatively small-scale study compared to other robust attitude studies, and future researchers are encouraged to include more participants to see if similar findings will be deduced. Female participants could not be included in this study due to institutional restrictions and thus findings should not be generalized to all Saudi population. Also, the present study looked at attitudes among a narrow band of Saudi speakers; students who shall graduate and work in the petrol sector. Future research should investigate attitudes toward SauE in other contexts such as education (e.g., Buckingham, 2014). SauE is a relatively new concept with a gradually growing descriptive literature that attempts to distinguish this variety from other varieties in the Arab world based on various linguistic features. That said, it should be acknowledged that SauE could be fairly similar to other Englishes spoken by Gulf nationals and that identifying a norm-based SauE can be challenging. Nevertheless, I argue that the same is true with IC varieties–variation is quite wide. Overall, the findings of this study provide a preliminary understanding of Saudis’ attitudes toward their own English and call for the inclusion of SauE in EFL education at Saudi Arabia.