A Pattern-Driven Investigation of Discourse-Pragmatic Markers in Kuwaiti English

1. Introduction

Factors such as British colonialism and globalisation (Seargeant, 2012; Jenkins, 2015) have made English the world’s foremost lingua franca. With its propagation around the globe, English has impacted the local languages with which it comes into contact and is in turn influenced by these languages, leading to the development of linguistic patterns specific to certain areas, as in Thailand (Trakulkasemsuk, 2012), Nigeria (Unuabonah, 2021) and the Netherlands (Edwards, 2016). This influence can be seen not only in language-contact varieties but also, through factors such as pop culture and social media, in languages that have no direct (physical) contact with native English speakers.

Sustained language contact eventually leads to what is known as nativisation: the development of linguistic patterns that reflect characteristics of the local language(s). This type of language change can ultimately result in the emergence of distinct varieties of English (Schneider, 2007: 5–6). Alenezi (2022: 59) points out that nativisation in language-contact settings may involve changes that come about as the result of contact-induced change or so-called internal or universal changes. Thus, nativisation highlights the properties of linguistic usage and a language system, but the origin/source of change […] may be due to either continuity, innovation, or contact.

In other words, nativisation can be driven by both language-internal and -external factors. Language-internal factors include the full or partial modelling of linguistic patterns on the L1(s) (Thomason & Kaufman, 1988; Migge, 2003), giving rise to features such as grammaticalisation (Bruyn, 1996) and semantic broadening or narrowing (Fromkin et al., 2011). External factors include (i) language maintenance, through which linguistic features are passed on from generation to generation; (ii) innovation, whereby speakers use English in a way that is not based on any variety of English or is partly based on another language; and (iii) language change, induced by contact between two or more languages, during which one language incorporates features from the other language(s).

Language contact studies (e.g. Migge, 2003) have found that the extent to which people borrow vocabulary from their L1 when in contact with others depends on the intensity of the contact and the nature of their interactions. When borrowing takes place, people tend to retain content words from their native language related to flora, fauna and culture. This type of identity marking (Schneider, 2017) may give rise to linguistic innovations that are not understood by speakers of standardised varieties. In essence, nativisation—“the process of creating innovative [varietal] characteristics”—requires indigenisation: “the sociocultural process which makes a language adapted to a new sociocultural system” (Alenezi, 2022: 63). This entails the adoption of certain features that reflect people’s social identities. In Kuwait, one example is the maintenance of rhoticity by a certain group of English users, locally referred to as McChickens.

This article zooms in on discourse-pragmatic markers (DPMs) in Kuwaiti English, an as yet neglected area of research. As pragmatic features in particular are subject to the influence of social, cultural and religious factors, their in-depth analysis in a setting such as Kuwait is expected to provide insight into processes of nativisation and language change. Quantitative and qualitative corpus analyses are conducted to investigate the frequencies, sentential placement and discourse-pragmatic functions of L1-sourced DPMs and their English equivalents in Kuwaiti English. The analyses also incorporate speaker metadata to explore the influence of sociodemographic factors and the associations between the DPMs under investigation with social groups self-identifying as either ‘modern’ or ‘traditional’ speakers.

The next section provides a sociohistorical overview of Kuwait and addresses previous research on English in Kuwait as influenced by Arabic. Section 3 introduces DPMs, including their forms and functions as well as the challenges they present for L2 speakers, while Section 4 describes the corpus and methodology. Section 5 presents the results, focusing on the Arabic-derived DPMs ya’ni and alhamdulillah as well as their English-sourced equivalents. Section 6 concludes.

2. The context of Kuwait

2.1. A sociohistorical overview

Kuwait is a Muslim country in the Middle East whose official language is Arabic. It has a relatively small surface area of 17,820 km² and lies on the western edge of the Arabian Gulf, overlooking the north of the Gulf, southern Iraq and western Iran. In the early 17th century, this location made it an appealing destination for nomadic tribes from the Arabian Peninsula engaged in conflicts over water resources. At the same time, it was surrounded by major colonial powers: the British Empire and Persia (Alhajeri, 2017; Alyousifi, 2018). As geopolitical issues escalated, Kuwait aligned itself with Britain and ultimately signed a protection treaty with the British Empire in 1899, marking the ‘official’ arrival of English in the country.

As a British protectorate, Kuwait’s politics, commerce and administration were ruled by Britain. Kuwaitis were initially appointed as translators, but English gradually became more widespread, particularly among elite Kuwaitis who received scholarships to study in the UK. The language was first introduced in formal education in 1921. Following World War II, with the US emerging as the new world leader and the discovery of commercial quantities of oil, English came to be used as a lingua franca with the expats working en masse in the oil industry and related fields. For political reasons, including the ramifications of the tripartite war against Egypt, there was some resistance to English. But even after independence in 1961, the use of English continued to increase. Kuwait joined the United Nations in the 1960s, and by 1991, following its liberation from the Iraqi invasion by a US-led international coalition, English was fully entrenched in the country. More private schools teaching in either British or American English were opened and young Kuwaitis began to adopt English for peer communication. Today, rap music, literature and social-media communication in the local variety of English are on the rise among younger generations, many of whom use English even at home as their first language or primary mode of communication with their siblings and parents.

Research on Kuwaiti society tends to distinguish between the ‘sedentary’ Hadar people and ‘nomadic’ Bedouins (Alqenaie, 2011; Alhajeri, 2017). Based on empirical research and fieldwork, Alenezi (2022) finds that the picture is more complex, identifying two main groups—one more modern, the other comparatively traditional—with each comprising both Hadar and Bedouins. Those classed as ‘modern’ associate English with modernity and a sense of sophistication or ‘coolness’. They can be considered English-dominant users whose English shows little influence from Arabic. While some linguistic features have acquired distinct meanings or distributions from their use in varieties such as British English (BrE), these innovations do not appear to be attributable to L1 influence. ‘Traditional’ people, in contrast, use more Arabic-influenced features, particularly in areas such as pragmatics, as well as English forms (e.g. hopefully) influenced distributionally by their Arabic counterpart (inshallah). Alenezi (2022) shows how different social identities and contexts give rise to different stylistic options (‘modern’ versus ‘traditional’) in Kuwaiti English (KE).

In short, Kuwait offers fertile ground for research on the evolution of English varieties. Children and young people receive significant exposure to the language. It is widely used as a medium of education at all levels as well as in the media (TV, podcasts). English is also used by the many non-Kuwaitis working in the country’s service sectors, as well as being a prerequisite for professional employment for Kuwaitis themselves (Alenezi, 2022). While the language has clearly gained a great deal of ground, scant research has explored its nature, status, role and development as a variety; instead, research usually focuses on issues relating to English as a second/foreign language.

2.2. Research on English in Kuwait

To the best of my knowledge, only one study conducted to date deals with the linguistic features of KE from a World Englishes (WEs) perspective (Alenezi, 2022). The features outlined in this section are thus derived from literature in the fields of English-language learning/teaching and second-language acquisition. Such research typically takes a ‘deficit’ or error-analysis approach, interpreting local characteristics as the result of ‘negative’ L1 transfer. These studies nevertheless provide insight into how Arabic has influenced the use of English in Kuwait.

The tables below group the features identified into three main categories: a) grammar, b) pragmatics and discourse, and c) collocations. For each feature, the tables indicate the type of influence to which it can be attributed; n.d. (‘no data’) means no information was provided in the relevant reference. All features are derived from studies of written genres by subjects with an intermediate or advanced level of English and use methods such as collocation judgement tasks.

Table 1 shows that substrate influence plays an important role for various grammatical features of KE. Only one feature, the use of gonna, was attributed to the force of globalisation. Perhaps most striking in this table is the aspect feature. Arabic—be it Classical, Modern Standard or Dialectal Arabic—has only three tenses (Khalil, 1981; Kharma, 1972). With 10 salient temporal and aspectual distinctions, the equivalent English system is ripe for simplification by speakers of KE.

As can be seen from Table 2, studies reporting pragmatic features of KE tend to focus predominantly on codeswitching, notably with Arabic greetings and terms of endearment. With respect to fillers, the use of like is attributed to the influence of globalisation (Algharabali & Taqi, 2018); all others are linked to substrate influence (Meinhoff & Meinhoff, 1976).

Table 3 shows, again, a strong substrate influence on the use of collocations. For example, the Arabic /gahwa/ (coffee) + /Ɵǝgi:la/ (heavy) yields heavy coffee instead of the standard ‘strong coffee’ (Alotaibi, 2014:6). Likewise, binomials tend to be translated literally from Arabic (Alotaibi & Alotaibi, 2015).

3. Discourse-pragmatic markers

As the tables above suggest, the majority of research on the use of English by Kuwaitis focuses on morphosyntactic features. Discourse-pragmatic features have been comparatively neglected, even as they have been identified as warranting close study in L2 and postcolonial contexts given their interplay with social, cultural and religious factors (Anchimbe & Janney, 2017). Broadly, the study of pragmatics examines how language is used and meaning constructed in different social contexts (Barron et al., 2017). It encompasses a vast body of literature on features ranging from discourse markers (Schiffrin, 1987) to impoliteness on social media platforms like Facebook (Andersson, 2022) and apologetic strategies among L2 speakers of English in Kuwait (Alqenaie et al., 2022).

Turning to DPMs, over the years various definitions been offered and a number of alternative names advanced, such as discourse particles (Aijmer, 2002) and discourse connectives (Blakemore, 2002). Pichler (2013: 4) defines DPMs as follows:

Discourse-pragmatic features constitute a formally heterogeneous category of syntactically optional elements which make little or no contribution to the truth-conditional meaning of their host units, and – depending on their scope, linguistic context, as well as sequential, situational, and cognitive context – perform one or more of the following macro-functions: to express speaker stance; to guide utterance interpretation; and to structure discourse.

Pichler’s broad conceptualisation enables researchers to deal with the heterogeneity of DPMs, which encompass, according to Peterson et al. (2022a: 1),

An unlimited but delineated set of features consisting of discourse markers, pragmatic particles, discourse particles, and so on but also other elements that fulfil pragmatic functions: epistemic parentheticals, quotative, general extenders, consequence markers, politeness markers, intensifiers, and filled pauses. [emphasis original]

Given this breadth of forms and functions, DPMs form a challenging category for L2 speakers (Tyrkkö et al. 2021). While they are important for discoursal coherence and cohesion (De Rooij, 2000) and for the interpretation of the speaker’s stance, attitude and emotions (Alenezi, 2022), they are context-specific in that their function can vary from one utterance to another, and as ‘free’ elements they are not governed by syntactic rules, which makes them difficult to predict (Buysse, 2012). At the same time, the position of a DPM is crucial to the meaning of the sentence. Tyrkkö et al. (2021: 534) argue that subtle differences in sentence position and prosodic stress can alter the pragmatic meaning of a discourse marker, and as a result, even fairly advanced L2 speakers can inadvertently change the intended meaning of a sentence by using the inappropriate discourse marker or placing it in an unorthodox sentence position.

Among L2 users, therefore, higher proficiency levels have been linked to more native-like usage (Tyrkkö et al. 2021). In this context, several L2 varieties of English have been found to have a stronger preference for the sentence-initial positioning of DPMs than native varieties, as with wallahi in Nigerian English (Unuabonah, 2021) and si in Kenyan English (Muro and Unuabonah, 2022). In both cases, the authors attributed the differences to L1 influence. In a similar vein, while English native speakers have generally been found to use significantly more DPMs than non-native speakers (Schiffrin, 1987; Maschler, 1994), Alenezi (2022) revealed that some English-sourced markers such as hopefully can be modelled on the L1, leading to more frequent use.

A number of researchers have explored the influence of English-based DPMs on other languages, such as the use of pliis ‘please’ in Finnish, which was found to denote solidarity and positive politeness among young women in particular (Peterson et al., 2022c). In his study of the influence of English on Norwegian pragmatics, Andersen (2014) developed pragmatic borrowing theory, which refers to the process of transferring DPMs between languages, taking into account the semantics, functions and sentence position of DPMs. While Unuabonah (2021) took up this theory in a corpus-based study of pragmatic markers such as the Arabic wallahi ‘honestly/sincerely’ in Nigerian English, studies examining the nativisation of DPMs in transplanted varieties of English remain rare.

In the scant literature on such contexts, DPMs broadly can be grouped into (i) English-based markers such as well (Aijmer, 2013) and (ii) L1-based markers such as inshallah ‘hopefully’ in KE (Alenezi, 2022), jare ‘please’ in Nigerian English (NigE) (Unuabonah & Oladipupo, 2020) and sawa ‘okay’ in Kenyan English (KenE) and Tanzanian English (TanE) (Unuabonah & Muro, 2022). Unuabonah and Gut (2018) reported that NigE speakers used fewer types of English-based DPMs and also used them less frequently than BrE speakers. In KenE, Muro and Unuabonah (2022) found that when the English equivalents failed to convey the full pragmatic meanings of local DPMs, the latter were used instead. However, borrowed DPMs appear to lose some of their meanings and functions compared to their use in the source language (Andersen, 2014; Hennecke, 2013). In KE, wallah ‘if Allah/God wills’ was found to be largely restricted to use as a mitigating device despite having more than 10 functions in Arabic (Alenezi, 2022). In the example below, Mishary, a medical doctor, is reluctant to impose his rules on others. His use of wallah thus serves to mitigate a threat to negative face:

Aki: I would like to ask you first about safety; how can we stay safe at home and virus-free?

Mishary: Wallah, [ah], first of all, we have to agree on something; there is a very hard environment that is really free of viruses (Alenezi, 2022: 312).

The use of DPMs in all cases appears to be influenced by contextual and sociolinguistic factors (Andersen, 1998; D’Arcy, 2005; Gupta, 2006; Laitinen et al., 2017; Leimgruber et al., 2021, 2013; Leppänen et al., 2011; Pichler, 2016). In their study of KenE, Muro and Unuabonah (2022) indicate that local-based DPMs are more frequent in informal than formal contexts. Factors such as age, gender and ethnicity also play a role (Schiffrin, 1987; Maschler, 1994; Leimgruber et al., 2020), with DPMs used to index identity—including youth and urbanicity—and signal belonging (Alenezi, 2022; Muro and Unuabonah, 2022; Peterson, 2017) are therefore also considered in the analyses below.

4. Data and methods

4.1. Corpora

Given the abundance of DPMs in natural speech, DPMs are best studied using spoken corpora (Alenezi, 2022). In this study, DPMs in KE are investigated by means of the Kuwaiti English National Corpus (KENC2021; Alenezi, 2022), a spoken corpus of almost 800,000 words spread over six subcorpora (Table 4). The subcorpora encompass data from interviews, participant observation, social media and other mediated data (podcasts, TV, radio and YouTube) in which English is the sole or one of the languages used. The word count across these corpora is not balanced; this is the result of practical challenges in collecting data in a country in which English is used in ad-hoc situations rather than as an official language.

A total of 564 people contributed to the corpus (Table 5). Kuwaiti contributors (n=467, 83%) can be subdivided into those who are self-reported ‘modern’ users of English versus ‘traditional’ users, whose English is heavily influenced by Arabic. Non-Kuwaitis or expats (n=97, 17%) include Americans, Britons, Indians, and Filipinos. Like Dubai, Kuwait has a large proportion of immigrants/expats and they should not be excluded from the corpus since they are an integral part of society but there were data access issues. A number of non-Kuwaiti nationals (e.g. rig oil workers and shop workers) live in areas separate from the Kuwaiti neighbourhoods. Thus, obtaining data required a different approach and methods. I kept and fed all data coming from non-Kuwaitis to the corpus which I will expand in the future. For more details on the contributors to the corpus and its representativeness of all major segments of Kuwaiti society, see Alenezi (2022).

As reference corpora, the spoken components of the British National Corpus 2014 (BNC2014) and the Open American National Corpus (OANC) were used to facilitate cross-linguistic analysis and gain deeper insight into the behaviour of DPMs in KE.

4.2. Data extraction and analysis

Drawing on Tyrkkö and Kopaczyk’s (2018) pattern-driven approach, I conducted ngram analysis, specifically unigrams (an ngram is sequence of words, a unigram a sequence of one word). First, to identify DPMs in the corpora, wordlists were imported from KENC2021, BNC2014 and OANC into Excel and the keyness formula (%DIFF) was applied; the keyness formula identifies/highlights unique words to the target corpus when compared with the reference corpus. Second, significance tests (Fisher’s exact test, p-value threshold 0.05) were computed to identify statistically significant DPMs in KENC2021 compared to the reference corpora. Two L1-sourced markers (ya’ni and alhamdulillah) and five English-sourced markers (it means, I mean, that is, thankfully and fortunately) were identified. After exclusion of cases in which DPMs such as alhamdulillah were not used for discourse-pragmatic functions, the data were subjected to quantitative and qualitative analyses. Frequencies, distributions and discourse-pragmatic functions were examined through corpus searches and concordance analyses using the SketchEngine toolkit. A percentage score was used to calculate the preferred position of the DPMs within clauses and the preferred clause types to which they belonged. The origin, meanings and functions of the DPMs were analysed..

5. Results and discussion

Table 6 shows the normalised frequency (per million words, pmw) of the L1-sourced DPMs ya’ni and alhamdulillah and their English equivalents as used in KENC2021. An additional English equivalent of alhamdulillah is thank God, but as this does not appear in the corpus it is not provided in Table 6. As can be seen, the substrate-derived markers are vastly more frequent than their English equivalents, with the exception of I mean, as showed in the table. Below, the equivalents of ya’ni are not discussed in detail, as (unlike the equivalents of alhamdulillah) they do not differ significantly in frequency in KENC2021 compared to the reference corpora.

5.1. Ya’ni in Kuwaiti English

Ya’ni is a highly productive loan DPM from Arabic that alternates with English DPMs such as I mean and that is, signifying elaboration on prior discourse units. It is used across almost all varieties of Arabic (e.g. Syrian Arabic; Habib, 2021) and has been reported in varieties of English such as KenE and TanE (Unuabonah & Muro, 2022). In Arabic, ya’ni is used to clarify an previous utterance or to provide additional information, filling what would otherwise be an empty space; it thus serves almost exclusively as a function word that has undergone considerable semantic bleaching (Habib, 2021; Clift & Helani, 2010; Persson, 2009). In KenE and TanE, ya'ni likewise signals elaboration and clarification as well as emphasis (Unuabonah & Muro, 2022). As in Arabic, it manifests semantic similarities, and thus alternates with, that is, really, honestly, actually and you know (Unuabonah & Muro, 2022). In addition to its use as a DPM, ya'ni can also be used as an interjection device, evoking admiration, surprise and other emotions, such as sadness. KenE and TanE show a propensity towards using ya'ni in clause-initial positions and, to a lesser extent, clause-final positions (with clause-medial uses strongly disfavoured); indeed, in KenE it appears to have shifted from clause-final to clause-initial position (Unuabonah & Muro, 2022).

In KENC2021, ya’ni is distributed across the entire corpus except for audio media; this may be due to the small size of that subcorpus. Of all instances of ya’ni, 41% occurred in the new media subcorpus, 25% in visual media, 16% in situated language use (coming from online posts), 14% in interviews and 4% in situated language use (coming from people I observed). It is used by Kuwaitis and non-Kuwaitis alike in both formal and informal settings, suggesting a considerable degree of entrenchment and nativisation. As in KenE and TanE (Unuabonah & Muro, 2022), it predominantly appears in clause-initial position (81% of occurrences, Example 1). Unlike in TanE and KenE, however, KE seems to favour clause-medial positions, at 17% (Example 2), over clause-final, at 2% (Example 3).

Dr Saleh: I think that making a law a common thing in many countries, including the developed ones. Ya'ni, there are laws in Britain and European countries which incriminate [INAUDIBLE] other individuals. (Example 1; KENC2021)

Ibrahim: if you leave it open-ended, it can get so ya' ni out of control very easily. (Example 2; KENC2021)

Speaker 1: Yeah, I love it.

Speaker 2: Yes, but I couldn't finish anything, ya'ni. (Example 3; KENC2021)

Unlike its diverse role in varieties of Arabic (Habib, 2021), ya’ni appears to be used in KE primarily to signal elaboration. In example 4, from the interviews subcorpus, the participant is reflecting on whether or not her English is affected by other Englishes present in Kuwait, such as Philippine English. After indicating that she personally spoke American English, she suggested that other people’s English is affected by Arabic—using the loanword ya’ni herself despite the negative view of L1 transfer apparent in her response.

Participant 1: [My English language is] not affected. [Theirs] is affected in our own language, ya'ni, let's say, if we said like apparently in Kuwaiti it's abbarently. So, that's how. It's affected by our own, like, our tongue language. (Example 4; KENC2021)

Something similar can be seen in Example 5, where ya’ni is used twice by a speaker discussing how she teaches her children English. The first ya’ni signals an elaboration and clarification of what she has just said, while the second is arguably a continuation of that thought, introducing an elaboration of an elaboration:

Speaker 1: My husband doesn’t like the language. My children also, yeah, they think it would be like a barrier between them and me. And they take a lot of time. Ya’ni, I need to understand [that] I almost give them words. Yeah, let’s say, take a shower, for example. Ya’ni, but yeah, yeah, this is, it takes time. (Example 5; KENC2021)

The English equivalents I mean, it means and that is also appear in KENC2021, acting as elaboration markers in the same way as ya’ni. In Example 6, the speaker uses I mean to further elaborate on what he meant by “not advanced speakers of English” when discussing how Kuwaitis pronounce English words.

Participant 3: [Kuwaitis are] not advanced speakers of English, Kuwaitis, I mean, and they talk with the Kuwaiti accent … . (Example 6; KENC2021)

There seems to be little difference in KENC2021 between the locally sourced DPM and its English variants in terms of function. As Table 6 above showed, however, there are differences in terms of frequency; while I mean is considerably more common that ya’ni, the latter is in turn vastly more frequent than it means and that is.

5.2. Alhamdulillah in Kuwaiti English

Islam and the Quran have a large impact on the lives of Arabs and Muslims, as can be seen in the myriad words in Arabic containing the word Allah (‘God’), such as inshallah, mashallah and alhamdulillah (Al-Saeedi, 2012). Alhamdulillah is used to supplicate God and thank him for his blessings. It is commonly used in the everyday interactions of Arabic speakers and Muslims of all languages (Al-Saeedi, 2012; Al-Rawafi & Gunawan, 2019). Unsurprisingly, it is also in evidence in their L2 English. In KENC2021 (Alenezi, 2022: 289–290), alhamdulillah alternated with fortunately and thankfully. While alhamdulillah occurred in four subcorpora—new media, audio and visual media, and both situated language use subcorpora—and was employed in reference to a wide range of topics, from politics to health, thankfully and fortunately only occurred in the audio and visual media data, and in very low frequencies (8 pmw and 6 pmw, respectively). Alhamdulillah thus appears to be the preferred DPM in KE, its usage far exceeding that of its English counterparts in KENC2021. This may be attributable to the inseparability of Islam and Muslims’ speech (Al-Saeedi, 2012).

The sentential placement of alhamdulillah in KENC2021 also appears to be influenced by the speakers’ L1. Alhamdulillah is frequently used in the clause-initial position in the Malay quotes from talk shows in Redzwan et al. (2020: 12151). Anecdotally, I have noticed that in Kuwaiti Arabic it is also widely used in sentence-final and, perhaps to a lesser extent, clause-medial positions. In line with these observations, alhamdulillah was commonly used among Kuwaiti and non-Kuwaiti speakers in KENC2021 in various positions: clause-initial 49% (Example 7), clause-medial 3% (Example 8), clause-final 33% (Example 9) and in a standalone position 15% (Example 10).

Tareq: Kuwait is a very good country, actually, in the modern technologies, and the Ministry of Health is actually one of the top.

Aliaa: Alhamdulillah, I agree with you. I think we are well-equipped to handle this situation, and hopefully, it will go smoother than this. (Example 7; KENC2021)

Alnaqwi: And he's avoiding sugar, which is Alhamdulillah good but, to be honest with you, I know I have known you for many years, but I don't know what's your favourite dish. (Example 8; KENC2021)

Rania: That's amazing to hear alhamdulillah. So, Ms Nada, tell us more about the conference itself. So, I understand there are several committees. (Example 9; KENC2021)

Speaker 1: How are you doing, bro?

Speaker 2: I am good. How are you?

Speaker 1: How is everything?

Speaker 2: Alhamdulillah

Speaker 1: Sorry for disturbing you. (Example 10; KENC2021)

Arabic influence also impacted the sentential placement of thankfully and fortunately, which occurred in KENC2021 exclusively in a clause-initial position (Examples 11 and 12), unlike in BrE (Examples 13 and 14) and American English (Examples 15 and 16). The prominence of these English-sourced DPMs in clause-initial positions in KE echoes earlier findings in African (Unuabonah, 2017; Unuabonah & Muro, 2022; Muro & Unuabonah, 2022) and Nordic (Tyrkkö et al., 2021) varieties of English.

Dr Ahmad: What if, in a situation, you're actually dealing with a minor or a person that is not able to really sign [the] consent form? What's the process?

Dr Salim: Thankfully, the law has, you know, put points and laws for this. You know, the parent would sign the consent and agree to treatment. (Example 11; KENC2021)

Participant 6: Fortunately, I think in the Kuwaiti context that there were there was, you know, kind of a peaceful colonization for a certain period of time. (Example 12; KENC2021)

Speaker 1: Because you need your overdraft, so if, er, cos if you're fifteen hundred pounds into your overdraft which, thankfully, I'm not anymore, and then they say, right well, actually the interest-free bit's over. (Example 13; BNC2014)

Doc 552: They know me well enough though, fortunately, and then the people that don't know you well don't know you. (Example 14; BNC2014)

Doc 1650: They used to, thankfully; they've just changed, um, you don't um-hum uh-huh, you only get a choice of doctors that are in your area. (Example 15; OANC)

Speaker 51: I haven't, fortunately, been in a situation where it's been applicable to me, but, well, I think I think—I—I mean, what we've turned the schools into now are just daycare centres. (Example 16; OANC)

While the clause-initial, -final and -medial occurrences of alhamdulillah were spread across different subcorpora and used by both ‘modern’ and ‘traditional’ speakers of KE, standalone ones occurred exclusively in Islamic TV shows (visual media subcorpus) and casual conversations (situated language use [coming from posts sent in social media platforms] subcorpus) and were used by traditional speakers only. Thankfully and fortunately, on the other hand, were observed in the new media, unscripted TV shows and observations subcorpora, and were used exclusively by modern speakers.

Alhamdulillah denotes an attitudinal epistemic stance in KENC2021, exhibiting similarities with the epistemic stance adverbs thankfully and fortunately in that it reveals the speaker’s attitudes towards the proposition stated by either the speaker or hearer. In example 7 above, Aliaa is reacting to Tariq’s comment that Kuwait is coping with the latest advancements in the technology and health sectors. She begins her response with alhamdulillah to demonstrate her agreement with Tariq’s proposition. Alhamdulillah largely parallels its use in Arabic, and thankfully and fortunately are modelled on their L1-sourced DPM, showing strikingly similar functions to alhamdulillah in KENC2021.

6. Conclusion

This study examined the impact of Arabic on discourse-pragmatic markers as used in Kuwaiti English in the context of language contact. It confirms the substrate influence identified in earlier research such as that of Algharabali and Taqi (2018), although they followed an error-analysis approach. In my data, Arabic-derived DPMs coexisted with their English equivalents, fulfilled similar roles and occurred in both formal and informal settings, suggesting that they have become an integral part of this variety of English. Indeed, for the most part ya’ni and alhamdulillah vastly surpassed their English equivalents in terms of frequency. At the same time, the semantics and functions of these DPMs appeared to be narrower than in L1 contexts. Similarly, DPMs in KE were more likely to be placed in clause-initial positions than in BrE and AmE, echoing findings for various African Englishes (Unuabonah & Muro, 2022).